Is an HIV Vaccine Possible

1
Is an HIV Vaccine Possible?

• David I. Watkins

2
Monkey Testing of Vaccines Problem
Almost all NHP SIV challenges use homologous
viruses - never going to happen with HIV.
3
Homologous Challenge
4
DNA/Ad5 Gag/Tat/Nef/Rev, Challenge SIVmac239
Peak of 4 million and Set Point of 5,000
Wilson et al., J.Virol. 2006 80 (12) 5875-85
5
Conclusions

• Vaccine-induced T-cell responses can control
  replication of SIV without neutralizing
  antibodies.

Caveats

• How much damage was inflicted by 106 copies/ml
  during acute phase?
• Vaccine and challenge virus exactly matched
  (homologous challenge).

6
The goal of a T-cell-based vaccine is to prevent
transmission in the chronic phase.
7
DNA/Ad5 Encoding All SIV Proteins (except for
Env) Induces High Frequency and Broad T-cell
Responses
8
Vaccination scheme
Challenge at 61 weeks
Mamu-A02
Mamu-B08- Mamu-B17- Mamu-A01-
SIVmac239 sequences
9
Peak vaccination responses SFCs per million
PBMC against each protein
Mean 12,360
Positive result SFCs/106 PBMC gt 100, and gt 2
SDs above background.
10
Number of epitopes recognized
Mean 20
Positive result SFCs/106 PBMC gt 100, and gt 2
SDs above background.
11
Challenge Choices

• SIVmac239/SIVmac251- homologous to vaccine
  inserts.
• SIVsmE660, heterologous swarm virus.

12
A Heterologous Challenge

• Animals were vaccinated using sequences from
  SIVmac239.
• Animals were challenged with the swarm virus
  SIVsmE660.
• Strains in the same subtype can differ by as much
  as 20 in their Env sequence.

Gashen et al., Science 296 2354-2360 (2002)
13
(No Transcript)
14
Challenge was heterologous and low dose to mimic
human HIV infection

• Keele et al. showed that the number of viral
  variants that cross a mucosal membrane is dose
  dependent.
• In humans, very few variants are present in the
  early acute phase.
• We titrated our SIVsmE660 stock so that we had
  1-3 viral variants cross the mucosal barrier.

Keele, BF, et al. PNAS 1047552-7557 (2008)
15
Vaccination and challenge scheme
16
The vaccine does not offer protection against
acquisition of SIV infection
17
Virus is undetectable in one of eight naïve
control animals six of eight had peaks greater
than one million and two have died of AIDS
18
Virus is undetectable in six of eight vaccinees
six had acute peaks of less than one million
19
Vaccinees Control Acute and Chronic Phase Viral
Replication of Heterologous SIVsmE660
vRNA copy Eq/ml plasma
Weeks post infection
20
An animal with peak viremia of 302 viral Eq/ml
has high frequency anamnestic responses
Assays done d21 post infection. Peak VL 302
21
Frequency of anamnestic whole PBMC responses
Mean 8,174
Positive result SFCs/106 PBMC gt 100, and gt 2
SDs above background.
22
Breadth of anamnestic whole PBMC responses
Mean 10
Positive result SFCs/106 PBMC gt 100, and gt 2
SDs above background.
23
Frequency of anamnestic CD8-depleted responses
Mean 2,151
Positive result SFCs/106 PBMC gt 100, and gt 2
SDs above background.
24
Breadth of anamnestic CD8-depleted responses
Mean 7
Positive result SFCs/106 PBMC gt 100, and gt 2
SDs above background.
25
Mucosal E660 Challenge

• After up to five mucosal challenges with 800
  TCID50, we infected five of eight vaccinees and
  six of eight controls.
• We then challenged our remaining (uninfected)
  vaccinees (3) and controls (2) with 4,000 TCID50
  of E660 every two weeks.
• We have now infected all three remaining
  vaccinees and both controls after up to 6
  additional challenges.
• The vaccinees averaged 31,600 copies/ml at peak
  and the controls averaged 2,500,000 copies/ml at
  peak.

26
Mucosal E660 Challenge Caveats

• Would Ad5 Gag/Pol/Nef (STEP protocol) control
  replication of SIVsmE660 after repeated mucosal
  challenge? (But Ad5 Gag is more immunogenic in
  monkeys than humans)
• Will SIVmac239?Nef vaccinated animals control
  replication of SIVsmE660 after repeated mucosal
  challenge?
• How long will control last?

27
Speculations

• Large vaccine vectors may give the host too many
  epitopes to choose from, thereby diminishing the
  insert-specific response.
• If vaccinees have seen the vector before, then
  pre-existing vector-specific T cells may expand
  preferentially drowning out the response to the
  HIV transgene.
• We need small vectors that have not been seen by
  the host previously.
• Perhaps we should also deliver large ORFs in
  small segments to overcome immunodominant
  responses drowning out the subdominant responses

28
Final Point It is Possible to Make a CD8-Based
Vaccine

• First, of course it would be ideal to induce
  neutralizing antibodies. A combination of both
  vaccine-induced antibodies and CD8 cells would be
  optimal
• BUT WE DO NOT HAVE ANY CANDIDATE ANTIBODY-BASED
  VACCINES YET
• Attenuated SIVmac239 likely works because of MHC
  class I-CD8 interactions
• Macaque vaccine regimens (no Env) based on
  induction of CD8 cells control viral replication
  in outbred macaques (not B08, B17 postive)
  after stringent SIV challenge
• The majority of elite controller humans and
  macaques express particular MHC class I alleles
  (B57, B27, B08, B17)
• SO CD8 CELLS CAN CONTROL VIRAL REPLICATION IN THE
  ABSENCE OF NEUTRALIZING ANTIBODIES

29
Acknowledgements
30
Acknowledgements
University of Wisconsin Nancy Wilson Jason
Reed David Ahn Matt Reynolds Nick Maness Caitlin
Mac Nair John Loffredo Thomas Friedrich Shari
Piaskowski Eva Rakasz Funded by NIH and IAVI
University of Alabama - Birmingham Beatrice
Hahn George Shaw Brandon Keele Funded by
CHAVI David B. Allison Stephen Erickson
Merck Research Labs John Shiver Danilo
Casimiro Andrew Betts Xiaoping Liang Fubao
Wang Elizabeth Thoryk
NCI-Frederick Jeff Lifson Mike Piatak