HIVe Home
Posted: 13 Apr 2012
2012 Keystone symposia on HIV Vaccines - Meeting Report

Categories: meeting

The 2012 Keystone joint symposia on "HIV Vaccines" and "Viral Immunity and Host Gene Influence" was held in Keystone, CO, 21-26 March.  The meeting kicked off with a keynote about hepatitis A and hepatitis C by Stanely Lemon and an update on the structure of the HIV-1 envelope trimer by Joe Sodroski.  This year's meeting featured a number of interesting advances in HIV vaccine research and development, some of which are summarized here.

Joe Sodroski (Dana-Farber Cancer Institute) used single particle cryoelectron microscopy to define the structure of unliganded and antibody-bound trimeric HIV envelope.  The trimer used is slightly more complete than the ones previously used for structural analysis as it has a longer C1 region and V3 loop and thus the complete inner domain of the trimer.  The density map has a <5Å resolution and is in agreement with Subramaniam’s trimer e-tomography image.  The unliganded trimer architecture, a triangular pyramid, is a loosely packed open interior structure that seems to be held in place by three main areas of contact between Env monomers: the trans-membrane region of gp41, the gp41 and gp120 inner domain interaction, and the gp120 association domain formed by the variable loops at the membrane-distal part of the structure.  In addition, the unliganded trimer seems to be in a high energy state, which moves to a lower energy state when bound to CD4.  After binding, the outer domain rotates as the relationship with inner domain changes and the trimer association domain splits into two parts.

Chris Miller (U. California) presented lessons learned when replicating the Step trial in the NHP model.  Using a mutant human Ad5 that replicates in macaques (create by growing Ad5 in monkey cells (hrAd5)), they infected macaques.  Infection with hrAd5 transiently and repeatedly increases pDC activation as well as IFNα and IFNγ levels and decreases TNFα and IL-2 levels.  The hrAd5 infection is localized and seems to persist in the gut.  Looking at hrAd5 positive macaques and comparing unvaccinated with those that were vaccinated with hrAd5-SIV Merck vaccine, hrAd5-empty vaccine or the SIV Merck vaccine alone, shows that none of the three options conferred protection from infection and some enhanced susceptibility to infection.  Enhancement was seen but only when vaccinating with hrAd5 SIV Merck vaccine (no enhancement with the hrAd5 empty vaccine construct).  Further studies are necessary to identify what aspects cause enhancement of susceptibility.

Jinghe Huang (Mark Connors’ lab at VRC) presented data on a newly isolated bnAb to the MPER region (10E8) that covers the 4E10 and Z13 epitopes but has no autoreactivity, doesn’t bind to phospholipids, is only 21% mutated from the germline and has similar broadness to 4E10 and similar potency to VRC01.

Amarendra Pegu (VRC) presented passive immunization studies with VRC01 in NHP. Not surprisingly, the antibody conferred sterilizing protection from mucosal challenge with SHIV.  Tissue distribution analysis showed that VRC01 associated with epithelial cells in the submucosa.

Brian Moldt (Dennis Burton’s lab at Scripps) showed that non-fucosylated b12 Ab has enhanced affinity to the Fc receptor and increased ADCC activity when compared to wt b12, but it didn’t improve protection against mucosal SHIV challenge in macaques, indicating that other mechanisms such as phagocytosis or viral trapping might have a more important role than ADCC.

Laura McCoy (Robin Weiss’ lab at UCL) presented a short talk on a potent llama VHH Ab (J3), comparable in breadth to other newly identified Abs and easier to produce because it consists of a single monomeric variable domain.

Jason McLellan (Peter Kwong’s lab at the VRC) presented nice work done at the VRC on two Abs, CH58 and CH59, identified in RV144 vaccinees.  Both Abs bind to the V1/V2 region, are poor neutralizers and likely mediate ADCC (antibody-dependent cellular cytotoxicity).  The difference with the PG9/16 Abs, which are broad and potent Nabs that bind the same region, is that CH58/59 recognize side chains of V1/V2 and apparently bind to V1/V2 when region 167-177 forms an α-helix, whereas PG9/16 recognize that region when it is in a β-strand conformation. In addition, they and only recognize the V1/V2 loop in the trimer and not in the monomer.  This suggests that the V1/V2 loop changes conformation when in monomer (α-helix) or trimer (β-strand) and will be important when designing immunogens to induce PG9-like antibodies.  

A number of people , including Samantha Hoot (on b12) and Gordon Joyce (VRC01), showed data on Ab germline precursors and how somatic hypermutation in HCDR region is needed to generate bnAbs. On the positive side, bnAbs are not as unusual as previously thought - they were present in all of the 16 patients screened (Jordan Willis).

Stephen Kent (U. of Melbourne) gave a nice presentation on the current challenges in measuring non-neutralizing antibody effector functions (ADCC, ADCVI [antibody-dependent cellular viral inhibition], ADCP [antibody-dependent cellular phagocytosis]) and reviewed the current assays, which all have some caveats.  Creating standardized assays that more accurately reflect what happens in vivo will be critical to define the role of non-neutralizing Ab effector functions in protection. 

David Watkins (U. of Miami) showed interesting data on the role of cellular immunity in protection.  He compared vaccination (using a yellow fever vector) of macaques with and without B*08 restricted CTL epitopes and challenged them with a high dose of SIVmac239.  There was no difference in immunogenicity after vaccination between the two groups, but animals with responses to immunodominant epitopes became elite controllers. 

Nelson Michael (US MHRP) updated on the latest RV144 correlate analysis showing that high IgG responses to V1/V2 correlated with increased vaccine efficacy and high env IgA correlated with risk. They found env IgA binds to gp120 C1 domain, which is a domain that likely is the target for ADCC Abs, suggesting that IgAs might have competed with ADCC mediating Abs and thus reduced protection.  However, there is no data correlating C1 with protection and the two Abs isolated from RV144 vaccinees that mediate ADCC (CH58/59) bind to the V1/V2 loop. 

Amy Chung (Galit Alter’s lab at Ragon) compared non-neutralizing Ab functions elicited by the VaxGen 003 and RV144 trials; both trials used the same env immunogen but only RV144 had an ALVAC prime.  She compared affinity of Abs to Fc receptors that mediate ADCC and ADCP or regulate immune activation noting that the trials induced different profiles.  It’s difficult to conclude what the contribution of these different Ab profiles was in modulating protection since the trial population and likely mode of transmission was different in the trials (low risk heterosexual vs high risk IDUs).

Bob Seder (VRC) presented data on studies comparing immune activation and seroprevalence profiles of different adenoviruses (human and chimp).  Ad26, Ad28 and Ad35 have lower seroprevalence than Ad5 but are less immunogenic.  He looked into the immunogenicity of ape adenoviruses which have low/no seroprevalence in humans and showed that chAd3 had the same potency as Ad5, chAd6 and Ad28 had a somewhat lower potency and Ad35 was the least potent of all the adenoviruses compared.  It also seems that Ad28 and Ad35 induce type I IFN which might explain the difference in immunogenicity as knocking out type I IFN increased the magnitude of CD8 responses.  rAd5 and chAd3 seem to work best when used as a boost to a DNA vaccine prime.

Michel Nussenzweig (Rockefeller) showed data on VRC01-like Abs they isolated from samples of the VRC01 patient, one of them being Ab 45-46 which is related to VRC01.  Introduction of a mutation on CDRH3 designed to increase the contact of the Ab to the gp120 inner domain/binding sheet (Phe43 pocket) made the Ab more broad and potent than VRC01 and all other known Nabs including the PGT Abs.  He also presented a new in vivo entry system in mice that might help measure neutralization more accurately in vivo.  Mice were humanized to express human CD4 and CCR5 molecules and neutralization is measured with luciferase.  Preliminary data shows that the in vivo system and the current neutralization assays give similar results.  More assays will be needed to ensure that only CD4 and CCR5 play a role in HIV entry into human cells.


Want to highlight more data presented at the meeting?  Please leave your comments below.