HIVe Home
Posted: 25 Feb 2013
2013 Keystone symposia on HIV Vaccines - Meeting Report
Author: Yegor Voronin

“HIV Vaccines”
Keystone Resort
Keystone, Colorado
10-15 February, 2013

The “HIV Vaccines” meeting this year was organized concurrently with the “B Cell Development and Function” meeting, reflecting the recent attention on antibodies and B cell responses in the HIV vaccine field. Both meetings were full of interesting presentations and new data. Below is a small sample of what’s been discussed.


Dr. Michel Nussenzweig expanded on his previous work, which showed that a combination of five broadly neutralizing antibodies (bnAbs) was required to suppress HIV in infected humanized mice. His group put the infected mice on ART to suppress the virus before adding the antibodies, and then stopped therapy. In this system, a single monoclonal was sufficient to preserve the undetectable viral load in most mice. Interestingly, this worked only with some antibodies and not others (e.g. it did not work with PG16). Viral suppression in this system could be achieved by passive immunization and also by delivering antibodies with an AAV vector. Dr. Nussenzweig pointed out that it’s not clear from the current data whether this approach would be applicable to humans. On one hand, humans usually have higher viral load and higher viral diversity than humanized mice. On the other hand, these mice lacked the complete immune system – they did not have the humoral and cellular responses to the virus that are present in infected humans.

Dr. Ruth Ruprecht presented data comparing antibodies of different isotypes that have the same specificity. The IgG1 antibody HGN194 was converted into IgA1 and IgA2 versions and all three were tested for the ability to protect passively-immunized macaques from SHIV challenge (the antibodies were applied topically prior to rectal challenge). IgA1 version of the antibody was significantly better in this assay than the other two antibodies. The binding to Env was similar for all three, but IgA1 version was significantly better at virion capture, which perhaps explained its higher efficacy. Dr. Ruprecht hypothesized that the effect was due to a much wider separation of the Ag-binding regions in this antibody, which allowed cross-linking of viral particles and prevention of transcytosis.


Dr. Genoveffa Franchini presented the results of testing RV144-like regimens in a NHP model. The vaccine most closely resembling pox-prime protein-boost (adjuvanted with Alum) of RV144 showed a 40% efficacy in the first 12 weeks. Changing the prime to a DNA-based vaccine did not significantly impact the efficacy, but priming with Ad26-delivered antigen resulted in complete loss of protection. Changing the boost to an envelope protein fused to CD4 receptor (believed to expose CD4-induced epitopes) led to reduced efficacy. Importantly, changing the adjuvant used in the vaccine from Alum to MF59 also led to a complete loss of efficacy in these studies. The differences in immunogenicity of the two regimens were carefully investigated in an effort to identify the mechanisms responsible for efficacy. The majority of assays showed either no difference in immunogenicity of the two vaccines or higher immune responses for MF59-adjuvanted vaccine. This included the levels of binding Abs, neutralization, ADCC activity, CD4 responses (CD8 responses were minimal in both arms). However, Alum-containing vaccine resulted in higher levels of binding IgGs in the mucosal compartments (both vaginal and rectal). Further studies are needed to understand whether these mucosal Abs were responsible for the vaccine efficacy in these experiments. Meanwhile, these results may affect the planned repeat of the RV144 trial in South Africa, which was going to use MF59 as an adjuvant.

More data on the comparison of Alum and MF59 (and also of Alum supplemented with a TLR7 agonist) in NHP models were presented by Dr. Karin Lore. Injecting HIV Env alone or in combination with each of these adjuvants, her group looked at a number of events that may affect immunogenicity of a vaccine. Looking at the infiltration of cells at the site of injection, they observed that main cells attracted by all adjuvants were neutrophils (monocytes, mDCs and pDCs were also found at higher frequencies) and that the effect did not depend on the presence of Env. However, many more neutrophils could be found in draining lymph nodes after vaccination with MF59 than with Alum or Alum+TLR7. Increased IFN expression in both muscle and draining lymph nodes (LNs), as well as maturation of mDCs and monocytes, was only observed in the Alum+TLR7 arm. Uptake of Env and transport of the antigen to LNs differed significantly between adjuvants: mDCs were the primary transporters when Alum or Alum+TLR7 were used, while in the MF59 arm the Env was mostly transported by monocytes and mDCs were not affected. Dr. Lore concluded by saying that Alum+TLR7 is a cell-type specific adjuvant, which increases IFN production and leads to more efficient Env uptake, transport and presentation, while MF59 provides a more general inflammation signals activating monocytes and neutrophils, which are less potent in presentation of the antigen to T cells, but are present in higher numbers.

Dr. Joseph Francica compared binding Ab titers elicited by a trimeric gp140 Env supplemented with various adjuvants. The titers were higher and more durable for MF59, pICLC, and ISCOMS than for Alum. Supplementing Alum with TLR4 or TLR7 agonists improved the titers to the level observed with other adjuvants. However, these responses were measured in plasma, while the results from Franchini group suggest that mucosal responses may be more important for protection.

Dr. David Weiner’s group is working on new adjuvants MEC and CTACK, which may stimulate B cell responses and target them to mucosal compartment. Adding plasmids for these cytokines to a DNA vaccine improves Ab binding titers and leads to protection against vaginal E660 challenge. Interestingly, the best correlate of protection for this vaccine is vaginal IgA responses, while the systemic IgA levels do not correlate with protection.

New Technologies

Dr. Ron Germain’s group developed a new technology to observe innate immune responses in mouse tissues. They pre-populate muscle tissue with various labeled immune cells and then can observe their behavior in response to different stimuli. For example, they can see neutrophils swarming to a local tissue damage or infection and by testing cells from various knockout animals can identify molecules important for this behavior. He also showed how T cells respond to an antigen by stopping their movements for a short period of time, during which they secrete cytokines. If the level of antigen remains stable, T cells eventually recover and resume their movement in tissues; higher levels of antigen are required to repeatedly elicit the arresting behavior. Currently the technology is limited by the fact that cells have to be pre-stained with one of several available markers. However, they are working on a multi-parameter confocal microscopy, which will allow them to visualize particular subtypes of cells with up to 10 different markers.

Dr. Mario Roederer presented some proof-of-concept work in using Fluidigm technology to identify various T cell subtypes. The technology allows transcriptomics analysis of a predefined set of genes for ether a bulk cell population or an individual cell. Using this assay his group was able to identify specific signatures in samples from subjects vaccinated in the RV144 trial, something that was not possible with more conventional assays because of the extremely low level of T cell responses in that trial. These signatures can be interrogated for new correlates of protection.

T-cell responses

Many different factors impact CD8-based immune responses, so Dr. Bruce Walker’s group developed an in vitro assay where some of these variables can be controlled. They isolate tetramer-sorted CD8 cells that target KK10 epitope and use HLA B27-transfected HIV-infected cells as targets. In this assay, CD8 cells from elite controllers are drastically better at suppressing the virus than CD8 cells from rapid progressors. The in vitro model allows to further dissect CD8 responses from the two groups of patients by looking at specific clonotypes of CD8 cells. Clonotypes obtained from controllers had variable activity against HIV, some being significantly better than others in inhibiting viral replication. On the other hand, all clonotypes from progressors had minimal activity. Moreover, clonotypes from controllers also had more breadth in recognizing different variants of the KK10 epitope, had higher proliferative responses, and exhibited faster kinetics of responses than clonotypes from progressors (although the final magnitude of responses was the same).

Innate immunity

Dr. Nina Bhardwaj’s talk focused on the role the microparticles (MPs) play in HIV pathogenesis. Normally the blood contains MPs derived from platelets, but in HIV infection other MPs appear, which are derived from cells undergoing apoptosis. While these MPs have signals on their surface that stimulate their absorption and digestion by macrophages, they continue to be present in the blood, perhaps because the macrophages are overloaded by the excess of apoptosis associated with HIV infection. Bhardwaj group showed that in vitro these MPs suppress DC functions, including secretion of IL12 and the ability to prime CD4 and CD8 cells. CD44 receptor on the surface of DCs seems to be important for this activity of MPs; blocking this molecule partially restores DC functionality.

Dr. Bhardwaj also showed data on plasmacytoid DCs (pDCs), which are activated by HIV via TLR7. pDCs secrete INF and other pro-inflammatory cytokines. While this activation persists for weeks and may contribute to chronic immune activation during HIV infection, the same cells according to Bhardwaj stimulate differentiation of Tregs from naïve T cell population.

Dr. Persephone Borrow looked at HIV evolution of sensitivity to IFNs. Previously it was shown that pro-inflammatory cytokines have highest levels very early after infection and their levels decrease in chronic infection. Borrow hypothesized that transmitted founder (TF) viruses may possess higher resistance to IFN effects. By cloning and/or reconstructing TF and chronic viruses from longitudinal samples and testing their ability to replicate in the presence of IFNs in vitro, her group indeed observed that TF viruses are more resistant to IFNs than chronic viruses, and that resistance is lost fairly quickly, within the first 6 months of infection.