AIDS Vaccine 2008, Cape Town – Vaccine Concepts and Design
Ed Rybicki, Institute of Infectious Disease and Molecular Medicine, University of Cape Town
The fallout cloud from the failed STEP and Phambili clinical trials of the Merck adenovirus 5-vectored vaccines cast a long shadow over the conference, and especially over the area of vaccine concepts and design. Inevitably, there was debate over whether or not T-cell response-based vaccines should ever be tested on a large scale again; and there appeared to be an intense and renewed interest in broadly-neutralising antibodies, and how to elicit them. It was understandable, then, that those of whose stock in trade is T-cell vaccines were a little apprehensive going into this meeting: however, there was much to excite and much to enthuse, and in particular, several lines of evidence suggesting that T-cell vaccines are not dead and should still be vigorously pursued.
The conference opening was memorable for a number of reasons: among these was the Sizophila Choir of HIV+ folk from Cape Town, who moved many to tears with their amazing harmonies and hymns to ARVs. Another, very important reason was the presence of South Africa’s new Minister of Health, Barbara Hogan: for the first time in years in a major forum, a senior member of the SA Government affirmed that HIV causes AIDS, and that the search for a vaccine was of paramount importance to SA and the rest of the world.
Arthur C Clarke’s Third Law states that “When an elderly and distinguished scientist says something is possible, he is almost certainly correct”: it was a pleasure, therefore, to hear the certainly distinguished Stanley Plotkin (Sanofi Pasteur / Univ Penn., PA) close the evening with a calm and reasoned explanation of why he thinks vaccines against HIV are possible. He noted that HIV is not the only vaccine to see major difficulties in its development – and cited measles and CMV as object examples. He suggested that multivalent vaccine(s) and regular boosters may be necessary; that the immune response needs neutralising Ab and CD4+ and CD8+ cells, in blood and mucosa – and pointed out that these are feasible to produce for other vaccines, so why not for HIV?
The most important Keynote/Plenary talks from the point of view of T-cell vaccines were those by Julie McElrath (Fred Hutchinson Cancer Res Inst, Seattle, WA; Plenary Session 1) on immune responses in the STEP trial; Tony Fauci (NIAID/NIH; Special Keynote) on future strategy, and Bruce Walker (Mass Gen / Harvard U; Plenary 2) on correlates of protective T-cell immunity. Julie McElrath’s analysis of the STEP data was sobering, and potentially depressing, but there was a positive message: she said that T-cell epitope recognition as a result of the Merck Ad5 vaccine was inadequate, so we needed to use different strategy – such as a protein vaccine, which should almost certainly be adjuvanted for increased immunogenicity. Tony Fauci summed up current strategic thinking very well, with his analogy of a radio dial, with Discovery and Development at opposite sides: he said that the failure of all large-scale vaccine trials to date meant we should turn the dial back to Discovery, with more focus on innate immunity, animal models and adjuvants, before any more large-scale trials were done. Bruce Walker’s message, after an exhaustive analysis of “elite controllers”, was that these people have weaker CD8+ T-cell responses to HIV antigens – but they are significantly more Gag-focussed, and that stronger Env responses are correlated with increased risk of progression to AIDS. His most important comment was that the Merck vaccine / STEP trial result was a failure of product, not of the concept, and that we are not barking the wrong tree with T-cell vaccines.
A novel introduction at the Conference was Special Session 02, Innovations in AIDS Vaccine Discovery: this was chaired by Wayne Koff (IAVI, NY), and had the objective of highlighting novel strategies for vaccine development. K Reed Clark (Nationwide Children’s Hospital, Columbus, OH) presented a case for “reverse immunisation”, or using a DNA construct to express a humanised neutralising mAb: he used rAAV1 DNA to express scFv-h-C2-C3 IgG2 constructs for sustained delivery of neutralising Ab in macaques. Sterilising immunity was achieved following NAb gene transfer in the face of a pathogenic SIV challenge, and he achieved sustained (1 yr) circulating levels of 200-400 ug/ml. As a possible downside, there was an idiotypic anti-NAb response in animals which became infected. Sanjay Phogat (IAVI, NY) spoke on the use of immune complexes as vaccines: he used neutralising and non-neutralising MAb complexed to gp120 with an adjuvant (AdjuplexLAP) to generate quick and durable neutralising antibody responses against the Env protein, with immune sera neutralising 6 out of the 10 clade B viruses tested – far better and at much higher titre than adjuvanted gp120 alone. Clayton Beard (Carolina Vaccine Inst, UNC, NC) had as his goal the use of a chimaeric live alphavirus (VEE) to create a simple self-replicating entity that presents the major antigens of HIV in vivo until an appropriate immune response suppresses its growth, leaving the recipient immune to HIV. His almost complete redesign of VEE resulted in a virus expressing SIV/HIV Env and a SIV Gag modified to bind the VEE genomic encapsidation signal, which replicates to titres of ~106/5 ml culture in Ghost cells. All in all, this session was a welcome addition to the programme, and very well received.
Session OA02 – T-Cell Vaccines and Animal Models – contained several interesting approaches to T-cell vaccines. Brad Jones (Univ Toronto) opened with a description of how T-cells specific for LINE-1 (long interpersed nuclear element) retrotransposon proteins were effective at eliminating HIV-1 and HIV-2-infected cells: apparently APOBEC-3 family proteins inhibit LINE-1 transposition, and HIV Vif interference with APOBEC allows aberrant LINE-1 expression in HIV-infected cells, which leads to MHC presentation of the LINE-1 proteome, and CTL killing of the affected cells. A LINE-1-specific T-cell clone recognised, and killed within 2 hours, cells infected with 42 HIV isolates (37 of them primary isolates) from all subtypes, and HIV-2 isolates. He argued that LINE-1 proteins represented a novel, stable vaccine target as they lacked variability, and speculated that anti-LINE-1 responses could be a part of natural control of HIV, as their T-cell clone was derived from an elite controller. David Garber (Emory Univ, GA) spoke on the optimisation of modified vaccinia virus Ankara (MVA) to reduce expression of irrelevant antigenic targets: his group had essentially reduced the vector to immediate-early expression only outside of cells used for propagation, as well as lessening its immune evasion capacity by targetted deletions. Modified vectors with gag and env genes performed 3-5 fold better than MVA in macaques, and it was possible to tune responses for better CD8+ or multifunctional responses. Tomáš Hanke (Univ Oxford) presented a “universal T-cell vaccine”, HIVconsv: this was a DNA vaccine encoding a spectrum of T-cell epitopes separated by junction regions, derived from the HIV-1 proteome, concentrating on Gag and Pol, with some Env and Vif epitopes. The vaccine potentially had 270 of a documented 1100 possible HIV-1 CD8+ T-cell epitopes. T-cells from HIV-infected subjects were stimulated by vaccine epitopes: 11 of 12 subjects reacted to 2 or more peptide pools (covering, indicating good coverage. Macaque immunisation resulted in a strong, broad response as assessed by ELISpot assay. His hope was that the vaccine would redirect responses compared to natural infection, so as to negate immunodominance of one or a few epitopes.
Symposium 03 – Next Generation Vaccine Vectors – was a highlight of the Conference, with a number of excellent presentations. Dan Barouch (Beth Israel Deaconess Med Ctre, Harvard) gave a tour de force talk on what amounted to a rerun of the Merck Ad5 vaccine efficacy trial in macaques, with a gag-only heterologous Ad26/Ad5 or Ad35/Ad5 vaccination regime. The Ad26/Ad5 combination was best, 2x the Ad35/Ad5 response, which was 2x the Ad5/Ad5 response. The Ad26/Ad5 regime gave long-term (500 day) durable partial protection against challenge, with a 3x greater breadth of epitope responses than to Ad5/Ad5. The 26/5 regime elicited a good memory Gag-specific response, and similar to what Bruce Walker had said for elite human controllers, there was a significant correlation of the height and breadth of the Gag–specific response, and reduction of viral load. Dan repeated Walker’s earlier comment, with some significant evidential weight to his iteration: the STEP trial was a failure of product, not of concept.
Louis Picker (Oregon Health & Science Univ, OR) discussed how a kinetic mismatch between replication and development of T-cell clones at the site of infection could result in infection taking hold – and further, that live attenuated SIV vaccines elicited mainly effector memory (EM) cells, whereas prime-boost vaccine regimes elicited mainly central memory (CM) cells. His group used rhesus CMV – known to elicit mainly EM-dominated responses, and which can infect and reinfect monkeys, which remain infected lifelong – to vector SIV rev, nef, tat and gag genes into macaques. In contrast to the CM response of Ad5-vectored genes, these elicited EM responses, enriched in bowel and lung and other mucosa. Very weak Ab responses with no NAb were seen. Protection against infection was seen in macaques challenged by repeat low-dose intrarectal SIV, with control groups infected at a median of two doses, and vaccinees taking 8: replication of virus was eliminated or controlled very early in infection, apparently by a CD8+ T-cell independent mechanism. His message was that the CMV vaccine and the EM cell response drastically cut down transmission.
Anna-Lise Williamson (IIDMM, Univ Cape Town) closed out the session with an account of the vaccine development efforts in Cape Town under the auspices of the SA AIDS Vaccine Initiative (SAAVI). Her group has brought a DNA and an MVA-vectored heterologous prime-boost multigene HIV vaccine combination to the point of human trial after successful broad-spectrum immunogenicity trials in baboons; however, they are also developing M bovis Bacillus Calmette-Guerin (BCG) auxotrophs and the limited host range Lumpy skin disease capripoxvirus (see also P16-02) as vectors, with very promising baboon and macaque immunogenicity results with HIV genes. Additionally, the group has gone a long way in developing Pr55Gag and chimaeric Gag virus-like budded particles (VLPs) as vaccines, with good evidence of significant T-cell response boosts by VLPs of DNA-or BCG-primed immunity in mice and baboons.
An important sub-theme at the Conference was DNA vaccines: there has been a lot of disparaging talk in recent years concerning their potential efficacy; however, in the post-STEP era, the usurping adenovirus vectors have lost some popularity, and it seems the original genetic vectors have a new lease on life.
George Pavlakis (NCI, Frederick, MD) in talk OA05-01 gave a masterly account of how electroporation of optimised DNA vaccines hugely enhanced humoral and mucosal responses. He made the point forcefully that DNA vaccines had the advantage that there was no immune response to the vector, and that their preparation was rapid, scalable and safe – and that “increased expression improves the DNA vaccine result”. He noted that different forms of antigen affect the immune response that electroporation as a means of delivery increases both antigen expression and immunogenicity; that natural cytokines delivered as DNAs were effective molecular adjuvants (eg: IL-12, IL-15), and that heterologous combinations such as DNA+protein, or DNA+viral vectors could be very effective. In macaque vaccination experiments using gag, pol, nef and vif genes in combination with chemokine fusions, the group was able to get very high (30 000 sfu/106 cells) Ifnγ ELISpot results, with ~0.3% of total circulating T-cells being Ag-specific, and high serum Ab response. They got a balance of central memory and effector CD4+/CD8+ cells, which is a shift of the type of response (central memory CD4+) obtained with previous DNA vaccines.
David Weiner (Univ Pennsylvania, PA) spoke in S03 on a very similar theme – electroporation and molecular adjuvants – and reiterated that optimisation of DNA yielded excellent results. He also added manufacture as an optimization parameter, noting that it was now possible to get ~15 mg/ml of plasmid DNA: this allowed much higher, less dispersed doses of DNA. He also noted that electroporation (EP) changes the phenotype of the response. They tested an SIV DNA vaccine with IL-12, IL-15 or RANTES DNA as adjuvants in macaques: challenge showed IL-12, RANTES gave viral loads 2 log less than IL-15 DNA adjuvanted or DNA vaccine alone. Concentrated DNA was as good an immunogen as Ad5, and much better when used with IL-12 DNA: there was increased magnitude of individual responses and increased polyfunctionality.
Fiona Tanzer (IIDMM, Univ Cape Town) in OA02-02 gave an excellent example of how to improve a DNA vector. She used elements from Porcine circovirus (PCV) to significantly enhance expression from and immunogenicity of an already good DNA vector, increasing HIV Ag-specific Ifnγ ELISpot scores by 3 – 5-fold in mice using only 172 bases from the capsid gene promoter of PCV inserted upstream of a HIV-1 polygene vaccine construct.
Taken together, these talks give an indication that DNA vectors for T-cell and other vaccines are alive and kicking, and on the verge of another growth phase in their deployment. Improvements in manufacturing, antigen expression levels and DNA delivery, and parallel advances in the use of co-expressed molecular adjuvants, all herald a new era in heterologous prime-boost studies for HIV and other vaccines.
Two posters – one an oral abstract – stood out for me as vaccine design highlights of the very rich and well-attended sessions. Darrin Martin (IIDMM, Univ Cape Town) spoke briefly on P19-08: designing a recombination-proof HIV vaccine. His bioinformatic approach identifies “cold spots” for recombination in the HIV-1 genome, and suggests targets for polyepitope-based vaccines. Champiat et al. (P12-15) found that APOBEC is a T-cell target in HIV+ people: this further extends the range of invariant, HIV-induced potential T-cell vaccine targets.
A few plant production-related HIV vaccine posters caught the eye, as my group has been involved in this for years, and it is a field with much promise but as yet, no few realised achievements. Cherni et al. (P02-03) presented interesting data on gp41 MPR on Hepatitis B virus core particles made in plants; Andersson et al. (P12-08) demonstrated that transgenic Arabidopsis expressing p24 is orally immunogenic; Meyers et al. (P12-10) showed that plant-produced vaccine-relevant HIV Ag boosted DNA primed T-cell responses; Regnard et al. (P18-08) showed that plant production of HIV antigens could be significantly increased by used of a replicating geminivirus-derived vector.
As for other posters, Welte and Walwyn (Univ Witwatersrand, SA) in P01-01 demonstrated elegant mathematical modelling of acute infection and vaccine design – and Guerbois et al. (LB-32) had truly excellent expression of budded Gag-ΔV1V2Env particles from a measles-vectored vaccine.
All in all, then, the HIV vaccine enterprise is battered but still functional.