Dr. Barton Haynes, Director of the Human Vaccine Institute, welcomed the meeting participants and announced three new working group members: Dr. Susan Buchbinder, Dr. Ian Wilson, and Dr. Joseph Sodroski (not present). Dr. Haynes stated that this meeting would focus on the state of vector research, examining the current vector research pipeline, evaluating gaps in the research, and determining whether certain vectors should receive increased research emphasis.
Dr. James Bradac of the National Institute of Allergy and Infectious Diseases (NIAID) stated that the meeting also would consider vector persistence and funding for new-vector research. Dr. Bradac reviewed aspects of vaccine strategies and listed research areas currently funded. He encouraged the meeting participants to consider how the many vectors stack up with regard to immune-response safety, preexisting immune response, maximum insert size, level of antigen expression, and more. There is a need to define persistence and to consider evidence that persistence leads to strong, durable immunoresponse. Dr. Bradac encouraged the AVRWG members to consider prime/boost modalities, the use of head-to-head comparative studies, and whether the field is receiving ample support.
Dr. Rafi Ahmed, Emory Vaccine Center
Rafi Ahmed presented recent data from his laboratory regarding the establishment of humoral and cellular memory using the mouse model of infection with LCMV.
Homing receptors were only expressed on effector cells found in the spleen, inguinal and mesenteric LN.
Experiment 2. Adoptive transfer naive, effector ,or memory transgenic CD8 T cells to naive mice and harvest IEL 16h post-transfer.
Only effector cells that express mucosal homing markers were found to migrate to gut mucosa.
Conclusion: Mucosal immunization is best for inducing mucosal immunity but effector cells in spleen also capable of homing to gut tissue
Goldilocks model- memory CD8 T cell differentiation
This model proposes that “just right” amount of antigen strength/signal and duration is required to maintain maximum T cell fitness and memory.
Too Much Signal During a chronic infection virus-specific CD8+T cells failed to acquire cardinal memory T cell properties of long-term antigen-independent persistence.
Other examples include HIV, HCV, HBV
Too Little Signal
Experiment. Compare live viral infection (LCMV) vs DNA immunization for memory, recall, and protection.
Prime TCR transgenic animals with DNA (LCMV gp33) or LCMV infection
At four months transfer equal numbers of P14TCR tg T cells from immunized mice into naive mice and boost/challenge with rVVgp33 and measure tetramer positive cells (Fig 1) for recall response or measure challenge virus titers in ovaries for protection.
Dr. Philip Johnson, Children’s Hospital of Philadelphia
Dr. Johnson described the biology, genomic structure, and viral particle structure of AAV. The AAV vector is devoid of any AAV sequences except approximately 400 nucleotides of inverted repeat sequences at both ends. His AAV vector contains CMV promoter and poly(A) signal sequences. He demonstrated that the vectors are highly efficient DNA delivery vehicle for transduction of both self genes (for gene therapy) and non-self genes (for vaccines).
Three SIV vaccines have been constructed: SIV rev-gag-PR-ΔRT-RRE, SIV rev-env, and SIV RT-IN. Experimental results show that the SIV rAAV vaccines elicit antigen-specific humoral and cellular immune responses in macaques. Pathogenic SIVsm/E600 was used to challenge vaccinated animals. Protection was demonstrated in low dose challenges and low viral loads were demonstrated in high dose challenges. He also developed HIV vaccines using both serotype 1 and 2 AAV for vaccines designed to clade C HIV-1. Immunogenicity studies to these recombinant vaccines were carried out in monkeys. Results showed that vaccine was well tolerated and integration was not detected. With the help of IAVI, a Phase I clinical trial of rAAV/HIV-1 in human was initiated in December 2003 and now was fully enrolled in Belgium and Germany.
Dr. David Knipe, Harvard Medical School
David Knipe gave a summary on the construct of replication-competent and replication-defective but latency competent HSV-1 vectors that express SIV env, SIV envelope + gag/pol, and SIV env + nef.
* Vector d106 (HSV-1) was designed to completely eliminate the toxicity associated with HSV infection. Expression of all 5 of the intermediate early gene (IE) was abrogated. These are the infected cell polypeptides (ICP) 0, 4, 22, 27, and 47 and have been shown to have regulatory functions that affect the coordinated expression of the HSV genome. ICP4 and ICP27 are absolutely essential for virus replication. The function of ICP47 may be more relevant in vivo, where it blocks the presentation of antigenic peptides to CD8+ cells and may help the virus escape immune surveillance.
Data was also shown for a replication-defective mutant candidate genital herpes vaccine (HSV-2 dl5-29).
In a guinea pig model he saw;
Dr. Ronald Desrosiers, Harvard Medical School
Dr. Ronald Desrosiers presented on recombinant herpes virus vectors with emphasis on rhesus rhadinoviruses (RRV) and rhesus lymphocryptovirus (LCV) and their potential in animal models for KSHV and EBV, respectively. He also presented on a single cycle SIV infection system.
Advantages of Herpes Viruses as vectors:
Potential advantages of KSHV (Kaposi sarcoma-associated virus)-HHV-8 as a vector:
Potential advantages of rhesus rhadinoviruses (RRV) as an animal model (RRV is a close relative of KSHV / HHV-8):
Rhesus lymphocryptovirus (LCV) as an animal model for EBV:
SIV single cycle infection system:
The single cycle SIV infection system is based on a gag-pol-complemented frame shift mutant at the ribosome slip site. The data indicate that they do not spread in vitro and are cleared quickly in animals. This single cycle system also generates neutralizing antibody responses to SIV251 isolate and induces EliSpot responses to SIV gag peptides. This mutant also reduces the viral set point in some of the challenged animals.
Dr. Matthias Schnell, Thomas Jefferson University
Attenuated rabies virus (RV), a member of the Rhabdovirus family of negative-stranded RNA viruses, offers several advantages as a potential vaccine vector: (1) The modular genome is organized with short transcription stop/start sequences flanking the genes, making it relatively easy to manipulate. (2) The genome can accommodate large (and multiple) foreign genes without affecting viral replication. (3) Since the genome is RNA, it replicates exclusively in the cytoplasm, eliminating any possibility of recombination, reversion, or integration. (4) There is negligible seropositivity to RV in the general population to interfere with vaccine “take”. (5) The virus is non-cytopathic in infected cells and expresses moderate to high levels of foreign proteins for an extended period of time. (6) Further attenuation of the vaccine strain of RV has resulted in several highly attenuated second generation RV vectors that are non-cytopathic, even after intracranial infection in mice. (7) RV replicates at mucosal surfaces, where it can generate mucosal immune responses.
Recombinant RV expressing HIV-1 Gag was able to generate gag-specific CTLs in mice, and recombinant RV expressing HIV-1 Env generated long-lasting envelope gp160-specific CTLs that were able to cross-kill target cells expressing different (non-vaccine) HIV-1 envelope proteins. The RV-HIV-1 env vector also generated anti-envelope binding antibodies in the mouse studies. In a preliminary study in four rhesus macaques, RV-based vectors were shown to be non-pathogenic and safe, with no clinical signs being observed. The nonhuman primates were infected with RV expressing SHIV89.6P Env and SIVmac239 Gag and generated substantial humoral immune responses to HIV-1 gp140 envelope by 4 weeks post immunization. The anti-envelope antibodies, which were still positive (by ELISA) in 3 of 4 animals, were boosted with a second inoculation of vaccine at 55 weeks. Viral specific cellular immune responses were not strong enough to be detected after either immunization.
Studies are under way to develop second generation RV vectors that are further attenuated and safer by adding mutations in the RV glycoprotein (G) gene. The attenuated viruses have good replication profiles, generate cellular responses in mice, and do not cause lethal infections in mice after intercranial inoculation. In other studies, in which cytolytic and non-cytolytic versions of RV were compared in mice, it was found that both viruses induce similar levels of cellular memory responses, but that the non-cytolytic RV gives better humoral immune responses, making it the preferred vector, since both cellular and humoral immunity will probably be needed for an effective HIV vaccine.
Dr. Marjorie Robert-Guroff, National Cancer Institute
The use of replicating adenovirus (Ad) vectors for AIDS vaccine offers several potential advantages. Ad-based vaccines induce both cellular and humoral immunity, and can target mucosal inductive sites in the upper respiratory tract and/or intestine, leading to the generation of immune responses at mucosal sites which are the targets in HIV transmission. The replication competent vector leads to prolonged expression of the inserted genes, giving the potential for a longer duration of immune responses, and there is a long history (25 years and approximately 10 million doses) of safety with the wild-type Ad4 and Ad7 oral vaccines.
In a comparative immunogenicity study of replicating and non-replicating Ad recombinants conducted in chimpanzees, the replicating Ad5DE3/HIVMN env/rev and the non-replicating Ad5DE1, E3/HIVMNenv/rev were administered intranasally to separate groups of 5 chimps at 0 and 13 weeks, and both groups of animals were subsequently boosted with HIVSF162 gp140DV2 protein in MF-59 adjuvant at weeks 37 and 49. The replicating Ad/HIV recombinants induced stronger cellular immune responses (as measured by IFN-g secreting cells after antigen-specific stimulation), stronger proliferative responses, and higher anti-envelope antibody titers than the non-replicating adenovirus vectors. Sera from some animals in both groups were able to neutralize primary isolates of HIV, with the chimps receiving the replicating vector showing better (but not statistically significant) neutralization. The replicating Ad-recombinants were significantly better at eliciting antibodies that mediated ADCC killing of target cells.
The protective efficacy of replication competent Ad vectors was demonstrated using Ad5 host range mutant-SIV recombinants for immunization of rhesus macaques. In this study (published in Malkevitch et al., J. Immunol. 2003; Patterson et al., J.Virol. 2003; and Patterson et al., J. Virol. 2004) animals were immunized with Ad-SIV recombinants orally and intranasally at week 0, intrathecally at week 12, then boosted with SIV envelope protein or envelope peptides by the intramuscular route at weeks 24 and 36. The Ad-SIV recombinants for each group were: (1) Ad5hr-SIVenv/rev, (2) Ad5hr-SIVenv/rev + gag, (3) Ad5hr-SIVenv/rev + nef, (4) Ad5hr-SIVenv/rev + gag + nef. These animals were boosted with SIVgp120 at weeks 24 and 36. Group (5) Ad5hr-SIVenv/rev-immunized animals were boosted with a peptide polymer mimicking the CD4 binding region of gp120. All the animals were challenged intrarectally with pathogenic SIVmac251 (the original strain, provided by R. Desrosiers). After challenge, all of the control animals were infected and showed the typical pattern of SIV replication, with acute virus peak loads in the range of 107 to 109, with set points at 20 weeks of 105 – 107. In contrast, 39% of the immunized animals (across all groups) showed strong control of SIV replication. Eight of the 30 immunized animals controlled virus replication to below or at the threshold of detection (6 of them doing so by 8 weeks post challenge), and 4 of the macaques had no detectable plasma viremia at any point after challenge, although viral DNA could be detected. Viremia control at set point was correlated with cellular responses (IFN-g ELISPOT) against Env and Rev peptides, and reduction of viremia during acute infection was correlated with Env-specific binding antibodies possessing ADCC activity.
When the CD8+ T cells of the animals that showed set point control of viremia were depleted with antibody, virus replication increased significantly, then dropped again as CD8+ T cells were replenished, demonstrating that control of viremia in the animals was associated with SIV-specific CD8+ T cell responses. The protection afforded by the course of immunization was long lasting. With no intervening immunization, 8 of 11 controller animals remained protected against a second SIVmac251 challenge one year later.
In summary, replicating adenovirus HIV vaccines appear capable of eliciting systemic and mucosal viral-specific cellular and humoral immune responses that are potent and long-lasting and can result in control of virus replication. Dr. Guroff is planning a Phase I to evaluate human immune responses to rAd-HIV vaccines, and is establishing collaborative studies in nonhuman primates to try to improve the level of protection afforded by this approach.
Dr. David Strayer, Jefferson Medical College
Dr. Strayer is exploring the use of recombinant SV40 vectors (rSV or rSV40) to generate adaptive immunity to HIV and SIV antigens. He reminded that protective adaptive immune correlates are yet unknown as to optimal viral antigens and what assays would predict or identify them. He believes that certain general assumptions may be misleading: among others, that CMI (with IFNg ELISPOT read-outs) is most important, that high levels of antigen insert expression lead to better CMI, that one needs a highly antigenic vector, and that immunity should be achieved with 1 or 2 inoculations. Several potential advantages of rSV40 vectors were cited: made at high titers (≈1011 IU/ml) and can be stored lyophilized at room temperature; transduce multiple immune cell types (macrophages, T/B cells, DCs, and monocytes) and integrate into resting or cycling cells rapidly, and despite this potentially advantageous property for antigenicity, their safety history is relatively high (but clearly still a concern); “gutless” SV40 vectors (with no SV40 genes) can carry foreign gene inserts of up to 5 kb and are made at similar titers. Dr. Strayer observed that lower levels of protein expression may not necessarily be a disadvantage for generating immune responses especially since the vector itself is relatively non-antigenic (elicits no neutralizing antibody responses), and thus multiple boosts with the same vector are possible (the explanation is that SV40 enters cells via caveolae, and a microtubular transit system takes the virion straight to the nucleus where it enters and un-coats; thus, at entry, SV40 avoids endocytic/phagocytic pathways and its capsid proteins capsid proteins are not expressed nor processed as antigens).
Dr. Strayer reported somewhat idiosyncratic but neatly simplified immune response assays on spleen cells obtained from HIV-1NL4-3 and SIVmac239 env-, gag- and tat-immunized BALB/c mice (e.g., in “direct” CTL assays, he used stably transfected, cloned, histocompatibility antigen-expressing P815 target cells, with single-cell suspensions of unstimulated/unselected effector cells added directly to 51Cr-labeled antigen-P815 cells). Data from his lab show the following: rSV40s express these HIV and SIV antigens, and immunogenicity results are largely comparable with these different inserts (for time reasons, no explicit expression data were shown, though HIV nef-transduced DCs were shown); multiple injections can boost responses; the immunity generated can be long-lived (rSVgp120 elicited memory CMI >1 year compared to HIV Env alone); IL-12 accelerated CMI, and IL-15 accelerated development of memory CMI; CMI responses were seen despite BALB/c mice typically having weak type 1 responses; rSVgp120 elicited stronger antibody responses vs. HIV gp120 envelope alone; repeated inoculations of rSVgag boosted serum anti-gag antibodies. Antibody data in mice multiply inoculated with rSV40 HBsAg lend support this approach (3 to 7 inoculations with the same rSV40 vector were required to show statistically different levels compared to controls at P<0.05).
Dr. Strayer appropriately credited his own lab associates and collaborators. Strayer lab: Hayley McKee, Sandra Calarota, Rumi Kondo, and Patricia T’sao; Feitelson Lab: Mark Feitelson and Bill Sun; Universidad de Navarra in Spain: Maria Vera, Puri Fortes, and Jesus Prieto; NIAID.
Dr. William Jacobs, Howard Hughes Medical Institute
Dr. Bill Jacobs presented on Strategies to Generate Recombinant Mycobacterial Vaccines for TB, Malaria, and HIV.
Advantages of Recombinant Mycobacteria
He showed considerable data on why he thinks a Mycobacterium tuberculosis derived vaccine is safer and can protect better than BCG. Toward this end he has developed three double deleted mutant strains of attenuated MTB, 2 non-replicating and one replicating.
As a proof of concept data with the replicating mutant mc26030 he showed
Dr. Jacobs also showed that the deletion of the anti-apoptotic factor NlaA, which enables M. tuberculosis to survive longer in macrophages, can enhance the immunogenicity of attenuated MTB vaccines, by exposing more antigens to the immune system.
For HIV he showed that constructs with mycobacterium smegmatis could express HIV Env as surface, secreted, or intracellular antigens and also could be used as a DNA delivery vector.
In work done with Norm Letvin, results from a single dose DNA/DNA prime/ boost experiment in mice using a recombinant M. smegmatis HIV-1 HXBc2 gp120 expressing plasmid showed the prime produced more CD8 T cell response than when followed by the boost. There was a slight dose response effect with 108 CFU being better than 106 CFU. In vitro results via an EliSpot assay showed that the same plasmid elicited an HIV-specific CD8 T cell response that was 27 fold better than the empty vector.
In collaboration with the Duke group recombinant M. smegmatis expressing Group M consensus (Con6) envelopes gp120 and gp140CF were made. These were used to prime (2x dose of 108 or 109 CFU) for a recombinant Con6 Env rgp140CF protein boost (50ug). The anti-HIV gp140 antibody response was statistically significant (p<.05) when compared to M. smegmatis empty plasmid prime/rgp140CF 50ug protein boost and the no prime protein boost alone.
Dr Jacobs also described his malaria vaccine work with recombinant BCG expressing the PfMSP1-19 protein of Plasmodium falciparum. Balb/c mice immunized with rBCG showed antibody and T cell responses to PfMSP1-19.
Dr. Jacobs believes that the recombinant mycobacterium vaccines can deliver triple protection against HIV, malaria, and tuberculosis.
Dr. Elizabeth Hohmann, Massachusetts General Hospital
Small exploratory clinical studies can provide safety and immunogenicity data that will help direct clinical development of bacterial vectors such as L. monocytogenes. There are several advantages to using Listeria as a vector: 1) protection is mediated by MHC class I-restricted CD8+ T cells in mice, 2) It is taken up by both macrophages and dendritic cells, 3) may be effective for both prophylactic and therapeutic applications, 4) does not contain LPS, 5) previous exposure may not be a problem for vaccines, and 6) Listeria only rarely causes disease even though it is present in food and water. Limitations to using Listeria include: 1) may cause bacteremia and CNS infection, 2) pregnant women are at risk of abortion, 3) HIV patients are at risk of infection, 4) human biology, shedding and immunology are not well understood.
For human studies the oral route may be used to induce mucosal immunity. Seroconversion for IgG and IgA is usually seen. Immune responses can also be measured by ELISPOT. The Listeria delta-actA/delta-inlB mutant vaccine strain infects phagocytes but not non-phagocytes. It is cleared rapidly from the liver and is less toxic. They have made a codon optimized HIV-Gag insert for Listeria and are working to make a more attenuated killed but metabolically active vaccine. In summary, this is not a highly persistent organism, there are good attenuation strategies available, and human studies are possible.
Dr. Stephen Udem, Wyeth Pharmaceuticals
Vesicular Stomatitis Virus (VSV) is an enveloped, negative strand RNA, rapidly replicating lytic virus that is a minor pathogen of livestock and which can rarely cause usually asymptomatic infections of humans. Infection with these viruses is limited to the oral mucous membranes and the respiratory tract. The virus encodes 5 proteins (nucleocapsid, phosphoprotein polymerase subunit, matrix, G protein glycoprotein, and L protein polymerase); this simple genome can be easily manipulated to introduce foreign genes, which are robustly expressed, and to shuffle the gene order producing attenuation; attenuation can also be produced by various gene deletions and temperature sensitive mutations. Even the attenuated viruses appear to induce strong immune responses (mucosal as well as systemic responses) to both viral and inserted antigens; and there is little pre-existing immunity in human populations. The virus grows efficiently in continuous cell lines and the RNA genome does not recombine nor does it integrate into the host genome. All of these properties make this virus a good candidate for a vaccine vector.
With respect to the topic of the meeting (persistent vector-driven immunity) it was stated that, while VSV induces persistent infection in insects and some mutant viruses can induce persistent infection of mammalian cells in culture, no evidence of wild type virus persistence in natural hosts (cattle, swine, horses) nor evidence of wt or attenuated VSV persistence in mice, ferrets, or NHPs following high dose inoculations exists. However, a similar, negative-strand RNA virus, measles, can cause rare persistent infections of humans (these have serious consequences because of the neurotropic nature of measles virus); this persistence seems to be enhanced in experimental animals by mutations that reduce production of infectious virus particles (including mutations in the measles matrix gene which is analogous to the VSV matrix gene or the F and H proteins which have some analogy to the VSV G protein). So it may be possible to create persistent VSV vectors by mutating the M or G proteins. Such defective viruses could easily be propagated by complementation in a packaging cell line. The potential advantages of persistent VSV vectored vaccines include: long duration of immune responses and the safety of no vector integration. However, it is not clear whether persistent exposure to vector and insert antigens will induce the most appropriate immune responses as well as whether persistent infection can be controlled limiting duration and dissemination, or what will be the long-term consequences of persistent VSV infection.
Dr. Haynes referred the group to a chart (Appendix 1) summarizing 6 key characteristics of each of the 12 vectors addressed in this meeting and indicating current gaps in knowledge. The characteristics were the following: providing a persistent immune response; affected by preexisting immunity; suitable as a prime vaccine; suitable as a boost; providing mucosal immunity from mucosal immunization; and providing mucosal immunity from systemic immunization. The group members reviewed the terms and responded “yes,” “no,” or “unknown” for each. Overall, the chart revealed a large number of unknowns, indicating the need for research in many areas.
Considering the portfolio of vectors currently being supported by DAIDS, the WG recommended that DAIDS consider support of promising vector systems that are currently not receiving substantial DAIDS’ support, including rAAV and replication-competent adenovirus vectors. The WG was also interested in hearing updates on other promising vector systems that were not presented at the workshop, including poliovirus and measles virus vectors.
Last Updated June 27, 2011
Last Reviewed June 20, 2011