Greenberg: Mouse Immunology Laboratory
For the first time in over 25 years of research, a modest but significant level of protection was achieved in a human clinical HIV vaccine trial. However, the RV144 trial, using a non-replicating poxvirus prime with a protein boost, raises many questions about the mechanism of protection observed. For instance, T cell responses were detected in only a fraction of vaccine recipients, and broadly neutralizing antibody responses were not observed. Despite the lack of a concrete understanding of the correlates of protection against HIV, these results remain consistent with the hypothesis that a successful vaccine will require cooperation between both T and B cell responses and designing more effective HIV vaccines would be greatly facilitated by a more comprehensive understanding of how vaccines stimulate these immune subsets. Mouse models that would allow researchers to monitor immune responses to vaccines could provide insights that would facilitate designing and refining preventative HIV vaccines.
Researchers in the Mouse Immunology Laboratory led by Phil Greenberg and Joe Blattman have developed sensitive mouse models to evaluate the ability of candidate HIV vaccines to engage HIV-specific B and T cells. Such models allow researchers to evaluate the abilities of vaccine candidates to elicit effector, long-term memory and mucosal T cell responses and/or functional and neutralizing B cell antibody responses. Moreover, these models can be used to determine the innate and adaptive immunologic mechanisms required to achieve these responses, to provide a comparative means for predicting which candidate vaccines will exhibit the greatest activity in humans, and to identify principles that can be incorporated into HIV vaccine vectors or formulations to augment the generation of potentially protective human immune responses. For T cells, these models provide a standardized platform for evaluation of strategies to increase the breadth of T cell responses as a means to increase coverage for circulating HIV strains and to prevent escape after infection. For B cells, these models can be used as part of strategies for generating broadly neutralizing Ab responses to evaluate the ability of HIV vaccines to engage precursors B cells capable of developing into a cell producing a selected neutralizing Ab but that do not recognize HIV with high avidity.
The development of mouse models that allow researchers to monitor CD4 and CD8 T cell responses has been the first step, and creating mouse models that allows the evaluation of broadly neutralizing antibody responses is now proceeding. Creation of these mice has required researchers to genetically modify mice to make components of the murine immune system more closely resemble the human system. The models being developed and those already developed should provide an accurate screen for determining what vaccines and formulations should be selected for testing in non-human primates and humans, and insights into how candidate vaccines might be improved to better achieve the desired immune responses.
- Identify immunodominant and subdominant CD4 and CD8 T cell responses to HIV proteins in vaccinated mice; create transgenic mice expressing TCR genes specific for the HIV epitopes recognized; and develop a mouse model utilizing transfer of T cells from TCR-transgenic mice to facilitate quantitation and in vivo tracking of responses to HIV vaccines.
- Employ this sensitive mouse model for examining HIV-specific T cell responses to evaluate and compare the immunogenicity of candidate HIV vaccines, as well as the ability of these vaccines to induce long-lasting memory responses specific for multiple epitopes, generate T cell responses that home to mucosal sites, and induce follicular helper T cells that promote generation of B cell responses.
- Adapt this mouse model for use with gene-deficient or transgenic mice to determine which cellular and molecular components of the innate and adaptive immune system, as well as which negative regulators of responses, are engaged during responses to candidate HIV vaccines. Use the lessons learned from such studies to improve HIV vaccine immunogenicity by addition/augmentation of identified positive innate and adaptive immune signals, or by disruption/removal of negative signals, to enhance the magnitude of responses and the homing of responding cells to mucosal sites induced by individual candidate HIV vaccines.
- Create knock-in, transgenic, and retrogenic mice that express immunoglobulin (Ig) genes that encode for broadly neutralizing antibodies in trackable naïve B cells; determine if immunogens are capable of eliciting such antibody responses and/or have the potential to induce even higher affinity and more strongly neutralizing antibodies.
- Identify unmutated (germline) immunoglobulin precursors of known neutralizing antibodies in human naïve B cells by deep sequencing. The analysis of the normal human naïve B cell repertoire can identify bnAbs for which unmutated germline precursors are commonly found, allowing prioritization of immunogens designed to elicit these bnAbs.
- Create knock-in and retrogenic mice that express unmutated (germline) immunoglobulin precursors of broadly neutralizing antibodies expected to be found in HIV negative vaccine recipients; identify immunogens capable of engaging these receptors in vivo; develop vaccination strategies which induce the maturation of these precursors into broadly neutralizing antibody responses.
Establishment of the mouse model, for studying T cell responses, which employs the administration of trackable HIV-specific T cell receptor (TCR)-transgenic CD4 or CD8 T cells into mice prior to vaccination, required the development of prototype vaccines expressing HIV proteins as a representative set of immunogens, identification of the epitopes derived from these HIV proteins recognized by host immunodominant and subdominant CD4 and CD8 T cell responses, cloning and expansion of T cells specific for these epitopes, and isolation and validation of the clonotypic TCR genes from these T cells for the subsequent generation of TCR-transgenic mice.
- Using DNA, MVA, and Ad5 prototype vaccines expressing consensus HIV clade B gag, env, pol, & rev genes, the researchers have identified immunodominant and subdominant HIV-specific CD8 and CD4 T cell responses elicited following immunization. T cell clones specific for each of these epitopes have been generated, the TCR genes from selected high or low avidity clones isolated, and the cloned TCR genes validated for reactivity and affinity for the HIV protein.
- The researchers have produced TCR-transgenic mice that express these isolated receptors on T cells. Constructs for efficient expression of TCR genes were developed and pronuclear injections with constructs encoding for corresponding pairs of TCRa and TCRb genes were performed resulting in multiple founder mice with integrated transgenes. The resulting mice were screened for expression of HIV-specific TCR transgenes on a majority of mature peripheral blood T cells, founders identified for: high, intermediate, and low avidity dominant CD8 gag-specific T cells; high and low avidity subdominant CD8 gag-specific T cells recognizing multiple epitopes; high and low avidity dominant and subdominant gag-specific CD4 T cells; dominant and subdominant env CD4 and CD8 T cells. These founders have been expanded to create strains useful for further studies.
- Low numbers of T cells, in the range of endogenous T cell precursor frequencies and which do not result in differences in the phenotype of resulting effector and memory populations, from the above mouse strains have been transferred into naïve B6 recipient mice and reactivity with DNA, MVA, and Ad5 prototype HIV vaccines confirmed. The resulting antigen-specific effector and memory T cell responses have been tracked in systemic and peripheral/mucosal sites and behave similarly to endogenous responding T cells.
The researchers have begun generation of B cell mouse models that will allow testing of the ability of candidate HIV vaccines to engage highly mutated broadly neutralizing antibodies or predicted germline precursors antibodies as well as the ability of such vaccines to generate B cell memory and antibody-producing plasma cells. These mice will be used in conjunction with the developed TCR transgenic mice to provide a standardized platform for HIV vaccine evaluation.
- The researchers are generating mice that will express the broadly neutralizing 4E10 antibody, which recognizes an epitope within the conserved membrane proximal external region (MPER) of gp41. 4E10 immunoglobulin heavy chain (IgH) knock-in mice have been generated by gene targeting in mouse embryonic stem cells (ES cells), and are being bred to 4E10 IgKappa light chain transgenic mice (see below). The researchers have generated transgenic mice expressing the 4E10 kappa light chain, in which cells expressing the transgene are marked by a cell-surface Thy1.1-PDGFR fusion protein. The majority of B cells in these mice express the transgenic light chain and the cell-surface Thy1.1 marker, detectable by flow cytometry. Crossing these mice with the 4E10 heavy chain knock-in mice will allow tracking of 4E10 antibody-expressing cells in vivo after adoptive transfer and immunization.
- Additionally, the researchers are generating mice expressing the CD4 binding site-specific bnAb VRC01, and the V3 loop-specific HIV-neutralizing 447-52D antibody. Transfected ES cells are being screened by southern blot to identify IgH-targeted clones. 447 IgLambda and VRC01 IgKappa light chain transgenic mice have been generated and will be crossed with corresponding IgH knock-in mice.
- Embryonic cells are being developed with integrated exchange cassettes that will serve as efficient universal recipients of knock-in heavy chain and light chain genes, so that mice can be rapidly made to evaluate new antibodies of interest.
- The researchers are testing constructs to allow the rapid generation of mice with marked B cells expressing any immunoglobulin of interest, through the use of retroviral gene transfer into Rag-knockout mouse hematopoietic stem cells (retrogenic mice). These mice will facilitate rapid in vivo screening of immunogens designed to activate particular neutralizing B cell receptors or their precursors.
The model systems under development are already being used to evaluate candidate HIV vaccines obtained from the VDC expressing model antigens as a means to test vector immunogenicity and the mechanisms engaged by such vectors. The following represent some of the ongoing studies:
- Parks VDC: Immunization with Adeno-Associated virus (AAV1) vectors expressing flagellin (or the minimal TLR5 binding domain of flagellin) fused to the N-terminus of the gag protein resulted in higher initial responses than did AAV expressing gag alone or flagellin fused to the C-terminus of gag. However, evaluation of memory T cell responses at later time points revealed similar magnitudes with all of the mice regardless of the AAV vector used. These memory T cell responses in AAV immunized mice were similar to those detected in Ad5, and higher than those found in DNA immunized mice. Moreover, recall responses by AAV-primed cells were no different than those primed by Ad5-primed cells, providing evidence that AAV vectors do not result in dysfunctional T cells as has been suggested.
- Letvin VDC: The safety of attenuated recombinant Mycobacterium smegmatis vectors has been tested in mice deficient in specific components of Toll-like receptor (TLR) signaling pathways. Specifically, although the wild-type M. smegmatis was lethal to all strains tested including wild-type mice, both the IKE and IKE-PLUS attenuated strains were only lethal in MyD88-/- mice, but not mice deficient in an individual toll like receptor or in the TRIF TLR adaptor protein.
- Letvin VDC: Immunization of wild-type mice with rare-serotype Adenoviral vectors (Ad5, Ad26, Ad35, or Ad5HVR48) results in robust CD8 T cell responses. Immunization of MyD88-/- mice, which have diminished TLR signaling capacity, with the same vectors results in lower T cell responses. However, a lack of any individual TLR had no effect on immunogenicity suggesting that these vectors engage multiple TLR pathways.
- Patterson VDC: Immunization with clade B gag immunogens result in a different immunodominant response than does immunization with a clade C gag immunogen. However, subdominant responses are shared between these two gag proteins during immunization of B6 mice. Therefore, we examined the effect of cross-clade prime/boost strategies on generation of HIV-specific T cell responses and found that the dominant response during priming remains dominant during boosting despite numerous differences between the epitopes in each gag protein.
- Weiss VDC: Carbopol and Polyethylenimine have been suggested as next-generation adjuvants for generating potent B cell and helper CD4 T cell responses. Protein immunization of mice with these adjuvants resulted in higher CD4 T cell responses and binding-antibody responses compared to traditional freund’s adjuvants.
- McElrath VDC: The Yellow fever virus 17D strain vaccine results in potent T and B cell responses in humans, and long-lived protection from YFV infection. Recently, YFV recombinants have been generated that express HIV immunogens. Immunization of mice with these YFV-HIV vectors resulted in T cell responses that were clearly detectable but lower in magnitude than Ad5-HIV responses. However, the phenotype of YFV-induced HIV-specific memory CD8 T cells were primarily TCM while the Ad5 induced HIV-specific memory T cell responses were primarily TEM, as has been reported.
Grant at a Glance
Phil Greenberg, MD, PhD
University of Washington, Seattle, USA
Mouse Immunology Laboratory
$10 million over 5 years, awarded June 2006
- Fred Hutchinson Cancer Research Center
External Scientific Advisory Board
- Larry Corey, Fred Hutchinson Cancer Research Center
- Robert Seder, National Institutes of Health, Vaccine Research Center
- Barton Haynes, Duke University
Progress to Date