Herpes simplex virus type 1 (HSV-1) not only causes painful recurrent oral-labial infections, it can also cause permanent brain damage and blindness. There is currently no HSV-1 vaccine. An effective vaccine must stimulate coordinated T cell responses, but the large size of the genome and the low frequency of HSV-1-specific T cells have hampered the search for the most effective T cell antigens for inclusion in a candidate vaccine. We have now developed what we believe to be novel methods to efficiently generate a genome-wide map of the responsiveness of HSV-1-specific T cells, and demonstrate the applicability of these methods to a second complex microbe, vaccinia virus. We used cross-presentation and CD137 activation-based FACS to enrich for polyclonal CD8 + T effector T cells. The HSV-1 proteome was prepared in a flexible format for analyzing both CD8 + and CD4 + T cells from study participants. Scans with participant-specific panels of artificial APCs identified an oligospecific response in each individual. Parallel CD137-based CD4 + T cell research showed discrete oligospecific recognition of HSV-1 antigens. Unexpectedly, the two HSV-1 proteins not previously considered as vaccine candidates elicited both CD8 + and CD4 + T cell responses in most HSV-1-infected individuals. In this era of microbial genomics, our methods -also demonstrated in principle for vaccinia virus for both CD8 + and CD4 + T cells -should be broadly applicable to the selection of T cell antigens for inclusion in candidate vaccines for many pathogens. IntroductionHerpes simplex virus type 1 (HSV-1) infects 60% of the US population and has a significant cumulative health care burden in addition to causing painful recurrent oral-labial infections. For example, brain and eye infections can cause permanent damage or blindness (1). HSV-1 also causes approximately 50% of clinical first-episode genital herpes in the United States. Vaccines for HSV that have been tested thus far have failed in clinical trials, including a recent phase III trial of an adjuvanted glycoprotein D (gD2) product (2). This vaccine elicits antibody and CD4 + T cell responses but fails to induce CD8 responses. Newer platforms can elicit CD8 + and CD4 + cells, but they require rationally selected T cell antigens. We therefore developed methods to permit measurement of both CD8 and CD4 responses to the complete HSV-1 proteome to begin rational prioritization of next-generation vaccine candidates.Several recent observations support the concept that an effective HSV vaccine will need to induce coordinated CD8 + and CD4 + T cell responses. HSV-1-specific CD8 + T cells localize to the site of HSV-1 infection in human and murine trigeminal ganglia (TG) (3-5), and both HSV-specific CD8 + and CD4 + T cells localize to acute and healed sites of skin infection in mice and humans, sug-

Herpes simplex virus type 1 (HSV-1) not only causes painful recurrent oral-labial infections, it can also cause permanent brain damage and blindness. There is currently no HSV-1 vaccine. An effective vaccine must stimulate coordinated T cell responses, but the large size of the genome and the low frequency of HSV-1-specific T cells have hampered the search for the most effective T cell antigens for inclusion in a candidate vaccine. We have now developed what we believe to be novel methods to efficiently generate a genome-wide map of the responsiveness of HSV-1-specific T cells, and demonstrate the applicability of these methods to a second complex microbe, vaccinia virus. We used cross-presentation and CD137 activation-based FACS to enrich for polyclonal CD8 + T effector T cells. The HSV-1 proteome was prepared in a flexible format for analyzing both CD8 + and CD4 + T cells from study participants. Scans with participant-specific panels of artificial APCs identified an oligospecific response in each individual. Parallel CD137-based CD4 + T cell research showed discrete oligospecific recognition of HSV-1 antigens. Unexpectedly, the two HSV-1 proteins not previously considered as vaccine candidates elicited both CD8 + and CD4 + T cell responses in most HSV-1-infected individuals. In this era of microbial genomics, our methods -also demonstrated in principle for vaccinia virus for both CD8 + and CD4 + T cells -should be broadly applicable to the selection of T cell antigens for inclusion in candidate vaccines for many pathogens. IntroductionHerpes simplex virus type 1 (HSV-1) infects 60% of the US population and has a significant cumulative health care burden in addition to causing painful recurrent oral-labial infections. For example, brain and eye infections can cause permanent damage or blindness (1). HSV-1 also causes approximately 50% of clinical first-episode genital herpes in the United States. Vaccines for HSV that have been tested thus far have failed in clinical trials, including a recent phase III trial of an adjuvanted glycoprotein D (gD2) product (2). This vaccine elicits antibody and CD4 + T cell responses but fails to induce CD8 responses. Newer platforms can elicit CD8 + and CD4 + cells, but they require rationally selected T cell antigens. We therefore developed methods to permit measurement of both CD8 and CD4 responses to the complete HSV-1 proteome to begin rational prioritization of next-generation vaccine candidates.Several recent observations support the concept that an effective HSV vaccine will need to induce coordinated CD8 + and CD4 + T cell responses. HSV-1-specific CD8 + T cells localize to the site of HSV-1 infection in human and murine trigeminal ganglia (TG) (3-5), and both HSV-specific CD8 + and CD4 + T cells localize to acute and healed sites of skin infection in mice and humans, sug-

CD8-only vaccines can be protective on some mouse models. HSV-1-specific CD8s localize to sites of lesion and healed skin, and to chronically infected ganglion. We enriched polyclonal HSV-1-specific CD8s from human blood using a novel cross-presentation/activation marker-based protocol. After expansion, these cell populations were quite oligoclonal as assessed by TCR beta chain CDR3 deep sequencing. To query fine specificity, artificial APC were created by transfecting Cos-7 cells with each HLA A or B heavy chain from the subject. Each HSV-1 ORF was cloned into an expression vector and separately co-transfected with each HLA allele. Activation of CD8+ cells in the polyclonal T-cell population to specific ORFs was detected by interferon-gamma ELISA. The complexity of the response ranged from 7 to 23 reactive ORF/HLA combinations per subject. Most such epitopes are HSV-1 type-specific, but some are identical within the related pathogen HSV-2. One type-common epitope was studied in HLA A*0201-transgenic mice, in which DNA vaccination with the HSV-1 full-length ORF lead to very strong epitope-specific responses. Overall, we observed a remarkably similar complexity for the CD8 response to HSV-1 by using either the deep TCR sequencing or antigenic specificity method. Novel candidate immunodominant HSV-1 CD8 antigens were identified, and no apparent skewing to virion input or immediate early proteins due to TAP inhibition was detected.