If a relatively low level of self-tolerance in the CD8+ T-cell and B-cell compartments were to prove generalizable, it would provide an even stronger rationale to expect that addition of foreign helper epitopes to cancer vaccines would allow potent CD8+ T-cell and B-cell responses. In this issue of the European Journal of Immunology, Snook et al. [18] test whether strong CD4+ self tolerance and weaker or absent CD8+ T-cell and B-cell tolerance is a generalizable principle that is widely applicable in the design of cancer vaccines. compound screening assay The authors refer to this state of differential tolerance as “split-tolerance,” akin to the split-tolerance

often seen in allogeneic bone marrow transplantation [19]. Snook et al. [18] begin by examining the response to a key target for colorectal cancer vaccines, guanylyl cyclase C (GUCY2C), using immunization with an adenovirus expressing GUCY2C alone or also expressing an MHC class II-restricted influenza hemagglutinnin helper Hydroxychloroquine epitope (S1) [18]. They show that CD4+ T cells are tolerant of self GUCY2C but that B cells and CD8+ T cells respond robustly to GUCY2C and generate CD8+ T-cell memory if provided the linked S1 helper epitope [18]; these responses were prevented by CD4+ T-cell depletion. As expected, in knockout mice lacking

GUCY2C the CD4+ T cells were not tolerant and the S1 epitope was not required in order to generate B- and T-cell responses to GUCY2C. Immunization of BALB/c mice with adenovirus containing both GUCY2C and the S1 helper epitope generated a CD8+ T-cell-dependent reduction in lung metastases arising from GUCY2C-expressing

CT26 colorectal cancer cells and substantially extended survival (nearly eightfold RG7420 solubility dmso longer) compared with survival following immunization without the S1 epitope. Surprisingly, this protective immunity did not result in any detectable autoimmunity to healthy self-tissues that express GUCY2C [18] and therefore identification of the mechanisms leading to differential recruitment of effector cells to tumors as opposed to healthy host tissues warrants substantial investigation. The ability to manipulate recruitment would alleviate the potential dangers of achieving a maximal antitumor response. Perhaps most importantly, Snook et al. show that their conclusions are generalizable based on similar findings with different mouse strains and tumors/tumor antigens (e.g. melanoma and breast cancer antigens Trp2 and Her2, respectively), as well as additional helper epitopes such as the synthetic pan DR epitope known as PADRE [18]. In addition to the potential clinical utility, these studies highlight the underappreciated concept of differences in the level of self-tolerance of lymphocyte subsets to specific self-antigens. A key conceptual feature of the T-cell help mechanism in general and employed here is that the foreign helper (CD4+) and effector (CD8+ and B-cell) tumor epitopes must be linked (Fig. 1), meaning that they must be presented by the same antigen-presenting cell.