In the recent past, many attempts have been made to construct preventive anti-tumor vaccines by using as a primary source of antigen whole tumor cells treated by different procedures to make them more immunogenic. Irradiated or genetically modified tumor cells have been used even in clinical trials .
As a different approach to obtain optimal triggering of the adaptive anti-tumor immune response, a vaccination strategy with tumor cells transduced with the AIR-1-encoded MHC-II transactivator CIITA [24, 25] has been explored in our laboratory. The rationale underlying this approach was that CIITA-transfected tumor cells may act as “surrogate APC” via MHC-II-restricted tumor-associated antigen presentation to tumor-specific TH cells for their optimal triggering.
Indeed, beside controlling MHC-II gene expression, CIITA acts on other crucial steps of the antigen processing and presentation mechanism, via upregulation of Invariant chain  and DM  expression. Moreover, crucial to this approach were the assumptions, verified by previous elegant studies, that endogenous proteins (as most tumor antigens are) could access the MHC class II pathway of antigen presentation [28, 29] and that peptides of these proteins could be recognized and serve as immunogen for TH cell triggering [30, 31].
Our studies have demonstrated that efficient rejection of CIITA-transfected tumor cells of distinct histological origin can be achieved in high percentage of injected immunocompetent syngeneic mice [26, 32, 33]. Importantly, CIITA-tumor vaccinated mice develop an anamnestic response allowing them to reject parental tumors very efficiently [26, 32, 33]. That CIITA-dependent MHC class II expression in tumor cells was instrumental to trigger a protective immune response against the parental tumor was also demonstrated by vaccination experiments with non replicating CIITA-transfected tumor cells . The immunological basis of tumor rejection upon vaccination with CIITA-transfected tumor cells was extensively investigated. Rejection and/or reduced tumor growth were mediated by tumor specific CD4+ TH and CD8+ CTL [26, 32–34]. Importantly, tumor-specific primed TH cells were long-living memory cells and they could adoptively transfer resistance to tumor growth even after many months from original stimulation , implying the stabilization of a protective phenotype of effector cells over time.
The comparative study of the tumor microenvironment and of tumor draining lymph nodes in mice injected with parental or with CIITA-transfected tumor cells provided critical insight on the mechanisms triggered by CIITA-transfected tumor cells and their possible role as surrogate APCs . Tumors derived from parental cells presented scarce infiltrate, represented mainly by macrophages and neutrophils, very few CD4+ T cells, absence of DC and CD8+ T cells. In contrast the site of CIITA-transfected tumor cells injection was rapidly infiltrated by CD4+ T cells. This was followed, days later, by the appearance of DC and CD8+ T cells and, immediately after, by the generation of extensive areas of tumor cell necrosis. The fact that CD4+ T cells colonized CIITA-tumor tissue before CD8+ T cells and DC, along with the capacity of CIITA-expressing tumor cells to process and present antigenic peptides to CD4+ T cells in vitro [32, 35], supports the hypothesis that much of the tumor-specific TH cell triggering and/or restimulation takes place in the tumor tissue itself and is directly mediated by tumor cell-derived MHC class II molecules, as previously suggested [36–38].
Interestingly, mice rejecting CIITA-tumors and mice vaccinated with CIITA-tumor cells rejecting a challenge with parental tumors displayed a polarized CD4+ TH1 cell phenotype in their tumor-draining lymph nodes, as compared to TH2-like cells found in parental tumor-bearing mice. Moreover, the fact that CIITA-transfected tumor cells could trigger a potent anamnestic and persistent anti-tumor T cell response without an apparent sequel of autoimmunity, suggests that most of the anti-tumor response was directed against tumor- and not self-derived antigens.
The success of this approach underscores the importance of the optimal MHC-II-restricted antigen presentation, suggests that most tumor cells produce sufficient amounts of tumor antigens and renders unnecessary to know a priori the nature, the identity and the immunogenic hierarchy of the tumor-associated antigen(s). It is of relevance that the idea of increasing the density of tumor antigen displayed by MHC-II molecules, as it results from our approach of rendering tumor cells MHC-II positive by stable expression of CIITA, has been pursued also by other approaches based, for example, on providing sustained antigenic epitope in endosomal compartments by constructing Invariant chain-antigenic epitope chimeras [39–41] or by treating tumor cells with epigenetic modifiers that increase the expression of tumor-specific shared antigens [42, 43].