Adoptive cell transfer for cancer immunotherapy - pitfalls and resolutions

Extensive efforts in specific cancer immunotherapy have been directed at increasing the number of tumor reactive cytotoxic T cells. This was done under the hypothesis that a major factor determining the success of these therapies is the number of activated antigen-specific T cells. However, it was shown in animal models and clinical trials that tumor escape occurs in such therapeutic settings. In order to test whether a large amount of antigen-specific T cells can overcome tumor escape, we injected the MO5 melanoma cell line, which expresses Ovalbumin (Ova) into TCR transgenic OT-I mice where >95% of the T cells recognize the Ova257-264 peptide TAA. We show that even at this extreme proportion of TAA-specific T cells, tumor escape is a dominant process when a large dose of tumor is injected (5x105 cells). When a smaller dose of tumor is injected, there is no tumor take, indicating that the immune system is capable of eliminating this tumor. We further show that the escape mechanism utilized by the tumor is a reversible, epigenetic silencing of the TAA gene. In order to demonstrate that this phenomenon occurred in vivo and was not the result of prolonged in vitro tissue culture handling, we repeated some of the assays with freshly sub-cloned MO5 with similar results. We hypothesized that the predominant factor which hindered tumor elimination is the monoclonal nature of the T cell immune attack. We therefore hypothesized that adding variability to the skewed T cell repertoire of the transgenic mouse would improve the anti-tumoral response. To this end we performed an adoptive transfer of wild type C57BL/6J mice into OT-I mice and this resulted in a delay in initial tumor growth rate. We therefore conclude that immunoediting, the molding of the immunogenic phenotype of the tumor cells [3, 4], is a dominant process in massive adoptive cell transfer therapy. We therefore suggest that future research into this mode of immunotherapy should give more attention to diversifying the target of the immune attack rather than focusing on one target TAA. We further suggest that a naturally occurring phenomenon which can be harnessed to this end is epitope spreading, a process by which the targets of the immune response spontaneously increase over time [5].;The success of cellular cancer immunotherapy is also affected by the ability of the transferred cells to migrate to the site of tumor. We therefore wished to study the migration pattern of tumor specific CD8 T cells. For this purpose OT-I CD8 T cells were transferred into C57BL/6J mice bearing an MO5 tumor and the localization, proliferation and activation markers of these T cells were monitored. We found that, when the tumor mass was large, very few T cells actually entered the tumor site. Rather these T cells remained in draining lymph nodes where they were activated and proliferated extensively. Upon analysis of expression of integrins LFA-1 and VLA-4, we found that there was a systemic reduction over time in the expression of LFA-1, but not VLA-4. Since IFN-beta is known to down-regulate LFA-1 expression, we hypothesized that cells within the tumor might secrete this cytokine. A PCR for whole tumor mRNA revealed IFN-beta expression and subsequent separation of populations in the tumor revealed that the main source for this cytokine was a population of macrophages. It is therefore possible that macrophages in large tumors in our experimental system express IFN-0 over extended periods of time and cause down-regulation of LFA-1 and therefore hinder infiltration of significant amounts of CD8 T cells into the tumor site. If this is indeed the case, blocking IFN-beta or depletion of macrophages in the tumor may improve infiltration of T cells into the tumor and facilitate tumor eradication. Alternatively, the use of IFN-beta as a CD8 T cell migration restraint could be of potential clinical benefit in inflammatory disease.;A limiting factor in T cell based cancer immunotherapy is the affinity of the TAA-specific TCR. This is inherent in the source of the TCR, which is usually cloned from TAA responsive T cells that had undergone thymic negative selection and thus express a TCR of intermediate to low affinity to self epitopes. We therefore wished to improve on the affinity of a TCR specific for a self antigen in order to achieve improved T cell responses. To this end, we are currently designing and building an iterative mutagenesis process into which TCR sequences can be input and mutated. The affect of these mutations will be screened by tetramer binding and the highest affinity TCR sequences will be used to iterate the process. We believe that by this method it would be possible to improve the affinity of most given TCRs and, as a result, their reactivity.