Regeneration of Tumor Antigen Specific CTLs Utilizing iPSC Technology for Off-the-Shelf Immunotherapy

In the tumor immunotherapy field tumor antigen specific T cell therapy is currently attracting attention. However, a critical issue still awaits resolution; it is difficult to efficiently expand tumor antigen specific T cells without exhaustion. To solve this problem, we are now utilizing induced pluripotent stem cell (iPSC) technology. When iPSCs are established from tumor antigen specific cytotoxic T cells (CTLs), T cells regenerated from these iPSCs (T-iPSCs) are expected to express the same T cell receptors (TCRs) as the original CTLs. Therefore, it will become possible to obtain de novo generated tumor antigen specific CTLs almost unlimitedly. In line with this concept, we have succeeded in establishing iPSCs from CTLs specific for the melanoma antigen MART-1 and regenerating MART-1 specific CTLs (Vizcardo et al, Cell Stem Cell, 2013).

The strategy mentioned above was supposed to be applied as autologous transfusion. Autologous transfusion setting, however, has some problems; i) high cost to establish T-iPSCs from each patient, ii) heterogeneity of the quality as iPSCs, iii) difficulty in getting T-iPSCs with high affinity TCRs. To solve these problems, we are now planning to apply this method to the allogeneic transfusion setting. In allogeneic setting, we can select high quality T-iPSC clones with high affinity TCRs and such clones can be utilized to many patients with the same HLA type. In this context, it is preferable to establish T-iPSCs from HLA haplotype homozygous donors, since CTLs generated from these T-iPSCs can be administered to the patients carrying the same HLA haplotype at least on one allele. Thus by utilizing allogenic T-iPSCs the cost would be reduced and the quality of T-iPSCs can be ensured.

Now we are trying to establish T-iPSCs from CTLs specific for various tumor antigen from healthy donors. We succeeded in establishing T-iPSCs from CTLs specific for EB virus related antigen LMP2 and cancer testis antigen WT1, both of which are restricted to HLA A2402, the commonest HLA-A in the Japanese. We improved in vitro culture condition and have succeeded in regenerating CD8 single positive cells expressing CD8 alpha-beta heterodimers. These regenerated CTLs exhibited very high antigen specific cytotoxicity comparable to the original CTLs. The CTLs derived from WT1 specific T-iPSCs were able to lyse some leukemia cell lines and primary leukemia cells which express endogenous WT1 protein. Administration of regenerated CTLs prolonged the survival of mice inoculated with leukemia cell line in in vivo xenograft model. In addition, we succeeded in transducing iPSCs derived from non-T cells with WT1-specific TCR genes (TCR-iPSCs) and generating CTLs from them. Generated CTLs from TCR-iPSCs exhibited comparable TCR specific cytotoxicity as regenerated CTLs from T-iPSCs. Using this method, it would be much easier to produce CTLs with defined high affinity TCRs specific for various tumor antigen.

We are currently endeavoring to apply this method clinically. In the future, we intend to establish an allogeneic transfusion system, in which various tumor antigen specific T-iPSCs/TCR-iPSCs from a wide range of HLA haplotype homozygous donors will be lined up as a "T-iPS/TCR-iPSCs cell bank", with the aim of making off- the-shelf tumor immunotherapy a reality.