Adoptive Transfer of CD8⁺ T Cells Generated from Induced Pluripotent Stem Cells Triggers Regressions of Large Tumors Along with Immunological Memory Free

Adoptive transfer of ex vivo–expanded autologous tumor-infiltrating lymphocytes (TIL) or peripheral blood T cells genetically modified to express a T-cell receptor (TCR) or a chimeric antigen receptor (CAR) specific for an antigen expressed on tumor cells can induce durable clinical responses in significant numbers of patients with various cancers (1–3). Much progress has been made over the past decade in defining the subsets of T cells mediating effective tumor regression following adoptive T-cell therapy (1, 4–6). Preclinical data have demonstrated that infusion of CD62L^high less-differentiated T cells results in great expansion and persistence of T cells (7) and tumor destruction (4–6) compared with CD62L^low more-differentiated T cells. Clinical data have shown that long-term engraftment and persistence of infused T cells have been highly associated with objective tumor responses (8, 9). However, TILs harvested for adoptive cell therapy are effector T cells that are functionally impaired and exhausted (10, 11), and further stimulation of TILs results in terminally differentiated cells that have poor survival and reduced proliferative capacity and antitumor efficacy.

This limitation of adoptive T-cell therapy can be overcome by using induced pluripotent stem cells (iPSC) that self-renew, maintain pluripotency (12–14), and can provide an unlimited source of autologous T cells for immunotherapy. Remarkable advances made in reprogramming technology over the past few years have facilitated the generation of human iPSCs from differentiated cells such as T cells (15–18). These advances now allow us to develop strategies for the use of reprogramming technology for autologous cell therapy in the future. Recent studies have shown that mature human T cells can be reprogrammed into iPSCs that can redifferentiate into functional T lymphocytes in vitro (19–22). These iPSC-derived T cells exhibit TCR gene rearrangement patterns identical to the parental T cells and harbor antigen-specific effector functions in vitro (19–21). Moreover, T-cell–derived iPSCs genetically engineered to express a CAR differentiate to CAR-expressing T cells that display antitumor immunity in a xenograft model of lymphoma (22). These studies suggest that iPSC-derived regenerated T cells can be potentially utilized for cancer immunotherapy. However, issues including whether or not T-cell–derived iPSCs can generate less-differentiated T-cell subsets that can escape immune rejection, mediate effective regression of large established tumors through their endogenous TCRs, and establish immunological memory are not known due to the lack of syngeneic mouse models.

Here, we reprogrammed TCR transgenic CD8⁺ T cells into iPSCs and established a preclinical model to determine therapeutic efficacy of iPSC-derived regenerated T cells in a mouse model for melanoma. Our studies demonstrated that adoptive transfer of less-differentiated iPSC-derived regenerated CD8⁺ T cells mediated potent in vivo antitumor reactivity and established antigen-specific immunological memory.

C57BL/6 mice and NOD/SCID mice were purchased from Harlan Laboratories and The Jackson Laboratory, respectively. Pmel-1 TCR-transgenic mice (B6.Cg Thy1^a-Tg(TcraTcrb)8Rest/J) were purchased from The Jackson Laboratory and were bred in-house. All mice were 7 to 10 weeks old at the beginning of each experiment and were housed in the Unit for Laboratory Animal Medicine at the University of Michigan (Ann Arbor, MI) in compliance with the Institutional Animal Care and Use Committee regulations.

The mouse embryonic stem cell (mESC) line R1 and MC38 murine colon adenocarcinoma cell line were gifts from Drs. Sue O'Shea and Weiping Zou (University of Michigan, Ann Arbor, MI), respectively. SNL and B16F10 cells were obtained from Cell Biolabs, Inc and ATCC, respectively. OP9 and OP9-DL1 cells were kindly provided by Dr. Juan Carlos Zúñiga-Pflücker (University of Toronto, Toronto, Canada). Cells were authenticated by morphology, phenotype, and growth, routinely screened for mycoplasma, and were maintained at 37°C in a humidified 5% CO₂ atmosphere.

Pmel-1 CD8⁺ T cells were isolated from a single-cell suspension of splenocytes using anti-CD8 beads and MACS columns (Miltenyi Biotec). The sorted 1 × 10⁵ Pmel-1 CD8⁺ T cells were pulsed with 10 μmol/L of human (h)gp100_25-33 peptide, KVPRNQDWL (GenScript) in the presence of mitomycin-C–treated 5 × 10⁴ splenocytes from B6 mice in T-cell reprogramming medium (23). After 24 hours of culture, the solution that contained Sendai virus (SeV) vectors that individually carried each of Oct3/4, Sox2, Klf4, and c-Myc was added to wells at multiplicities of infection 20. After 24 hours of infection, the cells were collected and transferred to a 10-cm dish that contained mitomycin-C–inactivated SNL feeder cells in iPSC medium. The iPSC medium was changed every other day until the colonies were picked.

Pmel-1 iPSCs and mESCs were cultured using OP9 and OP9-DL1 stromal cell cocultured system (24). In brief, 5 × 10⁴ cells were plated on a 10-cm dish containing confluent OP9 cell monolayer in OP9 medium. On day 5 of culture, cells were trypsinized off, and 5 × 10⁵ cells were transferred to a fresh culture 10-cm dish containing confluent OP9 cell monolayer in OP9 medium with the addition of hFlt3 ligand (5 ng/mL; R&D systems). On day 8 of culture and every 4 days thereafter, semiadherent stem cell–derived hematopoietic cells were collected, filtered through a 40-μm nylon mesh, and were transferred onto fresh OP9-DL1 monolayers with the addition of hFlt3 ligand and mIL7 (1 ng/mL; Peprotech). Semiadherent cells obtained from days 16 to 28 were used for further experiments.

Semiadherent Pmel-1 iPSC-derived cells on OP9-DL1 cells were harvested and filtered through a 40-μm nylon mesh. CD8-expressing cells, including CD4 CD8 double positive (DP) T cells and CD8 single positive (SP) T cells in Pmel-1 iPSC-derived cells, Pmel-1 splenocytes, and thymocytes, were isolated using anti-CD8 beads and MACS columns to eliminate OP9-DL1 cells during T-cell activation. These cells (2 × 10⁶ cells) were cultured with mIL7 (10 ng/mL) and mIL15 (10 ng/mL; Peprotech) for 2 days in the presence of 1 μmol/L of hgp100 peptide and mitomycin-C–treated splenocytes from B6 mice (5 × 10⁵ cells). These activated cells were cultured with IL7 and IL15 or IL7, IL15, and IL-2 from day 3 and used for further experiments on days 6 to 8.

Female C57BL/6 mice were injected subcutaneously with 3 × 10⁵ B16F10 cells. Mice (n = 5 for all groups) were treated 7 to 11 days later with intravenous adoptive transfer of in vitro–activated 3 × 10⁶ Pmel-1 iPSC-derived T cells or splenic T cells. We injected 20,000 IU rhIL-2 intraperitoneally once on the day of vaccination and twice a day on the two following days. In some experiments, mice were vaccinated with 100 μL of saline containing 100 μg of hgp100 peptide, 50 μg of CD40-specific mAb (clone FGK4.5, BioXcell), 50 μg of poly (I:C) (InvivoGen) s.c., and 50 mg of imiquimod cream 5% (Aldara, Fougera) applied on the vaccination sites after adoptive transfer. Tumor volumes were calculated by determining the length of short (l) and long (L) diameters (volume = l² × L/2). Experimental endpoints were reached when tumors exceeded 20 mm in diameter or when mice became moribund and showed signs of lateral recumbency, cachexia, lack of response to noxious stimuli, or observable weight loss.

Datasets were compared using 2-tailed unpaired Student t test. Survival was analyzed with the Kaplan–Meier method using GraphPad Prism 5.0 (GraphPad Software Inc.), and groups were compared using log-rank test. P < 0.05 was considered statistically significant. Data are presented as mean ± SEM.

To establish a syngeneic mouse model to evaluate in vitro and in vivo antitumor immunity of iPSC-derived antigen-specific T cells, we chose to reprogram Pmel-1 TCR transgenic CD8⁺ T cells that were developed as a system to model treatment of melanoma using adoptive T-cell therapy (25). The target antigen, pmel-17, is an ortholog of the melanocyte differentiation antigen gp100, which is often overexpressed in human melanomas (26). Adoptive transfer of in vitro–activated Pmel-1 CD8⁺ T cells expressing the anti-gp100 TCR effectively mediates the regression of large established B16 melanomas in combination with systemic administration of common γ-chain cytokine and antigen-specific vaccination with an altered peptide ligand (4, 5, 25, 27).

Reprogramming of murine lymphocytes into iPSCs to date has been successful for B cells (28) and for natural killer (NK) T cells (29). However, mature T cells are known to be among the most difficult cells to reprogram, and reprogramming of differentiated murine T cells thus far has required a p53 knockout or the addition of a doxycycline-inducible gene expression system for successful iPSC generation (30, 31). In this study, we chose the SeV vector reprogramming system (15, 19–21, 32) to generate iPSCs from Pmel-1 CD8⁺ T cells because SeV can transduce and express foreign genes efficiently into activated murine and human T cells (33) and have high reprogramming efficiency of human peripheral blood T cells (15).

First, Pmel-1 CD8⁺ T cells were stimulated with hgp100 peptide and IL2 for 24 hours before SeV infection (Fig. 1A). Efficient expression of an exogenous gene in Pmel-1 CD8⁺ T cells was confirmed by following the expression of GFP-carrying SeV (Fig. 1B). To generate iPSCs, we used SeV that individually expressed Oct3/4, Sox2, Klf4, and c-Myc. Twenty-four hours after gene introduction, the cells were replated onto feeder layers of SNL cells with mESC medium containing the self-renewal cytokine, LIF (34). We observed the appearance of mESC-like colonies at 14 to 21 days after infection with the SeV vectors. However, the colonies isolated and subcultured in conventional mESC medium quickly differentiated and lost typical mESC morphology.

A previous report found that a combination of small-molecule inhibitors, CHIR99021, a GSK3 inhibitor that can activate the Wnt signaling pathway, and PD0325901, an MEK inhibitor, could promote partially reprogrammed cells to full pluripotency (35). Therefore, we investigated whether combining dual inhibition (2i) of MEK and GSK3, LIF, and overexpression of the pluripotency transcription factors could generate iPSCs from Pmel-1 CD8⁺ T cells. Genetic transduction using the SeV system and 2i/LIF–based iPSC medium allowed us to establish two colonies with an mESC-like morphology (an efficiency of 0.004%; Fig. 1C). These iPSCs 1 and 2 were positive for alkaline phosphatase and mESC markers, SSEA1, Oct3/4, and Nanog (Fig. 1C). To further verify expression of pluripotency genes, we performed RT-PCR analysis by primer sets that distinguish endogenous from viral transcripts (Supplementary Table S1) and found that the iPSCs expressed mESC transcripts for Nanog, Oct3/4, Sox2, and Ecat1 (Fig. 1D). Transgenes, Oct3/4, Sox2, Klf4, and c-Myc, and SeV in the iPSCs were lost after several passages (Fig. 1E).

Cytogenetic analysis showed derived iPSCs maintained normal karyotype (Fig. 2A). To evaluate the differentiation potential of the iPSCs, we first performed in vitro differentiation assays. The iPSCs formed embryoid bodies efficiently in vitro, and upregulation of marker genes for all three germ layers was detected by immunostaining (Fig. 2B). Furthermore, we tested the ability of the iPSCs to form teratomas in immunodeficient NOD/SCID mice. These iPSCs differentiated into a variety of cell types from all three germ layers (Fig. 2C). Therefore, those colonies were confirmed as typical murine iPSCs. We then examined the status of TCR gene rearrangement in those iPSCs and confirmed that they retained the same rearranged configuration as the original Pmel-1 T cells (Fig. 2D). These results demonstrated that Pmel-1 CD8⁺ T cells were successfully reprogrammed to iPSCs. We hereafter referred these established Pmel-1 CD8⁺ T-cell–derived iPSCs as “Pmel-1 iPSCs.”

Directed differentiation of mESCs into T lymphocytes required the engagement of Notch receptors to the ligand DL1 expressed on the OP9-DL1 stromal cells (24). mESCs cultured on OP9-DL1 cells follow a thymocyte developmental progression from CD4 CD8 double negative stage, which can be further subdivided into four sequential stages of differentiation based on CD44 and CD25 expression to CD4 CD8 DP T cells (24). To determine whether Pmel-1 iPSCs can differentiate into mature T cells in vitro, we cultured Pmel-1 iPSCs on OP9 cells, followed by culture on OP9-DL1 cells (Fig. 3A). On day 8, a population of cells expressing CD45 began to emerge from Pmel-1 iPSCs cultured on OP9 cells (Supplementary Fig. S1A). By day 16 of culture, these CD45⁺ cells exhibited a typical thymocyte developmental progression similar to mESCs on OP9-DL1 cells, and CD4 CD8 DP T cells appeared (Supplementary Fig. S1B). On day 20, Pmel-1 iPSCs cultured on OP9-DL1 cells contained a large population of DP T cells (Fig. 3B), and the majority of DP T cells were CD8αβ T cells that expressed CD5 (Supplementary Fig. S1C). No NK cells or NK-T cells were observed in the Pmel-1 iPSCs on OP9-DL1 cells (Supplementary Fig. S1D). We found very few mESC-derived DP T cells expressed CD3 or TCRβ, and only 30% of natural DP thymocytes expressed CD3 and Pmel-1 TCRβ chain Vβ13. In contrast, virtually all (>95%) of the Pmel-1 iPSC-derived DP T cells expressed CD3 and TCR Vβ13, similar to CD8 SP Pmel-1 thymocytes and splenocytes (Fig. 3B). We further characterized TCR repertoires by high-throughput sequencing and found that a majority of regenerated Pmel-1 iPSC-derived DP and CD8 SP cells bear the same TCRβ chain gene as original Pmel-1 CD8⁺ T cells (Table 1). These results suggest that iPSCs with rearranged TCR genes efficiently differentiate into TCR-expressing cells at the preselection stage, which is consistent with human iPSCs bearing rearranged TCR genes (19, 20).

Next, we investigated whether we could differentiate Pmel-1 iPSC-derived DP T cells into CD8 SP T cells and thereby obtain large numbers of less-differentiated CD8⁺ T cells suitable for adoptive cell therapy (1, 4–6). Signaling via TCRαβ induces DP thymocytes to differentiate into mature SP T cells (36). Therefore, we stimulated Pmel-1 iPSC-derived DP T cells with their peptide epitope, hgp100_25-33 in the presence of IL7 and IL15, two cytokines that are responsible for cytotoxic lineage specification (37) as well as for generation of long-lived memory T cells (38, 39). After 2 days of activation, DP T cells disappeared with a simultaneous increase in CD8 SP T cells (Fig. 3C). Phenotypic analysis of the CD8 SP T cells revealed that expression of IL2 receptor α chain (CD25) and CD69 in Pmel-1 iPSC-derived CD8 SP T cells was higher than thymocytes but lower than splenocytes after 2 days of activation. On the basis of increased expression of CD25, we tested whether the addition of IL2 to the culture could increase the expansion of Pmel-1 iPSC-derived CD8 SP T cells. We observed a marked increase in the number of T cells after culture with IL2, IL7, and IL15 from day 3 as compared with IL7 and IL15 (Fig. 3D). As activation of T cells may be accompanied by differentiation to effector T cells and the loss of memory potential (1), we evaluated the phenotype of iPSC-derived CD8⁺ T cells after activation. Despite increased expansion, in vitro–activated iPSC-derived CD8⁺ T cells were comprised of more than 50% of CD62L^high less-differentiated T cells, retained expression of the memory T cell markers, CD122, chemokine (C-X-C motif) receptor 3 (CXCR3), and stem cell antigen-1 (Sca-1), and acquired similar phenotype as in vitro–activated pmel-1 splenic CD8⁺ T cells (Fig. 3E).

To characterize the effector function of Pmel-1 iPSC-derived CD8⁺ T cells, we examined intracellular IFNγ and TNFα levels after activation using anti-CD3/CD28 mAbs or with hgp100 peptide. Influenza nucleoprotein epitope NP_366-374 was used as an irrelevant control peptide. A high frequency of cytokine-producing Pmel-1 iPSC-derived CD8⁺ T cells was documented after stimulation with anti-CD3/CD28 mAbs or with the hgp100_25-33 antigen. Importantly, cytokine production occurred in response to cognate antigen, as evidenced by the failure to detect cytokine-producing T cells in response to the NP_366-374 antigen (Fig. 4A).

We next evaluated the cytolytic activity of Pmel-1 iPSC-derived T cells and Pmel-1 splenic T cells using peptide-pulsed tumor targets. As measured using an LDH release cytotoxicity assay, Pmel-1 iPSC-derived T cells showed antigen-specific cytolytic activity in vitro that was similar to Pmel-1 splenic T cells (Fig. 4B). Consistent with cytolytic activity, granzyme B was significantly enriched in Pmel-1 iPSC-derived T cells (Fig. 4C). Next, Pmel-1 iPSC-derived T cells and Pmel-1 splenic T cells were compared for their abilities to release cytokine against bona fide tumor targets, namely the poorly immunogenic B16 melanomas. MC38 colon adenocarcinoma cells were used as irrelevant targets. Despite very low levels of MHC class I molecules in B16 tumors (Fig. 4D), Pmel-1 iPSC-derived T cells were able to produce abundant IFNγ to B16 tumors but not to MC38 tumors, which was equivalent to IFNγ production using Pmel-1 splenic T cells (Fig. 4E). Collectively, these results indicate that Pmel-1 iPSCs are capable of generating less-differentiated antigen-specific T cells with potent effector function suitable for adoptive cell therapy.

Given the potent in vitro antitumor reactivity against B16 tumors, next we investigated the in vivo antitumor immunity of Pmel-1 iPSC-derived CD8⁺ T cells using an established subcutaneous B16 tumor model. In vitro–activated Pmel-1 iPSC-derived CD8⁺ T cells were adoptively transferred into C57BL/6 mice bearing subcutaneous B16 melanomas that had been established for 7 days. Systemic administration of IL2 was used to enhance antitumor immunity of transferred T cells. In vitro–activated Pmel-1 splenic CD8⁺ T cells were used as a comparison. B16 tumor growth was unaltered in C57BL/6 mice receiving IL2 without adoptive cell therapy. In contrast, significant growth delays (Fig. 5A) and prolonged survival (Fig. 5B) were observed when C57BL/6 mice were adoptively transferred with Pmel-1 iPSC-derived CD8⁺ T cells in combination with IL2. Consistent with the in vitro functional studies (Fig. 4A–D), the in vivo antitumor immunity of Pmel-1 iPSC-derived CD8⁺ T cells was comparable with Pmel-1 splenic CD8⁺ T cells.

Previous reports showed adoptive transfer of tumor-specific T cells in combination with T-cell stimulation through antigen-specific vaccination and coadministration of immunostimulatory molecules and/or T-cell growth and activation factors can provide to inducing maximal tumor regression (4, 5, 25, 27). Therefore, we tested whether antitumor immunity of adoptively transferred iPSC-derived CD8⁺ T cells can be further enhanced by subcutaneous vaccination with hgp100, anti-CD40 mAb, and toll-like receptor (TLR) agonists, poly (I:C), and imiquimod cream, agonists for TLR3 and TLR7, respectively, in combination with administration of IL2 in mice with large subcutaneous B16 melanomas established for 11 days. As shown in Fig. 6A, adoptive transfer of iPSC-derived T cells together with vaccination with cognate antigen, hgp100, immunostimulatory molecules, TLR agonists, and IL2 resulted in complete responses in approximately 40% of mice bearing large established tumors. These complete responses were found to be durable over an observation period of >80 days (Fig. 6B).

Given our findings of long-term responders after the treatment, we investigated whether adoptively transferred Pmel-1 iPSC-derived T cells also had the capacity to become long-lived memory T cells in surviving mice. Thirty days after adoptive transfer of Pmel-1 iPSC-derived T cells (CD90.1), splenocytes from surviving C57BL/6 mice (CD90.2) still contained transferred T cells, of which more than 20% were CD44^highCD62L^high central memory T cells similar to surviving mice treated with Pmel-1 splenic T cells (Fig. 6C).

Finally, we evaluated whether treated mice developed antitumor reactivity by rechallenging survivors with implanted B16 tumors in the contralateral flank. MC38 tumors were also injected on the back to determine the antigen-specific systemic immunity by adoptive transfer of Pmel-1 iPSC-derived T cells. Age-matched, untreated nontumor experienced mice were used as controls. As shown in Fig. 6D, significant growth delays of the B16 tumors but normal growth of the MC38 tumors were observed in surviving C57BL/6 mice that were previously treated with adoptive transfer of Pmel-1 iPSC-derived T cells or Pmel-1 splenic T cells, suggesting development of antigen-specific long-lasting systemic memory.

Adoptive transfer of antigen-specific CD8⁺ T cells has emerged as one of the most effective treatments for a variety of cancers (1–3). A major limitation of this approach is poor survival of T cells in vivo following infusion. Although less-differentiated T cells would be an ideal T-cell subset for adoptive T-cell therapy (1, 4–6), generating large numbers of these “young” T cells is problematic. Human iPSCs hold great promise as unlimited sources of autologous T cells and can potentially generate an infinite number of naïve antigen-specific T cells for adoptive cell therapy. Two approaches have been described to generate iPSCs that can produce antigen-specific T cells (19–22). One is to reprogram T cells with known antigen specificity into iPSCs (19–21). Recent studies have shown that T-cell–derived iPSCs retaining rearranged TCR genes from original T cell clone can differentiate into T cells that exhibited antigen-specific cytokine production and cytotoxicity in vitro (19–21). This strategy is especially relevant for TILs because nearly 80% of TILs possess autologous tumor reactivity (40) but exhibit a more differentiated phenotype and functional exhaustion (1, 10, 11). We have recently described that melanoma TILs can be reprogrammed to human iPSCs, albeit with low efficiency compared with peripheral blood T cells (23). Another approach to generate antigen-specific T cells from iPSCs is to genetically engineer T-cell–derived iPSCs to express a CAR specific for the known tumor antigen, such as CD19, which is expressed on B-cell leukemias and lymphomas, and differentiate them into CAR-expressing T cells (22). Although iPSC-derived, CAR-expressing T cells recognize the targeted antigen via a CAR not through their endogenous TCRαβ, in line with our findings, they mediate antitumor reactivity in vivo in a xenograft model of lymphoma (22). Of note, the latter approach is useful to target only known tumor antigens, whereas the former approach can be used to target both known and unknown antigens as long as antigen specificity of the T cell is identified.

Our study has provided insight into the in vivo immunogenicity and antitumor reactivity of CD8⁺ T cells derived from iPSCs bearing a rearranged TCR of known antigen specificity. A central finding was that iPSC-derived regenerated CD8⁺ T cells escaped immune rejection and mediated potent antitumor immunity in vivo comparable with splenic CD8⁺ T cells from TCR transgenic mice. This in vivo antitumor immunity of iPSC-derived CD8⁺ T cells was supported by the findings that they not only had antigen-specific cytokine-producing and cytolytic function against the cognate antigen, hgp100 but also had the ability to produce cytokine against bona fide tumor targets, poorly immunogenic B16 melanomas. Taken together, these results suggest that T cells not only retain antigen specificity via endogenous TCRαβ but also maintain functional integrity, proliferative capacity, and effector function after reprogramming and redifferentiation processes.

This study demonstrates for the first time that murine T cells can be reprogrammed without using gene knockout mice (30) or drug-inducible gene expression systems (31). We found the SeV reprogramming system that was shown to be highly efficient method for the generation of iPSCs from human T cells (15, 19–22) could also be used for reprogramming murine T cells. The reprogramming efficiency, however, was quite low (0.004%) even with a higher MOI of 20. Furthermore, conventional mESC medium containing LIF was not sufficient to generate iPSCs from Pmel-1 T cells. We found that dual inhibition of both prodifferentiation MEK and GSK3 pathways that was shown to support the establishment of mouse iPSCs from partially reprogrammed cells (35) was required for reprogramming of Pmel-1 T cells. The low efficiency of iPSC derivation remains one of the major limitations to apply this approach into clinical practice. Of note, we observed Pmel-1 iPSCs can be maintained in conventional mESC medium without dual inhibition once they are established, and they maintain pluripotency and their ability to differentiate in the T lineage.

Persistence of TCR rearrangements is a distinct feature of T-cell–derived iPSCs (15–18) and also plays a pivotal role in differentiating iPSCs to CD8 SP T cells (19, 20). Our findings further expand on effector function of iPSC-derived CD8⁺ T cells in comparison with naïve CD8⁺ TCR transgenic T cells. We demonstrated Pmel-1 iPSC-derived DP T cells in culture with OP9-DL1 expressed high levels of CD3 and TCRVβ13, efficiently upregulated CD25 and CD69 upon stimulation via TCR, and quickly became CD8 SP T cells, functionally similar to in vitro–activated Pmel-1 splenic CD8⁺ T cells. Although we observed a subset of mESC-derived CD8 SP T cells expressing CD3 and TCRβ (data not shown) as reported in a previous study (24), Pmel-1 iPSCs cultured on OP9-DL1 cells differentiated more efficiently than mESCs and expressed CD3 and TCRVβ13 at the DP stage.

The early expression of pre-rearranged TCRα and TCRβ genes may cause undesired outcomes. Serwold and colleagues showed that mice derived from a reprogrammed T cell developed spontaneous T-cell lymphoma, which originated in the thymus (41). Although these mice have pre-rearranged TCR genes in all cells, and are different from mice receiving adoptive transfer of iPSC-derived T cells in the current studies, potential tumorigenicity of iPSC-derived T cells needs to be determined before this strategy is translated into clinical practice. Notably, we did not find any mice that developed tumors, including lymphoma after adoptive transfer of iPSC-derived T cells.

Immunogenicity of autologous iPSC-derived cells is still not well understood. Recent studies have shown that certain iPSC-derived cells such as smooth muscle cells and cardiomyocytes are immunogenic, whereas other cell types such as retinal pigment epithelial, hepatocytes, and neuronal cells exhibit little to no immunogenicity (42–45). Although our study did not evaluate host immune response against infused iPSC-derived T cells, our findings of long-lived adoptively transferred T cells that establish immunological memory suggest iPSC-derived T cells might not cause immune rejection.

T cells with CD62L^high less-differentiated phenotype maintain proliferative capacity and produce effector progeny in vivo, resulting in a greater objective response upon adoptive transfer relative to CD62L^low more-differentiated T cells (1, 4–6). The strategy of reprogramming and regeneration of antitumor T cells provides a more meaningful rationale for adoptive T-cell therapy if we can generate T-cell subsets that possess robust proliferative and engraftment capacities. A recent report showed that human T-cell derived iPSCs can differentiate to T cells with a central memory phenotype (19). We further determined that iPSC-derived less-differentiated T cells expressing memory T-cell markers had long-term persistence and established antigen-specific immunological memory in vivo after adoptive transfer. Interestingly, we found potent in vivo antitumor reactivity of adoptive transfer of iPSC-derived T cells and IL2 without antigen-specific vaccination, which was not seen in the previous study (25). A possible explanation for this finding may reside in the different in vitro activation protocol of T cells where we used IL2, IL7, and IL15, wheeras others used IL2 only (25). IL7 and IL15 play integral role in CD8⁺ T-cell memory generation and maintenance (38). IL7 is involved in the survival and homeostatic expansion of naïve CD8⁺ T cells and also can contribute to memory CD8⁺ T-cell homeostasis (46, 47). IL15 induces proliferation of naïve and memory CD8⁺ T cells and is involved in the maintenance of long-lasting antigen-specific memory CD8⁺ T cells (48–50). We found the majority of Pmel-1 iPSC-derived and splenic T cells after activation with IL2 were CD62L^low more-differentiated T cells (data not shown), whereas ones activated with IL2, IL7, and IL15 contained more than 50% of CD62L^high less-differentiated T cells (Fig. 3E).

In summary, our results demonstrate that adoptive transfer of iPSC-derived regenerated T cells is an effective cancer immunotherapy that improves local tumor control and overall survival in a lethal murine model of melanoma. Importantly, an establishment of antigen-specific immunological memory reveals the feasibility of generation of long-lived tumor-specific T cells via reprogramming to pluripotency and redifferentiation. Applying techniques described herein may provide an unlimited number of phenotypically defined, functional, and expandable autologous antigen-specific T cells harboring characteristics necessary for in vivo effectiveness.

No potential conflicts of interest were disclosed.

Conception and design: F. Ito

Development of methodology: H. Saito, F. Ito

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): H. Saito, A.E. Chang, F. Ito

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): H. Saito, F. Ito

Writing, review, and/or revision of the manuscript: H. Saito, K. Okita, A.E. Chang, F. Ito

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): F. Ito

Study supervision: K. Okita, F. Ito

The authors thank Dr. Mamoru Hasegawa in DNAVEC Corporation for kindly providing Sendai virus vectors. The authors also thank Drs. Tomonori Hosoya and James Douglas Engel for scientific discussion; Dr. Cindy DeLong for her technical assistance; Dr. Richard Lieberman for histologic analyses; Dr. Ren Shimamoto in ETH Zürich and Dr. Noemi Fusaki in Keio University for their technical assistance; and Elizabeth LaPensee, Link Hope, and Katherine Wood for administrative assistance.

This work was supported by institutional funds from the University of Michigan and Roswell Park Cancer Institute and grants from the Central Surgical Association, the American College of Surgeons, the Melanoma Research Alliance, and the NIH/NCI K08CA197966 (F. Ito), and by the National Center for Advancing Translational Sciences of the NIH UL1TR000433.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.