Background

T lymphocyte (T cells) play a key role in cell-mediated immune responses and are involved in monitoring and killing tumor cells. Throughout the last decades, several strategies have been developed to redirect, culture and/or enhance T lymphocytes against cancer (Houot et al., 2015; June et al., 2018) and utilize them for T cell-based adoptive immunotherapies. Recent clinical trials have shown outstanding outcomes in relapsed and refractory lymphoma patients treated with chimeric antigen receptor (CAR)-T cells (Riviere and Sadelain, 2017).

Human pluripotent stem cells (hPSCs), including embryonic (hESCs) and induced (hiPSCs), provide a promising resource to produce T cells for adoptive cellular immunotherapies, which can be coupled with genetic engineering technologies to generate off-the-shelf supplies of CAR T cells. In addition, generating hPSCs from antigen (Ag)-specific cytotoxic T lymphocytes (CTLs) and redifferentiating them into functional CTLs could enable the scalable production of rejuvenated CTLs (Minagawa and Kaneko, 2014; Kaneko, 2016). Several reports have demonstrated T cell generation from hPSCs (Nishimura et al., 2013; Vizcardo et al., 2013) and the feasibility of hiPSC based CAR T cell therapies (Themeli et al., 2013). However, there is still a need to improve the efficacy of T cell generation and expansion from hPSCs. In addition, further advances in hPSC-based T cell therapies will require their preclinical evaluation in large animal models. Since macaques are physiologically and immunologically similar to humans, including possessing orthologous MHC genes (Adams et al., 2001), and similarities in killer cell immunoglobulin-like receptors (KIR) with humans (Bimber et al., 2008; Parham et al., 2010), nonhuman primates (NHPs) will be the most appropriate model to address the therapeutic efficacy, safety and immunogenicity of PSC-derived T cells.

Here, we describe an improved method for the derivation of T cells from human and NHP-PSCs with a higher efficiency and shorter time (as soon as 3 weeks) than existing protocols. Differentiation of T cells from hPSCs involves two major steps: induction of hematopoietic progenitor cells (HPs) from hPSCs and their subsequent differentiation into T cells. Our lab previously reported well-established protocols on the induction of hematopoietic lineages from hPSCs on OP9 feeders and in defined feeder- and serum-free conditions (Vodyanik et al., 2005; Vodyanik and Slukvin, 2007; Uenishi et al., 2014). We showed that hemogenic progenitors from different stages of differentiation or different sources were cocultured on OP9-DLL4 to differentiate into T cells (Kumar et al., 2019b). We have also reported a protocol for the induction of hematopoietic lineages from NHP-PSCs (D'Souza et al., 2016). T cells differentiation from both hPSCs or NHP-PSCs proceeds through a CD5⁺CD7⁺ progenitor stage that eventually transitions into CD8⁺CD4⁺ double-positive cells. Altogether, the protocol used for the PSC-derived T cells presents a platform for T cell production to evaluate their utility for adoptive immunotherapies and preclinical testing in large animal models.

Related Information
Hematopoietic differentiation from human PSCs in an OP9 co-culture system
Hematopoietic differentiation of hPSCs on mouse stromal OP9 feeders is performed in serum-containing medium without addition of any cytokines (Vodyanik et al., 2005). In this system, hPSCs undergo stepwise progression toward APLNR⁺PDGFRα⁺ primitive posterior mesoderm with hemangioblast colony forming cells (HB-CFCs) that reflects primitive hematopoiesis, KDR^hiPDGFRα^lo/-VEC^- hematovascular mesodermal progenitors with definitive hematopoietic potential, immature VE-cadherin (VEC)⁺CD43^-CD73^- HE, which specify into DLL4⁺ arterial hemogenic endothelium (HE) with definitive hematopoietic potential, and DLL4^- non-arterial-type HE with mostly primitive hematopoietic potential; and CD34⁺CD43⁺ hematopoietic progenitors (HPs) that include CD43⁺CD235a⁺CD45^+/- HPs, enriched in erythromegakryocytic progenitors and CD43⁺CD235a/41a^- multipotent HPs with a lin^-CD34⁺CD90⁺CD38^-CD45RA^- hematopoietic stem progenitor cells phenotype, and lymphomyeloid potential (Vodyanik et al., 2006; Choi et al., 2009a and 2009b; Choi et al., 2012; Kumar et al., 2019a and 2019b; Uenishi et al., 2018) (Figure 1). CD43⁺ HPs generated in this coculture can be collected on days 8-9 of differentiation and subsequently cultured on OP9-DLL4 in lymphoid conditions to generate T cells (Kumar et al., 2019b). We also demonstrated that T cells can be generated directly from definitive hemogenic progenitors collected from earlier stages of development (hematovascular mesoderm or HE stage) and cultured in lymphoid conditions on OP9-DLL4 (Kumar et al., 2019b).
 
Hematopoietic differentiation from human PSCs in a chemically defined system
We have reported defined feeder- and serum-free conditions for generation of blood from hPSCs in chemically defined conditions. In this differentiation system, hPSCs follow stages of development similar to those described in hPSCs cocultured on OP9 feeders, including the formation of VE-Cadherin⁺CD73^-CD235a/CD43^- HE and CD34⁺CD43⁺ HPs with myeloid and T lymphoid potential (Uenishi et al., 2014) (Figure 1). We typically use collagen IV coated plates for differentiation in defined conditions to reduce cost. However, more expensive matrix Tenascin C can be used instead of collagen IV to improve hematopoietic differentiation and promote development of HPs with T cell potential in defined conditions (Uenishi et al., 2014).

Genetic engineering of Doxycycline-inducible iETS1 hPSCs
We have recently reported a protocol for the generation of conditional gene expression of ETS1 under tetracycline responsive element (TRE) promoter along with M2rtTA (reverse tetracycline transactivator) introduced into hPSCs using PiggyBac transposons (Jung et al., 2016; Park et al., 2018a and 2018b). In one vector ETS1 is downstream from the TREtight promoter, along with the zeocin resistance gene driven by the EF1α promoter, subcloned between the ends of 2 ITRs of the transposon vector. For easy detection of transgene expression upon addition of doxycycline to the culture, ETS1 is linked with Venus through a 2A self-cleaving peptide sequence. The second vector has the M2rtTA promoter linked with a puromycin resistance gene through a 2A peptide sequence subcloned between the ends of 2 ITRs. Using 2 different antibiotic genes facilitates the selection of clones that incorporate both vectors in a single step. The use of a two-vector system allows flexibility to adjust the TRE/M2rtTA ratio to achieve robust doxycycline dependent gene expression in hPSCs while limiting transgene leakage. hPSCs are cultured and then transfected on matrigel plates in mTeSR1 medium. Single hPSC colonies can be picked up from low-density cultures of cell populations (Park et al., 2018a and 2018b).

T cell production from iETS1 hPSCs
We have demonstrated that T cell production from hESCs can be increased through activation of the arterial program at the mesodermal stage of development by overexpression of the transcription factor ETS1. Hemogenic progenitors generated following induction of ETS1 were more than 100-fold enriched in T cell precursors as compared to control (Park et al., 2018b). Doxycycline treatment of differentiation cultures from days 2 to 6, enhances the generation of CD34⁺ HPs with lymphoid potential. HPs collected from day 9 of differentiation cultures in the presence of doxycycline can subsequently be differentiated into T cells in coculture with OP9-DLL4. Although T cell cultures from DOX- and DOX+ conditions generate a similar percentage of CD5⁺CD7⁺ and CD4⁺CD8⁺ T cells, total numbers of T lymphocytes produced per 104 CD43⁺ cells from DOX-treated cultures are dramatically (> 8-fold) greater.

Hematopoietic differentiation of NHP-PSCs on OP9 coculture
Although the defined serum and feeder-free differentiation system described above works well with hPSCs, it does not support efficient generation of CD34⁺CD43⁺ HPs with T cell potential from NHP-PSCs. That is why we recommend to culture on OP9 feeders to induce efficient production of lymphoid progenitors from NHP-PSCs. OP9 coculture supports blood production from different NHP species, including ESCs and iPSCs derived from rhesus and cynomolgus macaques. However, in contrast to human PSCs, NHP-PCS/OP9 cocultures require addition of GSK3β inhibitor and VEGF to promote hemogenic mesoderm development and human hematopoietic cytokines to support blood development (D'Souza et al., 2016) (Figure 2). Similar to human, NHP-PSC-derived HPs can be differentiated into T cells in coculture with OP9-DLL4 (D'Souza et al., 2016).