This article describes a novel method to generate tumor antigen-specific induced pluripotent stem cell-derived thymic emigrants (iTE) by a three-dimensional (3D) thymic culture system. iTE are a homogenous subset of T cells closely related to naïve T cells with the capacity for proliferation, memory formation, and tumor suppression.
The inheritance of pre-rearranged T cell receptors (TCRs) and their epigenetic rejuvenation make induced pluripotent stem cell (iPSC)-derived T cells a promising source for adoptive T cell therapy (ACT). However, classical in vitro methods for producing regenerated T cells from iPSC result in either innate-like or terminally differentiated T cells, which are phenotypically and functionally distinct from naïve T cells. Recently, a novel three-dimensional (3D) thymic culture system was developed to generate a homogenous subset of CD8αβ+ antigen-specific T cells with a naïve T cell-like functional phenotype, including the capacity for proliferation, memory formation, and tumor suppression in vivo. This protocol avoids aberrant developmental fates, allowing for the generation of clinically relevant iPSC-derived T cells, designated as iPSC-derived thymic emigrants (iTE), while also providing a potent tool to elucidate the subsequent functions necessary for T cell maturation after thymic selection.
Adoptive T cell therapy (ACT) can be an effective treatment for some patients with advanced cancer. Unfortunately, many patients do not experience tumor regression, and transferred cells fail to persist after infusion. This may be due to the quality of the infused T cells. An ACT mouse model showed that compared to naïve or less differentiated central memory T cells, terminally differentiated effector cells are less potent due to poor in vivo persistence1, an observation also supported by clinical data2,3.
In an effort to improve the efficacy of current ACT, T cell-derived induced pluripotent stem cells (T-iPSC) have been studied extensively4,5. When T cells are reprogrammed into T-iPSC and re-differentiated into T cells, the rearranged configuration of TCR genes is inherited by T-iPSC, and subsequently the re-differentiated T cells. Therefore, the capacity of T-iPSC to undergo unlimited in vitro expansion permits the efficient reproduction of immature T cells carrying the neoantigen-specific T cell receptors (TCR) when such cells are engineered from tumor antigen-specific T cells6,7. However, the precise method for differentiation of T-iPSC into mature T cells, which would allow the production of cancer antigen-specific T cells with a less differentiated phenotype and better anti-tumor potency, remains to be elucidated.
T-iPSC differentiation employing the co-culture of OP9 murine stromal cells over-expressing human Notch ligand DLL1 is a well-established method to produce T cells in vitro6,7. In mice and humans, this co-culture system can consistently differentiate iPSC, thereby recapitulating developmental events from the blastocyst stage until the immature T cell lineage stage6,7. Despite these biotechnological advances, the physiological differentiation after the CD4+CD8+ double positive (DP) stage is still difficult to achieve. One of the reasons is that in vivo CD4+CD8– and CD4–CD8+ single positive (SP) T cells are generated in the thymus, an organ responsible for the maturation and selection of T cells that have foreign antigen-specificity but not auto-reactivity8. These selective processes are defined as positive and negative selection, respectively. However, most of the molecular mechanisms necessary to mature T cells in the thymus are still not fully understood, making it difficult to reconstruct this process in vitro. In an attempt to overcome this physiological hurdle, several groups have stimulated the TCR complex using anti-CD3 antibodies or agonist peptides. These in vitro techniques generate cell products which express key T cell markers, like CD3, CD8αβ, TCRαβ, and CD62L, while still retaining tumor antigen-specificity. Unfortunately, T cells generated by these extrathymic methods constitute a broad heterogenous population of cells characterized by incomplete positive selection, innate-like features, TCR non-specific killing, inability for memory formation, and non-persistent anti-tumor effects in vivo8,9,10,11. These abnormalities have raised concerns that such cells might trigger a variety of side effects, including lymphoma and both skin and bone abnormalities, if used for therapeutic applications12,13,14.
To recreate the physiological signals missing in current in vitro differentiation systems, tumor antigen-specific T-iPSC were differentiated using a harvested thymus. The classical fetal thymus organ culture (FTOC) system, which was designed to study the intra-thymic development of T cells, was improved by using a 3D culture system which successfully produced T cells that completed thymic education. These post-thymic T cells, which were designated as iPSC-derived thymic emigrants (iTE), exhibited naïve-like properties15. iTE showed proliferation, memory formation, and adequate anti-tumor effects in a mouse model against established B16 melanoma tumors. This article describes in detail the protocol of this novel FTOC system using a 3D culture system (Figure 1).
All the animal experiments were approved by the Institutional Animal Care and Use Committees of the National Cancer Institute (NCI) and performed in accordance with NIH guidelines.
1. Preparation of OP9/DLL1 Cells for Co-culture with iPSC
2. In Vitro Differentiation of iPSC into Immature T Cells
3. 3D Thymic Organ Culture to Generate iTE
4. Preparation of Antigen Presenting Cells (APC)
5. Pulsing APC with Antigen
Co-cultured fetal thymuses were sectioned to analyze whether iPSC-derived T lineage cells can migrate into the thymic lobes. Unseeded control lobes had a tissue architecture characterized by an astrocyte-like thymic epithelial web17, deployed of endogenous CD3+ cells. On the other hand, thymic lobes seeded with iPSC-derived immature T cells were repopulated with CD3+ mononuclear cells, indicating migration of iPSC-derived immature T cells into the lobes (Figure 2A).
T cells that migrated into and matured within the thymic microenvironment subsequently egressed as iTE. To test their phenotypic characterization, flow cytometric analysis of C57BL6 thymocytes, Pmel iPSC-derived immature T cells (extrathymic), and cells that egressed from thymic lobes (iTE) was performed. Extrathymic T cells on OP9/DLL1 showed CD4+CD8+ (DP) T cells and CD8αSP T cells without expression of the positive selection marker MHC-I, whereas iTE had a clear population of CD8αSP MHC-I+ T cell phenotype, indicating their successful passage through positive selection prior to egressing from the thymic lobes. iTE consistently express MHC-I and CD62L, which are markers associated with high proliferative competency, cytokine production, peripheral survival, and lymphoid homing18,19,20. This phenotype is consistent with M2 SP thymocytes that are the most mature population of single positive T cells in the thymus20, which suggests that iTE have transitioned through a normal thymic developmental program (Figure 3). To monitor the efficiency of iTE generation, cells that had egressed from individual thymic lobes were isolated. On day 7, thymic lobes generated an average of 1 x 103 live CD8SP CD45.1+ CD3+ iTE per day (Figure 3B). A similar rate of iTE production is observed from day 6 to day 12 of 3D thymic co-culture.
Antigen-dependent activation and secretion of cytokines were analyzed to observe the functional properties of thymically educated iPSC-derived immature T cells. In the presence of an irrelevant peptide (nucleoprotein), Pmel-iTE did not release significant amounts of TNF-α, IL-2, or IFN-γ. When stimulated with the cognate peptide for Pmel T cells (hgp100), Pmel-iTE released robust amounts of TNF-α and IL-2, while also producing low amounts of IFN-γ (Figure 4), indicating that thymically educated iTE can recognize their cognate peptide and secrete effector cytokines with a profile resembling that of natural recent thymic emigrants (RTE).
To examine the transcriptional differences between iPSC-derived T lineage cells differentiated on OP9/DLL1 with or without thymic education (i.e., iTE versus extrathymic T cells), RNA-seq analysis was performed on these two populations and compared to that of DP T lineage cells differentiated using OP9/DLL1 (DP) and primary naïve CD8+ Pmel T cells. The expression of 102 genes which play crucial roles in T cell ontogeny, thymocyte activation, and memory formation were analyzed15,20,21,22. A principal component analysis of those four studied populations demonstrated that extrathymically generated DP and CD8SP T cells clustered together, while iTE clustered closer to naïve T cells (Figure 5). Collectively, these data demonstrate that iTE have a phenotype closer to naïve T cells than do T lineage cells generated by extrathymic methods.
Figure 1: Schematic overview of the differentiation of iPSC to iTE using OP9/DLL1 and 3D thymic culture. The protocol involves three separate differentiation steps; (Left) from iPSC cells to hematopoietic lineage cells on OP9/DLL1 (day 0 to 6), (Middle) from hematopoietic lineage cells to immature T cells on OP9/DLL1 with cytokines (day 6 to 16–21), and (Right) from immature T cells (day 16–21) to iTE using a 3D thymic culture system. Please click here to view a larger version of this figure.
Figure 2: Immuno-histochemistry of thymic lobes seeded with iPSC-derived immature T cells. Top: H&E staining of a thymic lobe with and without seeding of iPSC-derived immature T cells. From second top to bottom: confocal images of the sectioned lobes stained with DAPI (nucleus), CD3 (T cell), and merge. Scale bars = 100 µm. Please click here to view a larger version of this figure.
Figure 3: iTE show a post-thymic T cell phenotype. (A) FACS analyses of thymocytes, extrathymic T cells (OP9/DLL1 co-culture system) and Pmel-iTE. Live cells were gated on congenic CD45+. CD8 SP populations were further analyzed for CD62L and MHC-I expression. (B) Average number of CD8SP CD45.1 iTE produced overnight per lobe 7 days after pre-seeding. Data were collected from 12 independent experiments. Please click here to view a larger version of this figure.
Figure 4: iTE produce various cytokines by antigen-specific stimulation. FACS analyses of intra-cellular production of cytokines by iTE. iTE were co-cultured with APCs pre-loaded with irrelevant (nucleoprotein) or cognate (hgp100) peptide for three days. The numbers shown in upper right quadrants indicate the percentages of iTE producing cytokine. Please click here to view a larger version of this figure.
Figure 5: Whole-transcriptome analysis reveals a shift in iTE gene expression toward a naïve CD8+ T cell program. Principle component analysis (PCA) of RNA-seq data from DP, extrathymic CD8 SP, iTE, and naïve T cells. (Analysis of 102 genes related to thymic differentiation using public database GSE105110)15. Please click here to view a larger version of this figure.
Using T-iPSC to regenerate tumor antigen-specific T cells may overcome many of the current obstacles of ACT by generating young cells with improved persistence. Although several methods using the OP9/DLL1 co-culture system have been reported to generate CD8 SP cells6,7,10,13 that express CD8 molecules and tumor antigen-specific TCRs, global gene expression patterns and functional analysis show that these extrathymically regenerated CD8 SP cells are different from naïve T cells (Figure 4). Here, we describe a 3D thymic culture system that can generate iPSC-derived thymic emigrants (iTE) with high fidelity and homogeneity from murine T-iPSC. iTE resemble naïve T cells in global gene expression pattern and in functionality, such as memory formation and in vivo anti-tumor effect against established tumor15.
The classical FTOC system is a way to recapitulate thymic selection in vitro. It has been used for studying intra-thymic development of thymocytes23, and there are a few reports of FTOC being used to generate RTE24. However, the FTOC system has several limitations. To deal with the lack of oxygen in an artificial organ culture, several groups have used either a semi-dry membrane based culture23, or high oxygen submersion culture systems25. However, no current methods can constantly generate a homogenous population of post-thymic T cells. To overcome the limitations of the classical FTOC system, we designed a 3D thymic culture system that provides technical improvements over conventional methods15. For example, using our 3D thymic culture method, maximal oxygen exchange and the absence of surface-lobe mechanical stress keep the thymic lobes in a more physiological environment. Additionally, long term culture permits mature T cells to egress naturally from the thymic lobes. Finally, real time observation and micro-manipulation enable media exchange and a constant collection of iTE without physically disturbing the thymic lobes. Thus, the 3D thymic culture method provides significant technical improvements as well as an avenue to study thymically selected naïve T-cells that was not previously available.
There are several key points for the successful generation of iTE using this 3D thymic culture system. The quality of the FBS and culture conditions is critical to maintain the expansion of OP9/DLL1 cells without losing their ability to support iPSC differentiation. Therefore, we recommend pre-evaluation of the FBS lot as well as consistently passaging at 80% confluency to prevent cell differentiation and senescence. Additionally, a confluent OP9/DLL1 culture is required for in vitro differentiation of iPSC into immature T cells, as differences in confluency can affect their efficiency. Finally, the embryonic age of thymic lobes is crucial for the generation of iTE. We recommend using E14.5 – 15.5 thymic lobes.
As with any new protocol, this method has limitations and is subject to improvement. The culture technique presented here generates approximately 1000 iTE per thymic lobe per day for a period of two weeks. Increased iTE generation may be possible with further modifications, including optimization of oxygen concentration, media volume, and type of 3D culture plate. Addition or removal of cytokines, as well as changes in cytokine concentration, may also contribute to improved iTE yield.
Given that the 3D thymic culture system presented here can generate thymic emigrants in a completely ex vivo system, this technique can be applied to a variety of immunological and adoptive cell transfer research projects including, but not limited to T cell differentiation, post-thymic T cell maturation, and generation of antigen-specific T cells from hematopoietic progenitor or stem cells. Although this method is not directly applicable to human samples, iTE and the 3D thymic culture system hold great potential for elucidating the molecular mechanisms of positive and negative selection and may facilitate the creation of a culture system that enables the generation of clinically relevant tumor antigen-specific naïve-like T cells for ACT.
The authors have nothing to disclose.
We thank Hiroshi Kawamoto and Kyoko Masuda for kindly providing the OP9/DLL1 cell line. We thank Alan B. Hoofring and Erina Z. He for graphical assistance. This research was supported by the Intramural Research Program of the US National Cancer Institute (ZIA BC010763) and the Cancer Moonshot program for the Center for Cell-based Therapy at the NCI, NIH. The work was also supported by the Milstein Family Foundation.
Chemicals, Peptides and Recombinant Proteins | |||
2-deoxyguanosine | Sigma-Aldrich | 312693-72-4 | |
2-Mercaptoethanol (1000X) | Thermo Fisher Scientific | 21985-023 | |
ACK Lysing Buffer | Gibco | A1049201 | |
Ascorbic acid | Sigma-Aldrich | A8960 | |
Blasticidin | Thermo Fisher Scientific | R21001 | |
FBS | Gemini | 100-500 | |
Flt-3 ligand | R&D Systems | 427-FL | |
GlutaMAX (100X) | Thermo Fisher Scientific | 35050-061 | |
hgp100 | Genscript | 282077-1, KVPRNQDWL | |
Interleukin-2 | R&D Systems | 402-ML | |
Interleukin-7 | R&D Systems | 407-ML | |
MEM Non-Essential Amino Acids Solution | Gibco | 11140050 | |
MEM powder | Gibco | 61100061 | |
Monothioglycerol | Sigma-Aldrich | M-6145 | |
Nucleoprotein | Global Peptides | ASNENMETM | |
Penicillin/streptomycin | Thermo Fisher Scientific | 15140-122 | |
Phosphate buffered saline pH 7.4 (1x) | Thermo Fisher Scientific | 10010-023 | |
Puromycin | Thermo Fisher Scientific | A1113803 | |
RPMI 1640 | Gibco | 11875093 | |
Sodium Pyruvate | Thermo Fisher Scientific | 11360-070 | |
Stem Cell Factor (SCF) | R&D Systems | 455-MC | |
Stemfactor LIF, Mouse Recombinant | STEMGENT | 03-0011-100 | |
Trypsin-EDTA (0.05%), phenol red | Thermo Fisher Scientific | 25300-062 | |
Cell Culture Vessels and others | |||
10 cm dish | Corning, Inc. | 353003 | |
12ML Syringe | Covidien Monoject | 22-652-090 | |
6 well plate | Corning/Coster | 3516 | |
Cell strainer 100um | Fisher Scientific | 22-363-549 | |
Cell strainer 40um | Fisher Scientific | 22-363-547 | |
Forceps | DUMONT | 0108-5PO | |
Lab soaker mat | Versi-Dry | Cat. EF2175CX 74018-00 | |
Membrane filters ( 0.8 μm, 47diam) | Whatman | WHA7408004 ALDRICH | |
Perfecta3D Hanging Drop Plate | Sigma-Aldrich | HDP1096 | |
U Bottom 96 well plate | Corning/Coster | 3799 | |
Experimental Cell lines | |||
CD3-iPSC | Vizcardo et al., Cell Report 2018 | N/A | |
MEF-iPSC | Vizcardo et al., Cell Report 2018 | N/A | |
Mouse Embryonic Fibroblasts (MEF) | ATCC | SCRC-1040; RRID:MGI:5007926 | |
OP9/N-DLL1 | Riken Bioresource center | Cat# RCB2927; RRID:CVCL_B220 | |
Pmel-iPSC | Vizcardo et al., Cell Report 2018 | N/A | |
Experimental mouse models | |||
B6.SJL-PtprcaPepcb/BoyCrCrl | Charles River | Strain Code 564; RRID:IMSR_CRL:564 | |
C57BL/6N | NCI/Charles River | N/A | |
Pmel-1 mice | Overwijk et al. | J Exp Med 198(4):569-80 | |
Antibodies | |||
Anti-aTCR | Biolegend | 109202; RRID:AB_313425 | |
Anti-CD3 | abcam | ab11089; RRID:AB_369097 | |
Anti-CD4 | BD Biosciences | 553730; RRID:AB_395014 | |
Anti-CD44 | BD Biosciences | 559250; RRID:AB_398661 | |
Anti-CD45.1 | BD Biosciences | 553775; RRID:AB_395043 | |
Anti-CD45.2 | BD Biosciences | 553772; RRID:AB_395041 | |
Anti-CD62L | BD Biosciences | 560516; RRID:AB_1645257 | |
Anti-CD69 | BD Biosciences | 552879; RRID:AB_394508 | |
Anti-CD8a | BD Biosciences | 557959; RRID:AB_396959 | |
Anti-CD8b | BD Biosciences | 550798; RRID:AB_393887 | |
Anti-H-2Kb | BD Biosciences | 553570; RRID:AB_394928 | |
Anti-IFN-g | BD Biosciences | 557998; RRID:AB_396979 | |
Anti-IL-2 | BD Biosciences | 554428; RRID:AB_395386 | |
Anti-TCRb | Thermo Fisher Scientific | 35-5961-81; RRID:AB_469741 | |
Anti-TCRVb13 | BD Biosciences | 553204; RRID:AB_394706 | |
Anti-TNFa | BD Biosciences | 557644; RRID:AB_396761 |