Human Dendritic Cell–Derived Induced Pluripotent Stem Cell Lines Are Not Immunogenic

Human Dendritic Cell−Derived Induced Pluripotent Stem Cell Lines Are Not Immunogenic

This information is current as Arvind Chhabra, I-Ping Chen and Deepika Batra of September 29, 2021. J Immunol 2017; 198:1875-1886; Prepublished online 23 January 2017; doi: 10.4049/jimmunol.1601676 http://www.jimmunol.org/content/198/5/1875 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2017/01/20/jimmunol.160167 Material 6.DCSupplemental

References This article cites 41 articles, 12 of which you can access for free at: http://www.jimmunol.org/ http://www.jimmunol.org/content/198/5/1875.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists by guest on September 29, 2021

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2017 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Human Dendritic Cell–Derived Induced Pluripotent Stem Cell Lines Are Not Immunogenic

Arvind Chhabra,* I-Ping Chen,† and Deepika Batra*

Donor-speciﬁc induced pluripotent stem cells (iPSC) can be used to generate desired cell types, including naive immune effectors, for the treatment of different diseases. However, a greater understanding of the inherent immunogenicity of human iPSC and their cellular derivatives is needed for the development of safe and effective cell-replacement therapies, given that studies in mouse models claimed that the syngenic mouse iPSC lines can be immunogenic. We report the characterization of the innate and adaptive immune mechanisms in human iPSC lines derived from peripheral blood–derived dendritic cells using a nonintegrating RNAvirus, Sendai virus. We show that these iPSC lines express mRNA of TLR molecules and the Ag-presentation pathway intermediates; however, these mRNA are not translated into functional proteins, and these iPSC lines do not induce TLR-mediated inﬂammatory

cytokine responses or inflammasome activation. We also show that these iPSC lines do not activate T cells in an allogenic MLR; Downloaded from however, they express low levels of MHC class I molecules that can efficiently acquire antigenic peptides from their microenvironment and present them to Ag-specific T cells. In addition, we show that these iPSC lines can be efficiently differentiated into hematopoietic stem cell precursors, as well as APC, under appropriate culture conditions. Taken together, our data show that the dedifferentiation of human dendritic cells effectively shuts down their immunogenic pathways and implicates transcriptional and posttranscriptional mechanisms in this process. The Journal of Immunology, 2017, 198: 1875–1886. http://www.jimmunol.org/

uman pluripotent stem cells (hPSC) are defined by their MART-127–35 as a model human tumor Ag (4–8). Donor-derived ability to self-renew and to differentiate into different immunogenic dendritic cells (DC) and engineered anti-tumor H cell lineages under appropriate culture conditions. Der- T cells produced encouraging clinical results in active specific ivation of human embryonic stem cell (hESC) lines from early- immunization and T cell adoptive-transfer–based cancer immuno- stage embryos opened up possibilities for the development of therapy settings (9–12); however, the efficacy of these approaches pluripotent stem cell–based cell-replacement therapies (CRT) requires significant improvement. In this context, donor-derived (1), and successful reprogramming of somatic cells into induced iPSC can be a valuable resource for generating autologous naive pluripotent stem cells (iPSC) (2, 3) addressed the ethical issues immune effectors; however, efficient models are needed to system- associated with the therapeutic use of hESC. We are working atically characterize the immunogenicity profiles of human iPSC by guest on September 29, 2021 on developing T cell–based cancer immunotherapy approaches lines and their differentiation potential. utilizing the human melanoma–associated antigenic epitope The need for a systematic analysis of the inherent immunogenicity of human iPSC lines was highlighted by recent studies in mouse models that used rejection of iPSC line–induced teratoma *Department of Medicine, School of Medicine, University of Connecticut Health in immune-competent mice as a model to examine the inherent † Center, Farmington, CT 06030; and Department of Oral Health and Diagnostic immunogenicity of iPSC lines. These studies claimed that the Sciences, School of Dental Medicine, University of Connecticut Health Center, Farmington, CT 06030 syngenic mouse iPSC lines, especially those derived with genome- ORCID: 0000-0002-7349-3091 (I-P.C.). integrating viral vectors, are immunogenic (13) and that the inherent immunogenicity of these syngenic mouse iPSC lines correlates Received for publication September 28, 2016. Accepted for publication December 30, 2016. with their Ag profiles (14). Although other groups reported min- This work was supported by Stem Cell Seed Grant 10-SCA-23 and Stem Cell Estab- imal or no immunogenicity of mouse syngenic iPSC lines and lished Investigator Grant 13-SCB-05 (both to A.C.) from the State of Connecticut, as their cellular derivatives (15–17), these reports highlighted the well as by Grant 1RR06192 from the National Institutes of Health to the Clinical need for an in-depth characterization of the inherent immunoge- Research Center, University of Connecticut Health Center. nicity of hPSC. hESC lines express low levels of MHC class I A.C. designed the study, conducted the experiments, and wrote the manuscript; D.B. conducted T cell–generation experiments and assisted A.C. in the execution of other molecules (18), exhibit immune-privileged properties (19, 20), experiments; and I-P.C. performed teratoma-generation experiments. and are rejected in immune-competent mice through a T cell– Address correspondence and reprint requests to Dr. Arvind Chhabra, Department of mediated allograft-rejection process (19, 21, 22); however, the Medicine, University of Connecticut Health Center, 263 Farmington Avenue, Farm- inherent immunogenicity of human iPSC lines has not been ex- ington, CT 06030. E-mail address: [email protected] amined carefully. In this article, we show that the terminally The online version of this article contains supplemental material. differentiated gold standard for human immunogenic cells, DC, Abbreviations used in this article: 5-Aza, 5-azacytidine; CRT, cell-replacement ther- can be successfully dedifferentiated into the iPSC state using apy; DC, dendritic cell; EB, embryoid body; Flu, influenza-associated MP58–66; hESC, human embryonic stem cell; hPSC, human pluripotent stem cell; hPSC- Sendai virus, a nonintegrating RNA virus. We used these iPSC APC, hPSC-derived APC; HSC, hematopoietic stem cell; iDC, immature DC; iPSC, lines to characterize the status of innate and adaptive immune induced pluripotent stem cell; iPSC-APC, iPSC-derived donor-specific APC; M1, mechanisms that are functional in human peripheral blood– human melanoma–associated MART-127–35; M3, MAGE-3271–279;mDC,matureDC; MEF, mouse embryonic fibroblast; SCF, stem cell factor; TCReng, TCR-engineered; derived DC and show that these iPSC lines are inherently non- TSA, trichostatin A; UCHC, University of Connecticut Health Center; UConn, Uni- immunogenic. We also show that these iPSC lines express low levels versity of Connecticut. of MHC class I molecules, and they can efficiently present antigenic Copyright Ó 2017 by The American Association of Immunologists, Inc. 0022-1767/17/$30.00 peptides, acquired from their microenvironment, to corresponding www.jimmunol.org/cgi/doi/10.4049/jimmunol.1601676 1876 DEDIFFERENTIATION AND REDIFFERENTIATION OF HUMAN DC

Ag-specific T cells; as a result, they are also killed by the CTL. In the indicated Ab, and surface phenotype was analyzed using a FACSCalibur addition, we show that these iPSC lines exhibit potent differentia- (BD Biosciences). Ab against the pluripotency markers Oct-4, Sox-2, tion potential, as reflected by their ability to generate embryoid Nanog, Tra 1-61, and Tra 1-81 for immunofluorescence staining were purchased from Abcam. Intracellular staining was performed using Cytofix/ bodies (EB) that can produce hematopoietic stem cell (HSC) pre- Cytoperm (BD Biosciences). cursors and can be further differentiated into functional APC. Our findings provide a plausible explanation for the rejection of trans- RT-PCR analysis planted syngenic iPSC lines in animal models and demonstrate that RT-PCR was performed to confirm removal of Sendai virus vectors from the DC-derived iPSC lines is an ideal system to characterize the im- iPSC lines generated and pluripotency of the hPSC lines, as well as to munogenicity profile of human iPSC lines. Furthermore, we believe characterize hPSC line–derived EB. In brief, hPSC cultures or hPSC- that the dedifferentiation and redifferentiation model of human DC derived EB were washed with PBS and lysed using TRIzol Reagent (Life Technologies), and the RNA was made according to the manufac- is also an ideal system to systematically characterize the molecular, turer’s instructions, as described previously (4). RNA was quantified and cellular, and functional profile of human iPSC-derived donor- cDNA was made with 500 ng of RNA per sample using the SuperScript III specific APC (iPSC-APC). single-strand cDNA synthesis kit (Life Technologies). Primers used for RT-PCR analysis were procured from IDT Technologies and are listed in Supplemental Table I. Materials and Methods Cell lines, culture media, and reagents Immunofluorescence analysis Human subject work was done in accordance with the University of For immunofluorescence staining, hPSC were cultured on feeder-coated Connecticut Health Center (UCHC) Institutional Review Board guidelines. glass coverslips in 24-well plates. Colonies were fixed with 4% parafor- HLA-A2+ individuals were enrolled in the study, with voluntary informed maldehyde for 30 min at room temperature, washed three times with PBS, Downloaded from consent, to donate blood for the isolation of human PBL. CD4+ and CD8+ permeabilized with PBS containing 0.1% Triton X-100 for 30 min, and T cells and CD14+ monocytes were purified from human PBL by Ficoll- washed again with PBS three times. Fixed and permeabilized cells were Hypaque gradient separation followed by magnetic-bead purification, as kept in PBS at 4˚C until used for staining. For staining, hPSC were in- described previously (5, 23). hESC line H9 was obtained from the UCHC cubated in blocking reagent (PBS with 10% BSA) for 30–45 min, followed Stem Cell Core Facility, with an appropriate material transfer agreement by a 45–60-min incubation with primary Ab (1:50 dilution in PBS con- from the WiCell Research Institute (Madison, WI). Mouse embryonic fi- taining 1% BSA) in a humidified chamber in a 37˚C incubator. Unbound primary Ab was removed by washing three times with PBS, followed by broblasts (MEF), CF-1, were procured from GlobalStem. Culture medium http://www.jimmunol.org/ IMDM, DMEM, MEM-a, FBS, and KnockOut Serum Replacement were staining with the secondary Ab (1:100 dilution) in a humidified chamber purchased from Life Technologies. Cytokines and growth supplements for for 45–60 min in a 37˚C incubator, followed by three washes with PBS. hPSC culture and differentiation, such as GM-CSF, IL-4, IL-2, IL-3, IL-15, Hoechst (1 mM final) was added at the second washing step. Slides were IL-7, Flt3, bone morphogenetic protein 4, stem cell factor (SCF), and washed with water to remove traces of PBS before mounting with VEC- vascular endothelial growth factor, were purchased from R&D Systems. TASHIELD Mounting Medium (Vector Laboratories). Analysis was per- Abs for FACS were purchased from BD Biosciences, BioLegend, and formed using a Zeiss LSM 780 confocal microscope at the UCHC Affymetrix eBioscience. Microscopy Core Facility. Sendai virus–mediated iPSC line generation Teratoma-generation assay The ability of hESC and iPSC to form three germ layers in vivo was A CytoTune-iPS 2.0 Sendai Reprogramming Kit (Life Technologies) was by guest on September 29, 2021 used to generate iPSC lines from human PBL, according to the manu- confirmed by a teratoma assay. In brief, hPSC cultures were treated with 1 mg/ml dispase, collected in hPSC culture medium, and washed once with facturer’s instructions, with some modifications. In brief, 50,000 whole 6 PBL, purified CD4+ T cells, CD8+ T cells, and CD14+ monocyte-derived PBS, and 2 3 10 hPSC were injected i.m. into immunodeficient NOD- DC were transduced with recombinant Sendai virus, a nonintegrating RNA SCID mice (Jackson Laboratory, Bar Harbor, ME) for teratoma induction. virus, expressing GFP to optimize the virus infection conditions. For the Teratomas were excised 10–16 wk postinjection, fixed in 10% formalin, generation of iPSC lines, cells were infected with recombinant Sendai and embedded in frozen embedding medium (Shandon Cryomatrix; virus vectors expressing the reprogramming factor genes c-Myc, Klf-4, Thermo Scientific). Tissue sections were cut using a refrigerated micro- Oct-3/4, and Sox-2. Transduced cells were cultured in MEF-plated six- tome (Leica) and stained with H&E, and images were taken using an well culture plates, and the generation of iPSC clones was monitored. Olympus IX71 microscope fitted with a SPOT Flex camera (SPOT Im- Potential iPSC clones generated 7–14 d later were transferred into fresh aging Solutions). MEF-containing culture plates, propagated, and used to make frozen stocks. The karyotype and pluripotency status of the generated iPSC lines Generation of DC from human peripheral blood–derived were established before characterizing their immunogenic profile and monocytes differentiation potential. To confirm that the iPSC lines generated were free of the Sendai virus vectors used for reprogramming, RT-PCR was Immature DC (iDC) were derived from human PBL and cultured in the g done using primers specific for Sendai virus backbone that were provided presence of LPS (100 ng/ml) and IFN- (1000 units/ml) for 12–16 h to with the kit, as described below. Uninfected human DC and the DC in- generate mature DC (mDC), as described previously (4, 5, 24). In brief, fected with GFP encoding Sendai virus were used as negative and positive PBL were put in six-well culture plates for 15–30 min to allow DC pre- controls, respectively. Primer sequences are listed in Supplemental Table I. cursors to adhere to the wells. Nonadherent cells were removed by washing the wells three to five times with PBS. Alternatively, CD14+ monocytes + hPSC culture and passage were purified from blood-derived PBL using a CD14 cell purification kit (STEMCELL Technologies) and were differentiated into iDC by culturing Irradiated MEF were plated in 0.1% gelatin–coated plates in DMEM them in IMDM containing the DC-differentiating cytokines GM-CSF containing 10% FBS a day before plating hPSC. The following day, MEF (1000 IU/ml) and IL-4 (1000 IU/ml) for 3–5 d. were washed three or four times, and hPSC clumps were plated in hPSC culture medium. Fresh hPSC medium was fed every other day. For pas- Dextran-uptake assay sage, the hPSC culture plate was washed with PBS to remove the floating cells and fed with fresh medium, hPSC colonies were cut using a 1-ml The phagocytic ability of iPSC-APC and human blood monocyte-derived 3– syringe with a 16-gauge needle, and the cut colony clumps were plated in 5-d-old iDC was examined by phagocytosis of FITC-tagged dextran beads plates containing fresh CF-1 MEF. (40,000 m.w.; Molecular Probes), according to the manufacturer’s instructions. In brief, cells were coincubated with FITC-dextran beads for 2 h Surface phenotype analysis of hPSC for pluripotency markers at 37˚C and washed with PBS to remove free beads, and phagocytosis was by FACS quantified by FACS. To examine phagocytosis by immunofluorescence, following coincubation with the FITC-dextran beads, cells were mounted To confirm the pluripotency status of hPSC by FACS, a single-cell sus- on glass slides using a Shandon cytospin centrifuge (700 rpm for 7 min), pension of hPSC was made by treating the hPSC cultures with 0.05% fixed using 4% paraformaldehyde, and imaged, following nucleus staining trypsin. Trypsin was neutralized by hPSC medium, cells were washed with with 1 mM Hoechst, using a Zeiss 780 confocal microscope at the UCHC PBS, and a single-cell suspension was made in PBS. Cells were stained with Microscopy Core Facility, as described previously. The Journal of Immunology 1877

Generation of Ag-specific natural and TCR-engineered T cells molecules, costimulatory molecules, and Ag-presentation pathway– associated molecules was quantified by FACS. Functional T cells specific for the human melanoma–associated MART-127–35 (M1) epitope and the influenza-associated MP58–66 (Flu) epitope were Generation of EB generated by coculturing peripheral blood–derived CD8+ T cells with mDC, as previously described (25, 26). In brief, human peripheral EB were generated by two methods. In the first method, single-cell sus- blood–derived CD8+ T cells were cocultured with autologous mDC in pensions of hPSC lines were made by treating hPSC cultures with colla- the presence of 100 U/ml IL-2. Once cells started to proliferate, cul- genase (1 mg/ml). Next, cells were washed with PBS and resuspended in EB tures were maintained in the presence of 1000 U/ml IL-15. M1 and Flu medium, followed by forced aggregation in AggreWell 400 plates epitope-specific CTL generated were quantified 7–10 d later by FACS (STEMCELL Technologies) by centrifugation in a swing bucket rotor at staining with the respective tetramer reagents (BD Biosciences). 2000 rpm for 12 min. The following day, hPSC clumps were gently TCR-engineered (TCReng) CD4 and CD8 T cells specific for the M1 transferred to low-adherence tissue culture plates in fresh EB medium. In epitope were generated by engineering human CD4+ and CD8+ T cells, as the second method, hPSC colonies were gently scraped off, washed twice described previously (7, 27). In brief, human PBL–derived T cells were with PBS, and cultured in low-adherence tissue culture plates in the EB activated with anti-CD3 and anti-CD28 Ab cross-linked on tissue culture medium. For time-kinetics experiments, EB were used at the indicated time wells (5 mg each per 24-well tissue culture plate well) and cultured in points, whereas the 20-d-old EB were used for the rest of the experiments. IMDM containing 100 IU/ml IL-2. Cells were transduced with retrovirus The quality of EB generated was validated by examining the expression encoding the M1-specific TCR at 2–3 d postactivation. The TCReng of germ layer–specific markers by RT-PCR. T cells generated were quantified by FACS analysis with an M1 epitope– specific tetramer. Phenotypic analysis of HSC derived from EB Functional characterization of Ag-specific T cells To confirm the generation of HSC precursors in hPSC-derived EB by FACS, single-cell suspensions of EB were generated by treating the hPSC-derived Functional characterization of Ag-specific T cells generated was done by EB with 0.05% trypsin. Single-cell suspensions were washed with PBS and Downloaded from coculturing them with T2-A2 cells, an HLA-A2 MHC class I molecule– stained with the indicated Ab, and surface phenotype was examined by expressing cell line described previously (6), human DC, hPSC (hESC and FACS staining using a FACSCalibur (BD Biosciences). iPSC), or hPSC-derived APC (hPSC-APC) alone or following a 15–20-min pulse with a cognate peptide (M1 or Flu peptides for M1 Ag– or Flu Ag– CFU assay specific T cells, respectively) or a control peptide (MAGE-3271–279 [M3]). Generation of HSC in hPSC-derived EB was confirmed by a CFU assay Cytokines released in the supernatants were quantified 16 h postcoculture using MethoCult medium (STEMCELL Technologies). In brief, a single- set-up by ELISA (R&D Systems), as described previously (6, 7, 23). cell suspension of hPSC-derived EB was made as described above, and http://www.jimmunol.org/ ∼ 3 4 3 4 Characterization of TLR-mediated innate immune mechanisms 2 10 to 5 10 cells were cultured in MethoCult medium in humidified six-well plates in a 37˚C incubator. Generation of CFU was fol- Expression of TLR mRNA in H9 hESC and human DC–derived iPSC lowed, and the colonies generated were counted using an Olympus IX71 lines was examined by RT-PCR using human blood–derived DC as microscope fitted with a SPOT Flex camera (SPOT Imaging Solutions). control. Primers used are listed in Supplemental Table I. Expression of functional TLR proteins in H9 hESC and DC-derived iPSC lines was Generation of APC from hPSC lines quantified by FACS staining. Expression of TLR3 and TLR9 was ex- For the generation of functional APC from hPSC, hPSC line–derived EB amined by intracellular FACS staining using Cytofix/Cytoperm (BD were cultured in tissue culture plates in IMDM containing 20% FBS Biosciences). For functional characterization of TLRs in H9 hESC and with stage-specific differentiation-supporting cytokines. In stage 1, EB 3 4 human DC–derived iPSC lines, 1 10 cells were exposed to TLR were cultured in the presence of vascular endothelial growth factor by guest on September 29, 2021 ligands (InvivoGen) in IMDM, as recommended by the manufacturer, (50 ng/ml), bone morphogenetic protein 4 (50 ng/ml), and SCF (50 ng/ and cytokines released in the supernatants were quantified 16 h later by ml) cytokines for 5 d to facilitate HSC precursor generation. At day 5, ELISA. Human blood–derived DC were used as control. TLR ligands 8 medium was supplemented with Flt3 (50 ng/ml), SCF (50 ng/ml), and used included 1 mg/ml Pam3CSK4 (TLR1/2), 10 cells per milliliter of GM-CSF (50 ng/ml) cytokines to facilitate the generation of APC. From heat-killed Listeria monocytogenes (TLR2), 0.1 mg/ml polyinosinic- day 10 onward, EB were cultured in GM-CSF (50 ng/ml) and IL-4 polycytidylic acid (TLR3), 1 mg/ml LPS (TLR4), 1 mg/ml Salmonella (20 ng/ml)–supplemented medium to generate mature APC. hPSC- typhimurium flagellin (TLR5), 1 mg/ml FSL-1, a synthetic lipoprotein derived APC were retrieved by straining cultures through 50-mm cell representing the N-terminal part of the LP44 lipoprotein of Myco- strainers (BD Biosciences) and photographed with an Olympus IX71 plasma salivarium (TLR6/2), 1 mg/ml imiquimod (TLR7), 1 mg/ml microscope equipped with a SPOT Flex camera; human blood–derived 20 mer ssRNA coupled with cationic lipid (TLR8), and 5 mMCpG DC were used as controls. Functional characterization of the APC oligonucleotides (TLR9). generated was done by examining their phagocytic ability, using FITC- MLR tagged dextran beads (40,000 m.w.; Molecular Probes), and antigenic peptide epitope-presentation efficiency, as discussed before, using hu- The primary MLR assay was set up using H9 hESC, DC-derived iPSC, or man blood–derived DC as positive controls. human peripheral blood–derived DC as stimulators; human PBL or human PBL purified CD4+CD252ve T cells and CD8+ T cells were used as re- sponders. In brief, 1 3 103 stimulator cells were cocultured with responder Results cells at different ratios (stimulatory responder ratios of 1:1, 1:10, 1:100, Generation of iPSC lines from human peripheral 1:200). Radiolabeled thymidine (1 mCi per well) was added 72 h post- blood–derived DC coculture, and thymidine incorporated in the cells was quantified 24 h later + using a liquid scintillation counter (Beckman Counter). For measure- Human PBL of HLA-A2 donors were infected with recombinant ment of cytokines in MLR experiments, supernatants were taken out be- Sendai virus vectors, a nongenome-integrating RNA virus, fore adding radiolabeled thymidine and analyzed by ELISA. Activated + 2ve + expressing the reprogramming factor genes c-Myc, Klf-4, Oct-3/4, CD4 CD25 and CD8 T cells were used as control in the cytokine- and Sox-2 to generate iPSC lines. Fig. 1A shows our iPSC- release assays. Activated T cells were generated by coculturing naive T cells in 24-well culture plates with cross-linked anti-CD3 and anti-CD28 generation schema (Fig. 1Ai) and the efficiency of Sendai vi- Ab (5 mg each), in the presence of 100 units/ml IL-2, for 5 d. For the rus infection in human PBL (Fig. 1Aii). To establish which cell cytokine-release assay, 1 3 103 activated T cells were re-exposed to anti- population in human PBL is most likely to be reprogrammed by CD3/CD28 Ab (1 mg each) in a 96-well plate, and the cytokines released Sendai virus, we first examined the efficiency of GFP encoding were quantified by ELISA. Sendai virus infection in magnetic bead–purified CD4+ T cells, Effect of IFN-g and epigenetic modulator treatment on CD8+ Tcells,andCD14+ monocyte-derived DC (Fig. 1B). In DC-derived iPSC lines subsequent iPSC line–generation experiments, DC were infected H9 hESC or DC-derived iPSC were treated with 5-azacytidine (5-Aza; with recombinant Sendai virus encoding GFP or the reprog- 10 ng/ml; Sigma-Aldrich) and/or trichostatin A (TSA; 10 ng/ml; InvivoGen) ramming factor genes. The first set of experiments was used to overnight (∼16 h), and the expression of MHC class I and MHC class II verify virus infection efficiency by FACS, and the second set of 1878 DEDIFFERENTIATION AND REDIFFERENTIATION OF HUMAN DC Downloaded from http://www.jimmunol.org/

FIGURE 1. (A) Schematic representation of derivation of iPSC lines from human PBL. (Ai) iPSC line derivation. (Aii) FACS analysis showing human PBL infected with recombinant Sendai virus expressing GFP. (Aiii) Bright-field images showing a potential iPSC clone derived from human PBL (left panel) and a well-established iPSC clone (right panel). Original magnification 310. (B) Purity of human peripheral blood–derived T cells and monocytes used in the study. (Bi) FACS analysis showing the purity of human PBL-derived CD4+ and CD8+ T cells used for iPSC line generation, using a magnetic bead–purification method. (Bii) Purity of human PBL-derived CD14+ monocytes used for iPSC line generation, before and after purification with magnetic beads. (C) Efficiency of Sendai virus infection. (Ci) FACS analysis showing efficiency of GFP encoding Sendai virus infection in human PBL-derived CD4+ T cells, CD8+ T cells, and CD14+ monocyte-derived DC (multiplicity of infection = 5). (Cii) iPSC clone derivation data from PBL-derived T cells and DC from three donors. (Ciii) FACS analysis showing the phenotype of human blood monocyte-derived DC. by guest on September 29, 2021 experiments was used to generate iPSC lines. Although we found to the characterization of their immunogenicity profiles and that the Sendai virus could infect all three cell populations with differentiation potential. comparable efficiencies (Fig. 1Ci), it was the DC fraction that Characterization of TLR-mediated innate immune mechanisms consistently generated iPSC lines, in multiple experiments from in human DC–derived iPSC lines different donors (Fig. 1Cii). Fig. 1Ciii shows the surface phenotype of human peripheral blood monocyte–derived DC used To characterize the immunogenicity profiles of human DC–derived for iPSC line generation. Fig. 2Ai shows three iPSC lines de- iPSC lines, we first examined the expression and functional rived from two individuals: a 37-y-old individual (NS02) and a status of TLR molecules using human peripheral blood–derived 75-y-old individual (NS07). As shown in Fig. 2Aii, DC-derived DC as a positive control. As shown in Fig. 3A, human DC ex- iPSC lines became free of Sendai virus vector following ∼20 press TLR mRNA that are efficiently translated into corre- passages. sponding functional proteins (Fig. 3E), and they can induce an inflammatory cytokine response upon exposure to corresponding Pluripotency profiles of human DC–derived iPSC lines TLR ligands (Fig. 3B). Interestingly, human DC–derived iPSC The pluripotency status of iPSC lines was characterized by ex- lines also express mRNA of TLR molecules (Fig. 3C); however, amining the expression of pluripotency markers by RT-PCR, these mRNA are not translated into corresponding proteins FACS, and immunofluorescence staining (Fig. 2B–D), using (Fig. 3E), and exposure of these iPSC lines to different TLR the H9 hESC line as a positive control. Expression of the coding ligands does not trigger an inflammatory cytokine response and noncoding areas of the pluripotency genes was examined (Fig. 3D). Although Fig. 3D shows TLR-mediated cytokine- by RT-PCR to rule out the possibility of PCR amplification release data for just one iPSC line (NS07#1), other lines also from open reading frames of pluripotency genes encoded by the did not produce any inflammatory cytokines upon TLR ligand reprogramming vectors (Fig. 2B). TaqMan quantitative RT- exposure (data not shown). PCR–based hPSC pluripotency scorecard (Life Technologies) We next examined the functional status of the inflammasome analysis was also performed on iPSC lines; it produced com- pathway in these iPSC lines by examining the production of IL-1b parable CT correlation plots against the H9 hESC line (Fig. 2E). in response to LPS signaling. As shown in Fig. 3F, human PBL The iPSC lines were also confirmed for alkaline phosphatase produced considerable amounts of IL-1b when exposed to LPS, activity (Fig. 2F). The karyotype of H9 hESC and human DC– whereas DC-derived iPSC lines did not. Treatment with deme- derived iPSC lines was also examined; all hPSC lines exhibited thylating and/or deacetylating agents had no effect on TLR protein normal karyotypes (Fig. 2G). Taken together, human DC– expression or inflammasome activation in these iPSC lines (data derived iPSC lines were thoroughly validated before proceeding not shown). The Journal of Immunology 1879 Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021

FIGURE 2. Generation and characterization of iPSC lines from human peripheral blood–derived DC of HLA-A2+ve donors. (Ai) Bright-field images of three iPSC lines derived from two donor-derived DC and the H9 hESC line. (ii) RT-PCR analysis for Sendai virus genome presence in DC-derived iPSC lines, ∼20 passages post-iPSC clone isolation. Human peripheral blood–derived DC that were infected with Sendai virus or were left uninfected were used as positive and negative controls, respectively. The iPSC lines generated were characterized for the expression of pluripotency markers by RT-PCR (B), FACS (C), and immunofluorescence microscopy (D), using the H9-hESC line as a positive control. In the RT-PCR analysis, coding and noncoding en- dogenous regions (depicted as Oct-E, Nanog-E, and Lin-28-E) were amplified. (E) Pluripotency characterization of iPSC lines by TaqMan quantitative RT- PCR–based hPSC pluripotency panels (Life Technologies) using H9 hESC as control. CT correlation analysis of the three DC-derived iPSC clones against the H9 hESC line. (F) Alkaline phosphatase staining of iPSC lines and H9 hESC line by VECTOR Red (Vector Laboratories). (G) Karyotype analysis of iPSC and H9 hESC lines. Data for hPSC line characterization are representative of at least three independent experiments each. Original magnification 310 (A and F); original magnification 320 (D).

Characterization of adaptive immune mechanisms in DC express a range of costimulatory and coinhibitory mole- DC-derived iPSC lines cules and Ag-presentation pathway intermediates that are neces- Because T cells are the key immune mediators of adaptive immune sary for optimum priming of Ag-specific T cell precursors (28, 29). responses, and DC are the best professional APC to orchestrate the Therefore, we next examined the expression of these molecules generation of a protective Ag-specific T cell immune response, we in DC-derived iPSC lines using human peripheral blood–derived next examined the ability of DC-derived iPSC lines to engage DC as control (Fig. 5). As shown in Fig. 5A, blood-derived DC effector T cells and induce T cell responses. To this end, we first express mRNA of all of these molecules that are translated into performed primary allogenic MLR by coculturing the human corresponding proteins (Fig. 5D), and the expression of these peripheral blood purified CD4+CD252 and CD8+ T cells with DC- molecules is upregulated in response to exposure to the DC derived iPSC lines, using human DC as a positive control. As maturation factors LPS and IFN-g (Fig. 5B). Interestingly, we shown in Fig. 4A, iPSC lines derived from human DC did not found that the DC-derived iPSC lines do not express mRNA of trigger proliferation of T cells, unlike human DC. As a secondary costimulatory molecules (Fig. 5C, left panels), but they do express readout, we examined the production of effector cytokines by mRNA of Ag-presentation pathway intermediaries (Fig. 5C, right T cells in the MLR cultures, using activated T cells as a positive panels). However, this could not be attributed to their DC origin, control. As shown in Fig. 4B, T cells activated with anti-CD3 and because the H9 hESC line also expressed mRNA of these Ag- anti-CD28 Ab, upon restimulation with anti-CD3 Ab, produced presentation pathway molecules (Fig. 5C). More importantly, significant amounts of cytokines; however, we did not detect any none of these mRNA, with the exception of LAMP-2, are trans- cytokines in the primary MLR cultures at 72 h post-MLR cocul- lated into corresponding proteins in iPSC and H9 hESC lines ture set-up. (Fig. 5D). 1880 DEDIFFERENTIATION AND REDIFFERENTIATION OF HUMAN DC Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021 FIGURE 3. TLR-mediated innate immune mechanisms in DC-derived iPSC lines. (A) RT-PCR analysis of TLR expression in 5–7-d-old human peripheral blood–derived iDC from three donors (D1, D2, and D3). (B) Functional characterization of TLR in blood-derived iDC was done by quantification of the inflammatory cytokines IL-6 (Bi), TNF-a (Bii), and IL-10 (Biii) released in the supernatants following exposure to different TLR agonist ligands. Untreated DC were used as control. Minimum value of the y-axis is 2100 (C). Expression of TLR mRNA in DC-derived iPSC lines was examined using blood-derived DC as control. (D) The functional status of TLR expressed in iPSC lines was examined by exposing them to different TLR agonist ligands and quantifying the cytokines IL-6 (Di), TNF-a (Dii), and IL-10 (Diii) released in the supernatant. Human blood–derived DC were used as positive control. (Control: untreated DC, H9 hESC, and iPSC). Minimum value of the y-axis is 2100. (E) Expression of TLR proteins in DC-derived iPSC lines was examined by FACS staining, using human monocytes as control. Line graphs of respective protein expression (lines) overlaid on isotypes (filled graphs). (F) Activation of inflammasome in response to LPS exposure was examined in DC-derived iPSC lines by measuring IL-1b released upon LPS exposure. Human PBL were used as positive control. Data shown are representative of three independent experiments.

We also found that these iPSC lines express low levels of MHC Ag-presentation pathway intermediates TAP-2 and LAMP-2 in class I molecules and significant levels of b-2 microglobulin, but iPSC lines or the H9 hESC line (Fig. 5Eiv). Fig. 5Eiv shows the they do not express MHC class II molecules, similar to the H9 effect of 5-Aza and/or TSA treatment on TAP-2 and LAMP-2 hESC line (Fig. 5Di), in agreement with previous reports (19, 30). expression in the NS07#1 iPSC line; a similar effect was ob- However, it should be noted that the expression of MHC class I served in other iPSC lines and the H9 hESC line (data not shown). and b-2 microglobulin molecules in hPSC lines was significantly We used HLA-A2+ donor-derived DC to generate iPSC lines to lower than in the human blood–derived monocytes. Although the examine the ability of these iPSC lines to present HLA-A2– human DC-derived iPSC lines and the H9 hESC line do not ex- restricted MHC class I peptide epitopes to corresponding Ag- press the MHC class I pathway–associated Ag-presentation specific T cells. For this purpose, we used M1 epitope– and Flu intermediary, TAP-2, they express the MHC class II pathway– epitope–specific natural CTL generated from human blood– associated molecule LAMP-2, although at much lower levels than derived CD8+ T cells (Fig. 6Ai) in an in vitro coculture system, as do human blood–derived monocytes (Fig. 5Dii). Expression of described previously (4, 5, 25). Because human blood–derived MHC class I molecules was inducible by IFN-g treatment CTL populations are heterogenous in their TCR repertoire, we (Fig. 5Ei), as previously reported in hESC (30), but expression of also used the M1 epitope–specific transgenic TCReng CD8+ CTL MHC class II molecules was not (Fig. 5Eii). Of further interest, (Fig. 6Bi), which were generated according to our previously pretreatment with the demethylating agent 5-Aza and/or the published methods (6, 7). As shown in Fig. 6, despite expressing deacetylating agent TSA had no effect on the expression of co- significantly low levels of HLA-A2–restricted MHC class I mol- stimulatory molecules (Supplemental Fig. 1) or MHC class I ecules (Fig. 6Aii), these iPSC and H9 hESC lines efficiently molecules (Fig. 5Eiii), nor did it affect expression of the presented the M1 peptide epitope to M1 epitope–specific natural The Journal of Immunology 1881

hematopoietic differentiation potential of EB-derived HSC. We found that EB derived from these DC-derived iPSC lines produced HSC precursors (Fig. 7D) that generated different types of colonies in CFU assay (Fig. 7E). We next examined the ability of these iPSC lines to generate human APC precursors by coculturing the iPSC-derived EB in appropriate cytokines (Fig. 7F–H) in a stage- speciﬁc differentiation system (Supplemental Fig. 3). As shown in Fig. 7Fi, we generated cells with DC morphology that are CD14+, CD11c+, and CD1c+ (Fig. 7Fii). To examine the phagocytic potential of hPSC-APC, cells were incubated with FITC-tagged dextran beads at 37˚C for 2 h, and phagocytosis of beads was examined by FACS and immunoﬂuorescence microscopy (Fig. 7G). hPSC-APC exhibit phagocytic potential (Fig. 7Gib), as do human blood–derived iDC (Fig. 7Gii). It should be noted that in a forward and side scatter plot analysis of human PBL (Fig. 7Gia, lower panel), monocytes (R2 gate) are located in the upper left portion of the plot since they are larger in size than the lymphocytes (R1 gate). Similarly, in the FITC-dextran bead phagocytosis

assay shown in Fig. 7Gib, most phagocytic iPSC-APC also were Downloaded from the larger-sized cells (Fig. 7Gia, upper panel, R2 gate). To examine the ability of iPSC-APC to present peptide epitopes to Ag- FIGURE 4. Activation of human peripheral blood–derived naive T cells specific T cells, we used human tumor Ag-specific TCReng CD4+ 2 in an MLR. (A) Allogenic human PBL and purified CD4+25 and CD8+ and CD8+ T cells (Supplemental Fig. 2). iPSC-APC efficiently T cells were cocultured with human DC–derived iPSC lines. Human pe- presented the peptide epitopes to Ag-specific CD4 (Fig. 7Hi) and ripheral blood–derived DC were used as positive control. Proliferation of CD8 (Fig. 7Hii) T cells. Human PBL–derived monocytes were http://www.jimmunol.org/ cells was quantified by measuring incorporation of radioactively labeled used as controls in the bright-field imaging (Fig. 7Fi), Giemsa thymidine. (B) Cytokines released in the MLR culture supernatants were staining (Fig. 7E), and phagocytosis assay (Fig. 7Gii) experiments. quantified by ELISA at 72 h postcoculture set-up. Data shown are representative of two or more independent experiments. The minimum value of These sets of data confirm the differentiation potential of human the y-axis is 2100. DC–derived iPSC lines. Taken together, we have found that the dedifferentiation of human DC into iPSC lines shuts down their innate and adaptive immune mechanisms (Fig. 8), and that these CTL (Fig. 6Aiii), as well as to M1 epitope–specific TCReng DC-derived iPSC lines are pluripotent and exhibit potent differ- monoclonal CD8+ antitumor T cells (Fig. 6Bii) and efficiently entiation potential. presented the Flu peptide epitope to corresponding Flu Ag-specific by guest on September 29, 2021 natural CTL (Fig. 6Aiv). M3 was used as a control peptide in Discussion functional assays in Fig. 6, and all of the CTL used exhibited Ag- Human DC express pattern recognition receptors on their cell specific effector cytokine responses. Peptide dose kinetics for the surface, as well as in intracellular compartments, to sense the presentation of M1 peptide epitope to M1 epitope–specific presence of foreign pathogens via recognition of pathogen- TCReng CD8 T cells showed that the iPSC lines exhibit peptide associated molecular patterns. In doing so, DC contribute to the dose kinetics that are similar to those of T2-A2 cells (Fig. 6C), a development of an immediate inflammatory innate immune re- human HLA-A2 MHC class I molecule–expressing cell line that is sponse, as well as undergo maturation to facilitate the subsequent routinely used in functional assays (4, 7, 27). However, in generation of adaptive immune responses. mDC present endoge- agreement with our data showing a lack of costimulatory molecule nously synthesized antigenic peptide epitopes via the MHC class I expression in human DC–derived iPSC lines (Fig. 5Di), their in- pathway and exogenously acquired Ags via the MHC class II + + ability to trigger T cell proliferation in MLR settings (Fig. 4), and pathway to CD8 and CD4 effector T cells, respectively, to the published reports showing immune-privileged properties in generate protective cellular immunity. In addition to Ag- hESC lines (19), peptide-pulsed iPSC lines could not generate M1 presentation machinery, DC express several other accessory epitope–specific functional CTL from autologous human periph- molecules, the costimulatory and coinhibitory molecules, which eral blood–derived naive CD8+ T cells in an in vitro culture assay play essential roles in the generation of a protective or an inhib- (data not shown). As expected, these peptide-pulsed iPSC lines itory immune response. These unique properties make DC a were also killed by the CTL in an epitope-specific manner (data central figure in immune-intervention strategies, including cancer not shown). immunotherapy (10, 29, 31). As discussed previously, studies in mouse models describing Differentiation potential of human DC–derived iPSC lines rejection of iPSC-based teratomas claimed that the syngenic mouse Generation of EB and formation of teratoma in mice are used to iPSC lines can be inherently immunogenic (13, 14). Other groups characterize the differentiation potential of hPSC lines by char- showed minimal or no immunogenicity of syngenic mouse iPSC acterizing the expression of ectoderm, mesoderm, and endoderm lines or their cellular derivatives (15–17). These reports high- markers in EB and the corresponding cellular structures in hPSC- lighted the need for a detailed characterization of the inherent induced teratomas. As shown in Fig. 7A, all of the iPSC lines immunogenicity of human iPSC lines for the development of safe generated EB that expressed characteristic ectoderm, mesoderm, and effective CRT. Because of the well-characterized immuno- and endoderm markers (Fig. 7B). Fig. 7C shows the histological logic properties of human DC, we generated iPSC lines from analysis of teratomas generated from an iPSC line, showing all human peripheral blood–derived DC (Fig. 2) and examined the three germ layer–associated structures. EB are known to produce status of innate and adaptive immune mechanisms in these HSC precursors, and the CFU assay has been used to test the iPSC lines, using human DC as controls, to address the issue of 1882 DEDIFFERENTIATION AND REDIFFERENTIATION OF HUMAN DC Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021

FIGURE 5. Analysis of costimulatory molecules and Ag-processing machinery in DC-derived iPSC lines. (A) RT-PCR analysis of the expression of costimulatory molecules and Ag-presentation pathway–associated molecules in the 5–7-d-old immature state (iDC). (B) FACS analysis for the expression of proteins corresponding to costimulatory molecules CD80 and CD86 and MHC class I molecules in human blood–derived DC and in iDC, as well as following maturation (mDC) with LPS and IFN-g. As shown, iDC express these molecules, and their expression is upregulated upon maturation. (C)RT- PCR analysis of the expression of costimulatory molecules and Ag-presentation pathway intermediates in DC-derived iPSC lines. (D) FACS-mediated characterization of the expression of MHC class I and MHC class II molecules and costimulatory and coinhibitory molecules (Di) and Ag-presentation pathway intermediates (Dii) in human blood monocyte-derived DC and DC-derived iPSC lines. Line graphs of respective protein expression (lines) overlaid on isotypes (filled graphs). (E) Effect of IFN-g, 5-Aza, and TSA treatment on DC-derived iPSC lines. FACS analysis of the effect of IFN-g treatment (1000 IU/ml) on the expression of MHC class I (Ei) and MHC class II molecules (Eii). FACS analysis of the effect of 5-Aza (10 ng/ml) and/or TSA (10 ng/ml) treatment on the expression of MHC class I molecules in DC-derived iPSC lines and the H9 hESC line (Eiii) and on the expression of Ag-presentation pathway intermediates TAP-2 and LAMP-2 in the NS07#1 iPSC line (Eiv). In (Eiii) and (Eiv), isotypes are shown as filled graphs; left panel shows the expression of MHC class I (Eiii) and TAP-2 and LAMP-2 (Eiv) in control hESC or iPSC lines. Effect of 5-Aza and/or TSA treatment on the expression of these proteins under different treatment condition (lines) and untreated control cells (filled graphs). Data shown are representative of at least three independent experiments. immunogenicity of human iPSC lines. We show that the dedif- effective translation of these mRNA, such as requirement of ferentiation of human DC effectively shuts down their innate lineage-specific molecular mediators/transcription factors. Taken immune response mechanisms, because no functional TLR pro- together, our data show that the human DC–derived iPSC lines do teins were detected in these iPSC lines, and the exposure to dif- not possess functional innate immune mechanisms. ferent TLR ligands does not trigger inflammatory cytokine release DC harbor MHC class I and MHC class II Ag-presentation or inflammasome activation in these iPSC lines, unlike in the pathways that are tasked with processing of internally synthe- peripheral blood–derived DC (Fig. 3). Although we found that sized and exogenously acquired Ags, respectively, with avenues for these iPSC lines express mRNA for the TLR molecules, this could cross-presentation (28). DC are well known as the best profes- not be attributed to their DC origin, because the H9 hESC line also sional APC to optimally prime Ag-specific T cell precursors for expressed TLR mRNA. Treatment with demethylating and/or the generation of protective T cell immunity, as a result of their deacetylating agents had no effect on the expression of func- ability to provide signal 1 (engagement of TCR with MHC class I tional TLR proteins or on the activation of inflammasomes in and MHC class II molecules bound antigenic peptide epitopes), human DC–derived iPSC lines (data not shown). The presence of signal 2 (costimulatory signals), and signal 3 (production of TLR mRNA but the absence of functional TLR proteins also immune-stimulatory cytokines, such as IL-12) (32). Therefore, we suggest the involvement of posttranscriptional mechanisms for an next examined the status of coreceptor molecules and Ag- processing The Journal of Immunology 1883 Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021 FIGURE 6. Presentation of HLA-A2–restricted peptide epitopes to Ag-specific T cells by human DC–derived iPSC lines. (Ai) M1 epitope– and Flu epitope–specific natural CTL were generated from HLA-A2+ve donor blood-derived CD8+ naive T cells and quantified by FACS using the respective epitope-specific tetramers. Isotype is overlaid on the upper left panel (upper left quadrant). (Aii) Expression of HLA-A2–restricted MHC class I molecules on DC-derived iPSC lines and the H9 hESC line was quantified by FACS using HLA-A2+ve and HLA-A22ve donor-derived PBL as positive and negative controls, respectively. HLA-A2 protein expression (lines) overlaid on isotypes (filled graphs). (Aiii) M1 epitope–specific CTL were cocultured with H9 hESC or DC-derived iPSC pulsed with control or cognate peptides; IFN-g and TNF-a cytokines released in the supernatants were quantified by ELISA. (Control cells, cells pulsed with control peptide, M3, or the M1 cognate peptide.) The minimum value of the y-axis in the graphs is 2100. (Aiv) Flu epitope–specific CTL were cocultured with peptide-pulsed H9 hESC or DC-derived iPSC as in (Aiii) using Flu as the cognate peptide, and IFN-g cytokine released in the supernatants was quantified by ELISA. Peptide-pulsed HLA-A2+ve donor-derived DC and/or HLA-A2+ surrogate target T2 cells were used as positive control in (Aiii) and (Aiv). The minimum value of the y-axis in the graphs is 2100. (Bi) M1 epitope–specific TCReng CD8+ CTL were generated according to our published methods (7, 27) and quantified by M1 epitope–specific tetramer staining. (Bii) TCReng CD8+ CTL were used in functional assays against the peptide-pulsed iPSC and H9 hESC lines, as in (A), for the natural CTL. TCReng CD8 T cells also effectively recognized peptide epitope presented on human DC-derived iPSC lines. (C) Different dosages of M1 peptide–pulsed H9 hESC, DC-derived iPSC, and T2-A2 cells were cocultured with M1 epitope–specific TCReng CD8 T cells; IFN-g and TNF-a released in the supernatants were quantified. The minimum values of the y-axis in the graphs in (Ci) and (Cii) are 220 and 2100, respectively. Data shown are representative of at least two independent experiments each. and -presentation intermediates in human DC–derived iPSC lines. intermediate TAP-2; however, this could not be attributed to their We found that the dedifferentiation of human DC results in dra- DC origin because the H9 hESC line exhibited a similar profile matically lower expression of MHC class I molecules in DC-derived (Fig. 5C). Of further interest, TAP-2 mRNA were not translated iPSC lines and no expression of MHC class II molecules into corresponding functional TAP-2 protein in these lines (Fig. (Fig. 5Di). Although the expression of MHC class I molecules is 5Dii), and treatment with demethylating and/or deacetylating induced by IFN-g treatment, it had no effect on the expression of agents had no effect on TAP-2 protein expression (Fig. 5Eiv). MHC class II molecules (Fig. 5Ei, ii). In addition, treatment with Although we found mRNA and protein expression for the lyso- demethylating and/or deacetylating agents had no effect on the somal pathway–associated intermediate LAMP-2 in these iPSC expression of functional MHC class I and MHC class II proteins lines (Fig. 5C, 5Dii), this also could not be attributed to their DC (Fig. 5Eiii, data not shown). These DC-derived iPSC do not ex- origin because the H9 hESC line also expressed it. As discussed press the mRNA or the corresponding proteins of costimulatory previously, engagement of TLR molecules on DC modulates the molecules (Fig. 5C, 5Di), and treatment with demethylating expression of Ag-processing and -presentation pathway interme- and/or deacetylating agents also has no effect on the expression diates and results in their functional maturation (33); however, of these molecules (Supplemental Fig. 1). Interestingly, these engagement of TLR ligands had no effect on the immunogenic iPSC lines express mRNA for the MHC class I peptide-loading profile of human DC–derived iPSC lines, because these cells do 1884 DEDIFFERENTIATION AND REDIFFERENTIATION OF HUMAN DC Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021

FIGURE 7. Characterization of differentiation potential of DC-derived iPSC lines. (A) EB derived from iPSC lines and the H9 hESC line. (B) RT-PCR analysis of iPSC and H9 hESC line–derived EB confirmed the expression of ectoderm, mesoderm, and endoderm markers. (C) Analysis of teratoma generated by the NS07#1 iPSC line with germ layer–associated structures. (D) FACS analysis of single-cell suspension of EB shows generation of CD34+ve HSC precursors. (E) Differentiation potential of EB-derived HSC precursors in a CFU assay. The CFU colonies generated (upper panels) were scored as CFU-GM, BFU-E, and CFU-GEMM (ref., Atlas on hematopoietic colonies, STEMCELL Technologies). Cells from these CFU colonies were stained with Wright–Giemsa stain (lower panels); insets show data from different respective planes. (F–H). Generation of phagocytic APC from DC-derived iPSC lines. (Fi) iPSC-derived EB were cultured in a three-step method (Supplemental Fig. 3) to generate APC (iPSC-APC). Bright-field images (upper panels) and Giemsa staining (lower panels) of iPSC-APC and human blood–derived DC. Insets show data from different areas of the stained slides. Cells with morphological features of human DC were generated. (Fii) FACS analysis of iPSC-APC cultures confirmed the presence of CD14+, CD1c+, and CD11c+ precursors. (Gi) Phagocytic ability of iPSC-APC was examined by phagocytosis of FITC-tagged dextran beads (40,000 m.w.; Molecular Probes). (Gia) Gating strategy (upper panel). (Gib) As expected, larger-sized cells (R2 gate) contained more CD11c+ve cells with more phagocytic ability than did the smaller-sized cells. (Gia) Scatter plot of human PBL, with the R1 gate corresponding to the smaller-sized lymphocyte fraction and the R2 gate corresponding to the larger-sized monocyte fraction (lower panel). Human peripheral blood–derived CD11c+ iDC were used as positive control. Phagocytosis was examined by FACS (Giia) and by immunofluorescence microscopy (Giib). M1 epitope–specific TCReng antitumor CD4+ (Hi) and CD8+ (Hii) T cells were generated (Supplemental Fig. 2) and used to recognize the M1 epitope on peptide-pulsed iPSC-APC (iAPC). iPSC-APC efficiently presented M1 peptide to TCReng CD4 and CD8 T cells, as did T2-A2 cells (Supplemental Fig. 2). The minimum value of the y-axis in the graphs is 2100. Data shown are representative of at least two independent experiments each. Original magnification 34(A, left, and E, top left); original magnification 310 (A, right, and E, top middle and top right); original magnification 340 (E, bottom, and Fi). not express functional TLR proteins. Taken together, our data expressing low levels of MHC class I molecules in comparison show that the coreceptor engagement ability, as well as the Ag- with DC (Fig. 6Aii), these iPSC lines can efficiently present the processing and -presentation machinery, of human DC is effec- peptide epitopes to Ag-specific T cells (Fig. 6Aiii, iv); in doing so, tively shut down during the course of their dedifferentiation into peptide-pulsed iPSC lines are also killed by cytolytic T cells (data the iPSC state. not shown). Taken together, our findings (summarized schemati- Our data showing that the DC-derived iPSC lines, unlike human cally in Fig. 8) demonstrate that the dedifferentiation of human DC, do not trigger T cell proliferation in an allogenic MLR, even at terminally differentiated immunogenic DC effectively shuts down an E:T ratio of 200:1 (Fig. 4A), is in agreement with previous their innate and adaptive immune mechanisms. Our data also findings in hESC (19). We used an HLA-A2+ donor population for suggest that the presentation of antigenic peptides acquired from iPSC derivation, which allowed us to examine their ability to the microenvironment to cytolytic T cells might result in the tar- present HLA-A2 MHC class I molecule–restricted peptide epi- geted killing of the transplanted iPSC and rejection of iPSC- topes to Ag-specific T cells (Fig. 6). We found that, despite induced teratomas in immune-competent animals. The Journal of Immunology 1885 Downloaded from

FIGURE 8. Diagram showing the immunogenicity profile of human DC and the iPSC lines derived from them. Human DC harbor well-developed innate http://www.jimmunol.org/ and adaptive immune mechanisms. The innate immune receptors (e.g., TLR) trigger an immediate response upon encountering pathogen-associated molecular patterns/danger-associated molecular patterns. Adaptive immune mechanisms process the internally synthesized and exogenously acquired Ag through MHC class I and MHC class II pathways and present the processed antigenic epitopes to CD8+ and CD4+ T cells, respectively, resulting in generation of CD8+ cytolytic responses and CD4+ helper responses. These immune mechanisms are shut down as a result of dedifferentiation of human DC; however, these iPSC express MHC class I molecules that can acquire antigenic peptide epitopes from the microenvironment and present them to Ag-specific T cells. The IFN-g produced by T cells further induces the expression of MHC class I on these iPSC lines in a positive-feedback loop, and the peptide- bound iPSC lines are also killed by cytolytic T cells in an Ag-specific manner.

To characterize the differentiation potential of the iPSC lines Acknowledgments by guest on September 29, 2021 generated, we examined their ability to generate HSC precursors We thank UCHC General Clinical Core staff members Susan Walters, and functional APC. Human DC were generated from the HSC Kathleen Curley, Harriet Zawistowski, and Tom Kiley for human subject precursors derived from different anatomical locations, such as blood samples. We also thank Leann Crandell (University of Connecticut peripheral blood, bone marrow, and umbilical cord blood, and they [UConn] Health Stem Cell Core Facility) for providing hESC lines and tech- were shown to be effective immune stimulators (34–36). However, nical insights, John Glynn (UConn Health Molecular Core Facility) for help donor-specific iPSC lines can be used as an off-the-shelf resource with real-time PCR for stem cell pluripotency analysis, Susan Krueger for deriving autologous immunologically matched HSC precursors (UConn Health Microscopy Core Facility) for help with confocal microscopy, Dr. Judy Brown (UConn Chromosome Core Facility) for karyotype to generate donor-specific cells of choice, including APC. Given analysis of hESC and iPSC lines, and Dr. Evan Jellison (UConn Health that APC can be immunogenic or tolerogenic (28, 37), incorpo- FACS Core Facility) for help with FACS analysis. We also thank the lab- ration of appropriate growth supplements in differentiating me- oratories of Dr. Alexander Lichtler and Dr. Ernst Reichenberger for help dium could also be used to generate naive APC with defined during the course of the study. immunological properties. It should be pointed out that APC were generated from hESC lines (38–40); however, efficient models are Disclosures needed for in-depth characterization of the molecular, cellular, and The authors have no financial conflicts of interest. functional profiles of human iPSC-derived APC. This is important given that the molecular, cellular, and functional profiles of cellular derivatives derived from iPSC lines, from diverse somatic References cell sources, and/or by various methods can be different (41). In 1. Thomson, J. A., J. Itskovitz-Eldor, S. S. Shapiro, M. A. Waknitz, J. J. Swiergiel, V. S. Marshall, and J. M. Jones. 1998. Embryonic stem cell lines derived from this context, our data showing that young and elderly donor DC- human blastocysts. Science 282: 1145–1147. derived iPSC lines exhibit comparable differentiation potential are 2. Takahashi, K., K. Tanabe, M. Ohnuki, M. Narita, T. Ichisaka, K. Tomoda, and S. Yamanaka. 2007. Induction of pluripotent stem cells from adult human fi- noteworthy, given that they yield EB that can produce HSC pre- broblasts by defined factors. Cell 131: 861–872. cursors that can be further differentiated into cells that exhibit DC 3. Yu, J., M. A. Vodyanik, K. Smuga-Otto, J. Antosiewicz-Bourget, J. L. Frane, morphology, are phagocytic, and can present antigenic peptide S. Tian, J. Nie, G. A. Jonsdottir, V. Ruotti, R. Stewart, et al. 2007. Induced pluripotent stem cell lines derived from human somatic cells. Science 318: 1917– epitopes to Ag-specific T cells (Fig. 7). Our findings suggest that 1920. the dedifferentiation and redifferentiation of human DC can be 4. Chhabra, A., N. G. Chakraborty, and B. Mukherji. 2008. Silencing of endoge- used as an efficient model to systematically characterize the cel- nous IL-10 in human dendritic cells leads to the generation of an improved CTL response against human melanoma associated antigenic epitope, MART-1 27-35. lular, functional, and molecular profiles of human iPSC-derived Clin. Immunol. 126: 251–259. APC, using autologous blood–derived DC as control. We believe 5.Chhabra,A.,S.Mehrotra,N.G.Chakraborty,B.Mukherji,andD.I.Dorsky. 2004. Cross-presentation of a human tumor antigen delivered to dendritic cells that our findings have implications for understanding the devel- by HSV VP22-mediated protein translocation. Eur. J. Immunol. 34: 2824– opment of human DC lineages and for DC-based CRT. 2833. 1886 DEDIFFERENTIATION AND REDIFFERENTIATION OF HUMAN DC

6. Chhabra, A., L. Yang, P. Wang, B. Comin-Anduix, R. Das, N. G. Chakraborty, 23. Mehrotra, S., A. Chhabra, S. Chattopadhyay, D. I. Dorsky, N. G. Chakraborty, and S. Ray, S. Mehrotra, H. Yang, C. L. Hardee, et al. 2008. CD4+CD252 T cells B. Mukherji. 2004. Rescuing melanoma epitope-specific cytolytic T lymphocytes transduced to express MHC class I-restricted epitope-specific TCR synthesize from activation-induced cell death, by SP600125, an inhibitor of JNK: implications Th1 cytokines and exhibit MHC class I-restricted cytolytic effector function in a in cancer immunotherapy. J. Immunol. 173: 6017–6024. human melanoma model. J. Immunol. 181: 1063–1070. 24. Mehrotra, S., A. Chhabra, A. Chakraborty, S. Chattopadhyay, M. Slowik, 7. Ray, S., A. Chhabra, N. G. Chakraborty, U. Hegde, D. I. Dorsky, T. Chodon, R. Stevens, R. Zengou, C. Mathias, L. H. Butterfield, D. I. Dorsky, et al. 2004. E. von Euw, B. Comin-Anduix, R. C. Koya, A. Ribas, et al; UCLA-CALTECH- Antigen presentation by MART-1 adenovirus-transduced interleukin-10- CHLA-USC-UCONN Consortium on Translational Program in Engineered polarized human monocyte-derived dendritic cells. Immunology 113: 472– Immunity. 2010. MHC-I–restricted melanoma antigen specific TCR-engineered 481. human CD4+ T cells exhibit multifunctional effector and helper responses, 25. Chhabra, A., S. Mehrotra, N. G. Chakraborty, D. I. Dorsky, and B. Mukherji. in vitro. Clin. Immunol. 136: 338–347. 2006. Activation-induced cell death of human melanoma specific cytotoxic 8. Chhabra, A. 2009. MHC class I TCR engineered anti-tumor CD4 T cells: im- T lymphocytes is mediated by apoptosis-inducing factor. Eur. J. Immunol. 36: plications for cancer immunotherapy. Endocr. Metab. Immune Disord. Drug 3167–3174. Targets 9: 344–352. 26. Mehrotra, S., A. Chhabra, U. Hegde, N. G. Chakraborty, and B. Mukherji. 2007. 9. Mukherji, B., N. G. Chakraborty, S. Yamasaki, T. Okino, H. Yamase, J. R. Sporn, Inhibition of c-Jun N-terminal kinase rescues influenza epitope-specific human S. K. Kurtzman, M. T. Ergin, J. Ozols, J. Meehan, et al. 1995. Induction of cytolytic T lymphocytes from activation-induced cell death. J. Leukoc. Biol. 81: antigen-specific cytolytic T cells in situ in human melanoma by immunization 539–547. with synthetic peptide-pulsed autologous antigen presenting cells. Proc. Natl. 27. Chhabra, A., and B. Mukherji. 2013. Death receptor-independent activation- Acad. Sci. USA 92: 8078–8082. induced cell death in human melanoma antigen-specific MHC class I–re- 10. Gilboa, E. 2007. DC-based cancer vaccines. J. Clin. Invest. 117: 1195–1203. stricted TCR-engineered CD4 T cells. J. Immunol. 191: 3471–3477. 11. Morgan, R. A., M. E. Dudley, J. R. Wunderlich, M. S. Hughes, J. C. Yang, 28. Schuurhuis, D. H., N. Fu, F. Ossendorp, and C. J. Melief. 2006. Ins and outs of R. M. Sherry, R. E. Royal, S. L. Topalian, U. S. Kammula, N. P. Restifo, et al. dendritic cells. Int. Arch. Allergy Immunol. 140: 53–72. 2006. Cancer regression in patients after transfer of genetically 29. Steinman, R. M. 2012. Decisions about dendritic cells: past, present, and future. engineered lymphocytes. Science 314: 126–129. Annu. Rev. Immunol. 30: 1–22. 12. Porter, D. L., M. Kalos, Z. Zheng, B. Levine, and C. June. 2011. Chimeric 30. Drukker, M., G. Katz, A. Urbach, M. Schuldiner, G. Markel, J. Itskovitz-Eldor, antigen receptor therapy for B-cell malignancies. J. Cancer 2: 331–332. B. Reubinoff, O. Mandelboim, and N. Benvenisty. 2002. Characterization of the Downloaded from 13. Zhao, T., Z. N. Zhang, Z. Rong, and Y. Xu. 2011. Immunogenicity of induced expression of MHC proteins in human embryonic stem cells. Proc. Natl. Acad. pluripotent stem cells. Nature 474: 212–215. Sci. USA 99: 9864–9869. 14. Zhao, T., Z. N. Zhang, P. D. Westenskow, D. Todorova, Z. Hu, T. Lin, Z. Rong, 31. Banchereau, J., and A. K. Palucka. 2005. Dendritic cells as therapeutic vaccines J. Kim, J. He, M. Wang, et al. 2015. Humanized mice reveal differential im- against cancer. Nat. Rev. Immunol. 5: 296–306. munogenicity of cells derived from autologous induced pluripotent stem cells. 32. Reis e Sousa, C. 2006. Dendritic cells in a mature age. Nat. Rev. Immunol. 6: Cell Stem Cell 17: 353–359. 476–483. 15. Araki, R., M. Uda, Y. Hoki, M. Sunayama, M. Nakamura, S. Ando, M. Sugiura, 33. Vyas, J. M., A. G. Van der Veen, and H. L. Ploegh. 2008. The known unknowns

H. Ideno, A. Shimada, A. Nifuji, and M. Abe. 2013. Negligible immunogenicity of antigen processing and presentation. Nat. Rev. Immunol. 8: 607–618. http://www.jimmunol.org/ of terminally differentiated cells derived from induced pluripotent or embryonic 34. Caux, C., C. Dezutter-Dambuyant, D. Schmitt, and J. Banchereau. 1992. GM- stem cells. Nature 494: 100–104. CSF and TNF-alpha cooperate in the generation of dendritic Langerhans cells. 16. Guha, P., J. W. Morgan, G. Mostoslavsky, N. P. Rodrigues, and A. S. Boyd. 2013. Nature 360: 258–261. Lack of immune response to differentiated cells derived from syngeneic induced 35. Szabolcs, P., M. A. Moore, and J. W. Young. 1995. Expansion of immunosti- pluripotent stem cells. Cell Stem Cell 12: 407–412. mulatory dendritic cells among the myeloid progeny of human CD34+ bone 17. Morizane, A., D. Doi, T. Kikuchi, K. Okita, A. Hotta, T. Kawasaki, T. Hayashi, marrow precursors cultured with c-kit ligand, granulocyte-macrophage colony- H. Onoe, T. Shiina, S. Yamanaka, and J. Takahashi. 2013. Direct comparison of stimulating factor, and TNF-alpha. J. Immunol. 154: 5851–5861. autologous and allogeneic transplantation of iPSC-derived neural cells in the 36. Strobl, H., E. Riedl, C. Scheinecker, C. Bello-Fernandez, W. F. Pickl, brain of a non-human primate. Stem Cell Rep. 1: 283–292. K. Rappersberger, O. Majdic, and W. Knapp. 1996. TGF-beta 1 promotes in vitro 18. Drukker, M. 2004. Immunogenicity of human embryonic stem cells: can we development of dendritic cells from CD34+ hemopoietic progenitors. J. Immu- achieve tolerance? Springer Semin. Immunopathol. 26: 201–213. nol. 157: 1499–1507.

19. Li, L., M. L. Baroja, A. Majumdar, K. Chadwick, A. Rouleau, L. Gallacher, 37. Steinman, R. M., D. Hawiger, and M. C. Nussenzweig. 2003. Tolerogenic by guest on September 29, 2021 I. Ferber, J. Lebkowski, T. Martin, J. Madrenas, and M. Bhatia. 2004. Human dendritic cells. Annu. Rev. Immunol. 21: 685–711. embryonic stem cells possess immune-privileged properties. Stem Cells 22: 448– 38. Zhan, X., G. Dravid, Z. Ye, H. Hammond, M. Shamblott, J. Gearhart, and 456. L. Cheng. 2004. Functional antigen-presenting leucocytes derived from human 20. Mohib, K., D. Allan, and L. Wang. 2010. Human embryonic stem cell-extracts embryonic stem cells in vitro. Lancet 364: 163–171. inhibit the differentiation and function of monocyte-derived dendritic cells. Stem 39. Slukvin, I. I., M. A. Vodyanik, J. A. Thomson, M. E. Gumenyuk, and K. D. Choi. Cell Rev. 6: 611–621. 2006. Directed differentiation of human embryonic stem cells into functional 21. Drukker, M., H. Katchman, G. Katz, S. Even-Tov Friedman, E. Shezen, dendritic cells through the myeloid pathway. J. Immunol. 176: 2924–2932. E. Hornstein, O. Mandelboim, Y. Reisner, and N. Benvenisty. 2006. Human 40. Tseng, S. Y., K. P. Nishimoto, K. M. Silk, A. S. Majumdar, G. N. Dawes, embryonic stem cells and their differentiated derivatives are less susceptible to H. Waldmann, P. J. Fairchild, J. S. Lebkowski, and A. Reddy. 2009. Generation immune rejection than adult cells. Stem Cells 24: 221–229. of immunogenic dendritic cells from human embryonic stem cells without serum 22. Swijnenburg, R. J., S. Schrepfer, J. A. Govaert, F. Cao, K. Ransohoff, and feeder cells. Regen. Med. 4: 513–526. A. Y. Sheikh, M. Haddad, A. J. Connolly, M. M. Davis, R. C. Robbins, and 41. Polo, J. M., S. Liu, M. E. Figueroa, W. Kulalert, S. Eminli, K. Y. Tan, J. C. Wu. 2008. Immunosuppressive therapy mitigates immunological rejection E. Apostolou, M. Stadtfeld, Y. Li, T. Shioda, et al. 2010. Cell type of origin of human embryonic stem cell xenografts. Proc. Natl. Acad. Sci. USA 105: inﬂuences the molecular and functional properties of mouse induced pluripotent 12991–12996. stem cells. Nat. Biotechnol. 28: 848–855.