Inhibition of T-Cell Activation by Retinal Pigment Epithelial Cells Derived From Induced Pluripotent Stem Cells

Abstract

Purpose.: The purpose of this study was to determine whether human retinal pigment epithelial (RPE) cells from induced pluripotent stem (iPS) cells could inhibit T-cell activation in vitro.

Methods.: Cultured iPS-derived RPE (iPS-RPE) cells were established from fresh skin tissues or dental pulp cells obtained from healthy donors or a retinal patient after informed consent was obtained. To confirm expression of the specific markers on iPS and iPS-RPE cells, immunohistochemistry, quantitative RT-PCR (qRT-PCR), and flow cytometry were performed. Target T cells were obtained from peripheral blood mononuclear cells of healthy donors. Target T cells were assessed for proliferation by incorporation of bromodeoxyuridine or carboxyfluorescein succinimidyl ester for production of cytokines such as IFN-γ. Expression of TGFβ and other candidate molecules by iPS-RPE cells was evaluated with flow cytometry, ELISA, multiplex cytokine array, immunohistochemistry, and qRT-PCR.

Results.: The RPE cells we established from iPS cells had many characteristics of mature RPE cells but no characteristics of pluripotent stem cells. Cultured iPS-RPE cells inhibited cell proliferation and production of IFN-γ by activated CD4⁺ T cells. In some bystander T cells, iPS-derived RPE cells induced CD25⁺Foxp3⁺ regulatory T cells in vitro. Induced pluripotent stem-RPE cells constitutively expressed TGFβ and suppressed activation of T cells via soluble TGFβ, because TGFβ-downregulated iPS-RPE cells did not inhibit this T-cell activation.

Conclusions.: Cultured iPS-derived retinal cells fully suppress T-cell activation. Transplantation of iPS-RPE cells into the eye might be a therapy for retinal disorders.

Introduction

Ocular pigment epithelial cells contribute to immune tolerance in the eyes.^1–3 In the posterior segment of the eye, the retinal pigment epithelium (RPE) cell monolayer is an immune-privileged tissue³ that functions as the principal mediator of immune privilege in the subretinal space.⁴ Retinal pigment epithelium, either as explanted tissues or cultured cells, is immunosuppressive. In addition, RPE cells secrete soluble factors such as transforming growth factor beta (TGFβ), thrombospondin-1 (TSP-1), and several inhibitory factors that inhibit immune inflammatory cells.^5–14 Thus, it is assumed that the immunoregulatory activities displayed by RPE cells in vitro reflect their activities in vivo. 

Recently, reprogramming of cells has become widely accepted as a tool for obtaining transplantation materials. There has been great interest in cell-based therapies, including retinal transplants, because there is a reduced risk of immune rejection. Stem cells have the capacity for self-renewal plus the capacity to generate several differentiated cells.¹⁵ They are derived from many sources including human embryonic stem (ES) cells or adult-derived induced pluripotent stem (iPS) cells and have found early application in the context of ocular disease.^16,17 Retinal pigment epithelium graft surgery may preserve some degree of macular function, but immune rejection can occur around the graft when allogeneic RPE cells are used.^18–20 It is unknown whether donor-derived RPE cells, especially stem cell–derived RPE cells, display immunosuppressive properties when transplanted for the treatment of ocular disorders. 

Therefore, the purpose of the present study was to determine whether human RPE cells derived from iPS cells could inhibit T-cell activation in vitro. 

Materials and Methods

Establishment of Human iPS Cells

This research study followed the tenets of the Declaration of Helsinki, and the study was approved by the Institutional Ethics Committee of the Center for Developmental Biology, RIKEN. After informed consent was obtained, dermal tissue cells from healthy donors were cultured for 2 weeks, and iPS cells were then established from skin fibroblasts (TLHD1) or dental pulp cells (454E2), using an episomal vector carrying six genes: OCT3/4, SOX2, KLF4, L-MYC, LIN28, and p53 shRNA (Fig. 1A). In addition, iPS cells (101G26 and 836B1) were also established from the dermal fibroblasts of a patient with retinitis pigmentosa (101G26) or a healthy donor (836B1), using an episomal vector carrying six genes: OCT3/4, SOX2, KLF4, L-MYC, LIN28, and GLIS1.²¹ Electroporation of human skin fibroblasts or dental pulp cells was performed with several combinations of episomal vectors. Transfected cells were maintained in embryonic stem cell medium (primate ES cell medium; ReproCELL, Yokohama, Japan). The cells adopted human-ES cell-like morphology after transfection. 

Figure 1

 

Establishment of human iPS cells from human skin fibroblasts and terminally differentiated human RPE cells from iPS cells. (A) Summary of establishment of human iPS cells. HD, healthy donor; RP, retinitis pigmentosa; CiRA, Center for iPS Cell Research and Application (Kyoto University). (B) Polymerase chain reaction analysis of integration of episomal vectors in the generated iPS cells (TLHD1). As a positive control, two kinds of DNA were used positive control (PC) lines, including plasmid integrated lines, and the pCXLE-hSK vector (2 ng). (C) Immunofluorescence analysis of expression of the pluripotency markers Oct3 and -4, TRA-1-60, and SSEA-4 in human iPS cells (836B1). Induced pluripotent stem cells were also stained with isotype control (mouse IgG). Cell nuclei were counterstained with DAPI. Scale bar: 200 μm. (D) Retinal pigment epithelium cells induced from iPS cells (836B1) clearly showed polygonal morphology, mostly hexagonal, and contained melanin. (E) Measurement of phagocytosis by iPS-RPE cells. Induced pluripotent stem-RPE cells (836B1) were cocultured with FITC-ROS at 37°C (blue) and analyzed using flow cytometry. Control cells, which were iPS-RPE cells cultured with FITC-ROS at 4°C, were used to obtain baseline fluorescence (green), and iPS-RPE cells without ROS at 37°C were also analyzed (red). (F) Retinal pigment epithelium cell–specific markers MiTF and ZO-1 in iPS-RPE cells (836B1) were detected by immunostaining. Scale bars: 50 μm. (G) Detection of RPE marker genes in iPS-RPE cells. Total RNA was extracted from iPS-RPE cells (836B1, 454E2, and 101G26). Human iPS cells, 836B1, were also prepared as a control. For PCR amplification, cDNA was amplified by using primers for human bestrophin (Best), RPE65, Pax6, TGFβ1, TGFβ2, TGFβ3, Lin28, Nanog, and β-actin. Results indicate the relative expression of these molecules (ΔΔC_t: control iPS cells = 1). ND, not detected.

To confirm the expression of iPS markers, cultured iPS cells were fixed with 4% paraformaldehyde (PFA) for 24 hours at 4°C and washed with PBS. Induced pluripotent stem cells were treated with 0.3% Triton X-100-PBS (Sigma-Aldrich Corp., St. Louis, MO, USA) for 30 minutes at room temperature, and the samples were incubated overnight at 4°C with primary antibodies for Oct3/4, SSEA-4, Tra-1-60, and Nanog or with control mouse immunoglobulin G (IgG), diluted in PBS containing 2% fetal bovine serum (FBS). After the primary antibodies were removed and the samples were washed, secondary antibodies were applied for 1 hour at 25°C. Cell nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI). After being washed, stained cells were observed by fluorescence microscopy. Antibody information is described in Supplementary Table S1. 

To confirm that the episomal vectors were not integrated into the generated iPS cell lines, we performed PCR analysis.²² The primers used for OriP-F were 5′-TTCCACGAGGGTAGTGAACC-3′; for OriP-R were TCGGGGGTGTTAGAGACAAC; for endo-Oct3/4-F were 5′-AGTTTGTGCCAGGGTTTTTG-3′; and for endo-Oct3/4-R were 5′- ACTTCACCTTCCCTCCAACC-3.′ As a positive control, two kinds of DNA were used: DNA from positive control lines, including plasmid integrated lines, and pCXLE-hSK vector DNA (2 ng). 

Preparation of Primary Cultures of iPS-Derived RPE (iPS-RPE) Cells

For differentiation into RPE cells, human iPS cells were cultured for 4 days on gelatin-coated dishes in Glasgow minimum essential medium (GMEM; Gibco/Life Technologies, Grand Island, NY, USA) supplemented with 1 mM sodium pyruvate, 0.1 mM nonessential amino acids, 0.1 mM 2-mercaptoethanol, and 20% knockout serum replacement (KSR) and then incubated for 6 days in GMEM supplemented with 15% KSR. Cells were then cultured for 20 days in GMEM plus 10% KSR. Signal inhibitors such as Y-27632 (10 μM; Wako, Osaka, Japan), SB431542 (5 μM; Sigma-Aldrich Corp.) and CKI-7 (3 μM; Sigma-Aldrich Corp.) were added to the GMEM for 18 days. After the appearance of RPE-like colonies (pigment-rich cells), the medium was switched to Dulbecco's modified Eagle's medium (DMEM) supplemented with B27 supplement (Invitrogen, Carlsbad, CA) and 2 mM l-glutamine (Sigma-Aldrich Corp.), and RPE colonies were transferred to CELLstart-coated dishes (Invitrogen) in B27 medium supplemented with 10 ng/mL human basic fibroblast growth factor (FGF; Wako) and SB431542 and cultured until confluence. 

To confirm phagocytic function, cultured iPS-derived RPE (iPS-RPE) cells were cultured with RPE medium in the presence of FITC-labeled porcine-shed photoreceptor rod outer segments (ROS, 10 μg/cm²) for 5 hours at 37°C. As a control, untreated RPE cells without ROS were also prepared. After incubation with or without FITC-ROS, RPE cells were treated with 0.25% trypsin-EDTA, and phagocytosis was evaluated by fluorescence-activated cell sorting FACS. Cultured iPS-RPE cells were fixed with 4% PFA-PBS for 15 minutes at room temperature, washed three times with PBS, and permeabilized with 0.3% Triton X-100-PBS. As primary antibodies, anti-MiTF (mouse IgG), anti-ZO-1 antibody (rabbit IgG), control mouse IgG and control rabbit IgG were used. Goat anti-mouse IgG or anti-rabbit IgG was used as secondary antibody. Cell nuclei were counterstained with DAPI. Antibody information is described in Supplementary Table S1. The controls used for iPS-RPE cells were human ES-derived RPE cells,¹⁷ human RPE cell lines (ARPE-19 cells),²³ and primary fetal RPE cells (Ronza, Tokyo, Japan). 

Preparation of Immune Cells

Target-activated T cells were established from autogeneic or allogeneic T cells (CD4 or pan-T) of peripheral blood mononuclear cells (PBMCs). T cells were prepared separately by using separation beads and columns (MACS cell isolation kits; Miltenyi Biotec, Auburn, CA, USA). More than 95% of these cells were CD4-positive (CD4⁺). T cells were cocultured in RPMI 1640 medium containing 10% FBS. 

T-Cell–RPE Cell Assay of Cell Proliferation and Cytokine Production

Purified T cells (2 × 10⁵ to 5 × 10⁵ cells/well in 96-well plates) from healthy donor PBMCs were stimulated with anti-human CD3 antibody (1 μg/mL) and were then incubated with iPS-RPE cells for 72 hours. For some experiments, anti-human CD28 antibody (1 μg/mL) or phytohemagglutinin-P (PHA-P; Difco, Detroit, MI, USA) was also used. Induced pluripotent stem-RPE cells were cultured separately in 96-well plates (1 × 10⁵ cells/well) or 24-well plates (5 × 10⁵ cells/well). T-cell activation was assessed in terms of proliferation by carboxyfluorescein succinimidyl ester (CFSE; Molecular Probes, Eugene, OR, USA) or bromodeoxyuridine (BrdU) incorporation (ELISA; Roche Diagnostics, Mannheim, Germany) or of IFN-γ production (R&D Systems, Minneapolis, MN, USA). Labeling of T cells with CFSE was performed.²⁴ After 72 to 96 hours, CFSE-labeled T cells were washed and analyzed by flow cytometry. 

Flow Cytometry

Expression of IL-1 receptor antagonist (IL-1RA), IL-6, IL-8/CXCL8, IL-9, IL-21, IL-22, TNF-α, TNFRI, IFN-γ, monokine induced by gamma interferon (MIG/CXCL9), monocyte chemoattractant protein-1 (MCP-1, CCL2), macrophage inflammatory protein 3 α (MIP-3 α/CCL20), VEGF, macrophage migration inhibitory factor (MIF), TGFβ1, TGFβ2, and thrombospondin (TSP) by iPS-RPE cells (836B1) was evaluated by FACS analysis. For analysis of cytokine expression, these RPE cells were treated with an intracellular staining material (Cytofix/Cytoperm kits; BD PharMingen, San Diego, CA, USA). Before being stained, these cells were incubated with a human Fc block (Miltenyi Biotec) at 4°C for 15 minutes. After human Fc block staining, these RPE cells were stained with anti-human antibodies (listed above) at 4°C for 30 minutes. Retinal pigment epithelium cells were also stained with FITC- or phycoerythrin (PE)-labeled anti-rabbit IgG or anti-mouse IgG at 4°C for 30 minutes. 

Expression of T-cell activation markers, such as CD154 (CD40 ligand), Th1-associated molecules (IFN-γ and T-bet), and TGFβ-associated molecules (Smad2/3 and TGFβ RII), in T cells exposed to iPS-RPE cells was assessed by flow cytometry. Cell surface staining was performed for the expression of CD154 and TGFβ RII, and intracellular staining was performed for the expression of IFN-γ, T-bet, and Smad2/3. Before cells were stained for IFN-γ, some cells were prestimulated with T-cell stimulation materials (CytoStim; Miltenyi Biotec) for 2 hours, and then cells were incubated for an additional 4 hours in the presence of a protein transport inhibitor (GolgiStop; BD PharMingen). After human Fc block staining, these T cells were stained with anti-human antibodies (listed above) at 4°C for 30 minutes or at room temperature for 30 minutes. T cells were also stained with FITC- or PE-labeled isotype control antibodies (mouse IgG or goat IgG) at 4°C for 30 minutes. 

To examine expression of CD25 on T cells exposed to iPS-RPE cells, PBMCs were added to culture wells in the presence of iPS-RPE cells. Peripheral blood mononuclear cells were harvested and stained with anti-human CD4 and anti-human CD25 antibodies at 4°C for 30 minutes. After CD4⁺CD25⁺ T cells were collected by a sorting system (FACSAria II; BD PharMingen), T cells were stained with antibodies against Foxp3, GITR (TNFRSF18), CD152 (CTLA-4), and CCR4 (CD194) and against cytokines (TGFβ1, IL-17, and IFN-γ) at room temperature or 4°C for 30 minutes. 

All samples were analyzed using FACSCanto II or FACSAria II flow cytometry (BD Biosciences, San Jose, CA, USA). Data were analyzed with FlowJo version 9.3.1 software (FlowJo, Ashland, OR, USA). All antibody information is described in Supplementary Table S1. 

Quantitative RT-PCR

Expression of mRNA levels for bestrophin (Best), RPE65, Pax6, TGFβ1, -2, and -3, and iPS cell markers (Lin28 and Nanog) in iPS-RPE cells were evaluated using quantitative RT-PCR (qRT-PCR). mRNA expression of IFN-γ, IL-10, TGFβ1, T-bet, Foxp3, Smad2, and Smad3 in T cells was also evaluated. Total RNA was isolated from human iPS-RPE cell lines (836B1, 454E2, 101G26, and TLHD1) or TGFβ1, -2, -3 small interfering RNA (siRNA)-transfected RPE cells. Total RNA was also isolated from CD4⁺ T cells from a healthy donor and from T cells exposed to iPS-RPE cells (836B1). After cDNA synthesis, the expression of various molecules and β-actin in triplicate samples was analyzed by qRT-PCR (LightCycler model 480; Roche Diagnostics), using qPCR Mastermix (Roche Diagnostics) and highly specific Universal ProbeLibrary assays (Roche Diagnostics). The tested primers and the Universal Probe are described in Supplementary Table S2. Quantitative RT-PCR was performed by denaturation at 95°C for 10 minutes, followed by 45 cycles of denaturation at 95°C for 10 seconds, annealing at 60°C for 30 seconds, and an extension at 72°C for 1 second. Relative mRNA expression was calculated with Relative Quantification software (Roche Diagnostics) by using an efficiency-corrected algorithm with standard curves and reference gene normalization against that of β-actin (delta delta cycle threshold [ΔΔC_t]). Results indicated the relative expression of the molecules (ΔΔC_t: control cells = 1). 

Transfection of siRNA

Small interfering RNA targeting human TGFβ1, -2, and -3 (Santa Cruz Biotechnology, Dallas, TX, USA) was transfected into iPS-RPE cells. On day 0, the cells were cultured in DMEM containing 5% FBS (antibiotic free). After overnight culture, RPE cells were transfected with TGFβ1, -2, and -3 siRNA reagent or control siRNA (Santa Cruz Biotechnology) at 37°C for 6 hours and then cultured in DMEM plus 10% FBS at 37°C for 24 hours. After incubation, the cells were harvested and examined for expression of TGFβ1, -β2, or -β3 mRNA by quantitative RT-PCR. 

Detection of TGFβ in iPS-RPE cells

Expression of TGFβ in iPS-RPE cells (454E2) was evaluated by immunohistochemistry. Cultured iPS-derived RPE or control RPE cell lines were fixed with 4% PFA-PBS for 15 minutes at room temperature, washed three times with PBS, and permeabilized with 0.3% Triton X-100-PBS. Anti-human TGFβ2 antibody was used as the primary antibody, and anti-rabbit IgG was used as the secondary antibody (Supplementary Table S1). Cell nuclei were counterstained with DAPI. The concentration of TGFβ2 in the supernatants of iPS-RPE cell lines (836B1, 454E2, 101G26, and TLHD1), control RPE cell lines (ES-RPE, ARPE-19), and primary fetal RPE cells (n = 2) and control cells (fibroblasts, iPS cells [836B1 and 454E2]) was measured as the active form by using TGFβ2 ELISA (R&D Systems). 

Multiplex Cytokine Array Assay

Supernatants of iPS-RPE cells (836B1, 454E2, 101G26, and TLHD1), control RPE cells (ARPE-19, primary fetal RPE, and ES-RPE), and control cells (fibroblasts and 454E2 iPS cells) were prepared for multiplex cytokine array assay (Procarta immunoassay kit; Filgen, Nagoya, Japan). The following 33 factors were measured: IL-1α, IL-1β, IL-1RA, IL-2, IL-4, IL-5, IL-6, IL-8/CXCL8, IL-9, IL-10/CSIF, IL-12 p40, IL-13, IL-17A/CTLA8, IL-21, IL-22/IL-TIF, IL-23 p19, TNF-α, TNFRI/TNFRSF1A, TNFRII/TNFRSF1B, IFNα2, IFNβ, IFNγ, MIG/CXCL9, IP-10/CXCL10, ITAC/CXCL11, MCP-1/CCL2, RANTES/CCL5, MIP-3α/CCL20, VEGF-A, MIF, TGFβ1, granzyme B, and fas ligand/TNFSF6. 

Recombinant Proteins

The following recombinant human proteins were used for T-cell experiments in vitro: IL-2 (BD Biosciences), IL-21 (eBioscience, San Diego, CA, USA), and IL-12, IL-1RA, IL-6, IL-8/CXCL, IL-22, MIF, thrombospondin-1 (TSP-1,), TNF-α, soluble TNFR1, MIG/CXCL9, MCP-1/CCL2, MIP3 α/CCL20, TGFβ1, and TGFβ2 (all from R&D Systems). 

Statistical Evaluation

All experiments were repeated at least twice, and data are means ± SEM. Student's paired or unpaired t-test was used for all statistical analyses (GraphPad Prism; GraphPad Software, San Diego, CA, USA). Values were considered statistically significant if P was less than 0.05. 

Results

Induction of iPS-Derived RPE Cells

To generate human iPS cells, fibroblasts derived from abdominal skin or dental pulp cells of healthy donors and from a donor with ocular disease were transfected with a plasmid carrying several genes by electroporation (Fig. 1A).²⁵ PCR analysis indicated no integration of episomal vectors in the tested iPS cell lines, whereas all lines were positive for endo-Oct3/4 (Fig. 1B). To examine the expression of ES cell marker genes in the iPS cells, immunohistological analysis was performed. These established iPS cells expressed Oct3/4, TRA-1-60, and SSEA-4. No staining was observed with isotype control mouse IgG (Fig. 1C). The cells were also positive for expression of Nanog (data not shown). The combined immunohistological staining data indicated that these iPS cells expressed ES cell markers. Established iPS cells were then cultured with RPE induction medium. Retinal pigment epithelium cells induced from iPS cells clearly showed polygonal morphology (mostly hexagonal) and contained melanin (Fig. 1D), even after the formation of RPE cell sheets. The RPE cells had the ability to phagocytose shed photoreceptor ROS (Fig. 1E) and expressed MiTF and ZO-1, which are constitutively expressed by primary RPE (Fig. 1F). Quantitative RT-PCR analysis also indicated that they expressed RPE-related genes (BEST, RPE65, Pax6, and TGFβ1, -2, and -3) but not iPS-specific genes (Lin28, Nanog) (Fig. 1G). In order to compare the four iPS cells and the RPE lines, we summarized the results of their quality control tests (Supplementary Tables S3 and S4, respectively). Thus, the established RPE cells derived from iPS cells had many characteristics of mature RPE in vivo but had no characteristics of pluripotent stem cells. 

Capacity of Human iPS-RPE Cells to Suppress T-Cell Activation

Cultured primary RPE cells can inhibit cell proliferation and IFN-γ production by T cells when the target T cells are stimulated with anti-CD3.^8,24 Accordingly, we next tested whether established iPS-RPE cells could also suppress T-cell activation (proliferation and cytokine production). CD4⁺ T cells were stained with CFSE and stimulated with anti-CD3/CD28 monoclonal antibody in the presence of iPS-RPE cells. The T cells were harvested at 72 hours, and flow cytometry was used to evaluate the extent of progressive cell division. Up to a few rounds of T-cell division were evident in positive control cultures without the RPE cells (Fig. 2A). When iPS-RPE cells, as well as human RPE cell lines (ARPE-19), were present, the T cells underwent no division, whereas T-cell suppression by fibroblasts (control) was poor. 

Figure 2

 

Capacity of cultured iPS-RPE cells to suppress activation of bystander T cells in vitro. (A) CD4⁺ T cells were cocultured with iPS-RPE cells (836B1 or 101G26) in the presence of anti-human CD3 abs and rIL-2 for 72 hours. As controls, human RPE cell lines (ARPE-19) and human fibroblasts were also prepared. After 72 hours, T cells were harvested for flow cytometric analysis. Values in the FACS histograms indicate CFSE-positive cell data representative of four experiments. (B) CD4⁺ T cells were cocultured with iPS-RPE cells (836B1 [black bars]) or without RPE cells (positive control [white bars]) in the presence of rIL-2 and PHA-P (left) or of rIL-2, PHA-P, anti-human CD3 abs, and anti-human CD28 abs (right) for 72 hours. Data are means ± SEM of three ELISA determinations (BrdU). *P < 0.05, **P < 0.005, as compared to the positive controls. Data are representative of three experiments with other iPS-RPE cells. (C) Histograms present expression of T-cell activation markers on CD4⁺ T cells stimulated by anti-CD3/CD28 antibodies and rIL-2 in the presence of iPS-RPE (836B1). Cells were stained with anti-CD4 & anti-CD154 and analyzed by flow cytometry. Values in the histogram indicate cells double-positive for CD4/CD154. Data are representative of three experiments. (D) Expression of inflammatory-related cytokines and genes in T cells exposed to iPS-RPE cells. Purified CD4⁺ T cells were cocultured with iPS-RPE cells (836B1 [black bars]) or without RPE cells (positive control [white bars]) for 48 hours and were then examined for expression of IFN-γ, IL-10, TGFβ1, T-bet, Foxp3, Smad2, and Smad3 mRNA by qRT-PCR. Prior to PCR, we purified the CD4⁺ T cells again after 48 hours culture with iPS-RPE cells. Results indicate the relative expression of these molecules (ΔΔC_t: T cell alone = 1). Data are representative of two experiments with other iPS-RPE cells. (E) Histograms represent the expression of Th1-related factors (IFN-γ and T-bet) on CD4⁺ T cells in the presence of iPS-RPE cells (836B1). T cells were stained with anti-CD4 and anti-IFN-γ and with anti-CD4 and T-bet. Numbers in the histograms indicate the percentage of cells double-positive for CD4/IFN-γ or T-bet. Data are representative of three experiments. (F) CD4⁺ T cells were cocultured with iPS-RPE cells (836B1) in the presence of rIL-2 only for 96 hours. For induction of Th1 cells, CD4⁺ T cells were cocultured with iPS-RPE cells (black bars) or without RPE cells (white bars) in the presence of rIL-2, rIL-12, rTNF-α, anti-IL-4 antibody, and anti-CD3/CD28 antibodies for 96 hours. Interferon-γ production by T cells was then assayed. Data are means ± SEM of three ELISA determinations. ***P < 0.0005, in comparison to the positive control. Data are representative of three experiments.

When BrdU-labeled T cells were stimulated with recombinant IL-2 and PHA-P with or without anti-CD3/CD28, their proliferation was significantly suppressed by exposure to iPS-RPE cells (Fig. 2B). T-cell activation was not suppressed in the presence of 5 × 10³ to 5 × 10⁴ iPS-RPE cells, whereas T-cell activation was significantly suppressed in the presence of 1 × 10⁵ to more than 5 × 10⁵ iPS-RPE cells, as well as control RPE cell lines (ARPE-19 [Supplementary Fig. S1A]). The extent of suppression of activated CD4⁺ T cells by autologous iPS-RPE cells was much greater than that by allogeneic iPS-RPE cells (Supplementary Fig. S1B). We also examined the expression of T-cell activation markers after exposure to iPS-RPE cells. T cells exposed to iPS-RPE cells expressed less CD154 on their surface than T cells not exposed to iPS-RPE cells (Fig. 2C). In addition, iPS-RPE cells also suppressed mixed lymphocyte reactions (MLRs) if PBMCs from two healthy donors were cocultured with iPS-RPE cells (Supplementary Fig. S1C). Thus, iPS-RPE cells fully suppressed activation of bystander T cells in vitro. 

We next examined the extent to which CD4⁺ T cells stimulated by rIL-2 and anti-CD3 in the presence of iPS-RPE cells produced IFN-γ (an effector T-cell-associated cytokine/Th1 cytokine), IL-10 and TGFβ1 (immunosuppressive cytokines), T-bet (Tbx21, a Th1-associated gene), Foxp3 (a regulatory T-cell gene), and Smad2/3 (TGFβ-associated genes). Based on qRT-PCR data, T cells exposed to iPS-RPE cells expressed TGFβ1, Foxp3, and Smad2/3 significantly compared to T cells not exposed to RPE cells, whereas T cells exposed to iPS-RPE cells poorly expressed Th1-associated molecules (Fig. 2D). In FACS analysis, the expression of IFN-γ and T-bet was profoundly reduced when CD4⁺ T cells were cocultured with iPS-RPE cells (Fig. 2E). In addition, iPS-RPE cells significantly suppressed production of IFN-γ by Th1 cells (Fig. 2F), indicating that iPS-RPE cells suppress the production of cytokines, including the Th1 cytokine IFN-γ. 

Ability of iPS-RPE Cells to Induce T-Regulatory Cells (Tregs)

Primary cultured RPE cells can convert T cells into T regulatory cells (Tregs), which have immunoregulatory properties in vivo and in vitro.^7,9,10,12 We next assessed the expression of CD25, Foxp3, TGFβ1, IL-17, IFN-γ , GITR (TNFRSF18), CD152 (CTLA-4), and CCR4 (CD194) in iPS-RPE cell-induced Tregs. We first checked the population of CD4⁺CD25⁺ T cells (including Tregs) in RPE-exposed peripheral blood cells because we wanted to know the population of CD4⁺CD25⁺ T cells in RPE-exposed PBMC compared with that of control PBMC without RPE cells. For this assay, PBMCs from a healthy donor were cultured with or without iPS-RPE cells. The population of CD4⁺CD25⁺ T cells in the PBMC exposed to iPS-RPE cells (21.5%) was higher than that of control cells (18.3%) (Fig. 3A). We then collected the CD4⁺CD25⁺ iPS-RPE cell–induced Tregs and the control CD4⁺CD25⁺T cells by flow cytometric sorting (Fig. 3A). CD4⁺CD25⁺ iPS-RPE cell-induced Tregs expressed high levels of Foxp3, TGFβ1, GITR, CD152, and CCR4 but not inflammatory cytokines such as IFN-γ compared to CD4⁺CD25⁺ control T cells (Fig. 3B). In addition, the CD4⁺CD25⁺ iPS-RPE-induced Tregs fully suppressed activated T cells in vitro, but CD25 negative T cells did not (Fig. 3C). Although conventional human RPE cell lines do not induce Tregs in vitro,²³ our iPS-derived RPE cells were able to induce Tregs. 

Figure 3

 

Phenotype of Tregs induced by iPS-RPE cells. (A) Separation of CD4⁺CD25⁺ iPS-RPE cell-induced Tregs. CD4⁺CD25⁺ T cells were collected from PBMCs with (lower histogram) or without iPS-RPE cells (836B1 [upper histogram]) by flow cytometric sorting. Data are representative of three experiments. (B) CD4⁺CD25⁺ T cells as a control and CD4⁺CD25⁺ T cells that were exposed to iPS-RPE cells (right) were stained with anti-human Foxp3, TGFβ1, CD152 (CTLA-4), GITR (TNFRSF18), CCR4 (CD194), IL-17, and IFN-γ abs. Percentages of double-positive cells are shown (CD4 [above the molecule]). Data are representative of two experiments. (C) Target CD4⁺ T cells in the presence of anti-CD3 abs were labeled with CFSE and were cocultured with iPS-RPE-induced CD4⁺CD25⁺ Tregs for 72 hours. As controls, CD4⁺CD25-negative T cells that were cocultured with iPS-RPE cells were also prepared. After 72 hours, T cells were harvested for flow cytometric analysis. Values in the FACS histograms indicate CFSE-positive cells. Data are representative of two experiments.

Capacity of iPS-RPE Cells to Suppress T-cell Activation by Soluble Factors

To clarify the role of cell-to-cell contact in the suppression of T-cell activation by iPS-RPE cells, these cells were first cultured separately in 24-well plates. Transwell cell inserts were then placed in these wells, and each transwell contained CD4⁺ T cells plus anti-human CD3/CD28. In RPE-T cell culture, CD4⁺ T proliferation was significantly suppressed in the absence of transwell cell inserts (Fig. 4A). On the other hand, iPS-RPE cells still suppressed T-cell proliferation across the transwell membrane. These findings imply that soluble inhibitory factor(s) were essential for T-cell suppression by iPS-RPE cells. Thus, the RPE cells secrete soluble immunosuppressive factor(s). 

Figure 4

 

Ability of iPS-RPE cells to suppress T cells via soluble inhibitory factors. (A) Induced pluripotent stem-RPE cells (454E2) or ARPE-19 cells were first cultured in 24-well plates. Transwell cell inserts were placed in these wells, and each transwell contained CD4⁺ T cells plus anti-human CD3/CD28 to block cell-to-cell contact between RPE cells and T cells. Values in the FACS histograms indicate CFSE-positive cells, and values in parentheses indicate levels of IFN-γ production in the culture supernatants (ng/mL). Data are representative of three experiments with other iPS-RPE cells. (B) Supernatants of iPS-RPE cells (836B1, 454E2, 101G26, and TLHD1), control RPE cells (ARPE-19, primary fetal RPE, and ES-RPE), and control cells (fibroblasts and 454E2 iPS cells) were prepared for a multiplex cytokine array assay. The significant concentration of each cytokine and chemokine is >10 pg/mL, and the undetectable level is <1.0 pg/mL. Data are representative of two individual experiments. (C) Histograms represent the expression of cytokines, chemokines, cytokine antagonists, cytokine receptors, and growth factors in iPS-RPE cells (454E2). Cells were analyzed by flow cytometry. Blue histograms represent isotype control staining (mouse or rat IgG). Data are representative of two experiments with other RPE cells.

Based on the findings that iPS-RPE cells exclusively suppress T-cell activation (e.g., production of IFN-γ) through immunosuppressive factor(s), we hypothesized that intracellularly expressed molecules in RPE cells were responsible for this suppression. To detect candidate molecules, iPS-RPE cells were subjected to multiplex cytokine array assay. For this assay, we prepared several supernatants from iPS-RPE cells and control cells (Fig. 4B). Of 33 cytokines and chemokines assayed, 4 iPS-RPE cell lines (836B1, 454E2, 101G26, and TLHD1) significantly produced 13 factors such as IL-1RA, IL-6, IL-8, IL-9, IL-21, IL-22, TNFRI, MIG, MCP-1, MIP-3 α, VEGF-A, MIF, and active TGFβ1 (Fig. 4B). Control RPE cells (ARPE-19, fetal RPE, and ES-RPE) and fibroblasts (which did not produce IL-1RA) gave similar results, whereas supernatants of iPS cells did not contain any cytokines or chemokines except for RANTES and MIF (Fig. 4B). 

To confirm the results of the above array assay, iPS-RPE cells were subjected to flow cytometry after staining with specific monoclonal antibodies against the above-listed factors and against the following candidate molecules: TNF-α, IFN-γ, TGFβ2, and TSP. The FACS results indicated that iPS-RPE cells expressed cytokines (IL-6, IL-21, TNF-α, and MIF), chemokines (IL-8, MIG, MCP-1, and MIP-3 α), a cytokine antagonist (IL-1RA), a cytokine receptor (TNFRI), and TGFβ family members (TGFβ1, TGFβ2, TSP) (Fig. 4C). Induced pluripotent stem-RPE cells did not express IL-9, IL-22, or IFN-γ. Thus, these molecules appeared to be candidates for a role in the suppression of T-cell activation by iPS-RPE cells. 

Next, we tested these candidate molecules in the form of human recombinant proteins to determine their T-cell suppression ability. Recombinant human TGFβ1 and TGFβ2 profoundly suppressed T-cell activation (i.e., IFN-γ production) in a dose-dependent manner (Supplementary Fig. S2). Recombinant IL-1RA and TSP-1 (50 ng/mL) also suppressed IFN-γ production by T cells. Other recombinant proteins such as IL-6, IL-8, IL-21, IL-22, TNF-α, TNFRI, MIG, MCP-1, MIP-3 α, and MIF did not suppress T-cell activation, suggesting that TGFβ is important for RPE cell–mediated suppression. 

Detection of TGFβ in iPS-Derived RPE Cells

Cultured human RPE cells constitutively express TGFβ.^4,7,8 We therefore confirmed expression of TGFβ isoform 2 in the established iPS-RPE cells. Transforming growth factor β2 was assayed because TGFβ2 is the dominant TGFβ expressed and is important for immune tolerance in the eye.^1,3 Several iPS-RPE cell lines constitutively expressed TGFβ2 by ELISA (Fig. 5A), immune staining (Fig. 5B), and flow cytometry (Fig. 5C). Supernatants from iPS-RPE cells but not from human iPS cells or fibroblasts contained significant levels of TGFβ2 (Fig. 5A). 

Figure 5

 

Detection of TGFβ in iPS-RPE cells. (A) Supernatants of iPS-RPE cells (836B1, 454E2, 101G26, and TLHD1) were collected to measure active TGFβ2 (pg/mL). As controls, the supernatants of ES-RPE and ARPE-19, primary fetal RPE (n = 2), iPS cells (836B1 and 454E2, n = 2), and fibroblasts were also collected. Data are the mean ± SEM of three ELISA determinations. ND, not detected. Data are representative of three experiments. (B) Detection of TGFβ2 in iPS-RPE cells by immunostaining. Purified cells derived from iPS-RPE sheets (454E2), as well as the RPE cell line ARPE-19, expressed TGFβ2. Cell nuclei were counterstained with DAPI. Scale bars: 100 μm. Data are representative of three experiments. (C) iPS-RPE cells (TLHD1) were harvested separately at different culture stages (p1, day 14; p2, day 30; p3, day 60; p4, day 90; p5, day 120) and then stained with anti-human TGFβ2 abs. Blue histograms represent isotype control staining. Data are representative of two experiments. (D) Detection of Smad2 and -3 molecules in T cells exposed to iPS-RPE supernatants (sup) by immunostaining. T cells were stained with anti-Smad2/3 abs. Cell nuclei were counterstained with DAPI. Scale bars: 50 μm. Data are representative of two experiments. (E) Induced pluripotent stem-RPE–exposed CD4⁺ T cells were stained with anti-Smad2/3 or TGFβ receptor II (RII) abs. Values in the histograms indicate the percentage of cells double-positive for CD4/Smad2/3 or TGFβ RII. Data are representative of three experiments.

We next examined whether target T cells expressed the receptors and the related transcription factors for TGFβ when exposed to iPS-derived RPE cells. Intracellular molecules belonging to the Smad family are responsible for TGFβ signals. By immune staining, both T cells and T cells exposed to iPS-RPE cells expressed intracellular Smad2/3 molecules (Fig. 5D). Staining using isotype control antibodies (goat IgG) was poor (data not shown). In addition, CD4⁺ T cells exposed to iPS-RPE cells, as well as T cells not exposed to RPE cells, expressed TGFβ receptors (RII) and Smad2/3 by flow cytometry (Fig. 5E), indicating that Smad2/3 molecules increase in response to iPS-RPE cells. 

Capacity of T-Cell Suppression by TGFβ-siRNA–Transfected iPS-RPE Cells

To determine whether TGFβ produced by iPS-RPE cells could suppress bystander T cells, we examined the effect of down-regulating the mRNA expression of these molecules by using siRNA. As revealed in Fig. 6A, TGFβ siRNA-transfected iPS-RPE cells poorly expressed mRNA for TGFβ1, -2, and -3, whereas control siRNA-transfected iPS-RPE cells were able to express this mRNA. Compared with control siRNA-transfected cells, supernatants of TGFβ siRNA-transfected iPS-RPE cells contained lower levels of TGFβ2 proteins (Fig. 6B). 

Figure 6

 

Effect of TGFβ produced by iPS-RPE cells on the suppression of T-cell activation. (A) TGFβ1, -2, and -3 siRNA-transfected iPS-RPE (454E2) or control RPE cells (ARPE-19) were harvested on day 3 and examined for expression of TGFβ1, -2, and -3 mRNA by qRT-PCR. As a control, control siRNA-transfected cells were also analyzed. Results indicate the relative expression of these molecules (ΔΔC_t). Data are representative of four experiments. (B) Supernatants of TGFβ1, -2, and -3 siRNA-transfected iPS-RPE were collected to measure active TGFβ2 (pg/mL). Data are means ± SEM of three ELISA determinations. *P < 0.05, compared to positive controls (white bar). n.s., not significant. Data are representative of two experiments. (C) CD4⁺ T cells were cocultured with TGFβ1, -2, and -3 siRNA-transfected iPS-RPE (or control siRNA-transfected cells) and evaluated by BrdU incorporation (left) or IFN-γ production (right) by the T cells. Data are means ± SEM of 3 ELISA determinations. *P < 0.05, **P < 0.005 compared to two groups. n.s., not significant. Data are representative of four individual experiments.

Subsequently, we examined whether TGFβ siRNA-transfected iPS-RPE cells were able to suppress bystander T cells. The results indicated that the siRNA-transfected iPS-RPE cells failed to suppress activation of CD4⁺ T cells (i.e., T-cell proliferation and IFN-γ production) (Fig. 6C). 

Discussion

In the present study, we showed that cultured iPS-RPE cells significantly inhibited cell proliferation and IFN-γ production by T cells when the target T cells were stimulated with anti-human CD3/CD28 antibodies, PHA-P, and recombinant IL-2. The iPS-RPE cells constitutively expressed and secreted TGFβ, and TGFβ siRNA-transfected iPS-RPE cells did not inhibit T-cell activation. Thus, cultured human iPS-derived RPE cells fully suppress T-cell activation in vitro. According to a previous report, transplantation of human fetal RPE cells into the monkey eye induced immune tolerance.²⁶ Transplantation of iPS-RPE cells into the eye may be a promising therapy for ocular diseases because these cells might reduce T-cell mediated immune rejection. 

In immune-privileged sites such as the eye, TGFβ expression on the ocular resident tissues/cells contributes to immune tolerance and ultimately prevents blindness. The iPS-RPE cells we established can induce T-cell inactivation, and the conversion of Treg cells to bystander T cells in vitro, suggestive of the expression of negative signal-inducing molecule(s) by RPE cells. Based on the present and previous findings that RPE cells are able to suppress T-cell activation through immunoregulatory signals, we hypothesized that intracellular molecules expressed by RPE cells were involved in the T-cell suppression. As shown in the present study, iPS-RPE cells produce high levels of TGFβ, a powerful mediator of inflammatory cell suppression. Retinal pigment epithelium cells that express less TGFβ due to siRNA transfection do not suppress T-cell activation, and human recombinant TGFβ proteins fully suppress the T cells. Thus, TGFβ is a critical mediator in the mechanisms of suppression. The level of the active isoform of TGFβ2 in ocular fluids is within the low picogram-per-milliliter range under normal circumstances.²⁷ The iPS-RPE cell lines we established produced approximately 400 to 3000 pg/mL active TGFβ2 in the culture supernatants (Fig. 5A). All iPS-RPE cells significantly produced TGFβ2, but the levels produced differed between the iPS-RPE cell lines. It is unclear why the production of TGFβ2 in iPS-RPE cells varied. However, these data are consistent with the data of primary RPE cells (human fetal RPE cells) whose production of TGFβ2 also varied between the lines (Fig. 5A). Moreover, these levels of human recombinant TGFβ2 suppressed the activation of CD4⁺ T cells (Supplementary Fig. S2). Based on these results, we suspect that iPS-RPE cells can suppress T-cell activation in the posterior segment of the eye by delivering active soluble TGFβ to migrating T cells that are targeted via TGFβ–TGFβ receptor interactions. The TGFβ produced by RPE cells is an important factor in the pathogenesis of ocular inflammation after transplantation. In addition, T-cell proliferation and IFN-γ production were partially restored in the presence of TGFβ-siRNA transfected iPS-RPE cells (Fig. 6C). It is assumed that other inhibitory soluble factors are involved in the suppression of bystander T cells. Indeed, RPE cells (including iPS-derived RPE cells) can secrete several soluble factors, as shown in Figure 4B. 

We currently have no definitive answers regarding the possibility of immune attack of iPS-RPE cells after transplantation. However, it is assumed that there would be no risk (or less risk) of immune rejection of iPS-RPE cells for several reasons. First, the transplantation site is the subretinal space, which is an immune-privileged site.^1–3 Immunogenic inflammation within the retina after transplantation into the subretinal space is suppressed, that is, immune privilege is present, which leads to graft survival. Retinal pigment epithelium cells are important for creating an immune-privileged site. Second, RPE cells can inhibit T-cell activation of both CD4 and CD8⁺ T cells. These T cells play a critical role in the pathogenesis of ocular inflammation and immune rejection after transplantation, and as shown in the present study, iPS-RPE cells strongly suppress activated CD4⁺ T cells. In addition, iPS-RPE cells significantly suppressed the activation of CD8⁺ CTLs, B cells, DCs and monocytes (unpublished observation). Third, the expression of immunoregulatory molecules such as TGFβ by RPE cells might reduce the risk of immune rejection. To the best of our knowledge, MHC-class II molecules (e.g., HLA-DR antigens) are not expressed on RPE cells,⁵ and there is little or no expression of positive costimulatory molecules (e.g., CD40, CD80, and CD86) by RPE cells under normal conditions.^5,9 On the other hand, RPE cells do express negative costimulatory molecules such as B7-H1 (PD-L1),⁵ suggesting that T cells infiltrating the graft site after transplantation might interact with these molecules and be inactivated. Fourth, cultured RPE cells can convert T cells into Tregs that suppress bystander immune cells.^7,9,10,12 Although conventional human RPE cell lines do not convert T cells into Tregs,²³ our iPS-RPE cells easily induced this conversion. Thus, some of the T cells exposed to iPS-RPE cells may be converted into immunosuppressive cells instead of losing effector function. For instance, CD4⁺ T cells exposed to iPS-RPE cells do not produce IFN-γ inflammatory cytokines (Fig. 3). Fifth, eye-specific systemic immune suppression, that is, anterior chamber-associated immune deviation (ACAID), can be induced even after subretinal transplantation.^28,29 The grafts in the subretinal space displayed no evidence of immune rejection. In contrast, RPE cells implanted in the subconjunctival space (outside of the eye) of mice elicited an intense RPE-specific delayed hypersensitivity associated with cellular infiltration of the graft. At the moment we have no evidence for ACAID with the use of iPS-RPE, but it is assumed that eye-specific systemic immune suppression can occur if the immune cells in the recipient recognize allogeneic RPE cells after transplantation. Over time, the risk of immune rejection is reduced because of systemic immune suppression. Sixth, the iPS cells themselves will not be used for transplantation, instead, terminally differentiated RPE cells derived from iPS cells will be used. Recently, immune attacks have been reported to occur after iPS cell transplantation, but not after ES cell transplantation.³⁰ Zhao et al.³⁰ mentioned in their study that ES cells derived from blastocyst embryos from a given genetic background grow into teratomas when transplanted into mice of the same genetic background. The immune system is therefore tolerant of autologous ES cells. On the other hand, transplantation of autologous iPS cells derived from fetal fibroblasts into matched mice resulted in the rejection of teratomas, because of the expression of minor antigens such as Zg16 and Hormad1.³⁰ However, more recently, Araki et al.³¹ reported that there were no differences in rates of transplantation success when skin and bone marrow cells derived from mouse iPS cells were compared with cells derived from ES cells. This group also observed limited immune responses to tissues derived from either iPS or ES cells. In addition, there was no increase in the expression of immunogenicity-related genes in regressing tissues. A careful investigation of the immunogenicity of iPS cell-derived tissue including retinal cells is critical because partial reprogramming and genetic instabilities in iPS cells could elicit immune responses in transplant recipients even when iPS-derived differentiated cells are transplanted. 

In conclusion, the iPS-RPE cells we established are high-quality RPE cells. Retinal pigment epithelium cells induced from iPS cells clearly showed polygonal morphology and contained melanin. Moreover, RPE cells derived from iPS cells had many characteristics of mature RPE cells in vivo but no characteristics of pluripotent stem cells. We plan to use RPE cell sheets to treat patients with age-related macular degeneration. The iPS-RPE cells significantly inhibited cell proliferation and inflammatory cytokine production by T cells when the target T cells were stimulated in vitro. We are now conducting experiments to determine whether allogeneic T cells can recognize iPS-RPE cells from HLA-3 locus homozygote donors.²⁵ The iPS bank³² might be useful as a source of allografts in retinal disorders, if the recipient T cells cannot respond to allogeneic RPE cells because of a match to some of the main HLA antigens. 

Acknowledgments

The authors thank Wataru Ohashi, Chikako Yamada, and Chikako Morinaga (Laboratory for Retinal Regeneration, Center for Developmental Biology, RIKEN) for expert technical assistance. 

Supported by Scientific Research Grant B, 25293357 from the Ministry of Education, Culture, Sports, Science and Technology of Japan; a grant from the Project for Realization of Regenerative Medicine from the MEXT; and the Charitable Trust Fund for Ophthalmic Research in Commemoration of Santen Pharmaceutical's Founder. 

Disclosure: S. Sugita, None; H. Kamao, None; Y. Iwasaki, None; S. Okamoto, None; T. Hashiguchi, None; K. Iseki, None; N. Hayashi, None; M. Mandai, None; M. Takahashi, None