Generation of Human Alloantigen-Specific Regulatory T Cells under Good Manufacturing Practice-Compliant Conditions for Cell Therapy

Human Samples and Ethical Approval

Leukapheresis samples were freshly collected from 37 healthy donors (13 females and 24 males, median age 39.5 years; range: 22-65) at Pitié Salpêtrière hospital (Etablissement Français du Sang, Paris, France). Approval for these studies was obtained from the institutional review boards of the Université Pierre et Marie Curie and the Centre National de la Recherche Scientifique. Written informed consent was provided according to institutional guidelines and to the Declaration of Helsinki.

Generation of DCs

Monocytes were enriched from the adherent fraction of mononuclear cells (MNCs) (32) or purified using CD14 immunomagnetic beads (Miltenyi Biotec, Paris, France) as previously described (14). DCs were generated from monocytes using good manufacturing practice (GMP)-grade cytokines and reagents and an already approved procedure for DC-based clinical trials (46). Briefly, monocytes were cultured for 8 days in RPMI medium (Invitrogen Corporation, Paris, France) containing 10% human AB serum (hSAB purchased from Etablissement Français du Sang, Paris, France), 2 mM L-glutamine, 100 U/ml penicillin, and 100 ug/ml streptomycin (all from Invitrogen), 1,000 IU/ml granulocyte macrophage colony-stimulating factor (Gentaur, Kampenhout, Belgium), and 200 IU/ml IL-4 (Miltenyi Biotec) in order to generate immature DCs (iDCs) as previously described (14). Mature DCs (mDCs) were obtained using a slightly modified method described by Feuerstein et al. (16). Briefly, tumor necrosis factor (TNF)-α (10 ng/ml; Peprotech, Neuilly, France), JL-1β (100 UI/ml; Invitrogen), IL-6 (1,000 UI/ml; Peprotech), and prostaglandin E2 (PGE-2; 1 μg/ml, Prostin® E2; Pfizer, Paris, France) were added to the culture at day 6. At day 8, iDCs and mDCs were harvested, characterized, and frozen until use. The maturation status of mDCs, compared to iDCs, was assessed by flow cytometry on the expression of human leukocyte antigen (HLA)-DR, CD40, CD80, CD83, and CD86 (data not shown) as previously described (14).

Selection of Tregs

Tregs were enriched from MNCs either by FACS (FACS Aria®; BD Biosciences, Pont de Claix, France) or by magnetic-activated cell sorting (MACS; Miltenyi Biotec). For FACS purification, CD25⁺ cells were selected by MACS using anti-CD25 microbeads (Miltenyi Biotec), then stained with CD4, CD25, CD127, and CD8 directly labeled monoclonal antibodies (mAbs) according to the manufacturer's instructions (see Table 1) and sorted into CD4⁺CD25^highCD127^low/-CD8^- cells. Thus, highly purified Tregs (>99%) were obtained. For MACS purification, Tregs cells were obtained by using anti-CD25 microbeads (Miltenyi Biotec), and three consecutive magnetic selections were applied to the cells to reproduce the 3.1 enrichment-selective program performed at large scale with the CliniMACS device (Miltenyi Biotec) as described elsewhere (20).

OPEN IN VIEWER

Table 1

List of Monoclonal Antibodies

mAbs and Fluorochromes	Clone's Names	Manufacturer's Names
CD3-V500	UCHT1	Becton Dickinson (BD), Pont de Claix, France
CD4-PerCP	VIT4	Miltenyi Biotec, Paris, France
CD8-FITC	B9.11	Beckman Coulter, Villepinte, France
CD8-APC	B9.11	Beckman Coulter
CD8-FITC	BW135/80	Miltenyi Biotec
CD8-vioBlue	BW135/80	Miltenyi Biotec
CD14-PE	M5E2	Beckman Coulter
CD19-FITC	LT19	Miltenyi Biotec
CD25-PE	4E3	Miltenyi Biotec
CD25-Pe-Cy7	M-A251	BD
CD39-PE	MZ18-23C8	Miltenyi Biotec
CD40-PE	MAB89	Beckman Coulter
CD56- Pe-Cy7	N901(NKH-1)	Beckman Coulter
CD80-PE	BB1	Beckman Coulter
CD83-PE	HB15	Beckman Coulter
CD86-FITC	FUN-1	Beckman Coulter
CD127-FITC	eBioRDR5	eBioscience, San Diego, CA, USA
CD127-Brilliant-Violet 421™	A019D5	BioLegend, San Diego, CA, USA
CD152-Brilliant-Violet 421™	BNI3	BD
CD278-PE	DX29	BD
HLA-DR-FITC	G46.6	Beckman Coulter
FOXP3-APC	PCH101	eBioscience
GARP-PE	7B11	BioLegend
Helios-FITC	22F6	eBioscience
LAP-PE	Clone 27232	R&D systems

CD3, cluster of differentiation 3; PerCP, peridinin–chorophyll–protein complex; FITC, fluorescein isothiocyanate; APC, allophycocyanin; PE, phycoerythrin; HLA, human leukocyate antigen; FOXP3, forkhead box P3; GARP, glycoprotein A repetitions predominant (garpin); LAP, latency-associated peptide.

Cell Division Kinetics of Tregs

FACS-sorted Tregs (0.5-5 × 10⁶ cells) were labeled with 1 μM CFSE (Invitrogen). The CFSE-labeled Tregs were cocultured with either allogeneic mDCs or iDCs at a 10:1 ratio. Cocultures were performed in complete culture medium that corresponds to Roswell Park Memorial Institute (RPMI)-1640 medium containing 10% hSAB, 2 mM L-glutamine, 100 U/ml penicillin, and 100 ug/ml streptomycin (Invitrogen). Tregs were stimulated by IL-2 (600 UI/ml, Proleukin®; Novartis, Rueil-Malmaison, France), or IL-15 (10 ng/ml; Miltenyi Biotec), or by both. As control, CFSE-labeled Tregs were also polyclonally stimulated in the presence of anti-CD3/anti-CD28-coated Dynabeads (Invitrogen) and IL-2 (see below). Cells were daily collected from day 3 to day 6, and cell division-based CFSE dilution was monitored by flow cytometry.

Culture of Poly-Tregs and Allo-Tregs

Poly-Tregs were cultured as described elsewhere (20). Briefly, MACS-sorted Tregs seeded at 5 × 10⁵ cells/ml in complete culture medium containing 100 μxM rapamycin (Sigma-Aldrich, Saint Quentin Fallavier, France) and 600 UI/ml IL-2 were stimulated by anti-CD3/anti-CD28-coated Dynabeads (Invitrogen) at day 0 and day 15 of culture, cultures being maintained until day 21.

Spe-Tregs were prepared as follows. MACS-sorted Tregs were seeded at 10⁶ cells/ml in complete culture medium containing 100 μM rapamycin, 600 UI/ml IL-2, 10 ng/ml IL-15, and cultured with 10⁵ allogeneic mDCs. Ten days later, medium was changed, and Tregs were restimulated with the same source of allogeneic mDCs at the same ratio and expanded under the same culture conditions until day 21. In some experiments, CD4⁺CD25^-CD127⁺ Teff cells were purified by FACS from the CD25^- cell fraction and used as controls. Teffs were cultured using either the same antigen-specific stimulation as for spe-Tregs or under polyclonal activation in order to generate spe-Teffs and poly-Teffs, respectively. For all culture conditions, medium changes were performed every 3 days. At the end of culture, CD3/CD28 Dynabeads were removed with a magnet (Invitrogen); immunophenotypic and functional characterizations of Tregs were then performed.

Antibodies and Flow Cytometry Analysis

The list of murine anti-human mAbs directly conjugated to fluorescein isothiocyanate, phycoerythrin (PE), PE-cyanine 7 (PE-Cy7), allophycocyanin, and peridinin–chorophyll–protein complex is described in Table 1, and the mAbs were used according to the manufacturer's instructions. Cell surface staining and intracellular staining (FOXP3 and Helios) were performed according to the manufacturer's instructions. Matched mouse isotype mAbs were used as negative controls. Flow cytometry analyses were performed on either a FACSCalibur® or a LSR II® (both from BD Biosciences). Data were analyzed with FlowJo® software (Treestar, Ashland, OR, USA).

Suppression T-Cell Assay

The suppressive activity of Tregs was assayed in round-bottomed 96-well tissue culture plates (Corning, Avon, France) before and after culture. After culture, CD3/CD28 Dynabeads were removed, cells were washed twice in phosphate-buffered saline, and rested in RPMI-1640 medium containing 10% hSAB and IL-2 (600 UI/ml) for 24 h before performing the suppression assay. Then Tregs were mixed at various cell ratios (1:1; 1:2; 1:4; 1:8; 1:16, 1:32, 1:64, 1:128) with 5 × 10⁴ autologous CD4⁺CD25^- cells (responder cells) in the presence of 10⁴ stimulator cells corresponding to either the original DCs used to expand spe-Tregs (DCs referred to as from donor a, DCa) or another batch of allogeneic DCs, used as a third party (DCs referred to as from donor b, DCb) in RPMI medium containing 10% hSAB. Autologous CD4⁺CD25^- cells were obtained by depleting CD25⁺ cells from MNCs using 10 μl of CD25 microbeads per 10⁷ cells and followed by CD4⁺ cell selection using the untouched CD4⁺ T-cell Isolation kit II (Miltenyi Biotec). CD4⁺CD25^- cells were used immediately or frozen in liquid nitrogen until use. After 4 days, cell proliferation was determined by incorporation of 2 μCi (0.037 MBq) of tritiated thymidine ([³H]TdR; Amersham, Little Chalfont, Buckinghamshire, UK) for 16 additional hours and measured using a ß-counter (counter-WALLAC, Ramsey, MN, USA). Results expressed in counts per minute (cpm) represent the mean ± SEM of triplicates.

Quantitative PCR of FOXP3 Demethylation

The percentage of the Treg-specific demethylated region (TSDR) within the FOXP3 locus in spe-Tregs and poly-Tregs after culture was measured by Epiontis GmbH (Berlin, Germany), as a subcontractor, using quantitative PCR (qPCR) as previously described (60). As controls, the percentage of the TSDR was measured on cultured spe-Teffs and poly-Teffs. Briefly, genomic DNA was isolated using the DNeasy blood and tissue Kit (Qiagen, Valencia, CA, USA). After bisulfite conversion of genomic DNA, PCR products were generated with methyl- and non-methyl-specific primers for FOXP3 TSDR. Comparison of the number of demethylated TSDR to methylated TSDR copies allows an accurate measurement of Tregs (60).

Murine Experiments

Twelve- to 20-week-old nonobese diabetic/severe combined immunodeficient/IL-2 receptor γ chain (NOD/SCID/γC^-/-; NSG) male mice (NOD.Cγ-Prkdc^scid IL2rg^tm1Wjl/Sz, stock 61/4 005557; The Jackson Laboratory, Bar Harbor, ME, USA) were bred in our own animal facilities in specific pathogen-free conditions (accreditation number from the Veterinary services: A75-13-10) with an enriched fat regimen and addition of Bactrim (Roche, Boulogne-Billancourt, France) in drinking water every other week. The colony was regularly checked for γ-c deficiency by PCR according to the Jackson Laboratory protocol. Experiments were performed according to the European Union guidelines and approved by our institutional review board (CRRA Ile de France No. 3).

Xenogeneic GVHD Model in NSG Mice and Control by Tregs

To induce a xenogeneic GVHD, human CD3⁺ T cells were first purified using the CD3 microbeads-II kit (Miltenyi Biotec). The CD3⁺ T cells (10⁷ cells/mouse) were intravenously injected into NSG mice in the presence of allogeneic DCs (2 × 10⁶ cells/mouse). The latter are helpful to enhance the GVHD clinical signs and to accelerate the mortality, compared to our previously published xeno-GVHD model (51). Under these conditions, NSG mice developed a xenogeneic GVHD within 4-5 weeks. In order to evaluate the ability of spe-Tregs or poly-Tregs to control xenogeneic GVHD, spe-Tregs or poly-Tregs (5 × 10⁶/mouse) were concomitantly injected with autologous CD3⁺ T cells referred to as Teffs. Mouse survival and symptoms of GVHD including hunched back, diarrhea, and body weight were monitored daily. Mice were sacrificed when they lost ≥30% body weight. Each group contained three mice.

Histopathological Examination

Livers, lungs, and skin were preserved in Bouin's fixative (Polysciences, Inc., Europe GmbH, Eppelheim, Germany) and embedded in paraffin. For these organs, 3-μm-thick sections were stained with hematoxylin and eosin (Sigma-Aldrich) for histological examination. One pathologist analyzed slides in a blinded fashion to assess the intensity of GVHD. GVHD lesions were scored in each liver sample according to a semiquantitative scoring system (22) using seven parameters (inflammatory cell infiltrate within the portal tracts, bile duct injury, venous endothelialitis, periportal hepatocyte necrosis, acidophilic bodies, confluent lobular necrosis, sinusoidal lymphocytosis). The total score was the sum of those parameters, which were scored as follows: 0 as normal, 1 as focal and rare, 2 as focal and mild, 3 as diffuse and mild, 4 as diffuse and moderate, and 5 as diffuse and severe. GVHD lesions in the lung were evaluated as previously described (11) according to the following score, which is the sum of a periluminal (around airways and vessels) lymphocytic infiltrate index (0: absence; 1: lymphocytic cuff of 1-3 cells in diameter; 2: lymphocytic cuff of 4-10 cells in diameter; 3: lymphocytic cuff >10 cells in diameter), an alveolar/interstitial lymphocytic infiltrates index (0: absence; 1: increased number of cells only visible at high magnification; 2: interstitial thickening; 3: interstitial thickening and alveolar infiltrates), and an index of lung injury extension quantitated according the percentage of tissue involved (1: 5-25%, 2: 25-50%, 3: >50%). GVHD lesions in the skin were evaluated according to the sum of three parameters as described by Kaplan et al. (27) that are vacuolization of the basal layer of the epidermis, necrotic keratinocytes, and lymphocytic infiltrate within the dermis. Each of those parameters was scored as follows: 0 as normal, 1 as mild, 2 as moderate, 3 as severe.

Statistical Analysis

When data were compared according to purification methods from different leukapheresis samples, statistical analyses were performed using the unpaired Student's t-test with Welch's correction when appropriate. Data were analyzed with GraphPad Prism software (GraphPad Software, Inc., San Diego, CA, USA), and expressed as mean ± SEM. Kaplan–Meyer survival curves were constructed, and log-rank comparison of the groups was used to calculate p values. Statistical significance was taken at the 5% level (p < 0.05), the degree of significance was indicated as: p < 0.05, p < 0.01, p < 0.001.

Determination of Culture Conditions for the Activation and Stimulation of Treg Cell Division

We first studied different conditions for the emergence and/or activation of alloantigen-specific Tregs from a polyclonal Treg population. As iDCs may be more tolerogenic than mDCs (30), we compared their ability to activate Tregs. CD4⁺CD25^highCD127^low/- Tregs were purified by FACS (purity > 99%) and labeled with CFSE. Then Tregs were cocultured with either iDCs or mDCs in complete culture medium containing IL-2, IL-15, or both cytokines. Indeed, IL-15, whose α chain receptor is expressed by activated Tregs (62), enhances Treg survival, FOXP3 expression, and Treg suppressive function particularly when associated with rapamycin (58). As positive control, CFSE-labeled Tregs were also polyclonally activated with anti-CD3/anti-CD28 Dynabeads and IL-2.

Figure 1A shows the cell proliferation of Tregs quantified by CFSE staining dilution. Early dividing cells were detected as soon as 3 days of stimulation in all conditions. However, mDCs exhibited higher ability to trigger Treg cell divisions than iDCs, whatever the cytokines used in the culture medium. Treg activation was further documented by FACS analysis based on FOXP3 intensity assessed on mean fluorescence intensity and percentage (Fig. 1B). FOXP3 expression was upregulated between day 0 and day 6 in all conditions. When Tregs were cocultured with mDCs, IL-2 and IL-15 were able to upregulate FOXP3 expression at comparable levels, while combination of IL-2 and IL-15 strongly enhanced FOXP3 expression. When Tregs were cocultured with iDCs, FOXP3 was not so well maintained. Furthermore, IL-15 alone or combined with IL-2 was not as good as IL-2 alone to preserve FOXP3 expression. Surprisingly, FOXP3 expression was also maintained in the presence of mDCs and even iDCs without adding cytokines. This can be explained by the production of low amounts of IL-2 by DCs as already reported by others in mice and humans (19,37,61). For further experiments, these findings were taken into account, and alloantigen-specific Tregs were cultured in the presence of mDCs, IL-2, and IL-15.

OPEN IN VIEWER

Figure 1

Culture conditions for the activation and cell division of Tregs. FACS Tregs were labeled with CFSE and stimulated with allogeneic iDCs or mDCs in the presence of IL-2, IL-15, or both cytokines. mDCs were obtained by adding 10 ng/ml TNF-α, 100 UI/ml IL-1β 1,000 UI/ml IL-6, and 1 μg/ml PGE-2 at day 6 of culture. As control, CFSE-labeled Tregs were polyclonally stimulated with anti-CD3/anti-CD28-coated Dynabeads and IL-2. (A) Proliferation was daily monitored on the basis of CFSE dilution and analyzed by flow cytometry from day 3 to day 6. Data presented as CD4/CFSE dot plots gated on CD3⁺CD4⁺ cells are representative of three separate experiments. (B) FOXP3 expression of Tregs was monitored by flow cytometry at day 0, day 3, and day 6. Histograms show FOXP3 expression (filled curves) presented as percentage and mean of fluorescence intensity (in brackets), and isotype controls (black lines). Data are representative of three separate experiments.

Generation of Attoantigen-Specific Tregs Under Clinically Compliant Conditions

For clinically compliant conditions, we designed a protocol where Tregs were purified by MACS based on CD25⁺ selection using CD25-magnetic beads as previously described (20). Since preliminary data, also confirmed by others (39), indicated that two allogeneic stimulations would be preferable to a single one in order to enrich antigen-specific Tregs, MACS-purified Tregs were then stimulated twice (at day 0 and day 10) with allogeneic mDCs. Tregs were cultured in complete culture medium supplemented with IL-2, IL-15, and rapamycin up to day 21. As a negative control of specificity, MACS-purified Tregs were generated under polyclonal stimulation (poly-Tregs) as previously described (20). Treg expansion under polyclonal or antigen-specific stimulation is shown in Figure 2A. While poly-Tregs were expanded 39.6 ± 10.5-fold after 21 days of culture, the number of spe-Tregs obtained at 3 weeks was only 8.3 ± 0.1-fold the initial number of Tregs at the onset of cultures. Interestingly, large-scale experiments performed in our clinical cell therapy center using two blood mass leukapheresis products obtained from healthy donors allowed us to obtain ~130 × 10⁶ cells containing ~70-80% of CD4⁺CD25^highFOXP3⁺ Tregs after CD25⁺ selection using the CliniMACS® device (Miltenyi Biotec). With eightfold expansions of the initial number of Tregs, we would be able to produce ~1 × 10⁹ spe-Tregs.

OPEN IN VIEWER

Figure 2

Cell expansion and functional study of spe-Tregs and poly-Tregs. (A) Cell expansion of spe-Tregs and poly-Tregs after 21 days of culture. MACS Tregs (10⁶ cells) were seeded in culture and expanded either under polyclonal stimulation or under alloantigen stimulation using allogeneic mDCs as described in the Materials and Methods section. Data from four independent experiments performed with Tregs from four different healthy donors are shown. (B) Suppressive activity of spe-Tregs and poly-Tregs on the proliferation of autologous CD4⁺CD25^- T cells stimulated by allogeneic mDCs after 21 days of culture. Tregs generated from MACS-sorted CD25⁺ cells were expanded under alloantigen stimulation in the presence of allogeneic mDCs (donor a, DCa, left panels) or under polyclonal stimulation (right panels). For suppressive activity, spe-Tregs or poly-Tregs were mixed at different ratio with autologous effector cells and stimulated either with the same DCa () or a third party of allogeneic mDCs (referred to as donor b, DCb,) for 4 days. Cell proliferation was determined by incorporation of tritiated thymidine ([³H]TdR) and percentage of suppression calculated as follows: % suppression = [1 - (Tregs + Teffs + DC/Teffs + DC)] × 100. Results are presented as the mean ± SEM of six separate experiments for left panels and of four to five separate experiments for right panels. Statistical analysis was performed using an unpaired Student t-test: **p < 0.01.

Functional analysis of MACS-purified Tregs expanded under specific conditions showed that spe-Tregs exhibited a stronger and significant suppressive activity, particularly at low Treg/Teff ratio (p < 0.001), when stimulated by DCa initially used to prime the Tregs compared to a third-party DC (DCb) stimulation (Fig. 2B, left panel). Interestingly, DCa-stimulated Spe-Tregs only began to decrease their suppressive function below a 1:32 Treg/Teff ratio and were still suppressive at a 1:128 Treg/Teff ratio compared to DCb-stimulated Spe-Tregs (data not shown). As expected, poly-Tregs stimulated by DCa or by DCb exert a comparable suppressive activity (Fig. 2B, right panel). These data demonstrate that the initial population of MACS-sorted Tregs, which was polyclonal, can be enriched into spe-Tregs after two rounds of stimulation by allogeneic mDCs. It is important to point out that we have used clinical-grade reagents for the production of spe-Tregs, and subsequently, our culture conditions can be easily scaled up in view of clinical applications.

Immunophenotypic Characterization of Alloantigen-Specific Tregs

The strategy we set up to generate spe-Tregs was based on the purification of Tregs using anti-CD25 immunomagnetic beads. In the initial mononuclear cell fraction, Tregs defined as CD4⁺CD25^highCD127^low/- cells or CD4⁺CD25^highFOXP3⁺ cells represented 3-4<% of CD3⁺CD4⁺ cells (Table 2), and CD3⁺CD4⁺ cells represented 59.5 ± 4.6% of total MNCs.

OPEN IN VIEWER

Table 2

Identification of Tregs in CD3⁺CD4⁺-Gated Cells Before and After CD25⁺ Selection by MACS

	CD4+CD25HighCD127Low/-	CD4+CD25HighFOXP3+	CD4+FOXP3+Helios+
Initial cell fraction (MNCs)	3.76 ± 0.4	3.4 ± 0.3	3.3 ± 0.4
Positive cell fraction (CD25+ cells)	78.9 ± 2.2	73.7 ± 3.9	69.7 ± 1.3

Data presented are means ± SEM of at least nine separate experiments. Tregs, regulatory T cells; MACS, magnetic-activated cell sorting; MNCs, mononuclear cells.

After CD25⁺ selection, CD25⁺ cells represented 95.5 ± 0.9% of total cells and contained 91.7 ± 3.6% of CD3⁺CD4⁺ cells. Contaminating cells were B cells (CD19⁺ cells, 5.3 ± 0.7%), natural killer cells (CD3^-CD56⁺ cells, 0.7 ± 0.2%), CD8⁺ cells (1.4±0.3%), and monocytes (CD14⁺ cells, 2.1 ± 0.2%). Tregs (CD4⁺CD25^highCD127^low/- cells or CD4⁺CD25^high FOXP3⁺ cells) within CD4⁺CD3⁺ cells represented 74-79% of the cells (Table 2). Helios has also been described as a transcription factor expressed by natural Tregs (18,50). We found that CD4⁺FOXP3⁺Helios⁺ cells represented before and after CD25⁺ enrichment 3.3 ± 0.4% and 69.7 ± 1.3% of the cells, respectively (Table 2). Interestingly, 88.3 ± 1.6% of CD4⁺CD25^high were CD127^low/-FOXP3⁺Helios⁺, suggesting that Helios and FOXP3 are both good markers of Tregs.

After 21 days of culture, final cell products corresponding to spe-Tregs and poly-Tregs were composed of 88 ± 2.6% and 94.1 ± 3.6% of CD3⁺ cells, respectively. The fact that spe-Tregs cell products only contained 88% of CD3⁺CD4⁺ cells was due to the presence of 4.9 ± 1.4% CD3⁺CD8⁺ cells, 3.5 ± 0.3% CD3^-CD56⁺cells, and some remaining DCs. Further immunophenotypic analyses of spe-Tregs and poly-Tregs within gated CD4⁺CD3⁺ cells showed that Treg populations expressed at similar and high levels FOXP3, Helios, glycoprotein A repetitions predominant (garpin, GARP), latency-associated peptide (LAP), CD152, and CD278 (Fig. 3). As control, CD4⁺CD25^-CD127⁺ effector cells purified by FACS were cultured under similar conditions in order to generate spe-Teffs and poly-Teffs, and their phenotype was studied (Fig. 3). Interestingly, Helios, GARP, LAP, and CD152 could strongly statistically discriminate (unpaired Student t-test, p < 0.001) spe-Tregs from spe-Teffs (Fig. 3A). Similar statistical differences were observed when poly-Tregs and poly-Teffs were compared, except for CD152 that was upregulated when Teffs were expanded under strong polyclonal activation (Fig. 3B). The ectonucleotidase CD39, albeit important for the Treg function, was not discriminative after culture. Indeed, CD39 only expressed by ~40% of human Tregs (8,45) is also expressed by activated Teffs (41). As others (1,2), we observed that CD25 and CD127 could not anymore be discriminative markers between Tregs and Teffs after ex vivo expansion. It is also known that activated human Teffs express FOXP3 (2,48,59). Thus, 20-25% of spe-Teffs and poly-Teffs expressed FOXP3, albeit at lower intensity levels than Tregs. The highly conserved region within the human FOXP3 gene, named TSDR, is fully demethylated in Tregs only, and can discriminate Tregs from activated FOXP3⁺ conventional T cells (5). Therefore, we analyzed the number of demethylated TSDR in comparison to methylated TSDR copies in spe-Tregs and poly-Tregs as well as in spe-Teffs and poly-Teffs obtained after 3 weeks of culture, as previously described (60). As shown in Table 3, spe-Tregs and poly-Tregs expanded from MACS-sorted CD25⁺ cells exhibited a similar average demethylation level of 80.3 ± 5.5% (range 70.1-95.2%). Using the same culture conditions as for Tregs, less than 0.2% of expanded spe-Teffs and poly-Teffs were demethylated in the FOXP3 TSDR. Altogether these data confirm that spe-Tregs that exhibited a similar immunophenotype than poly-Tregs can be generated and easily distinguished from cultured Teffs.

OPEN IN VIEWER

Figure 3

Immunophenotypic analysis of spe- and poly-Tregs comparatively to spe- and poly-Teffs after 3 weeks of culture. Spe-Tregs and poly-Tregs were cultured as described in Materials and Methods for 3 weeks. Spe-Teffs and poly-Teffs generated from FACS-sorted CD4⁺CD25^-CD127⁺ cells were cultured under the same conditions. At 3 weeks of culture, aliquots from each culture were stained with different combinations of mAbs and analyzed by FACS. Data presented are means ± SEM from three to five separate experiments, represent the percentage of positive cells within gated CD3⁺CD4⁺ cells. (A) Immunophenotypic analysis of spe-Tregs (■) and spe-Teffs (). (B) Immunophenotypic analysis of poly-Tregs () and poly-Teffs (). Statistical analysis was performed using an unpaired Student t-test: **p < 0.01; ***p < 0.001.

OPEN IN VIEWER

Table 3

Analysis of the FOXP3 TSDR in Tregs and Teffs After 3 Weeks of Culture

	Donor 1				Donor 2
	Poly-Tregs	Spe-Tregs Poly-Teffs	Spe-Teffs	Poly-Tregs	Spe-Tregs	Poly-Teffs	Spe-Teffs
% TSDR	95.2	81.9	0.1	0.2	70.1	74.1	0.1	0.1

Teffs, effector T cells.

Control of Xeno-GVHD by Allogen-Specific Tregs Using Humanized NSG Mice

In order to test and to compare the ability of spe-Tegs versus poly-Tregs to control GVHD, we set up a model in which immunodeficient NSG mice were infused intravenously (IV) with human purified effector CD3⁺ T cells (10 × 10⁶/mouse) and allogenic DC (2 × 10⁶/mouse) in the presence or not of spe-Tregs or poly-Tregs (5 × 10⁶/mouse) that were autologous to effector T cells. Clinical signs of GVHD were daily monitored on several parameters (see Materials and Methods). It is important to note that the same model has been previously used with success to study the regulatory function of ex vivo expanded human Tregs (i) in xeno-GVHD (21,38) or (ii) in other models of allogeneic immune response (41). It is likely that this model reflects more closely the effect of human Tregs on human T cells in vivo and the consequences on the disease process, compared to pure murine models of GVHD. As expected, mice injected with effector CD3 T cells developed a GVHD within 4-5 weeks. Animals were sacrificed for ethical reasons in case of strong clinical signs of GVHD or at 120 days. Kaplan–Meyer survival curves (Fig. 4A) showed that 66-77% of mice treated by Tregs were alive at 120 days, while only 22% of mice survived in the control group (nontreated animals). Thus, both spe-Tregs and poly-Tregs can significantly (p < 0.05) control xeno-GVHD induced by human effector T cells and allogeneic DCs. Histopathological examinations of different tissues (Fig. 4B) showed that spe-Tregs significantly diminished GVHD scores in the skin (p < 0.01), the lung (p < 0.01), and the liver (p < 0.05), while poly-Tregs only significantly diminished GVHD scores in the skin (p < 0.05). These data indicate that spe-Tregs exert, as good as poly-Tregs, a suppressive activity that can significantly (p < 0.05) control a xeno-GVHD induced by human effector T cells.

OPEN IN VIEWER

Figure 4

Control of xeno-GVHD by spe-Tregs and poly-Tregs. Effect of spe-Tregs and poly-Tregs on the control of GVHD in a model of xenogeneic transplantation. Xenogeneic GVHD was induced by adoptive transfer of 10⁷ CD3⁺ human T cells (Teffs) into NOD/SCID/γC^-/- mice. Mice received either Teffs alone (●), or cotransfer of Teffs + 5 × 10⁶ spe-Tregs (■) or cotransfer of Teffs + 5 × 10⁶ poly-Tregs (▲). Mice also received concomitantly 2 × 10⁶ of allogeneic mDCs initially used to stimulate Tregs. Clinical symptoms of GVHD were monitored daily up to 120 days. (A) Kaplan–Meier survival curve for mice grafted with Teffs alone (n = 9) or Teffs + Tregs (Teffs/Tregs ratio: 2:1, n = 9). Data were obtained from a pool of three independent experiments. Log-rank comparison of the groups was used to calculate p values. (B) Histopathological examinations on the skin, lung, and liver of each animal were performed by a pathologist in a blinded fashion, and GVHD lesions were as described by Kaplan et al. (27), Cooke et al. (11), and Hill et al. (22), respectively (see Materials and Methods section). Statistical analyses were performed using an unpaired Student t-test with Welch's correction: *p < 0.05; **p < 0.01.