Regulatory T cell (Treg) therapy is a promising approach for transplant rejection and severe autoimmunity. Unfortunately, clinically meaningful Treg numbers can be obtained only upon in vitro culture. Functional stability of human expanded (e)Tregs and induced (i)Tregs has not been thoroughly addressed for all proposed protocols, hindering clinical translation. We undertook a systematic comparison of eTregs and iTregs to recommend the most suitable for clinical implementation, and then tested their effectiveness and feasibility in rheumatoid arthritis (RA). Regardless of the treatment, iTregs acquired suppressive function and FOXP3 expression, but lost them upon secondary restimulation in the absence of differentiation factors, which mimics in vivo reactivation. In contrast, eTregs expanded in the presence of rapamycin (rapa) retained their regulatory properties and FOXP3 demethylation upon restimulation with no stabilizing agent. FOXP3 demethylation predicted Treg functional stability upon secondary TCR engagement. Rapa eTregs suppressed conventional T cell proliferation via both surface (CTLA-4) and secreted (IL-10, TGF-β, and IL-35) mediators, similarly to ex vivo Tregs. Importantly, Treg expansion with rapa from RA patients produced functionally stable Tregs with yields comparable to healthy donors. Moreover, rapa eTregs from RA patients were resistant to suppression reversal by the proinflammatory cytokine TNF-α, and were more efficient in suppressing synovial conventional T cell proliferation compared with their ex vivo counterparts, suggesting that rapa improves both Treg function and stability. In conclusion, our data indicate Treg expansion with rapa as the protocol of choice for clinical application in rheumatological settings, with assessment of FOXP3 demethylation as a necessary quality control step.

CD4+CD25highFOXP3+ regulatory T cells (Tregs) suppress immune responses of conventional T cell (Tconv) and are major players in the maintenance of peripheral immune tolerance (1). Tregs can either differentiate in the thymus or be induced in the periphery from naive T cells under tolerogenic conditions. In humans and mice, Treg differentiation is driven by the transcription factor FOXP3, which is also essential for their suppressive function (24).

The therapeutic potential of Treg therapy is currently under scrutiny in the settings of transplantation and severe autoimmunity (57). Several protocols for ex vivo expansion of Tregs (eTregs) (810) or in vitro induction of Tregs (iTregs) from naive T cells (11, 12) have been developed, in an effort to obtain sufficient numbers of suppressive cells to be manipulated and reinfused into patients. However, the field has yet to achieve firm consensus for clinical translation, as most evidence has been acquired in mice (13) and cannot be directly translated to humans because of crucial differences in the Treg compartment between these two species. For instance, TGF-β alone is sufficient to induce suppressive functions in mouse but not human T cells (14, 15). Furthermore, the proposed strategies to obtain stable eTregs in humans (810, 16, 17) were tested under diverse and noncomparable experimental conditions. The characterization is even more scattered for human iTregs (11). In the absence of a clear indication, ongoing clinical trials relying on cultured Tregs (at least four currently registered at www.clinicaltrials.gov: NCT01210664, NCT01624077, NCT01446484, and NCT01634217) are all based on different protocols. Such an unfortunate dispersion of efforts highlights the urgency of a systematic and exquisitely human assessment to inform clinical decisions.

In addition, several questions remain to be addressed in view of a safe and effective clinical translation. In autoimmune diseases, Tregs are thought to be either dysfunctional (7, 1821), or unstable in the inflamed peripheral site where self-reactivity occurs (2224). Functional immune imbalance may be further exacerbated by pathogenic Tconv, which become resistant to Treg-mediated suppression (21). This scenario would undermine the effectiveness of a therapeutic Treg-based approach. Moreover, although the mechanisms underlying Treg suppression are partially understood, it is currently unknown whether in vitro–manipulated Tregs suppress using similar molecular effectors. Overall, critical information to inform clinical translation of Treg therapy for autoimmunity is still missing.

In this paper, we address this unmet medical need by performing a comprehensive investigation of Treg expansion/induction strategies in humans, with the final goal of recommending the most efficient method in yielding functionally stable Tregs. We then demonstrate the feasibility and effectiveness of the selected protocol with cells isolated from rheumatoid arthritis (RA) patients. Finally, we show that Tregs expanded in vitro display superior suppressive ability and functional stability than their freshly isolated counterparts, thereby providing support for Treg therapy in autoimmune arthritis.

Fresh EDTA-anticoagulated blood was collected at The Scripps Research Institute (TSRI; La Jolla, CA) Healthy Blood Donor Service under Sanford-Burnham Medical Research Institute Institutional Review Board approval and upon informed consent given in accordance to the tenets of the Declaration of Helsinki. Blood and synovial fluid samples from RA patients were obtained with appropriate Institutional Review Board approval from the Rheumatology Clinic at TSRI and from the Department of Rheumatology at Singapore General Hospital (see Table I for patients’ details), and immediately processed. PBMCs and synovial fluid mononuclear cells were separated by density gradient with Histopaque-1077 (Sigma-Aldrich) and frozen in freezing medium (90% FBS and 10% DMSO). Synovial fluid was clarified with hyaluronidase (Sigma-Aldrich) before synovial fluid mononuclear cell separation.

Table I.
Patients’ demographics and clinical characteristics
Time fromClinical Activity
Patient IDGenderAge (y)Disease Onset (y)(DAS28-3)Treatment
45 12 5.03 Oxycodone, folic acid, methotrexate, hydroxyzine, clopidogrel, prednisone 
61 14 0.68 Etanercept, hydrocodone, hydroxychloroquine, methotrexate, prednisone 
44 3.25 Etanercept, cyclobenzaprine 
42 <1 3.76 Methotrexate, folic acid 
53 15 2.22 Infliximab 
66 4.99 Methotrexate, folic acid, sulfasalazine 
28 4.23 Etanercept, methotrexate, folic acid 
63 35 Methotrexate, folic acid, infliximab, prednisone, hydroxychloroquine 
62 3.47 Methotrexate, folic acid, celecoxib, hydroxychloroquine 
10 45 13 Diclofenac, methotrexate, folic acid, hydroxychloroquine 
11 61 24 3.44 Folic acid, adalimumab, methotrexate, prednisone 
12 66 13 5.49 Methotrexate, folic acid, prednisone 
Time fromClinical Activity
Patient IDGenderAge (y)Disease Onset (y)(DAS28-3)Treatment
45 12 5.03 Oxycodone, folic acid, methotrexate, hydroxyzine, clopidogrel, prednisone 
61 14 0.68 Etanercept, hydrocodone, hydroxychloroquine, methotrexate, prednisone 
44 3.25 Etanercept, cyclobenzaprine 
42 <1 3.76 Methotrexate, folic acid 
53 15 2.22 Infliximab 
66 4.99 Methotrexate, folic acid, sulfasalazine 
28 4.23 Etanercept, methotrexate, folic acid 
63 35 Methotrexate, folic acid, infliximab, prednisone, hydroxychloroquine 
62 3.47 Methotrexate, folic acid, celecoxib, hydroxychloroquine 
10 45 13 Diclofenac, methotrexate, folic acid, hydroxychloroquine 
11 61 24 3.44 Folic acid, adalimumab, methotrexate, prednisone 
12 66 13 5.49 Methotrexate, folic acid, prednisone 

PBMCs were enriched for CD4+ T cells with the Human CD4+ T cell enrichment kit II (Miltenyi Biotec), then sorted in Tregs (CD4+CD25highCD127low/−) or naive Tconv (CD4+CD25low/−CD45RO) with a FACSAria II (BD Biosciences). Sorting purity was consistently above 99%. Fluorochrome-conjugated Abs were from BioLegend, BD Biosciences, and eBioscience. Cells were cultured in complete medium: RPMI 1640 (HyClone) supplemented with 5% heat-inactivated human serum AB (GemCell), 2 mM l-glutamine (Life Technologies), and 100 U penicillin/100 μg streptomycin (Life Technologies). Cells were stimulated with anti-CD3/CD28–coated beads (Life Technologies) at 3:1 ratio (bead:cell) in the presence of 300 U/ml recombinant human (rh)IL-2 (Life Technologies). Naive T cells were cultured in complete medium and rhIL-2 only (−), or with the addition of the following: 10 mM all-trans–retinoic acid (ATRA; Tocris Bioscience), 5 ng/ml rhTGF-β (PeproTech), 100 nM rapamycin (rapa; LC Laboratories), 10 mM ATRA + 5 ng/ml rhTGF-β (ATRA/TGF), or 100 nM rapa + 5 ng/ml rhTGF-β (rapa/TGF). eTregs were cultured in complete medium and rhIL-2 only (−), or with the addition of 100 nM rapa or 100 nM rapa and 10 mM ATRA (ATRA/rapa). Cells were rested for 24 h in rhIL-2 (50 U/ml) after 14 d of culture before further analysis. A second round of stimulation was performed without differentiation factors (beads + rhIL-2 only) for 14 additional days.

Dead cell exclusion was performed with Live/Dead Fixable Aqua Dead Cell Stain (Life Technologies). Intracellular staining was performed using the anti-human FOXP3 staining set from eBioscience, following manufacturer’s instructions. Samples were acquired with a FACSAria II (BD Biosciences) and analyzed with FlowJo version 9 (Tree Star).

Cells were stimulated with 20 nM PMA and 200 nM ionomycin (PMA/ionomycin, both from Fisher Scientific) for 2 h. RNA was extracted using the ZR-Duet DNA/RNA MiniPrep kit (Zymo Research) and converted into cDNA with the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Singleplex qPCRs were run using TaqMan chemistry on a StepOnePlus Real-Time PCR system (Applied Biosystems).

PBMCs were thawed and allowed to rest for 16 h in complete medium with 20 U/ml rhIL-2. Cells were then sorted as Tregs (CD4+CD25highCD127low/−) or Tconv (CD4+CD25low/-) with a FACSAria II (BD Biosciences). Tconv were labeled with CellTrace Violet (Life Technologies) according to the manufacturer’s instructions. Five thousand Tconv were stimulated with 1000 anti-CD3/CD28–coated beads for 5 d, in the presence or absence of ex vivo Tregs or cultured iTregs/eTregs. The suppressor-to-responder ratio was 1:1 for most experiments, as that allowed to capture both the weak and strong suppressions abilities of Tregs differentiated, expanded, or isolated using different protocols. In selected experiments, we confirmed results by repeating the assay over a range of Treg:Tconv ratios.

Functional grade blocking Abs and the corresponding isotype controls were from R&D Systems (anti–IL-35, clone 27537; anti-TGF, clone 1D11), BioLegend (anti–IL-10, clone JES3-9D7; anti–IL-10R, clone 3F9) and BD Biosciences (anti–CTLA-4, clone BNI3). All were used at 10 μg/ml except for anti-TGF, which was used at 30 μg/ml. rhTNF-α (R&D Systems) was used at 50 ng/ml. Transwell inserts (0.4-μm pore size) were from Corning. In selected experiments, T cells were stimulated with 2500 allogeneic dendritic cells (DCs) activated with LPS (50 ng/ml; Sigma-Aldrich), rather than beads. Allogeneic DCs were derived from circulating monocytes, isolated with the Monocyte Isolation Kit II (Miltenyi Biotec) and cultured for 5 d in RPMI-1640 medium (HyClone) supplemented with 10% FBS (Sigma-Aldrich), 2 mM l-glutamine (Life Technologies) and 100 U penicillin/100 μg streptomycin (Life Technologies), with the addition of 100 ng/ml rhGM-CSF and 50 ng/ml rhIL-4 (both from PeproTech).

At day 5, dead cells were excluded with Sytox Red (Life Technologies), and samples were analyzed by flow cytometry. The proliferation index was calculated using ModFit LT (Verity Software House).

IL-2, IL-4, IL-6, IFN-γ, and TNF-α were measured on cell-free supernatants from suppression assays at day 5 with CBA (BD Biosciences).

Because the baseline of FOXP3 Treg–specific demethylated region (TSDR) demethylation in Tregs varies between males and females because of X-linked inactivation, gender consistency is necessary when reporting results. When comparing protocols of Treg differentiation/expansion, we used healthy male donors. When comparing RA patients to healthy donors (HD), because patient samples are rarer and focusing on a single gender is impractical, we showed TSDR methylation profiles separately for males and females. gDNA was isolated from either bulk cultures or sorted FOXP3+ cells, as indicated, with the ZR-Duet DNA/RNA MiniPrep kit, and bisulfite conversion was performed with EZ DNA Methylation-Direct kit (both from Zymo Research). FOXP3 TSDR methylation was assessed using a published protocol (25).

Bar and line graphs depict mean and SEM. Hypotheses were tested with two-tail t tests in single pairwise comparisons or multiple orthogonal comparisons. ANOVA posthoc tests were used for correction of multiple non-orthogonal comparisons: Dunnett’s when multiple conditions were tested against a single control condition; Sidak’s when a meaningful subset of conditions was selected for pairwise comparisons.

Unsupervised hierarchical clustering was performed on normalized mRNA expression (relative to GAPDH and β-actin) using the euclidian distance and complete linkage.

First, we compared the overall efficiency of published methods in either expanding or inducing suppressive FOXP3+ Tregs from HD. iTregs were generated by stimulating naive CD4+CD25low/−CD45RO T cells in the presence of IL-2 and TGF-β, with the addition of either ATRA (in combination ATRA/TGF) (12) or rapa (in combination rapa/TGF) (11). CD4+CD25highCD127low/− Tregs (>85% FOXP3+) were expanded with IL-2, in the presence or absence of rapa, to obtain eTregs (8, 10).

After 14 d, iTreg and eTreg yields were in line with published data (911), with iTregs displaying greater proliferation potential compared with eTregs. As expected, the controls cultured in IL-2 only (−) proliferated more vigorously than their treated counterparts (Fig. 1A) (8, 11). All cultures showed similar viability, irrespective of the protocol used (Fig. 1B).

FIGURE 1.

Phenotypic and functional characteristics of eTregs and iTregs. Naive CD4+CD25low/−CD45RO T cells (for generating iTregs) or CD4+CD25highCD127low/− Tregs (for generating eTregs) were cultured for 14 d in the presence of IL-2 and the indicated factors. (A) Fold expansion at day 14 on a logarithmic scale. (B) Viability at day 14. (C and D) Representative plots for FOXP3 staining (with numbers indicating the percentage of FOXP3+ cells gated on viable cells) and summary statistics at day 14. (E) Suppression of Tconv proliferation by ex vivo or rapa eTregs. (F) At day 14, cells were stimulated with PMA/ionomycin before measuring mRNA expression. Unsupervised hierarchical clustering was performed on averaged normalized relative mRNA expression, color-coded according to the legend. The red dotted line denotes the first split of the dendrogram. n = 4–14 per condition over two to five independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

FIGURE 1.

Phenotypic and functional characteristics of eTregs and iTregs. Naive CD4+CD25low/−CD45RO T cells (for generating iTregs) or CD4+CD25highCD127low/− Tregs (for generating eTregs) were cultured for 14 d in the presence of IL-2 and the indicated factors. (A) Fold expansion at day 14 on a logarithmic scale. (B) Viability at day 14. (C and D) Representative plots for FOXP3 staining (with numbers indicating the percentage of FOXP3+ cells gated on viable cells) and summary statistics at day 14. (E) Suppression of Tconv proliferation by ex vivo or rapa eTregs. (F) At day 14, cells were stimulated with PMA/ionomycin before measuring mRNA expression. Unsupervised hierarchical clustering was performed on averaged normalized relative mRNA expression, color-coded according to the legend. The red dotted line denotes the first split of the dendrogram. n = 4–14 per condition over two to five independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Close modal

We measured the efficiency of FOXP3 induction (iTregs) or maintenance (eTregs) after a 24 h resting period to allow activation-induced FOXP3 expression to subside (14, 15). Both ATRA/TGF and rapa/TGF iTregs acquired high FOXP3 expression, with the rapa/TGF condition displaying the highest proportion of FOXP3+ cells among iTregs. Rapa eTregs homogeneously expressed FOXP3, whereas IL-2–only eTregs displayed lower FOXP3 expression (Fig. 1C, 1D). Accordingly, the suppressive ability of IL-2–only eTregs was lower than that of rapa eTregs. Of note, rapa eTregs displayed more potent suppressive ability also compared with their ex vivo counterparts (Fig. 1E).

To achieve a more detailed characterization of iTregs and eTregs, we investigated their regulatory molecular signature. Expression data of genes classically linked to Treg function were fed to an unsupervised clustering algorithm. The algorithm partitioned iTregs from eTregs, with the former clustering with the untreated control (− iTreg), whereas the latter clustered with ex vivo Tregs (Fig. 1F).

Altogether, these data suggest that, although iTregs are suppressive, they do not acquire the typical ex vivo Treg signature. By contrast, eTregs do maintain the overall characteristics of ex vivo Tregs.

Permanent lineage commitment is crucial for an effective Treg therapy. To test whether Treg functionality was stably maintained by the progeny, iTreg and eTregs were polyclonally stimulated without differentiation factors (TGF-β, ATRA, and rapa).

After restimulation, the prevalence of FOXP3+ cells decreased in both ATRA/TGF and rapa/TGF iTregs, with the latter displaying a higher percentage of residual FOXP3+ cells (Fig. 2A, solid lines). This was mirrored by reduced suppressive ability (Fig. 2B, solid lines). These data suggest that a steady reduction in FOXP3 expression in iTregs may lead to complete loss of suppression ability upon multiple rounds of restimulation, which is likely to happen both in vitro and in vivo.

FIGURE 2.

Rapa eTregs are more stable than iTregs upon restimulation. At day 14, cells expanded as detailed in Fig. 1 were restimulated for additional 14 d in the presence of IL-2 only (day 28). (A and B) At the end of the culture, cells were either stained for FOXP3 (A) or tested in a suppression assay, performed as in Fig. 1E. (C) FOXP3 expression in eTregs expanded (day 14) and restimulated (day 28) in IL-2 only. Each line corresponds to an individual donor. Dotted lines indicate donors retaining relatively high FOXP3 expression, whereas solid lines indicate donors substantially losing FOXP3 expression (“low FOXP3”) at day 28. (D) Suppression of Tconv proliferation by IL-2–only eTregs, segregated by prevalence of FOXP3 at day 28. (E) Cells were stimulated with PMA/ionomycin before measuring mRNA expression. Unsupervised hierarchical clustering was performed on averaged normalized relative mRNA expression, color-coded according to the legend. White indicates not detected. The red dotted line denotes the first split of the dendrogram. n = 4–12 per condition over two to four independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001.

FIGURE 2.

Rapa eTregs are more stable than iTregs upon restimulation. At day 14, cells expanded as detailed in Fig. 1 were restimulated for additional 14 d in the presence of IL-2 only (day 28). (A and B) At the end of the culture, cells were either stained for FOXP3 (A) or tested in a suppression assay, performed as in Fig. 1E. (C) FOXP3 expression in eTregs expanded (day 14) and restimulated (day 28) in IL-2 only. Each line corresponds to an individual donor. Dotted lines indicate donors retaining relatively high FOXP3 expression, whereas solid lines indicate donors substantially losing FOXP3 expression (“low FOXP3”) at day 28. (D) Suppression of Tconv proliferation by IL-2–only eTregs, segregated by prevalence of FOXP3 at day 28. (E) Cells were stimulated with PMA/ionomycin before measuring mRNA expression. Unsupervised hierarchical clustering was performed on averaged normalized relative mRNA expression, color-coded according to the legend. White indicates not detected. The red dotted line denotes the first split of the dendrogram. n = 4–12 per condition over two to four independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001.

Close modal

Interestingly, rapa eTregs maintained both high FOXP3 expression and strong suppressive ability, whereas IL-2–only eTregs displayed reduced FOXP3 expression but their average suppressive function was minimally affected (Fig. 2A, 2B, dotted lines). We explained this seeming discrepancy with the high individual variability observed in this condition. Indeed, although FOXP3 decreased in all donors, some displayed a substantial residual expression (>30% of cells), whereas others almost completely lost it (Fig. 2C). When segregated by FOXP3 prevalence, the donors with sizable FOXP3 expression also retained high suppressive ability, whereas the donors with low FOXP3 expression were much less suppressive (Fig. 2D, dotted and solid lines, respectively). Such heterogeneity was only observed in IL-2–only eTregs. All other treatments generated cells with consistent features across donors.

We further investigated Treg stability at phenotypical level. mRNA expression of the same markers and cytokines reported in Fig. 2 was tested immediately after the first expansion/differentiation round (day 14), or after restimulation in the absence of exogenous factors (day 28). Again, the clustering algorithm partitioned iTregs from eTregs, regardless of the day of differentiation, with eTregs clustering together with ex vivo Tregs. The only notable exception to this pattern were eTregs expanded with IL-2 only after two rounds of stimulation (− eTreg day 28), which clustered with iTregs. This finding further supports the notion of eTreg instability in the absence of previous rapa conditioning (Fig. 2E).

Our data indicate that human eTregs expanded in the presence of rapa retain higher stability and resemblance to ex vivo Tregs compared with all other protocols.

Human (2629) and mouse (3035) ex vivo Tregs are characterized by DNA demethylation in the FOXP3 TSDR, also called “conserved noncoding sequence 2” (CNS2). Although there is limited information on TSDR methylation status in human eTregs (9, 17, 28), no data are available for human iTreg protocols proposed for cell therapy. We hypothesized that TSDR methylation might be directly linked to Treg stability (i.e., iTregs might lose FOXP3 expression and suppressive ability over time because of their inability to demethylate the TSDR). In line with this hypothesis, DNA methylation at the FOXP3 locus showed a complete polarity between iTregs and eTregs at the end of in vitro differentiation/expansion. iTregs were always hypermethylated at the TSDR and were comparable to Tconv, regardless of FOXP3 expression. Conversely, eTregs retained the typical TSDR hypomethylation. Of note, eTregs expanded in IL-2 only exhibited increased methylation when compared with rapa eTregs (Fig. 3A, 3B). Because IL-2–only eTregs contained fewer FOXP3+ cells at day 14 (Fig. 1B, 1C), this difference in methylation might be due to the expansion of the few contaminating FOXP3 cells, which are expected to be methylated. This hypothesis was supported by the methylation degree of FOXP3+ cells sorted from cultures of IL-2–only eTregs, which were completely demethylated, similarly to FOXP3+ cells from rapa eTregs (Fig. 3C).

FIGURE 3.

Rapa eTregs retain, although iTregs fail to acquire, stable DNA demethylation at the TSDR. At day 14, cells expanded as detailed in Fig. 1 were restimulated for additional 14 d in the presence of IL-2 only (day 28). At the end of the culture, TSDR methylation was evaluated. (A) Methylation percentage of each CpG site of the TSDR at day 14, color-coded according to the legend. Each column corresponds to a CpG site. Absolute coordinates according to the GRCh37/hg19 human genome assembly are indicated. One representative donor is shown. (B and C) Average TSDR methylation measured at day 14 in the bulk culture (B) or in sorted FOXP3+ cells (C). (D and E) Average TSDR methylation measured at days 14 and 28 in the bulk culture (D) or in sorted FOXP3+ cells (E) from eTreg cultures. For (A–E), n = 4–7 per condition over two to four independent experiments. (F) Data collected on the suppressive ability at d28 for all conditions were segregated based on TSDR methylation measured at day 14. Demethylated: <20% methylation; hypermethylated: >20% methylation. n = 11 per condition over four independent experiments. *p < 0.05, **p < 0.01, ****p < 0.0001.

FIGURE 3.

Rapa eTregs retain, although iTregs fail to acquire, stable DNA demethylation at the TSDR. At day 14, cells expanded as detailed in Fig. 1 were restimulated for additional 14 d in the presence of IL-2 only (day 28). At the end of the culture, TSDR methylation was evaluated. (A) Methylation percentage of each CpG site of the TSDR at day 14, color-coded according to the legend. Each column corresponds to a CpG site. Absolute coordinates according to the GRCh37/hg19 human genome assembly are indicated. One representative donor is shown. (B and C) Average TSDR methylation measured at day 14 in the bulk culture (B) or in sorted FOXP3+ cells (C). (D and E) Average TSDR methylation measured at days 14 and 28 in the bulk culture (D) or in sorted FOXP3+ cells (E) from eTreg cultures. For (A–E), n = 4–7 per condition over two to four independent experiments. (F) Data collected on the suppressive ability at d28 for all conditions were segregated based on TSDR methylation measured at day 14. Demethylated: <20% methylation; hypermethylated: >20% methylation. n = 11 per condition over four independent experiments. *p < 0.05, **p < 0.01, ****p < 0.0001.

Close modal

To investigate the link between stable FOXP3 expression and TSDR methylation in eTregs, we measured their TSDR methylation after restimulation without differentiation factors. Importantly, the bulk culture of rapa eTregs retained low TSDR methylation (Fig. 3D), whereas eTregs expanded in IL-2 only showed increased TSDR methylation, in line with their declining FOXP3 expression (Fig. 2A). Interestingly, FOXP3+ cells sorted from IL-2–only eTregs after restimulation showed increased TSDR methylation when compared with rapa eTregs (Fig. 3E), in contrast to what observed at d14.

These data indicate that eTregs previously exposed to rapa do not lose their epigenetic signature upon in vitro expansion. Conversely, IL-2–only eTregs, although expressing FOXP3, lose their regulatory epigenetic signature over time. The defect in DNA methylation at the TSDR may thus precede functional failure in suppressive ability.

To investigate whether TSDR (de)methylation can predict the suppressive ability of human Tregs, we segregated the suppression data collected at d28 for iTregs and eTregs based on the degree of TSDR methylation at d14, regardless of the conditioning protocol. The threshold of TSDR methylation was set at 20%, which is the highest level of natural methylation observed in ex vivo Tregs (Fig. 3A). All cultures with <20% methylation at the TSDR at d14 maintained stable suppressive ability also at d28, whereas cultures with higher TSDR methylation did not (Fig. 3F). These data support the use of TSDR methylation as a predictor of Treg functional stability.

It was recently shown that combining ATRA with rapa might better preserve CD45RA Treg functionality during one round of ex vivo expansion (9). Because most Tregs are CD45RA- (29), we tested whether ATRA/rapa eTregs showed enhanced suppressive function and stability compared with rapa eTregs. In accordance with previous reports (9, 36), the addition of ATRA did not affect the magnitude of Treg expansion (Fig. 4A), viability (Fig. 4B), and FOXP3 expression (Fig. 4C) compared with rapa alone. In our hands, both the suppressive ability (Fig. 4D) and the TSDR methylation profile (Fig. 4E) were superimposable between rapa and ATRA/rapa Tregs at day 14. All these characteristics were stably maintained upon restimulation (day 28). Altogether, these data demonstrate that ATRA/rapa eTregs are as suppressive and epigenetically stable as rapa eTregs.

FIGURE 4.

Rapa and ATRA/rapa eTregs are equally stable and suppressive. (A) Fold expansion of eTregs expanded in the presence of rapa or ATRA/rapa at day 14 from four donors. (B) Viability at day 14. (C) FOXP3 expression in eTregs expanded in the presence of rapa or ATRA/rapa at day 14 or after restimulation (day 28). Numbers indicate the percentage of FOXP3+ cells (gated on viable cells). (D) Suppression of Tconv proliferation by rapa eTregs expanded and restimulated as in (B). Top panel, CellTrace Violet (CTV) dilution profile of proliferated Tconv at 1:1 Treg:Tconv ratio. Bottom panel, Suppression efficiency over a range of Treg:Tconv ratios. (E) Methylation percentage of each CpG site of the TSDR at days 14 and 28, color-coded according to the legend. Each column corresponds to a CpG site. (C–E) A representative donor of four per condition is shown.

FIGURE 4.

Rapa and ATRA/rapa eTregs are equally stable and suppressive. (A) Fold expansion of eTregs expanded in the presence of rapa or ATRA/rapa at day 14 from four donors. (B) Viability at day 14. (C) FOXP3 expression in eTregs expanded in the presence of rapa or ATRA/rapa at day 14 or after restimulation (day 28). Numbers indicate the percentage of FOXP3+ cells (gated on viable cells). (D) Suppression of Tconv proliferation by rapa eTregs expanded and restimulated as in (B). Top panel, CellTrace Violet (CTV) dilution profile of proliferated Tconv at 1:1 Treg:Tconv ratio. Bottom panel, Suppression efficiency over a range of Treg:Tconv ratios. (E) Methylation percentage of each CpG site of the TSDR at days 14 and 28, color-coded according to the legend. Each column corresponds to a CpG site. (C–E) A representative donor of four per condition is shown.

Close modal

On the basis of the data reported above, Treg expansion with rapa is the best performing protocol yielding stable and suppressive Tregs. Therefore, we focused on rapa eTregs and further characterized their suppressive function. First, rapa eTregs were able to suppress the production of proinflammatory cytokines by activated Tconv, in addition to suppressing their proliferation. Importantly, such suppressive function was substantially improved after rapa conditioning as compared with ex vivo Tregs (Fig. 5A). Second, rapa eTregs suppressed Tconv through a variety of both contact-dependent and soluble mediators, similarly to ex vivo Tregs. Indeed, physical segregation of Tregs and Tconv with a Transwell system strongly reduced rapa eTreg suppressive ability (Fig. 5B), indicating that cell contact is essential for optimal suppression. In DC-dependent experimental settings, anti–CTLA-4 blockade effectively reverted the suppression (Fig. 5C). In addition, experiments with neutralizing Abs showed that rapa eTregs use IL-10, TGF-β, and IL-35 to suppress Tconv proliferation, similarly to freshly isolated Tregs (Fig. 5D).

FIGURE 5.

Rapa eTregs suppress via both surface and soluble factors and are stable in a proinflammatory environment. (A) Suppression of Tconv proinflammatory cytokine production by ex vivo Tregs or rapa eTregs, expanded as in Fig. 1. (BD) Suppression of Tconv proliferation by ex vivo or rapa eTregs in a Transwell system (B) or in the presence of the indicated neutralizing Abs (C and D). In (C), T cells were stimulated with allogeneic LPS-activated DCs. (E) Suppression of Tconv proliferation by ex vivo or rapa eTregs in the presence of TNF-α. n = 8–19 per condition over nine independent experiments. *p < 0.05, **p < 0.01, ****p < 0.0001.

FIGURE 5.

Rapa eTregs suppress via both surface and soluble factors and are stable in a proinflammatory environment. (A) Suppression of Tconv proinflammatory cytokine production by ex vivo Tregs or rapa eTregs, expanded as in Fig. 1. (BD) Suppression of Tconv proliferation by ex vivo or rapa eTregs in a Transwell system (B) or in the presence of the indicated neutralizing Abs (C and D). In (C), T cells were stimulated with allogeneic LPS-activated DCs. (E) Suppression of Tconv proliferation by ex vivo or rapa eTregs in the presence of TNF-α. n = 8–19 per condition over nine independent experiments. *p < 0.05, **p < 0.01, ****p < 0.0001.

Close modal

Although rapa eTregs are stable upon restimulation even in the absence of rapa (Fig. 2), which mimics in vivo reactivation, their stability might be affected by the proinflammatory microenvironment at the site of Ag recognition. Therefore, we tested the suppressive ability of rapa eTregs in the presence of TNF-α. Importantly, although this proinflammatory cytokine partially reverted the suppressive function of ex vivo Tregs, it had only a negligible effect on rapa eTregs (Fig. 5E).

In summary, rapa eTregs inhibit Tconv proliferation and cytokine production through a variety of mechanisms, and stably maintain their functionality under inflammatory conditions.

Our data demonstrate that Treg expansion with rapa is the protocol best suited to obtain functionally stable eTregs. However, a wealth of data indicates that Tregs might be impaired in their number and/or function in autoimmune diseases, including RA (7, 1821). Thus, ex vivo expansion of Tregs from autoimmune patients may result in low yields or functionally compromised cells. To investigate whether this expansion strategy is effective and feasible in arthritis, we expanded Tregs from RA patients using rapa. Importantly, the magnitude of expansion (Fig. 6A) and the viability (Fig. 6B) of eTregs from RA patients was indistinguishable from those obtained from HD. Moreover, as for HD rapa eTregs, FOXP3 expression (Fig. 6C), suppressive ability (Fig. 6D) and TSDR hypomethylation profile (Fig. 6E, 6F) were stably maintained during the whole culture, even when rapa was withdrawn (i.e., during the restimulation, days 14–28). Results were highly consistent regardless of differences in gender, age, time from disease onset, clinical activity, or treatment (Table I).

FIGURE 6.

Rapa eTregs from RA patients are stable and suppressive. (A) Fold expansion of eTregs cultured in the presence of rapa for 14 d. (B) Viability at day 14. (C) FOXP3 expression in eTregs expanded in the presence of rapa at day 14 or after rapa-free restimulation at day 28. Numbers indicate the percentage of FOXP3+ cells gated on viable cells. (D) Suppressive ability of eTregs expanded and restimulated as in (B). (E) Methylation percentage of each CpG site of the TSDR in Tregs expanded and restimulated as in (B) in representative male and female RA patients and HD. (F) Average TSDR methylation, segregated by gender. Female Tregs display ∼50% methylation because of X-linked inactivation. Throughout the figure, data from 9 to 12 RA patients and 8 to 17 HD collected over five independent experiments are shown.

FIGURE 6.

Rapa eTregs from RA patients are stable and suppressive. (A) Fold expansion of eTregs cultured in the presence of rapa for 14 d. (B) Viability at day 14. (C) FOXP3 expression in eTregs expanded in the presence of rapa at day 14 or after rapa-free restimulation at day 28. Numbers indicate the percentage of FOXP3+ cells gated on viable cells. (D) Suppressive ability of eTregs expanded and restimulated as in (B). (E) Methylation percentage of each CpG site of the TSDR in Tregs expanded and restimulated as in (B) in representative male and female RA patients and HD. (F) Average TSDR methylation, segregated by gender. Female Tregs display ∼50% methylation because of X-linked inactivation. Throughout the figure, data from 9 to 12 RA patients and 8 to 17 HD collected over five independent experiments are shown.

Close modal

Overall, these data demonstrate the feasibility of eTreg cell therapy in rheumatological autoimmune diseases.

We further investigated whether rapa eTregs from RA patients retained all the typical features of their counterparts from HD, such as mechanism of suppression and resistance to inflammatory mediators. Rapa eTregs from RA patients were able to suppress Tconv cytokine production (Fig. 7A), required the same contact-dependent and soluble signals for optimal suppression (Fig. 7B–D), and retained their suppressive function in the presence of TNF-α (Fig. 7E), similarly to rapa eTregs from HD. Thus, rapa eTregs from RA patients and HD are indistinguishable in terms of functionality and stability in a proinflammatory environment.

FIGURE 7.

Rapa eTregs from RA patients suppress via both surface and soluble factors and are stable in a proinflammatory environment. (A) Suppression of Tconv proinflammatory cytokine production by rapa eTregs, expanded from RA patients or HD as in Fig. 1. (BD) Suppression of Tconv proliferation by rapa eTregs in a Transwell system (B) or in the presence of the indicated neutralizing Abs (C and D). In (C), T cells were stimulated with allogeneic LPS-activated DC. (E) Suppression of Tconv proliferation by rapa eTregs in the presence of TNF-α. Throughout the figure, data from 7 to 10 RA patients and 8 to 17 HD collected over five independent experiments are shown.

FIGURE 7.

Rapa eTregs from RA patients suppress via both surface and soluble factors and are stable in a proinflammatory environment. (A) Suppression of Tconv proinflammatory cytokine production by rapa eTregs, expanded from RA patients or HD as in Fig. 1. (BD) Suppression of Tconv proliferation by rapa eTregs in a Transwell system (B) or in the presence of the indicated neutralizing Abs (C and D). In (C), T cells were stimulated with allogeneic LPS-activated DC. (E) Suppression of Tconv proliferation by rapa eTregs in the presence of TNF-α. Throughout the figure, data from 7 to 10 RA patients and 8 to 17 HD collected over five independent experiments are shown.

Close modal

To achieve a sizable clinical benefit, Treg cell therapy should be effective within the inflamed microenvironment where tissue damage occurs. Unfortunately, T cells from the inflamed tissues might be resistant to Treg-mediated suppression (21). Therefore, we tested whether rapa eTregs were able to suppress the proliferation of T cells isolated from the affected synovium. Ex vivo Tregs from RA patients almost completely lost their suppressive ability when synovial rather than blood Tconv were used as responders. By contrast, rapa eTregs equally suppressed synovial and blood Tconv (Fig. 8), demonstrating that rapa conditioning of Tregs can overcome synovial T cell resistance.

FIGURE 8.

Rapa eTregs from RA patients suppress synovial and blood Tconv equally. Inhibition of suppression of synovial Tconv by ex vivo Tregs or rapa eTregs, relative to blood Tconv. n = 7 per condition over two independent experiments. **p < 0.01.

FIGURE 8.

Rapa eTregs from RA patients suppress synovial and blood Tconv equally. Inhibition of suppression of synovial Tconv by ex vivo Tregs or rapa eTregs, relative to blood Tconv. n = 7 per condition over two independent experiments. **p < 0.01.

Close modal

Overall, our data indicate that Treg expansion using rapa is the most viable strategy to obtain functionally competent and stably committed suppressive eTregs for the treatment of autoimmune rheumatological diseases.

Treg therapy is an attractive approach to treat diverse severe conditions, including refractory autoimmune diseases. Ex vivo expansion is required to get therapeutically meaningful Treg numbers. However, inherent problems related to the stability of cultured Tregs and the lack of standardization of the technology represent significant hurdles to the successful application of this approach. This urgent issue is prompting large-scale studies such as the European The One (37), underscoring that currently available information is fragmented and insufficient. In addition, none of the proposed protocols has ever been tested in autoimmune rheumatological settings, notwithstanding the interest in the field.

To address these needs, we systematically compared candidate protocols for Treg expansion/induction using cells from human healthy controls. We found that expansion of ex vivo Tregs in the presence of rapa is far superior to other approaches, including de novo differentiation of iTregs, to obtain functionally stable cells. In addition, truly suppressive human iTregs do not acquire the typical TSDR demethylation signature of ex vivo regulatory cells, as shown in in vitro–derived mouse iTregs. Moreover, the degree of TSDR methylation predicts Treg stability upon secondary TCR engagement, thus providing an essential biomarker for the clinical development of this technology. Finally, Tregs can be efficiently expanded in the presence of rapa without compromising their regulatory potential and stability even from autoimmune samples (i.e., RA), where this subset might be compromised. Importantly, eTregs from RA patients retain all the characteristics observed in HD eTregs, including their mechanism of action and their resistance to TNF-α.

The first attempts at expanding clinical grade bead-enriched CD4+CD25high Tregs employed polyclonal stimulation in the presence of IL-2 (17). A clinical trial for the feasibility and safety of so-expanded cord blood Tregs has been completed with promising results (38). However, under these conditions, the few contaminating Tconv may outcompete eTregs during the culture, hindering their safety and efficacy. Although sorting of Tregs as CD4+CD25highCD127low/− is an obvious improvement (39, 40), the few FOXP3 contaminants may still overcome genuine Tregs during the culture (Figs. 1, 2). Rapa was added to counteract the outgrowth of contaminant Tconv, which are more sensitive to rapa-mediated immunosuppression than Tregs (41, 42). Importantly, no skewing of the TCR repertoire was observed upon Treg expansion in the presence or rapa (43). In addition, rapa eTregs preserved all the characteristics of ex vivo Tregs, including the suppression of both proliferation and cytokine production by Tconv, through both surface and secreted molecules (Fig. 5).

More recently, ATRA was investigated as an additional factor potentially improving eTreg functionality (9). In our hands, both rapa and ATRA/rapa eTregs retained functional and epigenetic stability (Fig. 4). Without any obvious benefit from ATRA, rapa-only Treg expansion would seem a more reasonable choice, in accordance with the parsimony principle. However, as ATRA confers a peculiar chemokine receptor profile to eTregs (36), it might be introduced whenever reinfused cells must be redirected to specific target organs, such as the intestine.

As an alternative to the ex vivo expansion of Tregs, de novo generation of iTregs from naive T cells has been proposed, yielding higher numbers of Tregs. Murine iTregs were first generated in vitro in the presence of TGF-β; however, TGF-β–derived iTregs do not seem stable in vivo (4446). Moreover, human TGF-β–treated T cells are not suppressive (47), yet they are sometimes referred to as iTregs, generating confusion in the field (26). Subsequent refinements added rapa (11) or ATRA (12) to TGF-β–only protocols, but data on functional and epigenetic stability of ensuing iTregs are lacking. In this study, we demonstrate that iTreg differentiation protocols are efficient in generating and expanding iTregs, but their regulatory features are rapidly lost without differentiation factors, strongly cautioning against their clinical application (Figs. 1, 2).

It has been suggested that critical quality control parameters of in vitro–generated or ex vivo–expanded Tregs should comprise FOXP3 expression and suppressive ability (48). However, at the end of primary differentiation, iTregs were comparable to rapa eTregs in terms of suppressive ability, only to deteriorate over the course of secondary expansion without differentiating factors. eTregs cultured with IL-2 only displayed a similar trend (Fig. 2). Therefore, high FOXP3 expression and suppressive ability are not sufficient to ensure long-term stability of cultured cells. By concurrent analysis of TSDR epigenetic profile, we demonstrate that TSDR hypermethylation in iTregs correlates with their instability. Specifically, a longitudinal analysis showed that a demethylated TSDR (<20% in bulk cultures) predicts long-term Treg functionality, making it a reliable readout for effective expansion of stable Tregs (Fig. 3). Therefore, we recommend assessing TSDR demethylation as a critical quality control step before reinfusion of eTregs into patients.

The TSDR methylation data also shed light on the mechanisms leading to loss of regulatory features by both iTregs and IL-2–only eTregs. Indeed, two distinct effects may cooperate in the gradual loss of FOXP3+ cells, which in turn would determine reduced suppressive ability: either FOXP3+ cells may progressively lose FOXP3 expression, or the contaminating FOXP3 cells may outcompete FOXP3+ cells during in vitro proliferation. In iTregs, FOXP3 expression is sustained by differentiation factors but it is epigenetically unstable because of hypermethylated TSDR. Therefore, it is conceivable that withdrawal of exogenous factors might lead to loss of both FOXP3 expression and suppressive ability. In eTregs expanded with IL-2 only, the higher methylation of the bulk culture compared with sorted FOXP3+ cells at the end of the first round of expansion indicates that FOXP3 cells are hypermethylated Tconv, expanded from the few contaminating cells introduced at the time of the initial sorting; after restimulation, FOXP3+ cells also acquire TSDR methylation, suggesting that even former FOXP3+ cells are unstable under these culture conditions, and may subsequently reduce both FOXP3 expression and suppressive ability. Therefore, in IL-2–only eTreg cultures FOXP3 loss is likely to occur as a result of both intrinsic FOXP3 downregulation by former FOXP3+ cells and extrinsic out-competition by FOXP3 cells. Because the rates of these processes varies from subject to subject, cultures of cells expanded without rapa are highly heterogenous in their final content of suppressive cells. Rapa preferentially hinders Tconv proliferation, thereby selecting for Tregs. Accordingly, all rapa-conditioned cultures consistently yielded highly suppressive, FOXP3-expressing Tregs. Ultimately, the addition of rapa improves both the potential clinical efficacy and the safety of Treg therapy.

Rapa counteracts both mechanisms leading to the loss of FOXP3+ Tregs in culture, as it both depresses the outgrowth of contaminating Tconv and stabilizes FOXP3 expression through epigenetic mechanisms. Indeed, rapa affects T cell proliferation and differentiation by blocking the activity of the “mammalian target of rapamycin” (mTOR), a critical molecule governing costimulation pathways. Upon TCR engagement, Tconv but not Tregs downregulate the expression of PTEN, an inhibitor of the PI3K/AKT/mTOR pathway, which enables Tconv to proliferate (41). By contrast, Treg proliferation is dependent on the FOXP3-driven constitutive expression of Pim 2 (42). The relative independence of Tregs from mTOR for their proliferation confers them a replicative advantage over Tconv in rapa-conditioned cultures. In addition, constitutive AKT activation hinders the stable induction of FOXP3 expression by sequestering the FOXO transcription factors into the cytoplasm and by antagonizing SMAD signaling. Rapa-mediated inhibition of the AKT/mTOR pathway may thus facilitate stable FOXP3 expression (49). Finally, rapa also reduces the activity of DNMT1 in in vivo peripherally derived Tregs in the mouse, thus potentially enabling stable TSDR demethylation (50).

In addition to its stabilizing role, rapa also boosts Treg functionality. Indeed, rapa eTregs from healthy controls are more potent suppressors, and retain much higher suppressive function in a proinflammatory (i.e., TNF-α-rich) environment, compared with ex vivo Tregs (Fig. 5). Importantly, this boosting effect helps bypass the impaired Treg function often reported in autoimmune diseases (7, 1821), as rapa eTregs from RA patients are functionally identical to those obtained from healthy controls (Fig. 7). In addition, rapa eTregs suppress synovial T cell proliferation much more efficiently than their ex vivo counterparts (Fig. 8). The finding that rapa eTregs are still able to suppress both in the presence of TNF-α and when synovial T cells are used as responders is of particular relevance in RA. Indeed, not only TNF-α is enriched in inflamed synovial fluids and plays a major role in disease pathogenesis (18), but synovial T cells are also generally considered resistant to Treg-mediated suppression (21). Rapa makes eTregs resistant to both factors, which might otherwise undermine attempts to an effective Treg-based therapy.

The stabilizing effect of rapa is also evident in vivo. Rapa is routinely used in the clinics to prevent allograft rejection (51) and is currently being tested as a therapy for autoimmunity (52, 53). In most reports, the clinical formulation of rapa (Sirolimus) boosts both number and function of Tregs (5355). This is especially important in view of recent mouse data showing that self-reactive Tregs might turn into pathogenic effector T cells upon Ag-specific stimulation in a proinflammatory environment (24). Although no signs of conversion or pathogenicity have been reported for expanded Tregs in either animal models or patients, and despite the resistance of rapa eTregs to TNF-α (Fig. 7), adjunct therapy with rapa or other drugs might be considered as an addition precaution to ward off this danger.

By using the very same protocol we selected for RA patients, Tregs can be restimulated multiple times in vitro to increase the yield to levels suitable for cell therapy applications (10, 43). However, extended exposure to rapa might impair Treg proliferation (8, 10, 55). Of note, in our hands a single round of rapa treatment was sufficient to obtain stable regulatory cells (Fig. 2). Thus, further protocol optimizations might include restimulating rapa eTregs with reduced or intermittent addition of rapa, or with no rapa at all, to improve yields and achieve sufficient numbers for reinfusion without impairing subsequent in vivo Treg expansion (55). For more demanding applications, such as pediatric rheumatology, where the amount of the starting material is limited, further improvements might be considered, such as replacing beads with APCs supporting more efficient Treg expansion (10).

In conclusion, we show that the degree of FOXP3 demethylation in the TSDR, an exclusive feature of eTregs as compared with iTregs, can predict Treg stability. In addition, Treg expansion with rapa overcomes the Treg impairment observed in autoimmune diseases, boosting Treg functionality in normal and proinflammatory conditions. As such, rapa eTregs should be preferred over iTregs in suppressive cell therapy efforts for rheumatological autoimmune diseases, and TSDR methylation should be implemented in clinical practice as a quality control check in Treg generation. By providing a standardized protocol for the immunotherapy of autoimmune diseases, this work bears an immediate translational value. The already demonstrated safety profile of Treg therapy (38, 56) may encourage its application beyond rare untreatable cases. Treg therapy in combination with currently deployed biologics may benefit less severe yet relapsing cases, thus opening a new frontier in personalized medicine.

We thank Drs. Gary W. Williams, Dana Copeland, and Ken Pischel for the recruitment of RA patients at TSRI.

This work was supported by National Institute of Arthritis and Musculoskeletal and Skin Diseases and the Bartman Foundation. J.v.L. was supported by the Dutch Arthritis Foundation. Clinical work was partially funded by the Estate of Tan Sri Khoo Teck Puat (Khoo Clinical Scholars Pilot Award).

Abbreviations used in this article:

ATRA

all-trans–retinoic acid

DC

dendritic cell

eTreg

expanded Treg

HD

healthy donor

iTreg

inducible Treg

RA

rheumatoid arthritis

rapa

rapamycin

rh

recombinant human

Tconv

conventional T cell

Treg

regulatory T cell

TSDR

Treg-specific demethylated region

TSRI

The Scripps Research Institute.

1
Josefowicz
S. Z.
,
Lu
L. F.
,
Rudensky
A. Y.
.
2012
.
Regulatory T cells: mechanisms of differentiation and function.
Annu. Rev. Immunol.
30
:
531
564
.
2
Hori
S.
,
Nomura
T.
,
Sakaguchi
S.
.
2003
.
Control of regulatory T cell development by the transcription factor Foxp3.
Science
299
:
1057
1061
.
3
Williams
L. M.
,
Rudensky
A. Y.
.
2007
.
Maintenance of the Foxp3-dependent developmental program in mature regulatory T cells requires continued expression of Foxp3.
Nat. Immunol.
8
:
277
284
.
4
Fontenot
J. D.
,
Gavin
M. A.
,
Rudensky
A. Y.
.
2003
.
Foxp3 programs the development and function of CD4+CD25+ regulatory T cells.
Nat. Immunol.
4
:
330
336
.
5
Leguern
C.
2011
.
Regulatory T cells for tolerance therapy: revisiting the concept.
Crit. Rev. Immunol.
31
:
189
207
.
6
Rosenblum
M. D.
,
Gratz
I. K.
,
Paw
J. S.
,
Abbas
A. K.
.
2012
.
Treating human autoimmunity: current practice and future prospects.
Sci. Transl. Med.
4
:
125sr121
.
7
Valencia
X.
,
Yarboro
C.
,
Illei
G.
,
Lipsky
P. E.
.
2007
.
Deficient CD4+CD25high T regulatory cell function in patients with active systemic lupus erythematosus.
J. Immunol.
178
:
2579
2588
.
8
Battaglia
M.
,
Stabilini
A.
,
Tresoldi
E.
.
2012
.
Expanding human T regulatory cells with the mTOR-inhibitor rapamycin.
Methods Mol. Biol.
821
:
279
293
.
9
Golovina
T. N.
,
Mikheeva
T.
,
Brusko
T. M.
,
Blazar
B. R.
,
Bluestone
J. A.
,
Riley
J. L.
.
2011
.
Retinoic acid and rapamycin differentially affect and synergistically promote the ex vivo expansion of natural human T regulatory cells.
PLoS One
6
:
e15868
.
10
Hippen
K. L.
,
Merkel
S. C.
,
Schirm
D. K.
,
Sieben
C. M.
,
Sumstad
D.
,
Kadidlo
D. M.
,
McKenna
D. H.
,
Bromberg
J. S.
,
Levine
B. L.
,
Riley
J. L.
, et al
.
2011
.
Massive ex vivo expansion of human natural regulatory T cells (T(regs)) with minimal loss of in vivo functional activity.
Sci. Transl. Med.
3
:
83ra41
.
11
Hippen
K. L.
,
Merkel
S. C.
,
Schirm
D. K.
,
Nelson
C.
,
Tennis
N. C.
,
Riley
J. L.
,
June
C. H.
,
Miller
J. S.
,
Wagner
J. E.
,
Blazar
B. R.
.
2011
.
Generation and large-scale expansion of human inducible regulatory T cells that suppress graft-versus-host disease.
Am. J. Transplant.
11
:
1148
1157
.
12
Wang
J.
,
Huizinga
T. W.
,
Toes
R. E.
.
2009
.
De novo generation and enhanced suppression of human CD4+CD25+ regulatory T cells by retinoic acid.
J. Immunol.
183
:
4119
4126
.
13
Ohkura
N.
,
Kitagawa
Y.
,
Sakaguchi
S.
.
2013
.
Development and maintenance of regulatory T cells.
Immunity
38
:
414
423
.
14
Allan
S. E.
,
Crome
S. Q.
,
Crellin
N. K.
,
Passerini
L.
,
Steiner
T. S.
,
Bacchetta
R.
,
Roncarolo
M. G.
,
Levings
M. K.
.
2007
.
Activation-induced FOXP3 in human T effector cells does not suppress proliferation or cytokine production.
Int. Immunol.
19
:
345
354
.
15
Wang
J.
,
Ioan-Facsinay
A.
,
van der Voort
E. I.
,
Huizinga
T. W.
,
Toes
R. E.
.
2007
.
Transient expression of FOXP3 in human activated nonregulatory CD4+ T cells.
Eur. J. Immunol.
37
:
129
138
.
16
Hoffmann
P.
,
Boeld
T. J.
,
Eder
R.
,
Huehn
J.
,
Floess
S.
,
Wieczorek
G.
,
Olek
S.
,
Dietmaier
W.
,
Andreesen
R.
,
Edinger
M.
.
2009
.
Loss of FOXP3 expression in natural human CD4+CD25+ regulatory T cells upon repetitive in vitro stimulation.
Eur. J. Immunol.
39
:
1088
1097
.
17
Ukena
S. N.
,
Höpting
M.
,
Velaga
S.
,
Ivanyi
P.
,
Grosse
J.
,
Baron
U.
,
Ganser
A.
,
Franzke
A.
.
2011
.
Isolation strategies of regulatory T cells for clinical trials: phenotype, function, stability, and expansion capacity.
Exp. Hematol.
39
:
1152
1160
.
18
Ehrenstein
M. R.
,
Evans
J. G.
,
Singh
A.
,
Moore
S.
,
Warnes
G.
,
Isenberg
D. A.
,
Mauri
C.
.
2004
.
Compromised function of regulatory T cells in rheumatoid arthritis and reversal by anti-TNFα therapy.
J. Exp. Med.
200
:
277
285
.
19
Lawson
C. A.
,
Brown
A. K.
,
Bejarano
V.
,
Douglas
S. H.
,
Burgoyne
C. H.
,
Greenstein
A. S.
,
Boylston
A. W.
,
Emery
P.
,
Ponchel
F.
,
Isaacs
J. D.
.
2006
.
Early rheumatoid arthritis is associated with a deficit in the CD4+CD25high regulatory T cell population in peripheral blood.
Rheumatology
45
:
1210
1217
.
20
Venken
K.
,
Hellings
N.
,
Broekmans
T.
,
Hensen
K.
,
Rummens
J. L.
,
Stinissen
P.
.
2008
.
Natural naive CD4+CD25+CD127low regulatory T cell (Treg) development and function are disturbed in multiple sclerosis patients: recovery of memory Treg homeostasis during disease progression.
J. Immunol.
180
:
6411
6420
.
21
Buckner
J. H.
2010
.
Mechanisms of impaired regulation by CD4+CD25+FOXP3+ regulatory T cells in human autoimmune diseases.
Nat. Rev. Immunol.
10
:
849
859
.
22
Dominguez-Villar
M.
,
Baecher-Allan
C. M.
,
Hafler
D. A.
.
2011
.
Identification of T helper type 1-like, Foxp3+ regulatory T cells in human autoimmune disease.
Nat. Med.
17
:
673
675
.
23
McClymont
S. A.
,
Putnam
A. L.
,
Lee
M. R.
,
Esensten
J. H.
,
Liu
W.
,
Hulme
M. A.
,
Hoffmüller
U.
,
Baron
U.
,
Olek
S.
,
Bluestone
J. A.
,
Brusko
T. M.
.
2011
.
Plasticity of human regulatory T cells in healthy subjects and patients with type 1 diabetes.
J. Immunol.
186
:
3918
3926
.
24
Bailey-Bucktrout
S. L.
,
Martinez-Llordella
M.
,
Zhou
X.
,
Anthony
B.
,
Rosenthal
W.
,
Luche
H.
,
Fehling
H. J.
,
Bluestone
J. A.
.
2013
.
Self-antigen-driven activation induces instability of regulatory T cells during an inflammatory autoimmune response.
Immunity
39
:
949
962
.
25
Spreafico
R.
,
Rossetti
M.
,
van den Broek
T.
,
Jansen
N. J.
,
Zhang
H.
,
Moshref
M.
,
Prakken
B.
,
van Loosdregt
J.
,
van Wijk
F.
,
Albani
S.
.
2014
.
A sensitive protocol for FOXP3 epigenetic analysis in scarce human samples.
Eur. J. Immunol.
44
:
3141
3143
.
26
Baron
U.
,
Floess
S.
,
Wieczorek
G.
,
Baumann
K.
,
Grützkau
A.
,
Dong
J.
,
Thiel
A.
,
Boeld
T. J.
,
Hoffmann
P.
,
Edinger
M.
, et al
.
2007
.
DNA demethylation in the human FOXP3 locus discriminates regulatory T cells from activated FOXP3+ conventional T cells.
Eur. J. Immunol.
37
:
2378
2389
.
27
Nagar
M.
,
Vernitsky
H.
,
Cohen
Y.
,
Dominissini
D.
,
Berkun
Y.
,
Rechavi
G.
,
Amariglio
N.
,
Goldstein
I.
.
2008
.
Epigenetic inheritance of DNA methylation limits activation-induced expression of FOXP3 in conventional human CD25CD4+ T cells.
Int. Immunol.
20
:
1041
1055
.
28
Stockis
J.
,
Fink
W.
,
François
V.
,
Connerotte
T.
,
de Smet
C.
,
Knoops
L.
,
van der Bruggen
P.
,
Boon
T.
,
Coulie
P. G.
,
Lucas
S.
.
2009
.
Comparison of stable human Treg and Th clones by transcriptional profiling.
Eur. J. Immunol.
39
:
869
882
.
29
Miyara
M.
,
Yoshioka
Y.
,
Kitoh
A.
,
Shima
T.
,
Wing
K.
,
Niwa
A.
,
Parizot
C.
,
Taflin
C.
,
Heike
T.
,
Valeyre
D.
, et al
.
2009
.
Functional delineation and differentiation dynamics of human CD4+ T cells expressing the FoxP3 transcription factor.
Immunity
30
:
899
911
.
30
Floess
S.
,
Freyer
J.
,
Siewert
C.
,
Baron
U.
,
Olek
S.
,
Polansky
J.
,
Schlawe
K.
,
Chang
H. D.
,
Bopp
T.
,
Schmitt
E.
, et al
.
2007
.
Epigenetic control of the foxp3 locus in regulatory T cells.
PLoS Biol.
5
:
e38
.
31
Kim
H. P.
,
Leonard
W. J.
.
2007
.
CREB/ATF-dependent T cell receptor-induced FoxP3 gene expression: a role for DNA methylation.
J. Exp. Med.
204
:
1543
1551
.
32
Lal
G.
,
Zhang
N.
,
van der Touw
W.
,
Ding
Y.
,
Ju
W.
,
Bottinger
E. P.
,
Reid
S. P.
,
Levy
D. E.
,
Bromberg
J. S.
.
2009
.
Epigenetic regulation of Foxp3 expression in regulatory T cells by DNA methylation.
J. Immunol.
182
:
259
273
.
33
Zheng
Y.
,
Josefowicz
S.
,
Chaudhry
A.
,
Peng
X. P.
,
Forbush
K.
,
Rudensky
A. Y.
.
2010
.
Role of conserved non-coding DNA elements in the Foxp3 gene in regulatory T-cell fate.
Nature
463
:
808
812
.
34
Polansky
J. K.
,
Schreiber
L.
,
Thelemann
C.
,
Ludwig
L.
,
Krüger
M.
,
Baumgrass
R.
,
Cording
S.
,
Floess
S.
,
Hamann
A.
,
Huehn
J.
.
2010
.
Methylation matters: binding of Ets-1 to the demethylated Foxp3 gene contributes to the stabilization of Foxp3 expression in regulatory T cells.
J. Mol. Med.
88
:
1029
1040
.
35
Ohkura
N.
,
Hamaguchi
M.
,
Morikawa
H.
,
Sugimura
K.
,
Tanaka
A.
,
Ito
Y.
,
Osaki
M.
,
Tanaka
Y.
,
Yamashita
R.
,
Nakano
N.
, et al
.
2012
.
T cell receptor stimulation-induced epigenetic changes and Foxp3 expression are independent and complementary events required for Treg cell development.
Immunity
37
:
785
799
.
36
Scotta
C.
,
Esposito
M.
,
Fazekasova
H.
,
Fanelli
G.
,
Edozie
F. C.
,
Ali
N.
,
Xiao
F.
,
Peakman
M.
,
Afzali
B.
,
Sagoo
P.
,
Lechler
R. I.
,
Lombardi
G.
.
2013
.
Differential effects of rapamycin and retinoic acid on expansion, stability and suppressive qualities of human CD4+CD25+FOXP3+ Treg subpopulations.
Haematologica
98
:
1291
1299
.
37
Leslie
M.
2011
.
Immunology: regulatory T cells get their chance to shine.
Science
332
:
1020
1021
.
38
Brunstein
C. G.
,
Miller
J. S.
,
Cao
Q.
,
McKenna
D. H.
,
Hippen
K. L.
,
Curtsinger
J.
,
Defor
T.
,
Levine
B. L.
,
June
C. H.
,
Rubinstein
P.
, et al
.
2011
.
Infusion of ex vivo expanded T regulatory cells in adults transplanted with umbilical cord blood: safety profile and detection kinetics.
Blood
117
:
1061
1070
.
39
Liu
W.
,
Putnam
A. L.
,
Xu-Yu
Z.
,
Szot
G. L.
,
Lee
M. R.
,
Zhu
S.
,
Gottlieb
P. A.
,
Kapranov
P.
,
Gingeras
T. R.
,
Fazekas de St Groth
B.
, et al
.
2006
.
CD127 expression inversely correlates with FoxP3 and suppressive function of human CD4+ T reg cells.
J. Exp. Med.
203
:
1701
1711
.
40
Marek-Trzonkowska
N.
,
Mysliwiec
M.
,
Dobyszuk
A.
,
Grabowska
M.
,
Techmanska
I.
,
Juscinska
J.
,
Wujtewicz
M. A.
,
Witkowski
P.
,
Mlynarski
W.
,
Balcerska
A.
, et al
.
2012
.
Administration of CD4+CD25highCD127 regulatory T cells preserves β-cell function in type 1 diabetes in children.
Diabetes Care
35
:
1817
1820
.
41
Zeiser
R.
,
Leveson-Gower
D. B.
,
Zambricki
E. A.
,
Kambham
N.
,
Beilhack
A.
,
Loh
J.
,
Hou
J. Z.
,
Negrin
R. S.
.
2008
.
Differential impact of mammalian target of rapamycin inhibition on CD4+CD25+Foxp3+ regulatory T cells compared with conventional CD4+ T cells.
Blood
111
:
453
462
.
42
Basu
S.
,
Golovina
T.
,
Mikheeva
T.
,
June
C. H.
,
Riley
J. L.
.
2008
.
Cutting edge: Foxp3-mediated induction of pim 2 allows human T regulatory cells to preferentially expand in rapamycin.
J. Immunol.
180
:
5794
5798
.
43
Tresoldi
E.
,
Dell’Albani
I.
,
Stabilini
A.
,
Jofra
T.
,
Valle
A.
,
Gagliani
N.
,
Bondanza
A.
,
Roncarolo
M. G.
,
Battaglia
M.
.
2011
.
Stability of human rapamycin-expanded CD4+CD25+ T regulatory cells.
Haematologica
96
:
1357
1365
.
44
Selvaraj
R. K.
,
Geiger
T. L.
.
2007
.
A kinetic and dynamic analysis of Foxp3 induced in T cells by TGF-β. [Erratum in 2007 J. Immunol. 179: 1390].
J. Immunol.
179
:
11 p following 1390
.
45
Chen
Q.
,
Kim
Y. C.
,
Laurence
A.
,
Punkosdy
G. A.
,
Shevach
E. M.
.
2011
.
IL-2 controls the stability of Foxp3 expression in TGF-β–induced Foxp3+ T cells in vivo.
J. Immunol.
186
:
6329
6337
.
46
Beres
A.
,
Komorowski
R.
,
Mihara
M.
,
Drobyski
W. R.
.
2011
.
Instability of Foxp3 expression limits the ability of induced regulatory T cells to mitigate graft versus host disease.
Clin. Cancer Res.
17
:
3969
3983
.
47
Tran
D. Q.
,
Ramsey
H.
,
Shevach
E. M.
.
2007
.
Induction of FOXP3 expression in naive human CD4+FOXP3 T cells by T-cell receptor stimulation is transforming growth factor-β dependent but does not confer a regulatory phenotype.
Blood
110
:
2983
2990
.
48
Hoffmann
P.
,
Eder
R.
,
Edinger
M.
.
2011
.
Polyclonal expansion of human CD4+CD25+ regulatory T cells.
Methods Mol. Biol.
677
:
15
30
.
49
von Boehmer
H.
,
Daniel
C.
.
2013
.
Therapeutic opportunities for manipulating T(Reg) cells in autoimmunity and cancer.
Nat. Rev. Drug Discov.
12
:
51
63
.
50
Daniel
C.
,
Wennhold
K.
,
Kim
H. J.
,
von Boehmer
H.
.
2010
.
Enhancement of antigen-specific Treg vaccination in vivo.
Proc. Natl. Acad. Sci. USA
107
:
16246
16251
.
51
Lange
C. M.
,
Tran
T. Y.
,
Farnik
H.
,
Jungblut
S.
,
Born
T.
,
Wagner
T. O.
,
Hirche
T. O.
.
2010
.
Increased frequency of regulatory T cells and selection of highly potent CD62L+ cells during treatment of human lung transplant recipients with rapamycin.
Transpl. Int.
23
:
266
276
.
52
Fernandez
D.
,
Bonilla
E.
,
Mirza
N.
,
Niland
B.
,
Perl
A.
.
2006
.
Rapamycin reduces disease activity and normalizes T cell activation-induced calcium fluxing in patients with systemic lupus erythematosus.
Arthritis Rheum.
54
:
2983
2988
.
53
Monti
P.
,
Scirpoli
M.
,
Maffi
P.
,
Piemonti
L.
,
Secchi
A.
,
Bonifacio
E.
,
Roncarolo
M. G.
,
Battaglia
M.
.
2008
.
Rapamycin monotherapy in patients with type 1 diabetes modifies CD4+CD25+FOXP3+ regulatory T-cells.
Diabetes
57
:
2341
2347
.
54
Akimova
T.
,
Kamath
B. M.
,
Goebel
J. W.
,
Meyers
K. E.
,
Rand
E. B.
,
Hawkins
A.
,
Levine
M. H.
,
Bucuvalas
J. C.
,
Hancock
W. W.
.
2012
.
Differing effects of rapamycin or calcineurin inhibitor on T-regulatory cells in pediatric liver and kidney transplant recipients.
Am. J. Transplant.
12
:
3449
3461
.
55
Singh
K.
,
Kozyr
N.
,
Stempora
L.
,
Kirk
A. D.
,
Larsen
C. P.
,
Blazar
B. R.
,
Kean
L. S.
.
2012
.
Regulatory T cells exhibit decreased proliferation but enhanced suppression after pulsing with sirolimus.
Am. J. Transplant.
12
:
1441
1457
.
56
Desreumaux
P.
,
Foussat
A.
,
Allez
M.
,
Beaugerie
L.
,
Hebuterne
X.
,
Bouhnik
Y.
,
Nachury
M.
,
Brun
V.
,
Bastian
H.
,
Belmonte
N.
, et al
.
2012
.
Safety and efficacy of antigen-specific regulatory T-cell therapy for patients with refractory Crohn’s disease.
Gastroenterology
143
:
1207
1217 e1201–e1202
.

The authors have no financial conflicts of interest.