This protocol describes the reproducible generation and phenotyping of human induced regulatory T cells (iTregs) from naïve CD4+ T cells in vitro. Different protocols for FOXP3 induction allow for the study of specific iTreg phenotypes obtained with respective protocols.
Regulatory T cells (Tregs) are an integral part of peripheral tolerance, suppressing immune reactions against self-structures and thus preventing autoimmune diseases. Clinical approaches to adoptively transfer Tregs, or to deplete Tregs in cancer, are underway with promising first outcomes.
Because the number of naturally occurring Tregs (nTregs) is very limited, studying certain Treg features using in vitro induced Tregs (iTregs) can be advantageous. To date, the best although not absolutely specific protein marker to delineate Tregs is the transcription factor FOXP3. Despite the importance of Tregs including non-redundant roles of peripherally induced Tregs, the protocols to generate iTregs are currently controversial, particularly for human cells. This protocol therefore describes the in vitro differentiation of human CD4+FOXP3+ iTregs from human naïve T cells using a range of Treg-inducing factors (TGF-β plus IL-2 only, or their combination with retinoic acid, rapamycin or butyrate) in parallel. It also describes the phenotyping of these cells by flow cytometry and qRT-PCR.
These protocols result in reproducible expression of FOXP3 and other Treg signature genes and enable the study of general FOXP3-regulatory mechanisms as well as protocol-specific effects to delineate the impact of certain factors. iTregs can be utilized to study various phenotypic aspects as well as molecular mechanisms of Treg induction. Detailed molecular studies are facilitated by relatively large cell numbers that can be obtained.
A limitation for the application of iTregs is the relative instability of FOXP3 expression in these cells compared to nTregs. iTregs generated by these protocols can also be used for functional assays such as studying their suppressive function, in which iTregs induced by TGF-β plus retinoic acid and rapamycin display superior suppressive activity. However, the suppressive capacity of iTregs can differ from nTregs and the use of appropriate controls is crucial.
CD4+CD25+FOXP3+ regulatory T cells (Tregs) suppress other immune cells and are critical mediators of peripheral tolerance, preventing autoimmunity and excessive inflammation1. The importance of Tregs is exemplified by the human disease immunodysregulation polyendocrinopathy enteropathy X-linked syndrome (IPEX), in which loss of Tregs due to mutations in the `master´ Treg transcription factor forkhead box P3 (FOXP3) leads to severe systemic autoimmune disease, lethal at an early age. However, Tregs act as a double-edged sword in the immune system as they can also hamper anti-tumor immunity in certain settings2. Therapeutic manipulation of Treg number and function is therefore subject to numerous clinical investigations. In cancer, depletion of Tregs can be desirable and some success of clinical approaches encourages further research3. In autoimmune and inflammatory diseases, in addition to therapeutic effects of Tregs in several mouse disease models, recent first in-man trials of adoptive Treg transfer to prevent graft-versus-host disease (GvHD)4–7 and to assess safety in treating type 1 diabetes8 showed very promising outcomes.
Naturally occurring Tregs (nTregs) comprise thymic-derived tTregs and peripherally induced pTregs, with non-redundant essential functions in maintaining health9–11. However, nTreg numbers are limited, encouraging the complementary approach of inducing Tregs (iTregs) in vitro from naïve T cell precursors12. Still stability of iTregs, presumably due to lack of demethylation in the so-called Treg-specific demethylated region (TSDR) in the FOXP3 gene locus13, remains a concern and several studies indicate that in vivo induced Tregs are more stable14.
To date, FOXP3 remains the best protein marker for Tregs but it is not absolutely specific because human conventional CD4+CD25- T cells transiently express intermediate levels of FOXP3 upon activation15,16. Although significant progress has been made in elucidating the regulation of FOXP3 expression, much remains to be discovered regarding the induction, stability and function of FOXP3 particularly in human cells. Despite differences to nTregs, in vitro induced FOXP3+ CD4+ T cells can be used as a model system to study molecular mechanisms of FOXP3 induction and as a starting point to develop protocols in the future that allow for generation of iTregs that are more similar to in vivo generated Tregs, which could be applicable for adoptive transfer strategies in the future.
There is no `gold standard´ protocol to induce human iTregs, and current protocols have been developed based on mimicking Treg-inducing conditions in vivo: interleukin 2 (IL-2) and transforming growth factor β (TGF-β) signaling are crucial for FOXP3 induction in vivo17, and all-trans retinoic acid (ATRA) — which is produced in vivo by gut-associated dendritic cells — is frequently used to enhance FOXP3 induction in vitro18–21. We have developed additional human Treg-inducing protocols using butyrate22, a gut microbiota-derived short-chain fatty acid that was recently shown to augment murine Treg induction23,24. We also recently established a new protocol for generation of iTregs with superior suppressive function in vitro by using a combination of TGF-β, ATRA and rapamycin22, the latter being a clinically approved mammalian target of rapamycin (mTOR) inhibitor that is known to promote FOXP3 maintenance during human Treg expansion25,26.
This method describes the reproducible in vitro generation of human CD4+FOXP3+ iTregs using a set of different conditions, and their subsequent phenotyping by flow cytometry and quantitative reverse transcription polymerase chain reaction (qRT-PCR) to reveal protocol-specific patterns of expression of FOXP3 and other Treg signature molecules such as CD25, CTLA-4, EOS, as well as repression of IFN-γ and SATB1 expression22. The generated cell populations can be used for functional assays regarding suppressive activity or for molecular studies, either concerning general FOXP3 regulators or to study effects specific to certain compounds such as butyrate or rapamycin. Further understanding of molecular mechanisms driving Treg differentiation is highly relevant for future therapeutic approaches in autoimmunity or cancer to specifically target molecules involved in Treg generation and function.
Human peripheral blood mononuclear cells (PBMCs) were freshly isolated from anonymized healthy donor buffy coats purchased from the Karolinska University Hospital, Sweden. Ethical permit for the experiments was obtained from the Regional Ethical Review Board in Stockholm (Regionala etikprövningsnämnden i Stockholm), Sweden (approval number: 2013/1458-31/1).
1. T Cell Isolation from Peripheral Blood
2. Treg Induction Culture
3. Phenotypic Analysis of iTregs by qRT-PCR
4. Phenotypic Analysis by Flow Cytometry
NOTE: This protocol is optimized for staining in 96U well plate format with 3 x 105-1 x 106 cells. If there are less cells per well, start with several wells and pool as described below.
Figure 1 shows a scheme of the experimental setup. Figure 2 shows a representative purity control staining for magnetically isolated naïve CD4+ T cells and nTregs.
Figure 3A shows the flow cytometry gating strategy and Figure 3B shows representative FOXP3 and CD25 flow cytometry stainings on day 6 of culture under the indicated iTreg or control conditions. Upon in vitro stimulation, most cells upregulate CD25, which is reduced in the presence of rapamycin. Only under addition of iTreg-inducing factors, a clear population of FOXP3+ cells becomes apparent, which is also enriched within CD25+ cells under iTreg conditions. Figure 3C depicts the phenotypic appearance of iTreg cultures in the microscope with proliferating cells seen as dark clusters. Each iTreg condition shows a specific and reproducible pattern of cell proliferation: Under these culture conditions, TGF-β increases proliferation slightly, and ATRA further increases proliferation. At the same time, the inhibition of proliferation by rapamycin is apparent by the strongly reduced cluster size. The microscopic appearance of proliferation strength also corresponds to total cell counts (data not included).
Figure 4 displays representative results of qRT-PCR analyses of FOXP3 mRNA induction in iTreg (and control) cultures at different time points. As for FOXP3 protein, FOXP3 mRNA expression is higher in all iTregs compared to control stimulated cells, with rapamycin-treated iTregs having relatively low levels compared to other iTregs. FOXP3 mRNA in nTregs is higher than in iTregs.
Figure 1: Experimental scheme for iTreg induction, nTreg isolation, and Treg analysis. Human naïve CD4+ T cells were isolated from buffy coats and stimulated in serum-free medium with anti-CD3 and anti-CD28 antibodies plus 100 U/ml IL-2 for up to 6 days, either in the absence (`mock control cells´) or presence of different Treg-inducing factors (`iTreg´) as indicated. nTregs are isolated and assayed in parallel. Phenotypic analysis can be done by flow cytometry and qRT-PCR. Modified from Schmidt, A. et al. 201622. Please click here to view a larger version of this figure.
Figure 2: Cell purity of starting populations. Naïve human CD4+ T cells were isolated by the Naïve CD4 T Cell Isolation Kit II, human. Upper panel: Naïve CD4+ T cell purity, based on CD4, CD45RA and CD45RO, was 94 to 98% and the purity for a representative donor of more than 30 is shown (`Tnaïve´). Ex vivo Tregs (`nTreg´) were isolated by using limited amounts of CD25 microbeads and used as a positive control for iTreg experiments. nTreg purity of a representative donor, based on CD25 and CD4, is shown. Lower panel: Intracellular FOXP3 staining (performed as in Figure 3 and pre-gated on live CD4+ cells) is shown for naïve CD4+ T cells and nTregs respectively, using cells from the same donor as in the upper panel. Modified from Schmidt, A. et al. 201622. Please click here to view a larger version of this figure.
Figure 3: Phenotypic appearance of iTregs and nTregs. Human naïve CD4+ T cells were cultured in serum-free medium under the indicated conditions. T cells were stimulated with anti-CD3 and anti-CD28 antibodies plus 100 U/ml IL-2 (`Stim.´). Where indicated, TGF-β1 (`TGF´), rapamycin (`Rapa´), all-trans retinoic acid (`ATRA´) or butyrate were added. nTregs (ex vivo isolated peripheral blood CD25++ cells) were left unstimulated (`unstim.´) and used as positive control. Unstimulated naïve T cells were used as negative control. (A) On day 6 of culture, cells were stained with surface antibodies including anti-CD25 and anti-CD4, then stained with fixable viability dye and subsequently fixed/permeabilized and stained intracellularly with anti-FOXP3 and anti-CTLA-4 or isotype control antibodies. Acquisition and compensation was performed on a Cyan ADP flow cytometer and the data were analyzed with the FlowJo software. The gating strategy is indicated by the red arrows, and the shown example is an iTreg sample induced with TGF+ATRA. (B) Cells were stained as in (A), and FOXP3 and CD25 expression in control T cells and iTregs induced by different protocols is shown. The pseudocolor plots show representative FOXP3 and CD25 stainings for one donor, pre-gated on singlet, live CD4+ cells as in (A). The isotype example is shown for an iTreg (stim.+TGF+ATRA) sample. (C) Representative microscopy images (40X magnification) of individual 96U-plate wells of cells cultured for 4 days. Dark clusters of cells represent proliferating cells. Please click here to view a larger version of this figure.
Figure 4: FOXP3 mRNA expression upon use of different protocols for iTreg differentiation. iTregs were generated as in Figure 3 under the indicated conditions, and at the given time points, cell samples were taken and RNA was extracted. Unstimulated naïve T cells, and unstimulated nTregs, were sampled on day 0. Samples were analyzed by qRT-PCR, and FOXP3 mRNA expression was normalized to RPL13A expression for each sample. FOXP3 mRNA expression in unstimulated naïve T cells was set to 1, and fold change of FOXP3 mRNA was calculated. Shown are mean values and range of PCR replicate wells for a representative donor. Please click here to view a larger version of this figure.
Compound | Stock concentration | Endconcentration | 4x premix concentration | Stock dilution to obtain 4x premix (dilute stock in T cell culture medium) |
Note |
IL-2 | 400 000 IU/mL | 100 IU/mL | 400 IU/mL IL-2 | 1:1000 | |
anti-CD28 | 1 mg/ml | 1 µg/mL | 4 µg/mL | 1:250 | Add within IL-2 4x premix |
TGF-β1 | 100 µg/mL | 5 ng/mL | 20 ng/mL | 1:5000 | |
ATRA | 10 mM | 10 nM | 40 nM | 1:250000 | Pre-dilute stock solution first to enable pipettable volumes |
Rapamycin | 1 mg/mL | 100 ng/mL | 400 ng/mL | 1:2500 | If used together with ATRA, dilute Rapamycin within the 4x ATRA stimulation mix |
Butyrate | 0.908 M | 0.1 mM | 0.4 mM | 1:2270 |
Table 1
The described protocol enables the robust induction of human CD4+FOXP3+ iTregs from human naïve CD4+ T cells. It includes a new protocol that we described recently, using a combination of TGF-β, ATRA and rapamycin, for induction of iTregs with superior in vitro suppressive function22. Compared to other published protocols, another advantage is the induction of different iTreg populations in parallel by different protocols, which enables the direct comparison of effects of certain iTreg-inducing factors, along with control cells that are activated in the presence of IL-2 alone. The described protocols enable reproducible induction of FOXP3 with low donor variation. Naïve CD4+ T cells in this protocol are isolated by magnetic-activated cell sorting, but fluorescence-activated cell sorting is also possible. The expected yield of naïve CD4+ T cells with this protocol is typically between 5-10%, but strongly depends on the donor (age) and appears also lower when high fractions of erythrocytes are present. If an estimate is needed, PBMCs can be stained (see step 1.4) during the monocyte depletion step. Typically, the yield of naïve CD4+ T cells is about half of the "percentage naïve CD4+ T cells of the lymphocyte gate" in the PBMC stain. This protocol uses limited amounts of CD25 beads to obtain CD25-high (nTreg) cells29. However, these are not pure Tregs, but these are just enriched in Tregs to be used as positive control. If pure Tregs are needed, other kits should be used (such as combined with CD4 enrichment and CD127-depletion) or alternatively, CD25+ cells pre-enriched with 8 µl CD25 beads per 107 cells and stained and sorted by fluorescence-activated cell sorting with stringent CD4+CD25++ gating. Also inclusion of other markers, such as CD127 exclusion, should be considered.
It is important to consider that FOXP3 is necessary, but not sufficient to confer Treg identity30. While iTregs can be used to study certain aspects of, for example, FOXP3 regulation, it is important however to note that iTregs differ from nTregs in several aspects. It is therefore crucial to culture nTregs (ideally derived from the same donor) in parallel to iTregs in all assays for comparison. The difference between nTregs and iTregs is exemplified by only partial overlap of the nTreg and iTreg transcriptome as measured in murine iTregs31. Other Treg signature genes in addition to FOXP3 should be measured for this reason, and to ensure discrimination from activation-induced FOXP3 expression in human cells. For example, iTregs should display higher expression of CD25, CTLA-4 and EOS compared to activated T cells while expression of IFN-γ and SATB1 should be low in nTregs and iTregs induced by the protocols described here, as published previously22. Another important major difference between nTregs and iTregs is the lack of stable FOXP3 expression in iTregs, which most likely corresponds to methylation of the TSDR region in the FOXP3 locus in iTregs13. Also on a genome-wide scale, it was described that epigenetic patterns of DNA methylation and histone modifications in murine iTregs do not reflect the patterns found in nTregs30. We previously described that iTregs induced by the here described protocols, in contrast to nTregs, did not exhibit TSDR demethylation. Accordingly, iTregs lost FOXP3 when restimulated, but maintained FOXP3 expression when further cultured in the presence of IL-2 and without restimulation22. Interestingly, iTregs induced by an alternative protocol using M2 macrophage supernatants displayed enhanced FOXP3 stability, despite lack of TSDR demethylation and TGF-β being causative for FOXP3 induction27.
Modification of iTreg-inducing protocols by addition of compounds (such as Vitamin C or hydrogen sulfide) that influence Ten-eleven Translocation (TET) methylcytosine dioxygenase enzymes, as described very recently32–34, may add in stabilizing FOXP3 by affecting DNA methylation. Also, the stimulation strength and timing has an influence on FOXP3 expression and stability35. Along these lines, the use of bead-coupled CD3/CD28 antibodies instead of plate-bound antibodies was shown to increase murine iTreg in vivo suppressive function and stability albeit independent of TSDR demethylation36. Another factor that needs to be considered as a potential source of variation is the use of serum, which contains undefined factors including TGF-β which even from bovine source is 100% cross-reactive with human cells. Also, the source and activity of IL-2 can drastically influence the results of FOXP3 induction.
An important feature of Tregs is their suppressive ability, which needs to be tested subsequently with iTregs generated in this protocol. It should be noted that suppression assays with iTregs are not trivial and the literature about suppressive abilities of iTregs is controversial. Several methods and protocols to assess suppressive function of human Tregs have been published elsewhere22,37–41, with in vitro proliferation assays based on flow cytometry readouts such as dilution of Carboxyfluorescein succinimidyl ester (CFSE) being the most commonly used. It is important to wash and rest iTregs before use in suppression assays, and it needs to be considered that iTregs, in contrast to nTregs, may not be anergic but proliferate themselves during suppression assays. We consider it extremely important to use `mock´ stimulated T cells as control suppressor cells to identify the degree of unspecific suppression (such as through CTLA-4 expression, IL-2 consumption and culture overgrowth by activated T cells) that is unrelated to FOXP3+ Treg-specific effects, and the frequent lack of this control may contribute to some controversies in the literature. Using this control, we defined that only TGF-β/ATRA/Rapa-induced iTregs displayed suppressive activity in vitro22. Thus, we conclude that regarding suppressive activity (albeit not highest fraction of FOXP3+ cells), TGF-β/ATRA/Rapa is the best combination of factors of the described protocols to induce iTregs. Nevertheless, suppressive activity in vitro does not necessarily reflect suppressive activity in vivo, and indeed we determined that human iTregs generated by these protocols did not suppress in a xenogeneic graft-versus-host disease model at least under the conditions tested22. This may be related to instable FOXP3 expression which was lost in iTregs upon restimulation, in line with a lack of TSDR demethylation22.
Depending on which aspect of iTreg features (high fraction of FOXP3+ cells, superior suppressive activity, FOXP3 stability) is most important for a particular research question, different protocols may be most suitable to study these questions, rendering it difficult to define the generally 'best' protocol for Treg induction. Furthermore, several above-described subtle experimental differences can influence results and may contribute to controversies with respect to phenotype and suppressive function of TGF-β-induced iTregs that appear between reports even with apparently similar protocols for iTreg generation42. For example, even within one laboratory, we observed that iTregs induced with TGF-β and IL-2 in serum-containing RPMI medium displayed some suppressive activity compared to control cells27, while TGF-β/IL-2-induced iTregs generated in defined, serum-free T cell culture medium did not22, despite similar levels of FOXP3.
Future applications based on these protocols should strive to further optimize Treg induction conditions to achieve a phenotype with stable FOXP3 expression, TSDR demethylation, stable phenotype without conversion to cytokine-producing effector T cells and optimal suppressive activity. Further development of iTreg induction protocols may be useful for adoptive transfer approaches in the future, in which therapy by Treg transfer is highly promising for the potential treatment of autoimmune and inflammatory diseases.
The authors have nothing to disclose.
Nina Nagel is gratefully acknowledged for technical assistance during the video shoot and experimental preparation. We thank Eva-Maria Weiss for help with the intracellular FOXP3 staining protocol and Elisabeth Suri-Payer and Nina Oberle for establishing the nTreg isolation protocol. Matilda Eriksson and Peri Noori are acknowledged for laboratory management.
Funding: A.S. was supported by a Marie Curie Intra European Fellowship within the 7th European Community Framework Programme, the Dr. Åke Olsson Foundation and KI research foundations; A.S. and J.T. were supported by a CERIC (Center of Excellence for Research on Inflammation and Cardiovascular disease) grant, J.T. was supported by Vetenskapsrådet Medicine and Health (Dnr 2011-3264), Torsten Söderberg Foundation, FP7 STATegra, AFA Insurance and Stockholm County Council.
All-trans retinoic acid | Sigma Aldrich | R2625-50MG | |||
anit-human Foxp3-APC clone 236A/E7 | eBioscience | 17-4777-42 | |||
anti-human CD25 microbeads | Miltenyi Biotec | 130-092-983 | |||
anti-human CD25-PE | Miltenyi Biotec | 130-091-024 | |||
anti-human CD28 antibody, LEAF Purified | Biolegend | 302914 | |||
anti-human CD3 Antibody, LEAF Purified | Biolegend | 317315 | |||
anti-human CD45RA , FITC | Miltenyi Biotec | 130-092-247 | |||
anti-human CD45RO PE clone UCHL1 | BD Biosciences | 555493 | |||
anti-human CD4-PerCP clone SK3; mIgG1 | BD Biosciences | 345770 | |||
anti-human CD8-eFluor 450 (clone OKT8), mIgG2a | eBioscience | 48-0086-42 | |||
anti-human CTLA-4 (CD152), clone BNI3, mIgG2ak, Brilliant violet 421 | BD Biosciences | 562743 | |||
anti-human IFN-g FITC clone 4S.B3; mIgG1k | eBioscience | 11-7319-81 | |||
Brefeldin A-containing solution: GolgiPlug | BD Biosciences | 555029 | |||
cDNA synthesis kit: SuperScript VILO(Reverse transcriptase) cDNA Synthesis Kit | Invitrogen | 11754-250 | |||
Density centrifugation medium: Ficoll-Paque | GE healthcare | 17-1440-03 | |||
DMSO 99,7% | Sigma Aldrich | D2650-5X5ML | |||
FBS, heat inactivated | Invitrogen | 10082-147 | |||
Fixable Viability Dye, eFluor 780 | eBioscience | 65-0865-14 or 65-0865-18 | |||
Foxp3 Staining Buffer Set | eBioscience | 00-5523-00 | Caution, contains Paraformaldehyde Can be also bought in combined kit with antibody; 77-5774-40 Anti-Human Foxp3 Staining Set APC Clone: 236A/E7 Set |
||
GlutaMAX (200 mM L-alanyl-L-glutamine) | Invitrogen | 35050-061 | |||
human naive CD4 T cell isolation kit II | Miltenyi Biotec | 130-094-131 | |||
|
|
130-094-131 | |||
Human serum albumin 50 g/l | Baxter | 1501057 | |||
Ionomycin from Streptomyces conglobatus >98% | Sigma Aldrich | I9657-1MG | |||
MACS LS-columns | Miltenyi Biotec | 130-042-401 | |||
mouse IgG1 K Isotype Control APC Clone: P3.6.2.8.1 | eBioscience | 17-4714-42 | |||
mouse IgG1 K Isotype Control FITC 50 ug | eBioscience | 11-4714-81 | |||
mouse IgG2a isotype control, Brilliant violet 421, clone MOPC-173 | BD Biosciences | 563464 | |||
Pasteur pipet plastic, individually packed | Sarstedt | 86.1172.001 | |||
PMA PHORBOL 12-MYRISTATE 13-ACETATE | Sigma Aldrich | P1585-1MG | |||
Rapamycin | EMD (Merck) | 553210-100UG | |||
Recombinant Human IL-2, CF | R&D | 202-IL-050/CF | |||
Recombinant Human TGF-beta 1, CF | RnD | 240-B-010/CF | |||
RNA isolation kit: RNAqueous-Micro Kit | Ambion | AM1931 | |||
RPMI 1640 Medium | Invitrogen | 72400-054 | |||
Sodium butyrate | Sigma Aldrich | B5887-250MG | |||
T cell culture medium: X-Vivo 15 medium, with gentamicin+phenolred | Lonza | 04-418Q | |||
TaqMan Gene Expression Assay, FOXP3 (Best Coverage) | Applied Biosystems | 4331182; assay ID: Hs01085834_m1 | Caution, contains Paraformaldehyde | ||
TaqMan Gene Expression Assay, RPL13A (Best Coverage) | Applied Biosystems | 4351370; assay ID: Hs04194366_g1 | Caution, contains Paraformaldehyde | ||
TaqMan Gene Expression Master mix | Applied Biosystems | 4369514 |