Circulating human CD4+CD25++CD127FOXP3+ T cells with a persistent demethylated regulatory T cell (Treg)-specific demethylated region Foxp3 gene are considered natural Tregs (nTregs). We have shown that it is possible to identify functional Ag-reactive nTregs cells for a range of different common viral and vaccination Ags. The frequency of these Ag-reactive nTregs within the nTreg population is strikingly similar to the frequency of Ag-reactive T effector cells within the CD4+ T cell population. The Ag-reactive nTregs could be recognized with great specificity by induction of CD154 expression. These CD154+ Ag-reactive nTregs showed a memory phenotype and shared all phenotypical and functional characteristics of nTregs. The isolated CD154+ nTregs could be most efficiently expanded by specific antigenic stimulation, while their Ag-reactive suppressive activity was maintained. After an in vivo booster Ag challenge, the ratio of Ag-reactive T cells to Ag-reactive Tregs increased substantially, which could be attributed to the rise in effector T cells but not Tregs. In conclusion, the nTreg population mirrors the effector T cell population in the frequency of Ag-reactive T cells. Isolation and expansion of functional Ag-reactive nTregs is possible and of potential benefit for specific therapeutic goals.

Regulatory CD4+ T cells (Tregs) are considered to be crucial for the suppression of excessive T cell responses that may cause damage to the organism. FOXP3 is a key molecule for identification of suppressive function of Tregs. Deficiency or dysfunction of FOXP3+ Tregs in humans and Foxp3 Tregs in rodents is associated with T cell-mediated autoimmune diseases, allergies, and inflammatory bowel disease. Additionally, Tregs may be important in establishing tolerance to organ transplants (1).

The ontogeny of Tregs may involve several differentiation pathways, with thymus-derived and induced Tregs (iTregs) being the main ones. Tregs generated within the thymus are called natural Tregs (nTregs) (2). Additionally, at least in vitro Tregs may be induced from a FOXP3 T cell population using the appropriate cytokine milieu, including TGF-β. These iTregs are phenotypically indistinguishable from nTregs, are FOXP3+, and can be generated as Ag-reactive T cells as opposed to nTregs (4). Recently, demethylation at a highly conserved region within the FOXP3 gene (Treg-specific demethylated region [TSDR]) was found to discriminate between natural Tregs (highly demethylated TSDR) and induced Tregs (highly methylated TSDR) (5, 6).

Within the peripheral blood of healthy humans a relative stable population of Tregs can be recognized that comprise ∼5–10% of the total CD4+ T cell population. These FOXP3+ cells express high levels of CD25 and low levels of CD127 (FOXP3+CD25++CD127CD4+ T cells) (7). Recently, it has been shown that this cell population is composed of Tregs with a naive and memory T cell phenotype based on differential expression of the CD45 isoforms CD45RO (memory phenotype) and CD45RA (naive phenotype) (8, 9). The memory Tregs behave similar to memory cells of effector CD4+ T cells (Teff), as they have short telomeres and a remarkably high turnover rate (10). Sequence-based T cell clonotype tracing of human peripheral blood and transfer of CD45RA+ Tregs into NOD/SCID/common γ-chain–deficient mice showed conversion of these CD45RA+ Tregs into CD45RA Foxp3+ cells (8).

An unresolved issue that remains is whether Ag-reactive Tregs are present within the human nTreg population, as such cells cannot be directly identified. Previous studies have indicated the existence of HIV and CMV Ag-reactive Tregs in the peripheral blood but failed to show their existence on the single cell level. It has been postulated that these Ag-reactive Tregs may originate from activated CD4+ Teff and are in fact iTregs (7).

Alternatively, Ag-reactive Tregs could stem from conversion of naive FOXP3+ nTregs into memory nTregs, in a similar fashion as the Ag-reactive memory CD4+ Teff develop from FOXP3 naive CD4+ T cells (11). Identification on the single cell level of Ag-reactive nTregs would allow for tracking Ag-reactive responses. Isolation and propagation of such Ag-reactive nTregs could be highly useful for immunotherapy, as this would avoid any unwanted generalized suppression of immune responses (12). Previously, we have used membrane-bound CD154 (CD40L) expression in a live cell assay to detect and isolate human Ag-reactive CD4+ Teff with high sensitivity and specificity (13, 14). In this study we followed this strategy to show the presence and characteristics of Ag-reactive nTregs.

PBMCs were isolated from healthy donor-derived buffy coats obtained from the local blood bank (Sanquin Blood Bank, Rotterdam, The Netherlands), as described before in detail (14). The vaccination of healthy volunteers was performed with standard doses of tetanus toxoid vaccine and hepatitis B surface Ag (HBs Ag) with weekly isolation of PBMCs thereafter, using the protocols published previously (15, 16). All individuals included gave informed consent, and the local Medical Ethical Committee approved the study. It was conducted according to the principles of the Declaration of Helsinki and in compliance with International Conference on Harmonization/Good Clinical Practice regulations.

PBMCs were enriched for T lymphocytes using the Pan T cell isolation kit (Miltenyi Biotec, Bergisch Gladbach, Germany). Enriched T lymphocytes were prepared for sorting experiments using the following mAbs: AmCyan-labeled CD3 (BD Pharmingen, Erembodegem, Belgium), Pacific Blue-labeled CD4 (BD Pharmingen), PeCy7-labeled CD25 (epitope A; BD Pharmingen), PE-labeled CD127 (BD Pharmingen), and inclusion of a live/dead marker ViaProbe (7-aminoactinomycin D [7-AAD]; BD Pharmingen). After staining, the cells were washed and resuspended at 20–25 × 106/ml and sorted (BD FACSAria II SORP; BD Biosciences) into CD3+CD4+ (in some cases, in particular for proliferation experiments, we selected for CD3+ only) T cells, Tregs (CD3+CD4+CD25++CD127), and Teff depleted for Tregs (CD4+ Teff). For proliferation experiments, we used CD3+ T cells depleted of Tregs (CD3+ Teff). The isolated fractions were tested for purity, that is, CD4+ T cells (99.7 ± 0.07%), Tregs (97.6 ± 0.34%), and CD3+(CD4+)Teff (98.7 ± 0.33%), and were either used immediately or, following an overnight rest at 37°C in standard culture medium containing 20% of heat-inactivated pooled human serum, frozen until further use in standard medium containing a final concentration of 10% of DMSO.

To compare the phenotype of Tregs with that of CD4+ Teff, we stained intracellularly for FOXP3 using the FOXP3 PE-staining kit (Bender MedSystems, Vienna, Austria) according to the manufacturer’s instruction. We also dissected Tregs and CD4+ Teff into the different T cell subsets, that is, naive (CD45ROCCR7+), central memory (CD45RO+CCR7+), and effector memory (CD45RO+CCR7) T cells (17), using allophycocyanin-labeled CD45RO (BD Pharmingen) and FITC-labeled CCR7 (R&D Systems Europe, Abingdon, U.K.).

Genomic DNA from total PBMCs and FACS-sorted fractions was isolated using the QIAamp DNA Micro kit (Qiagen Benelux, Venlo, The Netherlands) following the protocol for small volumes of blood. Bisulfite treatment using the EZ DNA Methylation-Direct kit (Zymo Research from BaseClear, Leiden, The Netherlands) according to the manufacturer’s instructions converted unmethylated cytosines into uracil, whereas methylated cytosines remained unchanged. The TSDR of FOXP3 was amplified by quantitative PCR using the StepOnePlus real-time PCR system (Applied Biosystems Europe, Nieuwerkerk aan den IJssel, The Netherlands) with methylation- and demethylation-specific primers and probes (6). The percentage of TSDR demethylation was calculated as the ratio of unmethylated DNA copy numbers and the sum of methylated and unmethylated DNA copy numbers.

Isolated Tregs and CD4+ Teff (1 × 105) were stimulated in triplicate/duplicate in 96-well plates for 6, 12, 18, 24, 30, 36, and 48 h with autologous PBMCs (1 × 105) either loaded with CMV Ag (34 μg/ml; Microbix Biosystems, Toronto, ON, Canada) or not (background). Stimulation was done in the presence of costimulation, that is, Abs directed against CD28 (1 μg/ml; Serotec/Morphosys, Duesseldorf, Germany) and CD49d (1 μg/ml; BD Pharmingen). Anti-CD40 (1 μg/ml, clone 14G7; Sanquin, Amsterdam, The Netherlands) was added when CD154 was stained on the cell surface. The autologous PBMCs that were used as APCs were labeled with 5 μM CFSE (Molecular Probes Europe, Leiden, The Netherlands) for 10 min at 37°C. At the different time points, cells were harvested and stained to determine percentages of CD154- and CD137-expressing cells, respectively. Percentages of CD154 or CD137, induced by CMV Ag-loaded autologous CFSE-labeled PBMCs, were corrected for the background (i.e., autologous CFSE-labeled PBMCs not loaded with Ags) (14).

Upon a 24-h stimulation with different viral and vaccination Ags, we determined percentages of Ag-reactive (CD154-expressing) Tregs and CD4+ Teff in individuals that were tested seropositive for CMV Ag (34 μg/ml), varicella-zoster virus (VZV) Ag (30 μg/ml; Microbix Systems), or were previously immunized against hepatitis B virus (5 μg/ml HBs Ag; Bioconnect, Huissen, The Netherlands) or tetanus toxoid (37.5 Lf/ml; Dutch National Institute for Public Health and the Environment, Bilthoven, The Netherlands). Similar as described above, the percentages of Ag-reactive CD154-expressing Tregs and CD4+ Teff were corrected for the background signal.

Tregs or the CD154-enriched and CD154-depleted Tregs (at 1 × 104/well for polyclonal expansion and 2.5 × 104/well for Ag-reactive expansion) were expanded for 7–10 d using either CD3/CD28 T cell expander dynabeads (Invitrogen Dynal, Oslo, Norway) at a 1:1 ratio or CMV Ag-loaded (34 μg/ml) irradiated (30 Gy) autologous PBMCs (1 × 105/well). Expansion was performed in standard culture medium supplemented with 25 IU/ml recombinant human IL-2 (Alloga, Veghel, The Netherlands) and 10 ng/ml recombinant human IL-15 (PeproTech, Rocky Hill, NJ). Every 3 d, fresh culture medium supplemented with cytokines was added to the wells.

Suppressive capacity was tested using the freshly isolated, the Ag-reactive, and/or the polyclonal expanded Treg fractions (unseparated or CD154-enriched and CD154-depleted). Briefly, Tregs were transferred in triplicate to the wells and used at different ratios (i.e., 1:5 to 1:2560). Subsequently, 5 × 104 CD3+ Teff (responder cells) were added in triplicate to the wells. Irradiated (40 Gy) autologous PBMCs either loaded with the different Ags—that is, CMV Ag (34 μg/ml), VZV Ag (30 μg/ml), HBs Ag (5 μg/ml), Tet-Tox (37.5 Lf/ml), and PHA (1 μg/ml; Roche Diagnostics, Mannheim, Germany)—or not were used as stimulators and used at a concentration of 5 × 104/well. At day 5, [3H]thymidine was added and following 16–18 h incorporation, the plates were harvested, and radioactivity was counted using a liquid scintillation counter. The cpm or proliferation in presence of Tregs was related to proliferation in the absence of Tregs and depicted as percentages.

Percentages of cytokine-producing cells were determined by stimulating both Tregs and CD4+ Teff either in the presence of costimulation only (anti-CD28 and anti-CD49d both at 1 μg/ml) or with the combination of PMA (50 ng/ml; Sigma-Aldrich, St. Louis, MO) and ionomycin (1 μg/ml; Sigma-Aldrich) for 24 h, of which the last 12 h were in the presence of GolgiPlug (BD Biosciences). Subsequently, the cells were harvested and the cell surface was stained using Abs directed against CD3, CD4, and a viability marker (7-AAD). Following fixation and permeabilization, intracellular staining was performed using mAbs directed against FOXP3 (Bender MedSystems), IL-2, IFN-γ, or TNF-α (the last three from BD Pharmingen) (18, 19). Percentages of cytokine-producing cells within the Tregs were performed in the context of FOXP3 expression.

For comparisons between groups, the t test, Mann–Whitney U test, one-way ANOVA, or Kruskal–Wallis test was used, as appropriate. Post hoc analysis was performed using the Bonferroni test for multiple comparisons or the Mann–Whitney U test. A p value of <0.05 for two sides was considered statistically significant.

Typically, Ag-reactive T cells are found at very low frequencies (1:10,000 or less of total circulating T cells) in the peripheral blood of healthy individuals. The exception to this rule is CMV Ag-reactive T cells that can be detected at a much higher frequency of 1:1000 to 1:100. Therefore, we started our study with CMV Ag, as this allowed for relatively easy detection and isolation of Ag-reactive T cells, using the CD154 live cell assay. PBMCs were enriched for T cells and the viable CD4+T cells were sorted in a Teff (CD4+ Teff/CD3+CD4+CD25−/dimCD127+/−) and Treg (CD3+CD4+CD25brightCD127) population based on the differential expression of CD25 and CD127 (Fig. 1). This resulted in a Treg population with on average 95 ± 2.4% FOXP3-expressing cells (mean fluorescence intensity [MFI], 4475 ± 220) (Fig. 1H) as compared with CD4+ Teff that contained on average 7 ± 1% FOXP3-expressing cells with an overall MFI of 318 ± 157 (Fig. 1G). Moreover, the Tregs showed highly demethylated TSDR FOXP3 as compared with CD4+ Teff that contained predominantly methylated TSDR FOXP3 (Fig. 1I). In the presence of CMV Ag, CD154+ T cells were readily and consistently detected in both T cell populations in CMV-seropositive donors (Fig. 2). In contrast, CMV-seronegative donors did not have a CD154 signal above background (data not shown) (14). The background was very low in both populations (<0.01%) when freshly isolated T cells were used. Kinetic analysis of Ag-reactive CD154+ T cells showed a typical difference between CD4+ Teff and Tregs (Fig. 2A–F). The CD4+ Teff (Fig. 2A, 2B, 2E) showed a rapid but biphasic response (maximum at 6 and 24 h) of cell surface CD154 expression, whereas in the Tregs (Fig. 2C, 2D, 2F) only a few CD154+ cells were observed at 6 h. However, sorting of both populations was necessary to allow for discrimination of the CD154+CD4+ Teff and CD154+ Tregs at 24 h because the activated CD154+CD4+ Teff increased their expression of CD25 (Fig. 2I) and downregulated expression of CD127. This resulted in an average 30–50% overlap between both populations of Ag-reactive T cells at 24 h. It is known that activated CD4+ Teff may become FOXP3+ (20), but at 24 h the CD154+ Tregs still showed a significant higher FOXP3 expression (Fig. 2G). Additionally, the percentage demethylated TSDR FOXP3 remained high in CD154+ Tregs as compared with CD154+CD4+ Teff (Fig. 2H), even after expansion of >80% of the TSDR FOXP3 of CD154+Tregs was demethylated (data not shown).

FIGURE 1.

Phenotype of sorted CD4+ Teff and Tregs. PBMCs were enriched for T cells that were labeled with Abs directed against CD3, CD4, CD25 epitope A, CD127, and a live/dead marker (7-AAD). The lymphocytes were gated (A) and viable CD3+ (B) and subsequently CD4+ T cells (C) were selected. After FACS sorting, we obtained a highly enriched (>98%) fraction of viable CD4+ T cells (D), CD4+ T cells depleted of CD4+CD25++CD127, termed CD4+ Teff (E) and Tregs (F). Immediately following sorting, the CD4+ Teff (G) and Tregs (H) were stained intracellularly for FOXP3 (the open histogram resembles the isotype control staining in each of the fractions). Additionally, the percentage of demethylated TSDR FOXP3 was also determined for the different fractions (I). Results are a typical example from 10–15 experiments except for I, which represents mean and SEM for each fraction.

FIGURE 1.

Phenotype of sorted CD4+ Teff and Tregs. PBMCs were enriched for T cells that were labeled with Abs directed against CD3, CD4, CD25 epitope A, CD127, and a live/dead marker (7-AAD). The lymphocytes were gated (A) and viable CD3+ (B) and subsequently CD4+ T cells (C) were selected. After FACS sorting, we obtained a highly enriched (>98%) fraction of viable CD4+ T cells (D), CD4+ T cells depleted of CD4+CD25++CD127, termed CD4+ Teff (E) and Tregs (F). Immediately following sorting, the CD4+ Teff (G) and Tregs (H) were stained intracellularly for FOXP3 (the open histogram resembles the isotype control staining in each of the fractions). Additionally, the percentage of demethylated TSDR FOXP3 was also determined for the different fractions (I). Results are a typical example from 10–15 experiments except for I, which represents mean and SEM for each fraction.

Close modal
FIGURE 2.

CD154 and CD137 expression following CMV Ag stimulation in CD4+ Teff and Tregs. CD4+ Teff and Tregs were stimulated with CFSE-labeled autologous PBMCs, either loaded with CMV Ag or not, in the presence of costimulation (anti-CD28 and anti-CD49d mAbs) and anti-CD40 to stain for CD154 on the cell surface (A–F) or without anti-CD40 to stain for CD137 (J–O). At 6 and 24 h CMV Ag stimulation, CD154 (A–D) or CD137 (J–M) expression (black dots) was determined within CD4+ Teff and Tregs. The kinetics of CD154- and CD137-expressing cells within CD4+ Teff (E, N) and Tregs (F, O) was determined immediately following isolation and upon stimulation with or without CMV Ag for different time points. One of three representative examples is shown. MFI of FOXP3 within total and CD154-expressing CD4+ Teff and Tregs following stimulation with CMV Ag for 24 h (G) was analyzed as well as the percentage of demethylated TSDR FOXP3 (H) and MFI of CD25 (I). Results are means ± SEM of three to five independent experiments. *p < 0.05.

FIGURE 2.

CD154 and CD137 expression following CMV Ag stimulation in CD4+ Teff and Tregs. CD4+ Teff and Tregs were stimulated with CFSE-labeled autologous PBMCs, either loaded with CMV Ag or not, in the presence of costimulation (anti-CD28 and anti-CD49d mAbs) and anti-CD40 to stain for CD154 on the cell surface (A–F) or without anti-CD40 to stain for CD137 (J–O). At 6 and 24 h CMV Ag stimulation, CD154 (A–D) or CD137 (J–M) expression (black dots) was determined within CD4+ Teff and Tregs. The kinetics of CD154- and CD137-expressing cells within CD4+ Teff (E, N) and Tregs (F, O) was determined immediately following isolation and upon stimulation with or without CMV Ag for different time points. One of three representative examples is shown. MFI of FOXP3 within total and CD154-expressing CD4+ Teff and Tregs following stimulation with CMV Ag for 24 h (G) was analyzed as well as the percentage of demethylated TSDR FOXP3 (H) and MFI of CD25 (I). Results are means ± SEM of three to five independent experiments. *p < 0.05.

Close modal

Membrane-bound CD137 has been used to detect Ag-reactive T cells (21), and CD137+ cells could be detected after CMV Ag stimulation in the sorted Tregs (Fig. 2M). However, in contrast to the CD154 live cell assay, the expression in time of CD137+ Ag-reactive T cells was similar for CD4+ Teff and Tregs and reached the maximum at 24 h (Fig. 2N, 2O). We continued the experiments with the CD154 live assay, as CD137 background signals were in general higher and could vary considerably among donors.

Within the CD4+ Teff population, cells can be classified in subsets based on the expression of CCR7 and CD45RO isoform on their cell membrane. This leads to a division into three major subsets—naive T cells (CCR7+CD45RO), central memory T cells (CCR7+CD45RO+), and effector memory T cells (CCR7CD45RO+)—that correlate with increasing T cell differentiation (16). Within the Treg population a similar subset distribution was observed as for the CD4+ Teff population (Fig. 3B and 3A, respectively). The CD154+CD4+ Teff and CD154+ Tregs were predominantly found within the memory compartment (Fig. 3C).

FIGURE 3.

Subset and cytokine expression of CD4+ Teff and Tregs. Immediately following sorting, CD4+ Teff (A) and Tregs (B) were dissected into subsets using CCR7 and CD45RO mAbs. Next, CD154+CD4+ Teff and CD154+ Tregs (C, right) were dissected into naive and memory subsets upon a 24-h CMV Ag stimulation. Additionally, CD4+ Teff and Tregs were incubated for 24 h, of which the last 12 h was with brefeldin A, in the presence of costimulation combined with PMA and ionomycin (D–G and H–K, respectively), to determine percentages of IL-2–, IFN-γ–, or TNF-α–producing cells in the context of CD4 and FOXP3, respectively. The signal of the isotype control for the mAbs directed against cytokines is given in D and H.

FIGURE 3.

Subset and cytokine expression of CD4+ Teff and Tregs. Immediately following sorting, CD4+ Teff (A) and Tregs (B) were dissected into subsets using CCR7 and CD45RO mAbs. Next, CD154+CD4+ Teff and CD154+ Tregs (C, right) were dissected into naive and memory subsets upon a 24-h CMV Ag stimulation. Additionally, CD4+ Teff and Tregs were incubated for 24 h, of which the last 12 h was with brefeldin A, in the presence of costimulation combined with PMA and ionomycin (D–G and H–K, respectively), to determine percentages of IL-2–, IFN-γ–, or TNF-α–producing cells in the context of CD4 and FOXP3, respectively. The signal of the isotype control for the mAbs directed against cytokines is given in D and H.

Close modal

As expected, substantially lower numbers of Tregs secreting IL-2 (6.4 ± 2.4%), IFN-γ (1.5 ± 0.6%), or TNF-α (8.7 ± 2.8%) were found compared with the CD4+ Teff population (IL-2+, 30.3 ± 3.3%; IFN-γ+, 9.9 ± 2%; and TNF-α, 22.8 ± 6.6%) after polyclonal stimulation with PMA/ionomycin (Fig. 3H–K and 3D–G, respectively). More than 98% of the cytokine-producing Tregs coexpressed FOXP3 (Fig. 3I–K).

Next, the CMV Ag-specific CD154+ Tregs were sorted (Fig. 4A–C) and compared with CD154-depleted Tregs for their suppressive function either immediately (Fig. 4D, 4E) or following expansion (Fig. 4I, where only results of CMV Ag-expanded CD154-expressing Tregs are displayed). The CD154+ Tregs were significantly better suppressors of CMV Ag-induced T cell proliferation compared with the CD154-depleted Tregs (Fig. 4D, 4F), although both cell populations showed a similar efficacy in suppressing polyclonal PHA-induced T cell proliferation (Fig. 4E, 4F).

FIGURE 4.

Functional analysis of CMV Ag-stimulated Tregs. Upon a 24-h stimulation with CMV Ag, Tregs (A, 0.21% CD154+) of a CMV-seropositive donor were sorted into a CD154-enriched (B, >90%) and -depleted (C, >95%) fraction. The total CMV Ag-stimulated Tregs were also sorted again (sham treated) to be able to compare with CD154-enriched and -depleted Tregs. The function of these different Treg fractions was analyzed immediately following sorting (D, E). F, Summary (means ± SEM) of five experiments with respect to the proliferation of CD3+ Teff, in presence of a 1:5 ratio of the different Tregs, to CMV Ag and PHA Ag, respectively. Additionally, the different Treg fractions were also expanded using CMV Ag or CD3/CD28 beads (polyclonal). The fold increase of Ag-specific and polyclonal expansion of the different Treg fractions is depicted in G. As a negative control, the Ag-specific expansion protocol was applied to Tregs of a CMV-seronegative individual. A typical example of the FOXP3 expression at day 9 CMV Ag-expanded CD154-expressing Tregs is given in H (the open histogram resembles the isotype control staining, whereas the closed one resembles the FOXP3 staining). CMV Ag-expanded CD154-enriched Tregs were also tested for their suppressive capacity. The suppressive capacity of the different Treg fractions was tested using a proliferation assay with [3H]thymidine incorporation as a readout, and proliferation in presence of Tregs was compared with that without Tregs (no Tregs). For this purpose, CD3+ Teff were stimulated with irradiated autologous PBMCs only or combined with CMV Ag (D) or PHA Ag (E). Different numbers of the various Treg fractions were added to the CD3+ Teff stimulated with CMV Ag or PHA Ag. In F, the CMV Ag- and PHA Ag-induced relative proliferation of responders (means ± SEM, five experiments) at a ratio of 1:5 (Tregs/CD3+ Teff) is depicted. *p < 0.05 when CD154-enriched Tregs are compared with CD154-depleted Tregs. In I, the proliferation of CD3+ Teff induced by different Ags in presence of CMV Ag-expanded CD154-enriched Tregs is depicted. Results are a typical example from three to five experiments, except for F and G, which represent mean ± SEM for each value.

FIGURE 4.

Functional analysis of CMV Ag-stimulated Tregs. Upon a 24-h stimulation with CMV Ag, Tregs (A, 0.21% CD154+) of a CMV-seropositive donor were sorted into a CD154-enriched (B, >90%) and -depleted (C, >95%) fraction. The total CMV Ag-stimulated Tregs were also sorted again (sham treated) to be able to compare with CD154-enriched and -depleted Tregs. The function of these different Treg fractions was analyzed immediately following sorting (D, E). F, Summary (means ± SEM) of five experiments with respect to the proliferation of CD3+ Teff, in presence of a 1:5 ratio of the different Tregs, to CMV Ag and PHA Ag, respectively. Additionally, the different Treg fractions were also expanded using CMV Ag or CD3/CD28 beads (polyclonal). The fold increase of Ag-specific and polyclonal expansion of the different Treg fractions is depicted in G. As a negative control, the Ag-specific expansion protocol was applied to Tregs of a CMV-seronegative individual. A typical example of the FOXP3 expression at day 9 CMV Ag-expanded CD154-expressing Tregs is given in H (the open histogram resembles the isotype control staining, whereas the closed one resembles the FOXP3 staining). CMV Ag-expanded CD154-enriched Tregs were also tested for their suppressive capacity. The suppressive capacity of the different Treg fractions was tested using a proliferation assay with [3H]thymidine incorporation as a readout, and proliferation in presence of Tregs was compared with that without Tregs (no Tregs). For this purpose, CD3+ Teff were stimulated with irradiated autologous PBMCs only or combined with CMV Ag (D) or PHA Ag (E). Different numbers of the various Treg fractions were added to the CD3+ Teff stimulated with CMV Ag or PHA Ag. In F, the CMV Ag- and PHA Ag-induced relative proliferation of responders (means ± SEM, five experiments) at a ratio of 1:5 (Tregs/CD3+ Teff) is depicted. *p < 0.05 when CD154-enriched Tregs are compared with CD154-depleted Tregs. In I, the proliferation of CD3+ Teff induced by different Ags in presence of CMV Ag-expanded CD154-enriched Tregs is depicted. Results are a typical example from three to five experiments, except for F and G, which represent mean ± SEM for each value.

Close modal

It was possible to expand unseparated or CD154-expressing (CMV Ag-reactive) Tregs of a CMV-seropositive donor in the presence of CMV Ag, IL-2, and IL-15. No proliferation of Tregs was observed when CMV Ag expansion was performed without IL-2 and IL-15 (data not shown) or when cells of a seronegative individual were chosen for expansion (Fig. 4G). These results could be replicated by using VZV Ag in cultures of Tregs derived from VZV-seropositive donors, but again not in VZV-seronegative donors (data not shown). The CD154-enriched Tregs could be expanded more efficiently (yielding higher cell numbers) when compared with unseparated Tregs (Fig. 4G) using CMV Ag. In contrast, CD154-depleted Treg cultures showed no significant expansion in response to CMV Ag, and expansion was comparable to cytokine-only conditions. Expanded CMV Ag-specific CD154+ Tregs showed stable high FOXP3 expression (Fig. 4I); that is, >95% of these Tregs still expressed FOXP3. These CMV Ag-expanded CD154-expressing Tregs were potent suppressors of CMV Ag-induced T cell proliferation (Fig. 4I). Additionally, they showed a modest suppression of T cell proliferation to other common Ags but only at the highest ratios of Tregs/CD3+ Teff (Fig. 4I). These results indicate that CMV Ag-reactive CD154+ Tregs, when expanded in an Ag-specific manner, remain predominantly Ag-reactive in their suppressor function.

Proliferation in cpm is given in Supplemental Fig. 1A–D. Additionally, a kinetic analysis did not reveal a shift in the proliferative response of CD3+ Teff to CMV Ag in the presence of 1:5 CMV Ag-expanded CD154+ Tregs. The strong suppression observed at day 6 was therefore unlikely to be caused by exhaustion of this culture (Supplemental Fig. 1E).

Not only CMV Ag, but a number of other common Ags such as VZV Ag, HBsAg, and Tet-Tox showed a CD154+ response in the Treg population (Fig. 5A). Independent of the frequency of Ag-reactive T cells, a strikingly similar ratio for the percentage of CD154+CD4+ Teff/percentage of CD154+ Tregs was found for the different Ags tested (Fig. 5B). This ratio was on average 10:1, and similar to the amount of Tregs in the circulation, when tested in the unstimulated healthy individual.

FIGURE 5.

CD154 signal in CD4+ Teff and Tregs for different Ags. The CD154 signal was determined for different Ags in CD4+ Teff and Tregs upon a 24-h stimulation with autologous CFSE-labeled PBMCs (ratio 1:1) with anti-CD40 and costimulation only or supplemented with CMV Ag, VZV Ag, HBs Ag, or Tet-Tox. Individuals were tested seropositive for CMV Ag and VZV Ag and immunized against HBV (HBs Ag) or tetanus (Tet-Tox), respectively. Percentages of CD154-expressing cells were corrected for the background (i.e., autologous CFSE-labeled PBMCs, anti-CD40, and costimulation only). In A, the ratio of CD154 signal for different Ags in CD4+ Teff per Tregs is corrected for the percentage of Tregs within peripheral blood and the percentage of Tregs present within the circulation. In B, the CD154 signal for the different Ags within CD4+ Teff is plotted against that of Tregs. In C, the CD154 signal within Tregs and CD4+ Teff at several time points following vaccination against tetanus is shown. Subsequently, the ratio of the CD154 signal within the CD4+ Teff over Tregs is corrected for the percentage of Tregs within peripheral blood and is depicted for the different time points following vaccination against tetanus and HBV and, as a control, CMV Ag ratios over time are given (D). One out of four representative experiments is shown.

FIGURE 5.

CD154 signal in CD4+ Teff and Tregs for different Ags. The CD154 signal was determined for different Ags in CD4+ Teff and Tregs upon a 24-h stimulation with autologous CFSE-labeled PBMCs (ratio 1:1) with anti-CD40 and costimulation only or supplemented with CMV Ag, VZV Ag, HBs Ag, or Tet-Tox. Individuals were tested seropositive for CMV Ag and VZV Ag and immunized against HBV (HBs Ag) or tetanus (Tet-Tox), respectively. Percentages of CD154-expressing cells were corrected for the background (i.e., autologous CFSE-labeled PBMCs, anti-CD40, and costimulation only). In A, the ratio of CD154 signal for different Ags in CD4+ Teff per Tregs is corrected for the percentage of Tregs within peripheral blood and the percentage of Tregs present within the circulation. In B, the CD154 signal for the different Ags within CD4+ Teff is plotted against that of Tregs. In C, the CD154 signal within Tregs and CD4+ Teff at several time points following vaccination against tetanus is shown. Subsequently, the ratio of the CD154 signal within the CD4+ Teff over Tregs is corrected for the percentage of Tregs within peripheral blood and is depicted for the different time points following vaccination against tetanus and HBV and, as a control, CMV Ag ratios over time are given (D). One out of four representative experiments is shown.

Close modal

After challenging the memory T cells with a booster vaccination of HBs Ag or Tet-Tox (Fig. 5C), a time-dependent increase of Ag-reactive (CD154+) T cells was observed for CD4+ Teff. The frequency of circulating Ag-reactive CD154+ Tregs did not significantly change, leading to a temporarily increase of the CD154+CD4+ Teff/CD154+ Treg ratio with a maximum at 2 wk (Fig. 5D).

In this study we show the presence of human Ag-reactive nTregs in the peripheral blood. Reactivity for CMV Ag was initially studied, as the frequency of CMV Ag-reactive T cells is remarkably high, thereby facilitating the identification and isolation of Ag-reactive Tregs. Activated CD4+ Teff or iTregs with (temporarily) increased FOXP3 expression and increased CD25 and decreased CD127 expression may mimick nTregs (8, 20). However, there are a number of arguments in favor of the conclusion that Ag-reactive CD154+ cells within the sorted CD25++CD127CD4+ T cell population are true nTregs.

First, the expression in time of CD154 on Tregs was different from CD4+ Teff, as the rapid biphasic expression pattern (14) was not observed. Instead, similar to Ag-specific induction of CD137 on the cell surface, the expression was detectable after 6 h and reached the maximal level after 24 h stimulation. This implies that after 6 h Ag-specific stimulation all CD154+ T cells are found within the CD4+ Teff, which potentially offers the possibility to discriminate Ag-reactive CD4+ Teff from Tregs. Unfortunately, at 24 h the differentiation between CD154+ Tregs and CD154+ Teff could not be made without previous sorting, as both CD127 and CD25 expression of CD154+CD4+ Teff substantially overlapped with Tregs at that point.

Second, the CD154+ Tregs were found in a homogeneous population of Tregs characterized by the simultaneous expression of CD127 at low levels, as well as CD25 and FOXP3 at high levels. Additionally, the FOXP3 expression in CD154+ Tregs was still significantly higher at 24 h compared with the unchanged FOXP3 expression in CD154+CD4+ Teff.

Third, the unchanged highly demethylated TSDR FOXP3 within CD154+ Tregs was in favor of the development of CMV Ag-reactive Tregs from natural occurring Tregs instead of a development from CD4+ Teff.

Fourth, the Treg population we isolated expressed low levels of inflammatory cytokines after polyclonal activation with PMA/ionomycin, as has been reported before (8, 9).

The final and most compelling argument is the observation that freshly isolated Ag-reactive CD154+ Tregs showed Ag-reactive T cell suppression and could be efficiently expanded in vitro by Ag-specific stimulation combined with exogenous IL-2 and IL-15. These Ag-expanded CD154+ Tregs were highly suppressive in an Ag-specific fashion while still displaying a nonspecific, but much less effective suppression when tested in an assay of polyclonal-stimulated T cells or other recall Ags.

Based on these findings we conclude that Ag-reactive CD154+ cells within the sorted CD4+CD25++CD127FOXP3+ T cell population are indeed nTregs and not activated CD4+ Teff or iTregs. This finding was not specific for CMV Ag, as we could subsequently show similar results using different Ags, including both Ags involved in natural infections and vaccination Ags.

A remarkable stable and similar ratio of Ag-reactive CD154+CD4+ Teff and CD154+ Tregs was found for the Ags tested. The formation of Tregs in a fixed ratio, irrespective of the Ag, indicates that these cells may actually be the brake on every Teff response. Such an intimate relation between Teff and Tregs was not observed within the peripheral blood after booster vaccination, as the frequency of CD154+ Tregs did not significantly change. However, this led to a highly increased CD154+CD4+ Teff/CD154+ Treg ratio that eventually returned to baseline, but only after 6–8 wk. The difference in kinetics between CD154+ Tregs and CD154+CD4+ Teff also argues against the induction of Tregs from the CD154+CD4+ Teff.

Our data obtained in human individuals are in accordance with the data from in vivo tracking of color-coded Teff and Treg in the allograft response against pancreatic islets transplants in the mouse (22). Using this elegant technique, it was shown that natural Tregs and Teff accumulate at the same pace within the transplant. This result is very similar to the observations made on the local T cell immune response after cutaneous inoculation with tuberculin-purified protein derivative in humans (11). However, similar to our vaccination data, numbers of Tregs within the peripheral blood hardly change after islet transplantation and do not show the steep increase as observed for Teff (22). Therefore, after vaccination it may be a possibility that Ag-reactive Tregs, having a memory lymphoid homing profile (CD45RO+CCR7+), are retained and active within the lymphoid organs in the weeks following antigenic exposure. Further research in different clinical settings such as infection and autoimmunity is required to understand the relationship between circulating Ag-reactive Tregs and disease activity. The present data indicate that findings for Tregs within the peripheral blood may not accurately reflect the actual Treg response at the site of immune reaction and should be interpreted with caution.

A major conclusion of our study is that by inference of the results with several unrelated Ags, the CD4+CD25++CD127FOXP3+ Treg compartment is filled with Ag-specific T cells as a mirror image of the memory CD4+ Teff population. This in fact supports the hypothesis, based on the recent work by Mayari et al. (8) that, similar to Teff, the natural Treg compartment consists of thymus-derived naive T cells that will differentiate into memory T cells after appropriate antigenic stimulation. Our findings are also in line with the results from TCR analysis of Teff and Tregs, which did not show any evidence for affinity for autoreactive Ag (23), and Vβ T cell repertoire analysis, which showed that human CD4+CD25++ Tregs share equally complex and comparable repertoires with CD4+CD25 counterparts (24). In the latter study it was also a remarkable finding that CMV seropositivity gave rise to a similar shift in Vβ repertoire in Teff and Tregs.

The possibility of direct recognition of Ag-reactive Tregs will facilitate further research in this direction, aiming a more complete understanding of the significance and interaction of Tregs in humane disease and immune responses. Additionally, isolation and propagation of Ag-reactive T cells identified on the single cell level is of potential benefit for future immunotherapy with Tregs.

N.H.R.L. designed and conducted the experiments and wrote the paper; K.B. conducted experiments; and M.G.H.B. designed the experiments and wrote the paper.

The online version of this article contains supplemental material.

Abbreviations used in this article:

7-AAD

7-aminoactinomycin D

HBs Ag

hepatitis B surface Ag

iTreg

induced regulatory T cell

MFI

mean fluorescence intensity

nTreg

natural regulatory T cell

Teff

effector T cell

Treg

regulatory T cell

TSDR

regulatory T cell-specific demethylated region

VZV

varicella-zoster virus.

1
Sakaguchi
S.
2005
.
Naturally arising Foxp3-expressing CD25+CD4+ regulatory T cells in immunological tolerance to self and non-self.
Nat. Immunol.
6
:
345
352
.
2
Maggi
E.
,
Cosmi
L.
,
Liotta
F.
,
Romagnani
P.
,
Romagnani
S.
,
Annunziato
F.
.
2005
.
Thymic regulatory T cells.
Autoimmun. Rev.
4
:
579
586
.
3
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
.
4
Zheng
S. G.
,
Wang
J.
,
Horwitz
D. A.
.
2008
.
Cutting edge: Foxp3+CD4+CD25+ regulatory T cells induced by IL-2 and TGF-β are resistant to Th17 conversion by IL-6.
J. Immunol.
180
:
7112
7116
.
5
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
.
6
Wieczorek
G.
,
Asemissen
A.
,
Model
F.
,
Turbachova
I.
,
Floess
S.
,
Liebenberg
V.
,
Baron
U.
,
Stauch
D.
,
Kotsch
K.
,
Pratschke
J.
, et al
.
2009
.
Quantitative DNA methylation analysis of FOXP3 as a new method for counting regulatory T cells in peripheral blood and solid tissue.
Cancer Res.
69
:
599
608
.
7
Hartigan-O’Connor
D. J.
,
Poon
C.
,
Sinclair
E.
,
McCune
J. M.
.
2007
.
Human CD4+ regulatory T cells express lower levels of the IL-7 receptor α chain (CD127), allowing consistent identification and sorting of live cells.
J. Immunol. Methods
319
:
41
52
.
8
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
.
9
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
.
10
Vukmanovic-Stejic
M.
,
Zhang
Y.
,
Cook
J. E.
,
Fletcher
J. M.
,
McQuaid
A.
,
Masters
J. E.
,
Rustin
M. H.
,
Taams
L. S.
,
Beverley
P. C.
,
Macallan
D. C.
,
Akbar
A. N.
.
2006
.
Human CD4+ CD25hi Foxp3+ regulatory T cells are derived by rapid turnover of memory populations in vivo.
J. Clin. Invest.
116
:
2423
2433
.
11
Vukmanovic-Stejic
M.
,
Agius
E.
,
Booth
N.
,
Dunne
P. J.
,
Lacy
K. E.
,
Reed
J. R.
,
Sobande
T. O.
,
Kissane
S.
,
Salmon
M.
,
Rustin
M. H.
,
Akbar
A. N.
.
2008
.
The kinetics of CD4+Foxp3+ T cell accumulation during a human cutaneous antigen-specific memory response in vivo.
J. Clin. Invest.
118
:
3639
3650
.
12
Brusko
T. M.
,
Koya
R. C.
,
Zhu
S.
,
Lee
M. R.
,
Putnam
A. L.
,
McClymont
S. A.
,
Nishimura
M. I.
,
Han
S.
,
Chang
L. J.
,
Atkinson
M. A.
, et al
.
2010
.
Human antigen-specific regulatory T cells generated by T cell receptor gene transfer.
PLoS ONE
5
:
e11726
.
13
Chattopadhyay
P. K.
,
Yu
J.
,
Roederer
M.
.
2005
.
A live-cell assay to detect antigen-specific CD4+ T cells with diverse cytokine profiles.
Nat. Med.
11
:
1113
1117
.
14
Litjens
N. H.
,
van de Wetering
J.
,
van Besouw
N. M.
,
Betjes
M. G.
.
2009
.
The human alloreactive CD4+ T-cell repertoire is biased to a Th17 response and the frequency is inversely related to the number of HLA class II mismatches.
Blood
114
:
3947
3955
.
15
Litjens
N. H.
,
Huisman
M.
,
Hijdra
D.
,
Lambrecht
B. M.
,
Stittelaar
K. J.
,
Betjes
M. G.
.
2008
.
IL-2 producing memory CD4+ T lymphocytes are closely associated with the generation of IgG-secreting plasma cells.
J. Immunol.
181
:
3665
3673
.
16
Litjens
N. H.
,
Huisman
M.
,
van den Dorpel
M.
,
Betjes
M. G.
.
2008
.
Impaired immune responses and antigen-specific memory CD4+ T cells in hemodialysis patients.
J. Am. Soc. Nephrol.
19
:
1483
1490
.
17
Litjens
N. H.
,
van Druningen
C. J.
,
Betjes
M. G.
.
2006
.
Progressive loss of renal function is associated with activation and depletion of naive T lymphocytes.
Clin. Immunol.
118
:
83
91
.
18
Betjes
M. G.
,
Weimar
W.
,
Litjens
N. H.
.
2009
.
CMV seropositivity determines epoetin dose and hemoglobin levels in patients with CKD.
J. Am. Soc. Nephrol.
20
:
2661
2666
.
19
Litjens
N. H.
,
Huisman
M.
,
Baan
C. C.
,
van Druningen
C. J.
,
Betjes
M. G.
.
2008
.
Hepatitis B vaccine-specific CD4+ T cells can be detected and characterised at the single cell level: limited usefulness of dendritic cells as signal enhancers.
J. Immunol. Methods
330
:
1
11
.
20
Ziegler
S. F.
2007
.
FOXP3: not just for regulatory T cells anymore.
Eur. J. Immunol.
37
:
21
23
.
21
Wehler
T. C.
,
Karg
M.
,
Distler
E.
,
Konur
A.
,
Nonn
M.
,
Meyer
R. G.
,
Huber
C.
,
Hartwig
U. F.
,
Herr
W.
.
2008
.
Rapid identification and sorting of viable virus-reactive CD4+ and CD8+ T cells based on antigen-triggered CD137 expression.
J. Immunol. Methods
339
:
23
37
.
22
Fan
Z.
,
Spencer
J. A.
,
Lu
Y.
,
Pitsillides
C. M.
,
Singh
G.
,
Kim
P.
,
Yun
S. H.
,
Toxavidis
V.
,
Strom
T. B.
,
Lin
C. P.
,
Koulmanda
M.
.
2010
.
In vivo tracking of “color-coded” effector, natural and induced regulatory T cells in the allograft response.
Nat. Med.
16
:
718
722
.
23
Pacholczyk
R.
,
Kern
J.
,
Singh
N.
,
Iwashima
M.
,
Kraj
P.
,
Ignatowicz
L.
.
2007
.
Nonself-antigens are the cognate specificities of Foxp3+ regulatory T cells.
Immunity
27
:
493
504
.
24
Kasow
K. A.
,
Chen
X.
,
Knowles
J.
,
Wichlan
D.
,
Handgretinger
R.
,
Riberdy
J. M.
.
2004
.
Human CD4+CD25+ regulatory T cells share equally complex and comparable repertoires with CD4+CD25 counterparts.
J. Immunol.
172
:
6123
6128
.

The authors have no financial conflicts of interest.