CD4+CD25+ human regulatory T cells (Treg cells), which express the transcription factor FoxP3, suppress T cell activation. In this study, we sought to define cellular and molecular mechanisms of human Treg cell differentiation. A subset of human naive CD4+ T cells that are CD25+ express high levels of FoxP3. We show that upon activation through the TCR, these FoxP3-expressing naive T cells (termed TNreg cells) greatly expand in vitro. Expanded TNreg cells acquire a full Treg phenotype with potent suppressive activity and display low IL-2 production upon TCR stimulation. TNreg cells in which FoxP3 expression was reduced through RNA interference lost their suppressive activity, but retained their low IL-2 secretion in response to TCR stimulation. Furthermore, in support of the notion that TNreg cells represent a separate lineage of naive cells, we found that they were more susceptible to HIV infection as compared with naive CD4+ T cells. Based on these findings, we propose that TNreg cells are precursors for human Treg cells and that these cells require a high level of FoxP3 expression to maintain their suppressive function. Accordingly, modulation of TNreg cell numbers during infections such as HIV may disrupt human Treg cell development, and contribute to chronic immune activation.
The adaptive immune system has evolved several mechanisms to prevent responsiveness to autoantigens, including elimination of self-reactive T cells during thymic selection or functional inactivation in the periphery (1). It is now clear that professional suppressor T cells called regulatory T cells (Treg cells)4(2) represent a major component of regulating unwanted T cell activation (2, 3). Treg cells are very potent in suppressing the activation and proliferation of T cells, through a mechanism that is likely cell-contact dependent (4, 5, 6, 7, 8, 9). Indeed, deletion or elimination of Treg cells in vivo results in T cell hyperactivation and severe autoimmunity (10, 11, 12).
In both humans and mice, the Treg subset is defined as CD4+CD25+ cells that express high levels of the transcription factor FoxP3, which is required for their development in vivo (13, 14). A natural mutation or targeted deletion of the mouse FoxP3 gene results in CD4+ T cell-mediated fatal lymphoproliferative disease (11, 13). Mutation of the FoxP3 gene in humans also results in similar T cell hyperactivation and immune dysregulation characterized by polyendocrinopathy and enteropathy, called IPEX syndrome (12, 15). FoxP3 has been shown to play a role in transcriptional repression (16, 17, 18, 19, 20) and induction (19, 20, 21) and appears to be necessary for mature murine Treg cell function (22, 23). The role of FoxP3 in human Treg cell development and function is not yet defined.
Treg cells appear to develop as a functionally distinct subset analogous to conventional T cell development in thymus (24). In mice, Treg cells are selected after recognition of self-Ag in the thymus, forming a distinct subset of T cells expressing high levels of CD25 (25). It is not yet clear whether human Treg cells undergo maturation in the thymus or whether they arise from a peripheral precursor subset.
Human Treg cells express the memory marker CD45RO, and several activation markers such as glucocorticoid-induced TNFR (GITR) and HLA-DR, in addition to high levels of CD25 and FoxP3 (26). Recent studies indicate that Treg cells lack expression of the IL-7 receptor CD127 (18, 27). However, Treg cells with memory markers are not observed in neonatal blood, suggesting that human Treg cells potentially develop postnatally from precursor cells in the periphery. A subset of human T cells that express the naive phenotype of CD45RO−CD45RA+ has been identified that expresses CD25 and possesses Treg characteristics, including FOXP3 expression and partial suppressive function (28, 29, 30). But detailed characteristics and proliferative capacity of this novel subset of T cells remains to be elucidated. It is also not yet known whether sustained expression of FoxP3 is required to maintain the phenotype and function of human Treg cells and whether they develop from naive precursors in vivo.
Human Treg cells have been shown to be highly relevant for understanding HIV pathogenesis (26, 31, 32, 33, 34, 35, 36, 37). We have previously shown that Treg cells are depleted in some HIV-infected individuals and that Treg cells are susceptible to HIV infection (26). This depletion of Treg cells in HIV-infected patients may contribute to the excessive T cell hyperactivation observed in patients by reducing the active suppression of conventional T cells (26, 31, 36). It is also possible that an excessive Treg response can blunt HIV-specific immune responses or reduce the effectiveness of potential vaccines (33, 35, 37). Therefore, deciphering human Treg cell biology would benefit our understanding of the perturbation of the immune system during HIV infection and potentially reveal new ways to modulate inflammation and disease.
In this study, we show that human Treg cells develop in vitro from a naive precursor subset, and that FoxP3 expression is required for maintenance of their suppressive function. In addition, our findings indicate that the precursor population of Treg cells, which we termed naive Treg cells (TNreg cells), are highly susceptible to HIV infection. These findings have important implications for understanding human Treg cell development and for developing therapeutic strategies to treat conditions associated with chronic immune activation such as HIV infection, or to enhance the effectiveness of vaccination.
Materials and Methods
Cell isolation and culture
The blood obtained from healthy donors for this study have been reviewed and approved by Vanderbilt and New York University School of Medicine Institutional Review Board committees. PBMC were isolated from blood of healthy donors through Ficoll-Hypaque (Pharmacia). Resting CD4+ T cells were purified using CD4+ Dynabeads (Dynal Biotech) as previously described (38) and were at least 99.5% pure as determined by postpurification FACS analysis. To purify naive, memory, Treg, and TNreg subsets, purified CD4+ cells were stained with CD25 and CD45RO Abs and CD45RO−CD25− (naive CD4+ T cells (TN)), CD45RO+CD25−/low (memory CD4+ T cells (TM)), CD45RO+CD25high (Treg), and CD45RO−CD25+ (TNreg) subsets were sorted on a FACS Aria flow cytometer. The culture medium used in all experiments was RPMI 1640 (Invitrogen Life Technologies) and was prepared as described previously (38). All cytokines were purchased from R&D Systems. Monocyte-derived dendritic cells (DCs) were generated as previously described (38). Superantigen, staphylococcal enterotoxin B (SEB; Sigma-Aldrich) was used to stimulate resting T cells in the presence of DCs (39).
T cells were stained with the relevant Ab on ice for 30 min (chemokine receptor staining performed at room temperature for 20 min to enhance staining) in PBS buffer containing 2% FCS and 0.1% sodium azide. Cells were then washed twice, fixed with 1% paraformaldehyde, and analyzed with a FACSCalibur or FACSAria flow cytometer. Live cells were gated based on forward and side scatter properties and analysis was performed using FlowJo software (Tree Star). The following anti-human Abs were used for staining: CD4, CD45RA, CD45RO, HLA-DR, CD25, GITR, CTLA-4, CD127, CCR5, CXCR4 (all from BD Biosciences), and CCR4 (R&D Systems). TCR V region staining was done using the IOTest β Mark TCR Vβ Repertoire kit (Beckman Coulter) according to the accompanying protocol.
FoxP3 intracellular staining
PBMC or purified CD4+ T cells were stained with appropriate cell surface markers, as described above. Cells were washed and resuspended in 1× eBioscience fixation/permeabilization buffer, and incubated at 4°C for 30 min. Cells were washed once with PBS buffer containing 2% FCS and 0.1% sodium azide, and washed twice with 1× eBioscience permeabilization buffer. Cells were stained with Ab against FoxP3 (clone PCH110; eBioscience) for 30 min at 4°C. Cells were washed twice with 1× eBioscience permeabilization buffer, and analyzed with either FACSCalibur or LSR-II flow cytometers.
T cell proliferation and cytokine assay
To assess proliferation and quantify cell divisions, purified CD4+ T cell subsets were labeled with CFSE (Molecular Probes). Purified cells were first washed and resuspended in (PBS). While vortexing the cells, CFSE was added at a final concentration of 5 μM. The mixture was vortexed for an additional 15 s and incubated at 37°C for 3 min. Labeling was quenched by addition of 50% FCS in PBS. Cells were washed once more with 50% serum PBS, followed by two washes with RPMI 1640-supplemented medium. CFSE-labeled T cells were stimulated with either anti-CD3 (OKT-3; American Type Culture Collection) coated plates in the presence of soluble anti-CD28 (1 μg/ml; BD Biosciences), or with either allogeneic or autologous DCs in the presence of SEB (10-0.1 ng/ml) for 18 h. IL-2 in the supernatants were assayed using a commercially available cytometric bead array (CBA; BD Biosciences) (40) and analyzed using CBA six-bead analysis software (BD Biosciences).
T cell suppression assay
Purified CD4+ T cells were FACS sorted, as described above, based on expression of CD25 and CD45RO. Target CD25−CD45RO− TN were CFSE labeled. Effector, target, and mature allogeneic or autologous DCs were cultured in a round-bottom well of a 96-well plate, in the presence of SEB (10 ng/ml-0.1 pg/ml). Cells were collected, fixed with 1% paraformaldehyde, and analyzed on day 4 postactivation. Suppression experiments with short hairpin RNA (shRNA)-expressing T cells were performed as above, except target cells were labeled with the red dye SNARF-1 (Molecular Probes) to discriminate from GFP-expressing (shRNA+) effector T cells. Cells were incubated with 10 μM SNARF-1 and remaining protocol was the same as CFSE labeling described above.
RNA was isolated from T cell subsets using an RNeasy kit (Qiagen). To synthesize cDNA, 100 ng of RNA was treated with Superscript II Reverse Transcriptase (Invitrogen Life Technologies). TaqMan Assays-on-Demand Gene Expression Primers (Applied Biosystems) were used in real-time PCR analyses: β-actin primer mix assay ID Hs99999903_ml; FoxP3 primer mix assay ID Hs00203958_m1. Real-time PCR was performed using the ABI 7700 apparatus (Applied Biosystems). The reaction mixtures (20 ml total volume) contained 2 μl of serially diluted cDNA, 10 μl of TaqMan Universal PCR Master Mix (Applied Biosystems), and 1 μl of either FoxP3 or β-actin primer mix. The reactions were amplified as follows: 50°C for 2 min and 95°C for 10 min, followed by 40 cycles of 95°C for 1 min and 65°C for 1 min. Expression of FoxP3 mRNA was normalized to β-actin expression in each sample.
HIV production and infections
HIV pseudotyped with vesicular stomatitis virus glycoprotein (VSV-G) envelope (VSV-G.HIV) was generated as previously described (41). Replication competent CCR5-tropic HIV (R5.HIV) was prepared similarly by transfecting 293T cells with HIV that encodes R5-tropic (BaL) envelope and enhanced GFP (BD Clontech) in place of the nef gene as previously described (41). Typically, viral titers ranged from 1 to 5 × 106 infectious units (IFU)/ml for replication-competent viruses and 10–30 × 106 IFU/ml for VSV-G.HIV. T cells were infected at varying multiplicities of infection (MOI), and infection was quantified by GFP expression using flow cytometry. Viral replication in T cell cultures was determined by measuring p24 levels within supernatants by an ELISA (38).
shRNA construct design and virus production
shRNAs were designed using Invitrogen Life Technologies’s RNAi Designer online program (https://rnaidesigner.invitrogen.com/rnaiexpress/setOption.do?designOption = shrna). FoxP3 (accession number NM_014009) was used as the target sequence to design the single-stranded oligonucleotides against the open reading frame. A custom loop sequence (AGAGCTTG) was used. shRNA sequences used were the following: 5′-GTTATTCGCGCGAATAACGTTAGAGCTTGAACGTTATTCGCGCGAATAAC-3′ (control shRNA); 5′-CACCGGACCATCTTCTGGATGAGAAAGAGCTTGTTCTCATCCAGAAGATGGTCC-3′ (FoxP3 sh1), and 5′-CACCGCATGTTTGCCTTCTTCAGAAAGAGCTTGTTCCTGAAGAAGGCAAACATGC-3′ (FoxP3 sh2).
Double-stranded oligos were generated and cloned into pENTR/U6 using the Block-iT U6 RNAi Entry vector kit (Invitrogen Life Technologies). Ligation products were transformed into One Shot TOP10 competent Escherichia coli. Minipreps were performed on selected colonies and shRNA fidelity was verified by nucleotide sequencing. Gateway LR clonase reaction (Invitrogen Life Technologies) was performed to transfer the pENTR/U6-FoxP3 shRNA constructs into the GFP-expressing lentiviral vector system (destination vector) according to the protocol that accompanies the pENTR/U6 cloning kit. Lentiviruses encoding shRNAs were generated by cotransfecting 293T cells with the lentiviral vector along with plasmids encoding VSV-G and Gag/Pol as described (38). Viral supernatants were concentrated using Amicon Ultra Centrifugal Filter Devices 100 K (Millipore). Typically, concentrated viral titers ranged from 5 to 30 × 106 IFU/ml.
Phenotypic characterization of CD25+ human naive T cells
Recently, a subset of human naive T cells that express CD25 were shown to also express high levels of the transcription factor FoxP3 and to display some of the characteristics of Treg cells (28, 29, 30). To better establish that these FoxP3-expressing naive CD4+ T cells (termed TNreg cells) are precursors for Treg cells, we determined their phenotype and function both ex vivo and upon expansion in vitro. We defined this TNreg subset as CD4+CD45RA+CD25+, which comprises ∼1–3% of adult CD4+ T cells and is enriched in neonatal cord blood compared with Treg cells (Fig. 1 A).
TNreg cells expressed high levels of FoxP3 mRNA (Fig. 1,B) and protein (Fig. 1,C), comparable to Treg cells. To compare the phenotype of TNreg cells to Treg cells, purified CD4+ T cells were stained for various cell surface molecules expressed or down-regulated on Treg cells including GITR (42, 43), CTLA4 (44, 45, 46), CCR4 (47), and CCR5 (26). Similar to Treg cells, TNreg cells expressed high levels of CTLA4 and low levels of CD127 (Fig. 1,C). However, expression of other surface markers preferentially expressed on Treg cells, such as HLA-DR, GITR, CCR4, and CCR5, were much lower or negative on the TNreg subset (Fig. 1,C). The expression profile of these markers on TNreg cells more closely resembled TN cells (Fig. 1 C). The surface phenotype of TNreg cells can therefore be classified as an intermediate between TN cells and Treg cells.
Characterization of functional and proliferative capacity of TNreg cells
Compared with conventional T cells, Treg cells exhibit a lower proliferative capacity and produce less cytokines in response to TCR stimulation (4, 6, 48) both of which have been attributed to higher expression of FoxP3 in this subset (13, 26). Therefore, we first determined the cytokine secretion profile and proliferative ability of TNreg cells in response to TCR signaling induced by DCs in the presence of the superantigen SEB (Fig. 2,A). Similar to Treg cells, TNregs secreted much lower levels of IL-2 compared with TN or TM cell subsets (Fig. 2 A). Stimulating TNreg cells through Abs against TCR and CD28 did not increase their IL-2 secretion (data not shown). We also determined whether TNreg cells secreted different levels of suppressive cytokine IL-10. We found that IL-10 levels produced by TNreg cells upon activation was not higher than other T cell subsets cells and was variable from donor to donor (data not shown).
To compare the proliferative capacity of the TNreg cells to other T cell subsets, we labeled cells with CFSE and stimulated with DCs pulsed with SEB. Proliferation of the T cell subsets was determined 4 days postactivation through CFSE dilution. Interestingly, the TNreg subset exhibited a proliferative response that was slightly less than that of the TN subset, but much greater than that of Treg cells (Fig. 2 B).
To determine whether TNreg cells could be expanded long-term in culture similar to TN cells, we activated these cells with DCs and SEB. The activated T cells were then cultured in IL-2-containing medium for about 2 wk and the cell numbers were assessed by counting the expanded T cells. We found that TNreg cells had a high capacity to expand in vitro, in contrast to Treg cells (Fig. 2,C), despite expression of very high levels of FoxP3 (Fig. 1, B and C). TNreg cells proliferated and expanded comparably to TN cells, while proliferation of Treg cells was significantly less compared with TN cell proliferation (Fig. 2 C).
We next evaluated the ability of TNreg cells to suppress the activation and proliferation of CD4+ TN cells. For this experiment, highly purified TN cells were labeled with CFSE and cocultured with either unlabeled Treg, TNreg, or control TN cells. The coculture was stimulated with allogeneic DCs and varying concentrations of SEB for 4 days, because T cells were activated with a strong TCR stimulus (SEB) overall contribution of allogeneic activation in these experiments during the course of 4 days is minimal as determined by proliferation with DCs alone (data not shown). TNreg cells displayed a suppressive function compared with the TN subset, however, this function was not as potent as that obtained using Treg cells (Fig. 2 D). Efficiency of suppression by resting TNreg cells ranged from ∼50% to <10% suppression of target cells (data not shown), though Treg cells consistently suppressed 70–90% target cells (data not shown). To confirm that cell death in the TNreg and Treg conditions was not playing a role in the lack of target cell proliferation, we assessed cell death by percent of live cells based on forward vs side scatter plots. Approximately 85–90% of cells were within the live cell gate in all conditions (data not shown). Thus, we conclude that lack of proliferation is not due to excessive cell death. Together, these results show that TNreg cells possess an intermediate functional profile as compared with Treg cells and, importantly, that they have the capacity to expand upon Ag stimulation.
Phenotypic characterization of expanded TNreg cells
Because TNreg cells can be expanded >100-fold in IL-2 after TCR stimulation (Fig. 2,C), we sought to address whether their surface phenotype and suppressive capacity is altered by in vitro expansion. To ensure that TNreg cells maintained expression of FoxP3, and to exclude the possibility that the expanded cell population was contaminated with conventional T cell subsets, we performed surface phenotype analysis in conjunction with intracellular staining of FoxP3. Almost all of the expanded TNreg cells expressed high levels of FoxP3, whereas expanded TN cells remained low or negative for FoxP3 expression (Fig. 3 A).
We next analyzed the Treg cell-associated surface markers on expanded TNreg cultures 2 wk postactivation. Expanded TNreg cells expressed higher levels of CD25, GITR, HLA-DR, CCR4, and CTLA-4 as compared with expanded TN cells (Fig. 3,B). Similar to TN cells, TNreg cells also lose the expression of CD45RA and express CD45RO after activation and expansion in vitro (data not shown). This differentiated cell surface phenotype of TNreg cells is consistent with a Treg cell profile, ex vivo (Fig. 1,C). In addition, expanded TNreg cells expressed high levels of the HIV coreceptors CCR5 and CXCR4 (Fig. 3 C), similar to Treg cells (26) suggesting that these cells could be potentially susceptible to HIV infection.
Functional characterization of expanded TNreg cells
We asked whether activation and expansion of TNreg cells resulted in full acquisition of Treg cell functions. We first determined the responsiveness of expanded TNreg cells to TCR restimulation by measuring IL-2 secretion. Expanded TNreg cells stimulated with DCs and SEB continued to secrete low levels of IL-2 compared with the TN subset (Fig. 4 A), indicating that they maintain this Treg cell characteristic even after expansion in vitro.
We then evaluated whether expanded TNreg cells acquired more potent suppressive function compared with freshly isolated cells. We found that expanded TNreg cells had displayed more potent immunosuppressive activity (Fig. 4,B) than freshly purified TNreg counterparts (Fig. 2,D). At suboptimal TCR stimulation (0.01–0.001 ng/ml SEB) the suppression by expanded TNreg cells was similar (∼80–90% inhibition of target cell proliferation) to Treg cells (Fig. 4 B). Taken together, these findings suggest that activation and expansion of TNreg cells induces a program leading to their differentiation into fully competent Treg cells.
TCR Vβ repertoire of TNreg and Treg cells
Our data support the hypothesis that, upon their activation and expansion, TNreg cells may contribute to the pool of Treg cells upon activation and expansion. As a corollary to this, we hypothesized that the TCR repertoire of TNreg cells would be similar to that of Treg cells. To investigate the breadth of the TCR Vβ repertoire of the naive and memory T cell subsets, we stained purified CD4+ T cells with CD25 and CD45RO as subset markers, and costained with Abs directed against ∼80% of the mature human TCR Vβ repertoire of T cells (49). This analysis indicated that TCR Vβ repertoire of TNreg cells was very similar to that of other T cell subsets (Fig. 5). Furthermore, we noted that the relative expression of different Vβ families was very similar between Treg and TNreg subsets (Fig. 5). This finding supports a model whereby TNreg differentiate into Treg cells upon activation with self or nonself Ags in vivo, acquiring a suppressive capability while maintaining the capacity to recognize similar Ags as conventional T cell subsets.
Is FoxP3 required for maintaining TNreg cell phenotype and function?
To determine whether sustained expression of FoxP3 is required to maintain the capability of TNreg cells to differentiate into fully suppressive Treg cells, we reduced expression of FoxP3 in these subsets by RNA interference (RNAi). Accordingly, shRNAs targeting the FoxP3 gene were stably expressed in activated TN or TNreg cells using a lentiviral delivery system in which the coexpression of GFP allows for the identification of transduced cells (50).
Using this system, we specifically and stably reduced expression of FoxP3, as shown by intracellular staining with a FoxP3 Ab (Fig. 6,A). We next assessed the surface phenotype of the expanded TNreg cells in which FoxP3 expression have been reduced by RNAi. Knockdown of FoxP3 was associated with lower expression of CD25, CD62L, CTLA4, CCR4, and HLA-DR as compared with TNreg cells transduced with control shRNA vector (Fig. 6,A). Similar results were obtained by stably expressing a second distinct shRNA against FoxP3 in TNreg cells (Fig. 6 A).
We next evaluated the effects of FoxP3 knockdown in activated TNreg cells on their cytokine production and suppressive function. For this experiment, we first sorted the shRNA transduced expanded TNreg cells, restimulated them through the TCR, and measured secreted IL-2. TNreg cells with reduced FoxP3 were unchanged in their inability to produce IL-2 compared with control TNreg cells expressing high levels of FoxP3 (Fig. 6,B). However, reduction of FoxP3 levels in TNreg cells greatly diminished their ability to suppress the proliferation of conventional TN cells (Fig. 6 C). These results indicate that a high-level expression of FoxP3 is indispensable for suppressive activity of Treg cells.
Are TNreg cells susceptible to HIV infection?
Treg cells were shown to potentially play an important role in modulating the immune response to HIV and pathogenesis associated with HIV infection (26, 33). We have shown that TNreg cells are capable of significant proliferation upon TCR stimulation and express high levels of HIV coreceptors CCR5 and CXCR4. We therefore studied the susceptibility of TNreg cells to HIV infection in vitro. T cell subsets were purified and activated with DCs and SEB and at the time of activation, these cells were also infected with either replication-defective viruses pseudotyped with VSV-G.HIV or replication-competent HIV viruses, that use CCR5 (R5.HIV) for entry, both of which also express GFP as a marker of infection (38). TNreg cells were highly infected by VSV-G.HIV and were also more susceptible to infection with R5.HIV compared with TN cells (Fig. 7,A). To determine the level of HIV replication in the TNreg cells as compared with activated TN cells, subsets were infected with R5.HIV for 3 days and washed to remove input virus. Supernatants from infected cultures were then monitored for HIV replication by measuring the GFP expression from infected T cells and the level of HIV p24 Ag levels using an ELISA, as described (26). R5.HIV spread within TNreg cell cultures over a 2-wk period and replicated at levels comparable to memory T cells, and importantly higher than TN cells during the early stages of the infection (Fig. 7, B and C).
Although neither resting TN nor TNreg cells expressed detectable levels of CCR5 by Ab staining (Fig. 2,C) or mRNA by quantitative real-time PCR (data not shown), we hypothesized that TNreg cells more rapidly up-regulate CCR5 upon activation compared with TN cells. To test this hypothesis, we sorted TN and TNreg cell populations and stained each subset for the expression of CCR5 on days 5 and 7 postactivation. We found that TNreg cells contained ∼8- to 10-fold higher CCR5+ T cells at these earlier time points (Fig. 7,D). This result was corroborated by ∼5-fold higher CCR5 mRNA expression, by real-time PCR analysis upon activation, on TNreg cells compared with TN cells (data not shown). CCR5 expression was detectable on activated TN cells by surface staining 8–10 days postactivation (Fig. 3 C), thus lagging several days compared with activated TNreg cells.
In this study, we demonstrate that a population of naive T cells coexpressing CD25 and FoxP3, which we called TNreg cells, has the capacity to differentiate into fully functional Treg cells in vitro. In contrast to Treg cells, TNreg cells exhibit a high proliferative capacity upon TCR stimulation, despite expressing high levels of FoxP3. In our culture system, the proliferation of TNreg cells was higher compared with a prior report (28), which could be due to our activation conditions using DCs pulsed with SEB that we use to provide a more physiological cell-cell interaction. Several groups have reported expanding CD25+ Treg cells using a multitude of stimulation conditions, including CD28-coated beads (51, 52, 53). Upon expansion, these cells retain their high expression of FoxP3. However, in these studies, the CD25+ T cells were not subdivided into naive and memory subsets. Because we have not been able to expand mature Treg cells in vitro to any significant numbers, our present data raise the possibility that in vitro-expanded Treg cells obtained by others were derived from the TNreg cell population. This would be consistent with a recent report showing that in vitro expansion of only the CD45RA+ population of CD25+ T cells gives rise to a pure Treg cell line (54).
Our results indicate that the TNreg cell population does not adopt a full Treg phenotype or exert its function until it has undergone activation and expansion and reactivation through the TCR. This observation agrees with FoxP3 overexpression studies performed by our group (26) and others (13, 55, 56) to gain insights into its possible functions in naive T cells. Mere overexpression of FoxP3 in naive CD4+ T cells does not render them fully suppressive and hyporesponsive until FoxP3 overexpressing cells have been restimulated through their TCR (26). This leads to the intriguing hypothesis that the high expression of FoxP3 is not sufficient for Treg suppressive activity in the absence of other cell differentiation programs, some of which are possibly activated by FoxP3 after TCR stimulation.
In our experiments, we found that TNreg cells were consistently less suppressive relative to Treg cells with memory phenotype, but acquired equally potent suppressive activity after they are expanded in vitro. Prior publications (28, 29, 57), found that cells similar to TNreg cells were more suppressive, quantitatively compared with our findings. It is conceivable that these quantitative differences are due to the type of stimulus used in these previous studies. Indeed, we have tried anti-CD3/28 stimulations, similar to published reports, but the suppressions from these assays were not as consistent. This is not surprising because Treg cells require cell-to-cell contact and triggering only the TCR may have the caveat of not forming an immunological synapse with an APC and recruiting Treg cells to the site of activation. We believe our method to stimulate Treg cells and T cells through DCs and SEB is more physiological way to activate them which allows three cell (DC plus T cell plus TNreg or Treg) clusters to form, and could be a better reflection of the Treg-mediated suppression in vivo.
In this study, we demonstrated for the first time that reduction of FoxP3 expression in human Treg cell precursors results in an altered phenotype and is required to maintain their suppressive function. Similar findings were reported for murine Treg cells, which showed that induced ablation of FoxP3 disrupted Treg suppressive functions in vivo (22). In addition, it was demonstrated that attenuated FoxP3 expression subverted Treg cell functions in vivo (23). The suppressive functions of FoxP3-attenuated mice were abrogated, but the cells retained their hypoproliferative nature (23). We show that cells with reduced expression of FoxP3 are still hyporesponsive to TCR stimulation as measured by the production of IL-2. Williams et al. (22) showed a rescue of IL-2 production upon controlled deletion of FoxP3 in vivo. In our shRNA system, FoxP3 expression in Treg cells is greatly reduced but not completely ablated, leaving open the possibility that residual FoxP3 levels may be sufficient to render T cells hyporesponsive to TCR signals, but insufficient to maintain their suppressive function. It is also possible that IL-2 secretion in Treg cells is independent of FoxP3 expression, but rather due to epigenetic modifications that regulate gene expression in T cells, such as in the histone tails.
The nature of the Ag recognized by Treg cells remains controversial. Studies conducted using transgenic mouse models suggest that once activated through the TCR, suppressor effector function was completely Ag independent (58, 59). These studies, however, did not exclude the possibility that Treg cells exert their suppressive function in an Ag-specific manner. We show here that both TNreg cells and Treg cells express a full repertoire of TCR Vβ receptors. It was shown that Ag-mediated expansions within specific TCR Vβ families were associated with equal increases of the same TCR Vβ in the Treg cell population (60). Therefore, it is possible that Treg cells are replenished from a precursor population in the periphery. The precursor population is likely to be the TNreg subset as suggested by our study and others (28, 29, 54, 61). Accumulating evidence in animal models indicates that Treg cells are selected by self Ag, (25, 62, 63), although it is not clear whether a naive precursor subset for Treg cells described in the present study also exists in mice. It appears that, unlike the mouse system, human FoxP3 expression can be induced in human CD4+CD25− T cells following TCR stimulation (64, 65). We propose that, at least in the human system, acquisition of suppressive function by the TNreg subset is similar to that of the acquisition of effector functions of conventional TN cells. This suggests that during an immune response, TN cells undergo activation and differentiation to gain effector functions, and TNreg cells undergo a parallel differentiation process leading to clonal expansion. This scenario would predict that for exogenous Ags, Treg suppression would be weaker for primary immune responses and stronger after secondary or subsequent exposures, thereby functioning to control chronic immune activation.
The role of the Treg compartment of T cells during HIV infection and pathogenesis remains a very important question and is not yet resolved. We clearly demonstrate an enhanced ability of HIV to replicate and infect TNreg cells. Our findings also show a 10-fold higher HIV infection rate of TNreg cells compared with TN cells with R5-tropic viruses, which use CCR5 coreceptor for entry, during the early stages of infection (Fig. 7). We found that TNreg cells up-regulated CCR5 upon activation several days before up-regulation of CCR5 in TN cells (Fig. 7,D). We believe this finding could partly explain the difference in infection in the early days after activation. This finding could be important during HIV infection, because during an immune response the preferential infection of TNreg cells early after activation could contribute to lesser control of immune activation, and thus more cellular targets for HIV. This data also suggests that CCR5 expression is differentially regulated in TN and TNreg cells during the early stages of T cell activation. This is also consistent with the our results that HIV spreads to larger portion of activated TN cells after day 8 (Fig. 7 B), at which time they begin to up-regulate CCR5 expression. It has been suggested that the expression of FoxP3 enhances HIV long-terminal repeat activity (21), it is therefore possible that TNreg cells, that express high levels of FoxP3, also provide a more optimal intracellular environment for HIV gene expression, similar to cells that overexpress FoxP3 during early stages of infection (21). A contradictory report demonstrated that FoxP3 represses retroviral transcription (66). Differences in these two studies may be due to the model systems to used to test HIV long-terminal repeat activity, such as transient transfection assay (66) vs a luciferase reporter HIV infection assay (21). Thus, the effect of FoxP3 on HIV infection remains to be further clarified in future studies.
The effect of Treg disruption during HIV infection has been associated with both beneficial (26, 31, 36) and detrimental outcomes (26, 32, 33, 34, 35). It has been suggested that HIV-immune dysfunction is associated with Treg cells that may target HIV-specific effectors, leading to an inefficient immune response against HIV (32, 33, 34, 35, 37). We and others have proposed that Treg cells may serve a protective function by limiting the overall immune activation that is a hallmark of HIV infection (26, 31, 36). In this context, the depletion of Treg cells could potentially hasten the progression to acquired immunodeficiency syndrome. At the same time, the specific depletion of TNreg cells may prevent insufficient regeneration of the Treg cell pools during primary immune responses, and could be one of the reasons for immune hyperactivation associated with HIV infection. This is supported by data showing HIV-positive individuals with higher relative levels of FoxP3 have lower immune activation as measured by expression of HLA-DR (26, 31). We demonstrate that when compared with TN cells, TNreg cells are highly susceptible to infection by R5.HIV during early stages post-TCR activation. This result suggests that TNreg cells could be preferentially targeted during HIV infection. How TNreg cells are modulated in vivo during different stages of HIV infection remains to be determined. Furthermore, it will be important to determine whether a similar subset of TNreg cells can be identified in monkey models of SIV infection and whether this can be correlated with the presence or absence of immune activation in these models, which highly correlate with disease progression (67). Interestingly, it has been reported that administration of IL-2 to HIV-positive individuals leads to the expansion of TNreg subset (30). Thus, enhancing this subset by IL-2 treatment could be additionally beneficial in regulating T cell homeostasis during HIV infection by dampening chronic immune activation.
In summary, our results indicate that the TNreg cell population acquires a full Treg phenotype upon activation, and that this differentiation process is dependent upon sustained, high-level FoxP3 expression. This subset also appears to be more susceptible to HIV infection, when compared with TN cells. The TNreg cell population could provide an exciting target cell for potential therapeutics. The ability to manipulate this population of cells could allow for the generation of homogenous, Ag-specific Treg cells that could be used for the prevention and treatment of T cell-mediated diseases, including graft-vs-host disease or reducing chronic immune activation during HIV infection. It has been shown that the memory subset of Treg cells is highly susceptible to apoptosis and has limited replicative capacity in vivo (60, 68), suggesting that the modulation of Treg cells for therapeutic purposes may not be long-lasting. The therapeutic manipulation of TNreg cells may provide a more tractable way to obtain specific and long-term protection against aberrant T cell activation, and permit effective immune modulation for vaccine responses.
We thank Dr. Luc Van Kaer, Dr. Scott VanCompernolle, Brenna Simons, Micheal Vetter, Frances Mercer, and Dr. Simge Akbulut for critical reading and valuable suggestions. We thank Dr. Chris Aiken for p24 ELISA reagents and Aimee El Hed for help with CCR5 staining.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was support by National Institutes of Health Grant R01 AI065303 (to D.U.).
Abbreviations used in this paper: Treg, regulatory T cell; GITR, glucocorticoid-induced TNFR; TN, naive CD4+ T cell; TNreg, FoxP3+CD25+ naive T cell; TM, memory CD4+ T cell; DC, dendritic cell; SEB, staphylococcal enterotoxin B; shRNA, short hairpin RNA; CBA, cytometric bead array; VSV-G, vesicular stomatitis virus glycoprotein; MOI, multiplicity of infection; RNAi, RNA interference; IFU, infectious unit.