| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Laboratory of Molecular Immunology, Center for Neurologic Diseases, Brigham and Women’s Hospital, and Harvard Medical School, Boston, MA 02115
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
CD4+CD25+ regulatory T cells (Treg) play a fundamental role in immune homeostasis via their ability to down-modulate the activation of T cells. Mouse CD4+CD25+ T cells have been shown to be central in the control of self responses in a number of mouse models of autoimmunity. Deletion of Treg by thymectomy in 3-day-old mice results in a spontaneous multiorgan autoimmune disease that is prevented by the adoptive transfer of CD4+CD25+ T cells (9, 10, 11, 12, 13). Mouse CD4+CD25+ Treg also inhibit the proliferation of responder T cells (Tresp) in established in vitro coculture assays that provide an in vitro model to assess their in vivo suppressive ability (14). Similar in vitro experiments using CD4+CD25+ cells isolated from the circulation or thymus of humans were critical in providing evidence that human CD4 cells expressing high levels of CD25 (15, 16, 17, 18, 19) were functionally similar and corresponded to the Treg identified in the mouse (20, 21). These studies demonstrated that in vitro CD4+CD25+ cells from both species suppress IL-2 gene transcription by cocultured Tresp via a mechanism involving cell contact. Additional similarities between human and mouse Treg, such as the expression of CD62Lhigh and the abrogation of anergy and suppression by providing exogenous IL-2, also have been demonstrated (18, 22, 23). Thus, the similar in vitro characteristics between mouse and human CD4+CD25+ T cells suggest that the in vitro investigation of Treg isolated from human autoimmune and infectious diseases can reflect the in vivo function of Treg (24).
We have demonstrated previously that, in humans, highly homogeneous Treg are best purified if only the 
2–3% of CD4 cells that express the highest levels of CD25 are isolated (15). Restricting the isolation to CD25high cells increases the homogeneity of the Treg population, as indicated by function and by surface expression of CD45RO, CD122, and CD62Lhigh (15, 18). Furthermore, using newly described mAbs specific for human Foxp3 via intracellular staining, the cells expressing the highest levels of intracellular Foxp3 protein were shown to again be represented by the population of CD4 T cells expressing the highest levels of CD25 (25). Yet our initial work indicated that the CD4+CD25high regulatory population is still heterogeneous, because 20–30% also express DR (15). These DR+ Treg represent over one-third of all circulating MHC-II+ CD4 cells. Because contact-dependent in vitro suppression is not MHC restricted, it seemed unlikely that T cell expression of DR was directly involved in Ag presentation as a mechanism of regulation. This led us to ask whether circulating DR+ Treg represented a separate lineage of Treg, Treg in a transitional maturation state, or contaminating activated effector cells.
In this study, we demonstrate that MHC-II expression by human CD4+CD25high T cells identifies a functionally distinct population of mature Treg and not a separate lineage within the CD4+CD25high subset. The DR expressing Treg inhibit T cell proliferation and cytokine production via an early contact-dependent mechanism that is associated with induction of Foxp3. In striking contrast, MHC-II– CD4+CD25high T cells (DR– Treg) do not induce early contact-dependent suppression. Instead, in coculture analysis, DR– Treg initially enhance secretion of IL-10 and IL-4, and subsequently induce a late suppression of proliferation that is accompanied by Treg induction of Foxp3 mRNA. Furthermore, while both the DR+ Treg and DR– Treg subsets suppress via a cell contact-mediated mechanism, the DR– Treg also can suppress via the secretion of IL-10. Examining clones derived from DR expressing and nonexpressing, CD25high or CD25– CD4 T cells demonstrated that, while clones derived from DR+ CD25–/low cells were not suppressive and no longer expressed DR after culture, a high frequency of clones derived from the CD25high and either DR+ or DR– populations exhibited both suppression and an increased frequency of DR expression. In total, these data indicate that MHC-II expression on circulating CD4+CD25high T cells identifies functionally distinct Treg.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Cells were cultured in RPMI 1640 medium supplemented as described previously (17) with 5% human AB serum (SeraCare Life Sciences) in 96-well U-bottom plates (Costar) or Transwell plates (Nunc). The anti-CD3 (UCHT1 for bead and Hit3a for soluble stimulation) and anti-CD28 (28.2) Abs were purchased from BD Pharmingen. The anti-CD2 Ab (BMA 0111) was purchased from Dade Behring. To generate different stimuli, tosyl-activated beads (Dynal Biotech) were covalently bound with anti-CD3 alone, anti-CD3 and anti-CD2, or anti-CD3 and anti-CD28 at 1 µg/107 beads per the manufacturer’s instructions and were used at 1.5 x 104 beads/well. In certain experiments, anti-IL-10 (JES3–9D7; BD Pharmingen) was added at 5 µg/ml. Staining for intracellular cytokines was performed on cells stimulated with anti-CD3/anti-CD2 for 48 h, during which GolgiStop (BD Pharmingen) and PMA and ionomycin were added for the last 12 h before they were fixed and permeabilized (Cytofix/Cytoperm reagents; BD Pharmingen), and stained separately with allophycocyanin-labeled anti-IL-4, anti-IL-10, or anti-IFN-
Abs. Nonspecific background levels (control cytokine staining) were determined for each cytokine by incubating the permeabilized cells with each unlabeled cytokine specific mAb for 10' before adding the same, but fluorochrome-labeled anti-cytokine mAb. Staining for intracellular Foxp3 protein was performed on cells directly after ex vivo isolation with the PCH101 anti-Foxp3 Ab (eBioscience) per the manufacturer’s instructions.
Cell isolation
Whole mononuclear cells were isolated from human blood drawn (from 16 different healthy control individuals) in green-capped, heparinized tubes by Ficoll-Hypaque (Amersham Biosciences) gradient centrifugation. Total CD4 T cells were isolated via the CD4+ T cell isolation kit II (Miltenyi Biotec). CD4 T cells were stained for four-color sorting with anti-CD4 (clone RPA-T4) (optional), anti-CD25 (clone M-A251), anti-CD62L (clone Dreg 56), and anti-HLA-DR (clone L243). All Abs were purchased from BD Pharmingen and used after they had been dialyzed to remove azide. The specific DR+ and DR– Treg populations were isolated by sorting with a FACS Aria (BD Biosciences) to typically >98% purity in postsort analysis. The sorted cells were gated first on small lymphocytes by forward and side scatter, then on CD4 (optional, due to their sample being preselected to be 99% CD4 T cells) and CD62Lhigh expression, and then on CD25 and HLA-DR expression. In samples not directly stained for CD4, a separate control was analyzed with CD4 to demonstrate the typical 95–99% purity for CD4.
In vitro micrococulture assay
In all assays using cell populations isolated directly ex vivo, the CD4+CD25– (Tresp) cells were plated at 2.5 x 103/well, while the DR+ or DR– Treg were plated at 1.25 x 103/well, and cocultures were at a 2:1 ratio. No irradiated accessory cells were added. A total of 1.5 x 104 beads coated with anti-CD3 alone, anti-CD3/anti-CD2, or anti-CD3/anti-CD28 was added per well. The day on which the triplicate assays were set up was considered day 0. To determine proliferation on days 3, 4, and/or 5, half of the culture supernatant (100 µl) was replaced with 1 µCi of [3H]thymidine (NEN) at 24-h intervals for a 24-h pulse so that the entire final 3 days of the culture would be monitored. All assays exhibited <10% SEM and were repeated a minimum of three times using blood from different donors. In experiments using CFSE and/or fluorescence analysis of protein expression, the Tresp were labeled with 0.25 µM CFSE as described (26), and fluorescence was assessed on a FACSCalibur (BD Biosciences) using CellQuest Pro software (BD Biosciences).
Generating clones
For the single-cell cloning experiments, CD4+CD62Lhigh cells that also gave the profiles of DR– CD25– (Tresp), DR+ CD25–/low, DR+ Treg, and DR– Treg were sorted at one cell per well and stimulated with soluble anti-CD3 (Hit3a, 2.5 µg/ml) and anti-CD28 (28.2, 1 µg/ml), 1–5 x 104 irradiated (3000 rad) PBMC, and 100 U/ml IL-2 (rhIL-2, Tecin; NCI), as modifications to previously published procedures (18, 27). Half of the medium was replaced with fresh medium containing 100 U/ml IL-2 starting at day 10 and every 3–4 days thereafter. The Treg clones were generated by stimulation with anti-CD3/anti-CD28/IL-2, because it gave a significantly higher cloning frequency than PHA (data not shown). After 21 days, the cells were restimulated once with PHA-P (REMEL) at 0.5 µg/ml with a mixture of irradiated PBMC (25% allogeneic, 75% autologous) at 1–5 x 104/well. The clones were again given fresh IL-2 medium every 3–4 days and were analyzed after 21 days.
Testing clones.
All clones were tested for regulatory function by coculture analysis at 1:1, 1:1/2 and 1:1/4 (Tresp:clone) ratios with 4 x 103 Tresp/well via 
CD3 stimulation (bead bound) in the absence of accessory cells. Clones were classified as suppressive if they resulted in coculture proliferation that was equivalent or below the proliferation of the Tresp alone (<2500 cpm), and if decreasing numbers of cloned cells resulted in increasing proliferation. The clones that satisfied both of these criteria were classified as suppressive. These two criteria were applied so that all clones could be tested regardless of their levels of expansion to minimize the effects of errors in counting small numbers of cells.
Clone mRNA analysis.
RNA also was isolated from those clones that had grown to sufficient levels to be in excess of the number required for the functional suppressive assay, representing 6–10 clones from each T cell subset and including both suppressive and nonsuppressive clones. These RNA samples were interrogated by real time PCR for expression of Foxp3, IL-10, and TGF
, relative to 2-microglobulin (2m) and to each other. Although only a low number of clones were selected for RNA analysis, the frequency of suppressive clone RNA samples in each group basically mirrored the frequency of suppressive clones seen in the larger group of clones, even though they had been selected before functional characterization.
Cytokine analyses by cytometric bead array (CBA)
The supernatants that were removed before addition of [3H]thymidine were analyzed without dilution for the presence of IFN-, IL-10, and IL-4 via the Human Th1/Th2 Cytokine CBA kit (BD Pharmingen) per the manufacturer’s instructions. The beads were analyzed on a FACSCalibur using the CBA software purchased from BD Pharmingen. Upon initial comparison with ELISA analysis of the cytokine levels, the CBA data were very similar, highly reproducible, but much more sensitive.
Real-time PCR analysis
RNA, isolated from the indicated cell populations either directly ex vivo or after 3 and 5 days of stimulation by the RNeasy Micro, RNase-free DNase procedure (Qiagen), was converted to cDNA via RT by random hexamers and Multiscribe RT, via the TaqMan Gold RT-PCR kit (Applied Biosystems), diluted 1/10, and used in triplicate 25 µl of TaqMan PCR to amplify human Foxp3, TGF, or IL-10 and the 2m endogenous control on an ABI 7700 machine (Applied Biosystems). The primer and probe combinations for all of these transcripts were purchased from Applied Biosystems. The values are presented as the difference in Ct values normalized to 2m for each sample as per the following formula: Normalized expression = (power(2,–
Ct)) x 1000.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
To address the potential significance of MHC II expression on CD4+CD25high Treg, highly purified DR+ and DR– subsets of CD4+CD25highCD62Lhigh Treg were isolated from negatively selected CD4 peripheral blood T cells and examined for functional activity in vitro. CD4 cells were first gated on high expression of CD62L (L-selectin), as described previously (18, 22) (Fig. 1A). The CD62L– population, representing in vivo-activated CD4 T cells (13% of total CD4 cells), express intermediate levels of CD25. In contrast, the CD62Lhigh subset, representing 82% of CD4 T cells that have presumably not been activated in vivo (23, 28), contains the CD4+CD25high population that exhibits in vitro regulatory activity (15, 18). In this analysis, 2.5 and 32% of the peripheral blood CD4+CD62Lhigh T cells are CD25high and CD25low, respectively. The DR+ (30%) and DR– (70%) subsets were sorted from the total CD4+CD25highCD62Lhigh Treg population and are designated as Treg. The DR+ Treg selected by this gating represent approximately half of the DR+ cells in the CD4+CD62Lhigh population and correspond to 0.6% of total CD4 T cells. The remaining DR+ CD4+CD62Lhigh are CD25–/low. In contrast, while there are no CD25high DR+ T cells in the CD62L– population, the DR+ CD62L– cells are CD25low and represent 0.7% of total CD4 T cells. Thus, while two-thirds of all the peripheral blood CD4+ DR+ T cells are CD62Lhigh, approximately one-half of them reside in the CD25high regulatory population.
|
2m to control for differences in cDNA amount, Foxp3 mRNA was expressed at significantly higher levels in both the DR+ and DR– Treg, compared with the Tresp and the other potentially in vivo-activated populations (CD62L–). From similar analyses on 14 different donor-derived Treg subsets, the DR+ Treg population always expressed a significantly higher level of Foxp3 message than the DR– Treg (p < 0.0067, Wilcoxon matched-pairs test). In total, these data demonstrated that both the DR+ and DR– Treg contained >40-fold greater levels of Foxp3 message, compared with ex vivo CD4+CD25– T cells.
To assess Foxp3 expression at the protein level, DR+ Treg and DR– Treg were permeablilized and stained for Foxp3 directly after their ex vivo isolation (Fig. 1C). The FACS profile histograms in the top row demonstrate that the DR+ Treg are highly homogeneous for Foxp3 as all are Foxp3+, while 12% of the DR– Treg population do not stain over background. When the levels of Foxp3 protein expressed by the DR+ and DR– Treg populations are compared (fourth histogram), not only are the DR+ Treg more homogeneous for Foxp3 expression, they also express higher amounts of the protein. Two-parameter dot-plot analysis comparing the expression of HLA-DR and intracellular Foxp3 indicated that, although the DR+ Treg exhibit a range of HLA-DR expression, they all contain the same high level of Foxp3 protein (Fig. 1D). Thus, although all DR+ Treg express high amounts of Foxp3 protein, there is no relationship between the extent of HLA-DR expression and the extent of Foxp3 expression. In further characterizations, although the DR+ Treg are highly homogeneous for Foxp3, 40% of the DR+ Treg are CD45RA+RO–, while <10% of the DR– Treg are CD45RA+RO– (data not shown). Thus, these data suggest that the DR– Treg (CD4+CD25highCD62Lhigh) population is different from a newly described DR– CD45RA+RO– regulatory population (32).
Distinct regulatory activity of DR+ and DR– Treg populations
We then examined whether the DR+ and DR– Treg populations were both capable of suppressing the proliferation of CD4+CD25– Tresp in the in vitro coculture assay. In these experiments, because the total CD25highCD62Lhigh T cells that represent only 2.5% of the total CD4 population were further sorted into separate DR+ and DR– populations, the cell separations typically produced 0.8 x 105 DR+ Treg and 2 x 105 DR– Treg from 2 x 108 total PBMC. As a result, these cell populations were tested for regulatory function in a highly defined in vitro coculture microassay system that lacked APCs and included beads coated with CD3 alone (Fig. 2C), CD3/CD28 (Fig. 2B) to model interactions with APCs expressing CD80/86, or CD3/CD2 (Fig. 2A) to model signals generated through CD2 by interaction with the high levels of CD58 that are expressed on many types of cells, including accessory cells, endothelial cells, and activated T cells (33, 34). Proliferation and cytokine secretion were measured by [3H]thymidine incorporation and by a highly sensitive cytokine bead array. In response to CD3 stimulation, the 2500 CD4+CD25– Tresp proliferated, while the DR+ and DR– Treg subsets did not (Fig. 2C). In contrast, cocultures of Tresp stimulated in the presence of half as many DR+ or DR– Treg exhibited no proliferation. Because this CD3 stimulation alone is weak, proliferation could be detected only at day 5 and was accompanied by the secretion of IFN- only. These data demonstrate that both DR+ and DR– Treg subsets can exhibit in vitro suppressive function under conditions of weak strength of signal associated with CD3 cross-linking alone.
|
CD2 or CD28 costimulation (Fig. 2, A and B). In contrast with the day 5 measurement of proliferation in cultures stimulated with CD3 alone, the presence of a costimulatory signal resulted in an earlier Tresp proliferation that started at day 3 and was accompanied by increased cytokine secretion.
The DR+ and DR– Treg exhibited a striking difference in their kinetics of suppression subsequent to the engagement of CD2. Compared with the strong proliferation of Tresp, CD3/CD2-stimulation did not induce the proliferation of either the DR+ or DR– Treg subsets (Fig. 2A). The cocultures established with the DR+ Treg under CD3/CD2 stimulation exhibited a robust and early suppression (day 3) as indicated by the strong inhibition of proliferation and secretion of all measured cytokines, including IFN-, IL-10, and IL-4. In marked contrast, cocultures of DR– Treg and Tresp did not induce a similar early contact-dependent suppression, but rather, induced secretion of IL-4 and IL-10. This was followed by a late contact-dependent suppression (day 5) of both proliferation and IFN- secretion but not IL-4 or IL-10 production. Importantly, the late suppression was shown not to be due to culture exhaustion, because control cultures of Tresp alone were still exhibiting increasing proliferation at this point. Furthermore, the difference in kinetics of suppression by the DR+ and DR– Treg subsets was not simply a reflection of a delay in DR expression by DR– Treg as <5% of the DR– Treg expressed DR during the course of these 5-day cultures.
To determine the source of the increase in secreted IL-4 and IL-10 in supernatants from DR– Treg cocultures, CFSE-labeled Tresp, DR+ Treg, DR– Treg, or the specific Treg/CFSE Tresp cocultures were stimulated with CD3/CD2 and exogenous IL-2 and stained for intracellular cytokine production (Fig. 2D). IL-2 was added to augment cytokine production, although it also abrogated coculture suppression as expected. Also, although the DR+ Treg do not typically secrete any detectable cytokines, they were similarly analyzed to allow the comparison between the DR+ and DR– Treg populations. These data demonstrate that, while both the DR– Treg and Tresp subsets can produce IL-4 and IL-10 under these conditions, a greater proportion of DR– Treg produced IL-4 (25%) and IL-10, while a greater proportion of the Tresp produced IFN- (18%). Thus, the DR– Treg population was more prone to produce Th2 cytokines, while the Tresp population was primarily responsible for the production of Th1 cytokines. In additional studies, half of the Th2 (IL-4) cytokine-producing DR– Treg were found to express Foxp3 protein (data not shown), indicating that at least a significant portion of the Th2 cytokines produced in these cultures are derived from DR–, Foxp3+ natural Treg.
The qualitatively different and stronger signals associated with CD28 costimulation again altered the behavior of these functionally distinct Treg populations. With CD28 costimulation, the Tresp exhibited greater levels of proliferation and cytokine production (Fig. 2C), while the DR+ and DR– Treg populations still remained unresponsive. In contrast with CD2 costimulation, DR+ Treg cocultures established with CD3/CD28 stimulation did not induce early suppression. Rather, CD3/CD28 stimulated cocultures containing either DR+ or DR– Treg subsets exhibited the same delayed suppression of Tresp proliferation. Importantly, while these CD3/CD28 cocultures exhibited the same late suppression of proliferation, only the cocultures established with DR– Treg also induced an increase in Th2 cytokine secretion. Because the Tresp-only cultures continued to increase their proliferation during this time, the decrease in proliferation in the 4- and 5-day cocultures was not due to culture exhaustion.
Differential expression of IL-10 and kinetics of superinduction of Foxp3 by stimulated DR+ and DR– Treg subsets
Due to the distinct kinetics of inhibition by DR+ and DR– Treg, we examined whether these regulatory populations alter their expression of Foxp3 mRNA over time in relation to their temporal changes in suppressive activity. Because CD3/CD2 costimulation produces a striking difference in the kinetics of suppression by these two Treg subsets, cultures were set up in which Tresp alone, DR+ Treg alone, or DR– Treg alone were stimulated with CD3/CD2 in the presence or absence of IL-2. The cells were recollected after 3 and 5 days of stimulation with and without IL-2. Real-time PCR analysis of the mRNA isolated from these subsets indicated that, after 3 days of stimulation in the absence of IL-2, the levels of Foxp3 mRNA did not change appreciably from direct ex vivo levels (Fig. 3A). In contrast, when stimulated in the presence of IL-2, the DR+ Treg exhibit an additional 8-fold superinduction of Foxp3 message on day 3 when they also exhibit strong suppression. Yet, after 5 days of stimulation, the level of DR+ Treg Foxp3 expression is greatly decreased. In comparison, in response to the same IL-2 supplemented stimulation, the DR– Treg do not exhibit their maximal induction of Foxp3 mRNA until day 5, which correlates with their kinetics of maximal suppressive activity. Low-level Foxp3 mRNA also was induced in similarly stimulated Tresp, although the levels were significantly lower than those found in the Treg populations. Thus, these data indicate that, while Foxp3 mRNA can be expressed in activated, human nonregulatory CD4 T cells, as reported by others (35), it is expressed at >20-fold higher levels by activated Treg populations in an IL-2-dependent fashion, and its induction correlates with early suppression in DR+ Treg cocultures and delayed regulation in DR–Treg cocultures. The similar analysis of RNA isolated from DR+ and DR– Treg that had been reisolated from suppressive CD3/CD2 stimulated cocultures (no IL-2) showed a highly similar further induction of Treg Foxp3 message (data not shown).
|
2m by real-time PCR (Fig. 3B). As indicated, CD3/CD2-stimulated DR+ Treg did not express IL-10 mRNA regardless of the presence or absence of exogenous IL-2, while both the DR– Treg and Tresp populations produced message for IL-10 within 3 days of stimulation. IL-10 is a component of the suppressive repertoire of DR– Treg
Studies were then performed to assess the contact dependence and potential involvement of IL-10 in DR+ or DR– Treg suppression because the classical CD4+CD25+ Treg are known to suppress via a cell contact-mediated mechanism, while the ability of the DR– Treg to produce IL-10 suggests that they could suppress via its secretion. Contact dependence was assessed by Transwell assays in which the Tresp were cultured in the lower chamber and the DR+ or DR– Treg were stimulated either separately in the upper chamber or together with the Tresp in the lower chamber in the presence (open) or absence (filled) of neutralizing IL-10 mAb (Fig. 4). Both upper and lower chambers received CD2/CD3 stimulation, and proliferation in the lower chambers was assessed at day 5 when both the DR+ and DR– Treg cocultures show inhibition. The addition of IL-10 had no effect on the proliferation of the Tresp alone. The separate stimulation of DR+ Treg in the upper chamber also had no affect on the proliferation of Tresp in the lower compartment. However, if the DR+ Treg were stimulated in the lower chamber in contact with the Tresp, proliferation was suppressed by >50%. Thus, these data indicated that the DR+ Treg require cell contact to exert suppression.
|
IL-10. However, when IL-10 was neutralized in replicate wells, the ability of the DR– Treg to suppress across the membrane was abrogated as it resulted in full Tresp proliferation. Thus, the DR– Treg subset can directly suppress the proliferation of Tresp via the secretion of IL-10. However, these data also show that DR– Treg can induce a more vigorous suppression via an IL-10-independent, cell contact-dependent mechanism, because a significantly greater suppression occurred in the cocultures where cell contact was allowed that was only marginally reduced by neutralizing IL-10 (Fig. 4). Thus, suppression by DR+ Treg absolutely requires cell contact, while DR– Treg can suppress by either the secretion of IL-10 or via a cell contact-dependent mechanism. T cell clones from DR+ and DR– Treg populations exhibit an increased propensity to constitutively express CD25 and HLA-DR
A fundamental question in the biology of human T cells is whether the expression of MHC-DR on directly ex vivo-isolated CD4 T cells is a stable phenotype of a subset of T cells that would maintain DR expression with longer-term culture or whether DR is simply acquired with activation. It has been reported previously that suppressive human CD4+CD25+ Treg clones, generated by PHA/IL-2 stimulation, maintain high levels of CD25 on their surface 2 wk after in vitro stimulation, while CD25 expression is at baseline levels on non-Treg 2 wk after activation (27). We previously generated clones of CD4+CD25high cells under the same stimulatory conditions and observed similar findings (18). In this study, T cell clones were generated by single cell cloning of CD4+CD62Lhigh T cells sorted on the basis of CD25 and DR expression. Single-cell cloning was undertaken to enable the coordinate assessment of function and changes in DR expression after in vitro culture by individual T reg subsets representing the ex vivo DR+ and DR– populations.
DR+ Treg, DR– Treg, DR+ CD25–/low, or DR– CD25– (Tresp) cells from the CD4+CD62Lhigh population were sorted into single wells and expanded as described. Three weeks after a single restimulation, regardless of their level of expansion, every clone was tested for its ability to suppress the proliferation of freshly isolated autologous Tresp and was thus classified as suppressive (filled) or nonsuppressive (open) (Fig. 5). The two Treg populations exhibited markedly different cloning capacity as the DR+ Treg consistently grew less well, while over two-thirds of wells seeded with DR– Treg produced clones (Table I). When tested for functional activity, the DR+ and DR– Treg subsets gave rise to the highest frequencies of suppressive clones. However, in contrast with the assay of bulk ex vivo DR– Treg, the suppressive clones derived from the DR– Treg population no longer induced IL-4 or IL-10 during culture (Fig. 5B). Furthermore, when tested for surface expression of CD25 and DR 24 days after their last activation (Fig. 5D), 40% of the suppressive clones derived from the DR– Treg population now expressed DR (13 of 32), while 44% of those from the DR+ Treg population expressed DR. In striking contrast, no clones derived from the DR+ CD25–/low population and only 2% (1 of 50) of the DR– CD25–-derived clones expressed DR. Thus, the initial ex vivo expression of high CD25 levels strongly correlated with subsequent clonal expression of DR, while the initial expression of DR on CD25–/low cells did not predict long-term expression of DR. In addition, although both the DR+ and DR– CD25high populations gave rise to suppressive and nonsuppressive clones that did not express DR, virtually all the DR expressing clones were suppressive (4 of 5 and 13 of 14, respectively). Thus, DR expression is associated with a strong suppressive activity by ex vivo CD4+CD25high cells and by roughly half of the suppressive clones generated from either the DR+ or DR– Treg populations. These clone data indicate that the functional distinctions between the DR+ Treg and DR– Treg populations are observed only ex vivo, as further differentiated DR+ Treg or DR– Treg no longer function in a distinct fashion. Thus, these data lead us to hypothesize that DR may identify a mature and highly active regulatory CD4+CD25high cell and not a separate T cell lineage.
|
|
was performed on RNA isolated from a set of representative clones from each group (Fig. 6). Thus, although the DR+ and DR– Tresp ex vivo populations express little Foxp3 message as a whole, 18% of the clones derived from these populations expressed high levels of Foxp3 mRNA (Fig. 6A) after their clonal expansion, even though they were not suppressive in the in vitro assays. In contrast, 85% of the clones derived from the DR+ and DR– Treg populations were Foxp3high. These data also indicate that high levels of Foxp3 are maintained in the clones derived from the CD25high populations regardless of their functional suppressive ability. The amount of TGF mRNA (Fig. 6B) produced by each clone did not correlate with either its suppressive ability or its original cloning group. In contrast, the results of the IL-10 expression analysis (Fig. 6C) not only corroborate the lack of IL-10 production in cocultures with Treg clones, but also suggest that at least a subset of the suppressive clones derived from the DR– Tresp population may be Tr1-like in their mechanism of action, as they express IL-10 but lack Foxp3.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Because Treg contact-dependent suppression in rodents and in humans is not MHC restricted, what is the role of MHC-II expression on these cells? When we examined whether DR Abs that block MHC-dependent Ag presentation would block contact-dependent suppression by DR+ Treg, we found that coculture suppression proceeded in the presence of DR (data not shown). Nevertheless, because it seems unlikely that MHC-II expression is simply a "marker" of highly activated Treg, it is possible that other invariant counterligands of DR, such as CD4 or LAG-3, may be important. Alternatively, as expression of class II molecules has been shown to be critical in maintaining the pool of CD4+CD25+ T cells in mice in vivo (36) and CD4+CD25+ Treg have an increased propensity for self-peptide recognition (37), DR-expressing Treg may be involved in homeostatic maintenance of Treg cells in vivo by presenting self Ags to other Treg.
Whatever functional role MHC-DR may play in contact-mediated in vitro suppression, we postulated that DR expression may identify a distinct population of in vivo CD25high Treg. This was examined by single-cell cloning populations of CD4+CD62Lhigh T cells sorted on the basis of DR or CD25high expression. Surprisingly, high ex vivo expression of CD25 determined whether CD4 clones would express DR in their resting state. That is, suppressive CD25high clones expressed DR whether or not they expressed DR at the time of cloning, while initial expression of DR only (DR+ CD25–/low cells) did not predict the later, long-term constitutive expression of DR on differentiated T cell clones. Furthermore, DR+ T cell clones generated from the DR– Treg subset functioned similarly to the ex vivo DR+ Treg population, as they suppressed in the absence of the induction of IL-4 or IL-10.
Almost all of the clones that exhibited constitutive DR expression had been derived from the CD25high Treg populations. Although the DR+ Treg exhibit much lower cloning efficiency, approximately half of the suppressive DR+ Treg-derived clones and a similar percentage of the suppressive DR– Treg-derived clones expressed HLA-DR constitutively. In contrast, only 1 of 50 clones generated from the DR– Tresp population constitutively expressed DR or exhibited in vitro suppressive activity. The clones derived from the DR+ Tresp population did not subsequently exhibit constitutive DR expression, suggesting that their original expression of DR may have been related to their state of activation. Thus, even though ex vivo DR expression does define differences in the ex vivo function of Treg populations, the clonal analysis indicates that, while all suppressive clones do not necessarily express DR, all DR expressing clones are suppressive. The real-time PCR analysis of representative clones from these T cell subsets revealed that suppression by the clones derived from the DR+ and DR– Treg subsets did not correlate with production of IL-10 or TGF, suggesting that these inhibitory cytokines are not involved in their suppressive mechanism. In contrast, because two of the suppressive clones generated from the Tresp contained high levels of IL-10 and lacked Foxp3 mRNA, these clones may represent Tr1 regulatory cells. More surprising was the observation that 90% of the clones derived from DR+ Treg population maintained high expression of Foxp3 regardless of their in vitro suppressive activity or their expression of DR. These data indicate that nonsuppressive DR+ Treg-derived clones did not originate from activated effector cells contaminating the original Treg population. These data also confirm that within the CD25high population, DR expression identifies a population of distinct contact-dependent suppressor cells that maintain high levels of Foxp3.
We also observed a strong correlation between the kinetics of induced Foxp3 transcription and contact-dependent suppression in both DR+ and DR– Treg populations. Foxp3 was typically expressed at slightly higher levels ex vivo in DR+ than in the DR– Treg. When stimulated in the presence of exogenous IL-2, the two Treg subsets exhibited differences in the kinetics of additional Foxp3 mRNA induction. Because the levels of Foxp3 message did not change if either type of Treg was stimulated in the absence of IL-2, IL-2 may play a role in maintaining the suppressive activity of these cells as has been suggested previously (38). Therefore, in the typical CD3/CD2-stimulated coculture, low amounts of IL-2 may be provided by the activated Tresp that is used by the CD25high DR+ and DR– Treg to induce distinct kinetics of Foxp3 superinduction and suppressive function.
Although a fundamental role for Foxp3 in regulation was demonstrated by studies showing that transfection of the Foxp3 gene into mouse or human CD4+CD25– T cells resulted in acquired regulatory phenotype and function (30, 31, 39), it is not present in all types of Treg. The ability of DR– Treg to produce IL-10 and suppress the proliferation of other T cells is reminiscent of Tr1 regulatory cells, while the strong induction of Th2 cytokines, the contact-dependent component of DR– Treg suppression (see Fig. 4), and the elevated Foxp3 expression by the DR– Treg distinguish them from Tr1 cells (35). Rather, the analysis of clonal gene expression may suggest that Tr1 regulatory cells originate from within the DR– Tresp population. Regardless, because the cell contact-dependent suppression by freshly isolated DR+ and DR– Treg subsets is not abrogated by the neutralization of IL-10 or TGF (data not shown), they are truly distinct from Tr1 regulatory cells.
IL-10 has been shown to be important for CD4+CD25+ Treg function in a number of in vivo disease models in mice, although the corresponding in vitro assays often showed no role for IL-10. Specifically, an in vivo but, not in vitro, dependence on IL-10 has been shown in studies addressing Treg activity in autoimmune colitis (12, 40, 41), expansion of the memory CD4 T cells in lymphopenic hosts (20), and the development of experimental autoimmune encephalomyelitis (42), while there are other in vivo autoimmune models that show no IL-10 involvement in Treg activity (11, 40, 43, 44). The in vivo and in vitro discrepancies could be due to the study of heterogeneous Treg populations and the fact that Treg likely act individually in vivo but are assayed as total populations when tested in vitro. We have found that the individual types of Treg (DR+ and DR–) alter each other’s function when assayed together. These data and previous studies indicate that the individual characteristics of the DR+ and DR– Treg populations are lost when they are assayed together for in vitro function (data not shown and Refs.15 , 17 , 18 , 45). Thus, the additional interactions that occur in the study of mixed Treg populations may obscure the mechanisms of action of specific types of Treg cells, indicating that, in future human studies, subfractionation of these Treg populations will be required.
In summary, we demonstrate that MHC-DR expression defines two functionally distinct CD4+CD25+ Treg populations in humans. The DR+ Treg population exhibits early contact-dependent suppression, while the DR– Treg population initially skews the immune response toward the production of Th2 cytokines before initiating a late, Foxp3-associated, contact-dependent suppression. Single-cell cloning of CD25high Treg reveals that DR expression is associated with suppressive function. These data indicate that MHC-II expression on CD4+CD25high Treg identifies a mature, distinct Treg involved in contact-dependent in vitro suppression. Thus, future analysis of human Treg will require examination of these functionally distinct subsets found in the total CD4+CD25high Treg population.
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 This work was supported by the National Institutes of Health Grants U01DK6192601, R01NS2424710, P01AI39671, and P01NS38037; National Multiple Sclerosis Society Grants RG2172C9 and RG3308A10; and grants from the 2004 Federation of Clinical Immunology Society Centers of Excellence Amgen Award. 
2 Address correspondence and reprint requests to Dr. Clare Baecher-Allan, Harvard Medical School, 77 Avenue Louis Pasteur, Boston, MA 02115. E-mail address: callan{at}rics.bwh.harvard.edu
3 Abbreviations used in this paper: MHC-II, MHC class II; DR, determinant; Treg, regulatory T cell; Tresp, responder T cell; 2m, 2-microglobulin; CBA, cytometric bead array.
Received for publication September 2, 2005. Accepted for publication January 31, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
, and CTLA4. J. Immunol. 171: 5012-5017. , but not interleukin 10, and are distinct from type 1 T regulatory cells. J. Exp. Med. 196: 1335-1346. . J. Immunol. 174: 1446-1455. -dependent mechanism. J. Immunol. 170: 2516-2522. 1 production and responsiveness. J. Exp. Med. 196: 237-246. This article has been cited by other articles:
|
|
 |
|
  C. W. Ashley and C. Baecher-Allan Cutting Edge: Responder T Cells Regulate Human DR+ Effector Regulatory T Cell Activity via Granzyme B J. Immunol., October 15, 2009; 183(8): 4843 - 4847. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. M. Costantino, H. L. Ploegh, and D. A. Hafler Cathepsin S Regulates Class II MHC Processing in Human CD4+ HLA-DR+ T Cells J. Immunol., July 15, 2009; 183(2): 945 - 952. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Mjosberg, J. Svensson, E. Johansson, L. Hellstrom, R. Casas, M. C. Jenmalm, R. Boij, L. Matthiesen, J.-I. Jonsson, G. Berg, et al. Systemic Reduction of Functionally Suppressive CD4dimCD25highFoxp3+ Tregs in Human Second Trimester Pregnancy Is Induced by Progesterone and 17{beta}-Estradiol J. Immunol., July 1, 2009; 183(1): 759 - 769. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Pyne, X. Hu, K. Wang, E. Rossin, T.-I Lin, L. M. Maier, C. Baecher-Allan, G. J. McLachlan, P. Tamayo, D. A. Hafler, et al. Automated high-dimensional flow cytometric data analysis PNAS, May 26, 2009; 106(21): 8519 - 8524. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Beriou, C. M. Costantino, C. W. Ashley, L. Yang, V. K. Kuchroo, C. Baecher-Allan, and D. A. Hafler IL-17-producing human peripheral regulatory T cells retain suppressive function Blood, April 30, 2009; 113(18): 4240 - 4249. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. L. De Jager, C. Baecher-Allan, L. M. Maier, A. T. Arthur, L. Ottoboni, L. Barcellos, J. L. McCauley, S. Sawcer, A. Goris, J. Saarela, et al. The role of the CD58 locus in multiple sclerosis PNAS, March 31, 2009; 106(13): 5264 - 5269. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. F. Hutton, T. Gargett, T. J. Sadlon, S. Bresatz, C. Y. Brown, H. Zola, M. F. Shannon, R. J. D'Andrea, and S. C. Barry Development of CD4+CD25+FoxP3+ regulatory T cells from cord blood hematopoietic progenitor cells J. Leukoc. Biol., March 1, 2009; 85(3): 445 - 451. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Gobert, I. Treilleux, N. Bendriss-Vermare, T. Bachelot, S. Goddard-Leon, V. Arfi, C. Biota, A. C. Doffin, I. Durand, D. Olive, et al. Regulatory T Cells Recruited through CCL22/CCR4 Are Selectively Activated in Lymphoid Infiltrates Surrounding Primary Breast Tumors and Lead to an Adverse Clinical Outcome Cancer Res., March 1, 2009; 69(5): 2000 - 2009. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Monti, M. Scirpoli, P. Maffi, L. Piemonti, A. Secchi, E. Bonifacio, M.-G. Roncarolo, and M. Battaglia Rapamycin Monotherapy in Patients With Type 1 Diabetes Modifies CD4+CD25+FOXP3+ Regulatory T-Cells Diabetes, September 1, 2008; 57(9): 2341 - 2347. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Venken, N. Hellings, T. Broekmans, K. Hensen, J.-L. Rummens, and P. Stinissen Natural Naive CD4+CD25+CD127low Regulatory T Cell (Treg) Development and Function Are Disturbed in Multiple Sclerosis Patients: Recovery of Memory Treg Homeostasis during Disease Progression J. Immunol., May 1, 2008; 180(9): 6411 - 6420. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Swiatek-de Lange, W. Rist, H. F. Stahl, A. Weith, and M. C. Lenter Comment on "MHC Class II Expression Identifies Functionally Distinct Human Regulatory T Cells" J. Immunol., March 15, 2008; 180(6): 3625 - 3625. [Full Text] [PDF]
|
|
|
|
|
|
C. Baecher-Allan and D. A. Hafler Response to Comment on "MHC Class II Expression Identifies Functionally Distinct Human Regulatory T Cells" J. Immunol., March 15, 2008; 180(6): 3626 - 3626. [Full Text] [PDF]
|
|
|
|
 |
|
 B. Santner-Nanan, N. Seddiki, E. Zhu, V. Quent, A. Kelleher, B. F. de St Groth, and R. Nanan Accelerated age-dependent transition of human regulatory T cells to effector memory phenotype Int. Immunol., March 1, 2008; 20(3): 375 - 383. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. J. D. van Mierlo, H. U. Scherer, M. Hameetman, M. E. Morgan, R. Flierman, T. W. J. Huizinga, and R. E. M. Toes Cutting Edge: TNFR-Shedding by CD4+CD25+ Regulatory T Cells Inhibits the Induction of Inflammatory Mediators J. Immunol., March 1, 2008; 180(5): 2747 - 2751. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. M. Tiemessen, A. L. Jagger, H. G. Evans, M. J. C. van Herwijnen, S. John, and L. S. Taams CD4+CD25+Foxp3+ regulatory T cells induce alternative activation of human monocytes/macrophages PNAS, December 4, 2007; 104(49): 19446 - 19451. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Borsellino, M. Kleinewietfeld, D. Di Mitri, A. Sternjak, A. Diamantini, R. Giometto, S. Hopner, D. Centonze, G. Bernardi, M. L. Dell'Acqua, et al. Expression of ectonucleotidase CD39 by Foxp3+ Treg cells: hydrolysis of extracellular ATP and immune suppression Blood, August 15, 2007; 110(4): 1225 - 1232. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Brusko, C. Wasserfall, K. McGrail, R. Schatz, H. L. Viener, D. Schatz, M. Haller, J. Rockell, P. Gottlieb, M. Clare-Salzler, et al. No Alterations in the Frequency of FOXP3+ Regulatory T-Cells in Type 1 Diabetes Diabetes, March 1, 2007; 56(3): 604 - 612. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Kinter, J. McNally, L. Riggin, R. Jackson, G. Roby, and A. S. Fauci Suppression of HIV-specific T cell activity by lymph node CD25+ regulatory T cells from HIV-infected individuals PNAS, February 27, 2007; 104(9): 3390 - 3395. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Nadkarni, C. Mauri, and M. R. Ehrenstein Anti-TNF-{alpha} therapy induces a distinct regulatory T cell population in patients with rheumatoid arthritis via TGF-{beta} J. Exp. Med., January 22, 2007; 204(1): 33 - 39. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Seddiki, B. Santner-Nanan, J. Martinson, J. Zaunders, S. Sasson, A. Landay, M. Solomon, W. Selby, S. I. Alexander, R. Nanan, et al. Expression of interleukin (IL)-2 and IL-7 receptors discriminates between human regulatory and activated T cells J. Exp. Med., July 10, 2006; 203(7): 1693 - 1700. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Liu, A. L. Putnam, Z. Xu-yu, G. L. Szot, M. R. Lee, S. Zhu, P. A. Gottlieb, P. Kapranov, T. R. Gingeras, B. F. de St. Groth, et al. CD127 expression inversely correlates with FoxP3 and suppressive function of human CD4+ T reg cells J. Exp. Med., July 10, 2006; 203(7): 1701 - 1711. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) based on coculture proliferation. Long-term DR expression depended on initial ex vivo expression of CD25high and was associated with suppression by ex vivo CD4+CD25high cells and by long-term T cell clones.