Natural CD4+CD25+ regulatory T (CD4+CD25+ T reg) cells play a key role in the immunoregulation of autoimmunity. However, little is known about the interactions between CD4+CD25+ T reg cells and autoreactive T cells. This is due, in part, to the difficulty of using cell surface markers to identify CD4+CD25+ T reg cells accurately. Using a novel real-time PCR assay, mRNA copy number of FoxP3, TGFβ1, and interleukin (IL)-10 was measured in single cells to characterize and quantify CD4+CD25+ T reg cells in the nonobese diabetic (NOD) mouse, a murine model for type 1 diabetes (T1D). The suppressor function of CD4+CD25+CD62Lhi T cells, mediated by TGFβ, declined in an age-dependent manner. This loss of function coincided with a temporal decrease in the percentage of FoxP3 and TGFβ1 coexpressing T cells within pancreatic lymph node and islet infiltrating CD4+CD25+CD62Lhi T cells, and was detected in female NOD mice but not in NOD male mice, or NOR or C57BL/6 female mice. These results demonstrate that the majority of FoxP3-positive CD4+CD25+ T reg cells in NOD mice express TGFβ1 but not IL-10, and that a defect in the maintenance and/or expansion of this pool of immunoregulatory effectors is associated with the progression of T1D.

T1D is an autoimmune disease that is characterized by the selective destruction of the insulin-producing β cells which reside in the islets of Langerhans (1, 2). Studies in the NOD mouse, a spontaneous model of T1D, have demonstrated that CD4+ and CD8+ T cells are the primary mediators of β cell destruction (1, 2). The critical events that contribute to the breakdown of self-tolerance to β cells are not well-understood, although defects in central and peripheral tolerance have been implicated in the development and differentiation of pathogenic T effector cells (37).

The progression of T1D in NOD mice is marked by two general “checkpoints” (8). The first checkpoint is associated with infiltration of the islets by macrophages, DCs, and B and T cells. This insulitis begins at ∼3 wk of age and is well-established in 8–10-wk-old NOD mice. Despite extensive insulitis, β cell destruction is limited and NOD mice remain diabetes-free during this period. The second checkpoint corresponds with a shift from “benign” to “aggressive” insulitis that begins at ∼12 wks of age. At this time, β cells are destroyed efficiently to promote overt diabetes. Typically, by 30 wk of age, ∼80% and ∼20% of NOD female and male mice, respectively, have developed diabetes. How the progression from nondestructive to destructive insulitis is regulated, or why the disparity in overt diabetes exists between NOD female and male mice, are not clear. However, interrelated events, including islet recruitment of high avidity/affinity β cell–specific T cells that exhibit a type 1 phenotype (9) and failing peripheral immunoregulatory mechanisms, are believed to contribute (57).

Aberrant immunoregulation in NOD mice initially was viewed in the context of a functional imbalance between types 1 and 2 T cells. Various studies have demonstrated that induction of type 2 effectors, which are characterized by the secretion of IL-4 and IL-10, effectively prevents and/or suppresses the diabetogenic response in NOD mice (5, 1012). However, it is now apparent that immunoregulation of autoreactive T cells in general is highly complex and likely involves a heterogeneous group of immunoregulatory T (and non–T) cell effectors. Indeed, induction of type 1 regulatory T (T reg) cells, IL-10–only–producing CD4+ T cells, and TGFβ1-secreting Th3 cells have been reported to prevent tissue-specific autoimmunity (1315).

Another subset of immunoregulatory effectors consists of a naturally occurring population of CD4+CD25+ T reg cells. CD4+CD25+ T reg cells upon TCR-mediated stimulation are characterized by a potent in vitro and in vivo capacity to suppress proliferation and effector function of CD4+ and CD8+ T cells, in addition to modulating the antigen-presenting function of DCs (6, 1619). Typically accounting for 5–10% of murine peripheral CD4+ T cells, CD4+CD25+ T reg cells play key roles in modulating immune responses in a variety of models (6, 1619). For example, a reduced number of CD4+CD25+ T reg cells was associated with exacerbation of T1D in NOD mice which lacked CD80/CD86 or CD28 expression (6).

Current in vivo and in vitro studies strongly support the view that CD4+CD25+ T reg cells represent a unique lineage of immunoregulatory cells. However, accurate identification, characterization, and quantification of CD4+CD25+ T reg cells have been problematic. The IL-2Rα chain (CD25) has been used widely for the identification and isolation of CD4+CD25+ T reg cells, yet this marker also is expressed by activated CD4+ Th and memory T cells. Other cell surface markers that are used to define CD4+CD25+ T reg cells, including CD62L, cytotoxic T lymphocyte–associated antigen 4 (CTLA-4), glucocorticoid-induced TNF receptor gene (GITR), CD45RB, and CD103, also are shared by “conventional” CD4+ Th cells (20). Recently, reports demonstrated that the transcription factor, FoxP3, is expressed specifically by CD4+CD25+ T reg cells (2123). Furthermore, the Rudensky and Sakaguchi groups demonstrated that FoxP3 is a master regulator of CD4+CD25+ T reg cells differentiation. Transfer of FoxP3 cDNA into naive CD4+ Th cells was sufficient to induce differentiation that phenotypically and functionally resembled CD4+CD25+ T reg cells (21, 23). The precise mode of suppression by CD4+CD25+ T reg cells is not clear, but includes direct cell–cell contact and/or production of inhibitory cytokines, such as TGFβ1 and IL-10 (2426). Cell surface and secreted TGFβ1 have been associated with cell–cell contact dependent and independent mechanisms of CD4+CD25+ T reg cell–mediated suppression (24, 25). These different mechanisms of suppression may reflect distinct subsets of CD4+ CD25+ T reg cells and/or other types of immunoregulatory effector cells.

With this in mind, we established an assay that was based on real-time PCR to quantitate and characterize accurately FoxP3 expressing CD4+CD25+ T reg cells at the single cell level. This assay was used to test the hypothesis that progression of T1D in NOD mice is due to defects in CD4+CD25+ T reg cell numbers and/or effector function. Evidence is provided that demonstrates that the progression of benign to aggressive insulitis corresponds with a temporal decrease in peripheral tolerance as measured by reduced frequencies of FoxP3 and TGFβ1 coexpressing CD4+CD25+ T reg cells residing in the PLNs and islets of NOD female mice.

Results

The suppressor function of CD4+CD25+ T reg cells decreases in an age-dependent manner

Studies have suggested that defects in CD4+CD25+ T reg cell function may contribute to the progression of β cell autoimmunity in NOD mice (6, 7). Accordingly, the suppressor function of CD4+CD25+ T cells that were prepared from PLNs of 4-, 8-, and 16-wk-old NOD female mice—representing different stages of β cell autoimmunity—was investigated in vitro. PLNs are reported to be a key site for activation and tolerance induction of β cell specific T cells (27). PLN CD4+CD25+ T cells that were isolated from 4- and 8-wk-old NOD female mice inhibited proliferation of CD4+CD25- T cells similarly in cultures that were stimulated with anti-CD3 mAb and irradiated syngeneic splenocytes depleted of T cells (Fig. 1 

A). In contrast, a significant decrease in the suppressor function of PLN CD4+CD25+ T cells that were isolated from 16-wk-old mice versus 4- and 8-wk-old NOD mice was observed at limiting cell numbers (P ≤ 0.02) (Fig. 1 A).

Next, an in vivo model was used to test CD4+CD25+ T cell suppressor function. CD4+CD25+ T cells (5 × 105) were cotransferred with BDC2.5 CD4+CD25- T cells (5 × 106) into NOD.scid mice, and the development of diabetes was monitored. NOD BDC2.5 mice are transgenic for the diabetogenic BDC2.5 clonotypic TCR, and upon adoptive transfer into NOD.scid mice cause a rapid onset of disease (28). Coadoptive transfer of PLN CD4+CD25+ T cells that were isolated from 16-wk-old NOD female mice had no significant effect on the time of diabetes onset relative to recipients of CD4+CD25 BDC2.5 T cells alone (Fig. 1 B). However, a significant delay in diabetes onset, beginning at day 15 after transfer, was detected in NOD.scid recipients that were coinjected with PLN CD4+CD25+ T cells that were prepared from 8 wk-old NOD female mice (P = 0.007). The development of diabetes was delayed further in NOD.scid mice that received PLN CD4+CD25+ T cells that were isolated from 4-wk-old versus 8 wk-old NOD female mice (P = 0.009) (Fig. 1 B).

To determine the phenotype and effector function of CD4+CD25+ T reg cells more accurately, the in vitro suppressor capacity of varying numbers of PLN CD62Lhi and CD62Llo T cells was examined. The cell surface profile of CD62Lhi versus CD62Llo is used commonly to discriminate between CD4+CD25+ T reg cells and activated CD4+ Th cells, respectively (29, 30). Similar to that observed in the coadoptive transfer experiments (Fig. 1 B), PLN CD62Lhi T cells that were prepared from 4-wk-old NOD female mice proved to be the most effective at suppressing CD4+CD25 T cell proliferation (Fig. 2 

A). An intermediate level of suppression was detected for CD62Lhi T cells that were isolated from 8-wk-old NOD female mice, whereas CD62Lhi T cells that were prepared from 16-wk-old NOD female mice were the least effective (Fig. 2 A). The suppressor activity of PLN CD62Llo T cells, albeit markedly reduced relative to CD62Lhi T cells, exhibited a similar age-dependent decline (Fig. 2 B). These results demonstrate that the capacity to mediate suppression by PLN CD4+CD25+ T cells, specifically CD62Lhi T cells, declines in a temporal manner.

An increase in PLN CD62Lhi T cells is detected as β cell autoimmunity progresses

Quantitative and/or qualitative changes may account for the temporal decrease in CD4+CD25+ T reg cell function that was detected. To distinguish between these possibilities, the number and frequency of PLN CD62Lhi T cells were determined in prediabetic and diabetic NOD female mice. Notably, the number of CD62Lhi T cells in the PLNs of prediabetic NOD female mice increased with age (Table I)

. An approximate threefold increase in the number of CD62Lhi T cells was detected in the PLNs of 16-wk-old (147 ± 10) versus 4-wk-old (54 ± 5) NOD female mice (P < 10−4; Table I). In parallel, the number of CD62Llo T cells also increased with age, so that the overall percentage of CD62Lhi T cells was maintained during the progression of preclinical β cell autoimmunity (Fig. 3 

, Table I). In the PLNs of recent onset diabetic NOD female mice, no significant change in the number of CD62Lhi T cells was detected relative to 16-wk-old mice, although the percentage of these T cells was reduced as a result of increased numbers of CD62Llo T cells (Fig. 3, Table I). Finally, within the pool of PLN CD62Lhi T cells from prediabetic or diabetic NOD female mice, no significant difference was detected in the number or percentage of T cells which expressed additional markers that were associated with CD4+CD25+ T reg cells, including CTLA-4, GITR, CD69, CD103, and CD45RB (data not shown). Together, these results indicate that the number of CD62Lhi T cells in the PLNs of NOD female mice increases as β cell autoimmunity progresses, and is maintained at the onset of diabetes. Furthermore, an increase in the number of CD62Llo T cells, and not a decrease in CD62Lhi T cells in the PLN, correlates with the onset of diabetes.

A temporal decrease in FoxP3-expressing CD62Lhi T cells is unique to NOD mouse PLNs

The above findings indicated that despite an increase in the number of CD62Lhi T cells up to 16 wk in age, suppressor activity diminished within this pool. To further define the defect in PLN CD4+CD25+ T reg cells, mRNA expression of FoxP3, TGFβ, and IL-10 were measured. For this purpose, a real-time PCR assay was used to measure RNA transcript copy number in individual cells. Single CD62Lhi, CD62Llo, and CD4+CD25 T cells were sorted from PLNs that were prepared from pools of four prediabetic or recent diabetic NOD female mice; FoxP3, TGFβ1, and IL-10 transcript number were measured in the same individual cells. Consistent with previous findings, only a small percentage (<3%) of CD4+CD25 T cells expressed low copies (∼100–200) of FoxP3 mRNA per cell (Fig . 4 A and B; refs. 29, 30). In contrast, the majority of PLN CD62Lhi T cells (70 ± 5%) that were prepared from 4-wk-old NOD female mice expressed an average of 105 copies of FoxP3 mRNA per cell (Fig. 4, A and B)

. Strikingly, a marked decrease in FoxP3-expressing CD62Lhi T cells was observed in the PLNs of 8-wk-old NOD female mice (48.8 ± 6.3%; P = 0.04; Fig. 4 A). Furthermore, a progressive decrease in the percentage of FoxP3-expressing CD62Lhi T cells was detected at later stages of β cell autoimmunity. In the PLNs of 16-wk-old and diabetic NOD female mice, only 28.8 ± 3.8% and 25 ± 5% of CD62Lhi T cells, respectively, expressed FoxP3 (Fig. 4 A). Despite a progressive loss of CD62Lhi T cells that expressed the transcription factor, mRNA copy number did not vary significantly among FoxP3-positive T cells that were prepared from the respective groups of NOD female mice (Fig. 4 B).

A significant percentage of CD62Llo T cells also was found to express FoxP3 at comparable levels to CD62Lhi T cells (Fig. 4, A and B). Within the CD62Llo pool, the highest percent of FoxP3-expressing T cells was observed in PLNs of 4-wk-old NOD female mice (35.0 ± 1.0%); this decreased with age (Fig. 4 A). In 8-wk-old NOD female mice for example, 22.5 ± 5.0% of CD62Llo T cells expressed FoxP3 (Fig. 4 A).

To determine whether the age-dependent decrease in FoxP3-expressing T cells was unique to NOD mice, CD4+ T cells were prepared from the PLNs of B6 and NOR female mice that varied in age, and were analyzed as above. The NOR mouse is a NOD-related MHC syngeneic recombinant strain that contains ∼12% C57BL/KsJ-derived genes, exhibits peri-insulitis and only minimal intrainsulitis, and fails to develop overt diabetes (31). Approximately 40% and 20% of CD62Lhi and CD62Llo T cells, respectively, expressed FoxP3 independent of the age of the NOR or B6 female mice (Fig. 4, C and E). Furthermore, FoxP3 transcript copy numbers that were expressed by PLN CD62Lhi T cells were equivalent between NOR and B6 mice (Fig. 4, D and F). At 16 wk of age, CD62Lhi T cells that were prepared from NOD female mice expressed on average ∼10-fold less FoxP3 transcripts relative to that detected in NOR and B6 mice (P ≤ .0003; Fig. 4, B, D, and F). These data demonstrate that despite an expanding number of PLN CD62Lhi (and CD62Llo) T cells, the percentage of FoxP3-expressing T cells declines as β cell autoimmunity progresses in NOD female mice. In contrast, a temporal decrease in the percentage of FoxP3-expressing CD62Lhi (and CD62Llo) T cells was not detected in NOR and B6 mice (Fig. 4, C and E, and Table S1). Importantly, the percentage of FoxP3-expressing cells in NOD female mice at 16 wk of age, when destructive insulitis has begun, was significantly lower than in NOR and B6 mice (Fig. 4, A, C, and E; Table I; and Table S1). Although the actual number of FoxP3+CD62Lhi T cells in NOD female mice remained constant (Table I), the suppressive capacity of this population diminished with age (Fig. 1). Together, these data indicate that an optimal ratio of CD4+CD25+ T reg cells/effector T cells is needed to suppress disease progression.

The percentage of TGFβ1- and FoxP3-coexpressing PLN CD62Lhi T cells decreases as β cell autoimmunity progresses

Studies have shown that TGFβ1 and IL-10 are key cytokines that are expressed by CD4+CD25+ T reg cells (3234). Analysis of transcript copy number demonstrated that none of the FoxP3-positive CD62Lhi (or CD62Llo) T cells expressed detectable levels of IL-10 mRNA (data not shown). Conversely, >80% of FoxP3-expressing CD62Lhi T cells expressed TGFβ1, regardless of the age of the NOD female mice (Fig. 5 

A). In addition, TGFβ1 expression was detected in ≤7% of FoxP3-negative CD62Lhi (or CD62Llo) T cells (Table S2). TGFβ1 transcript copy number varied with age (Fig. 5 A). For example, in 4- and 8-wk-old NOD female mice, FoxP3-positive CD62Lhi T cells were defined by two distinct subsets based on high (1.2 to 1.3 × 106) versus low (2.1 to 4.9 × 103) TGFβ1 transcript copy number per T cell. However, in 16-wk-old or diabetic animals, high copy TGFβ1-expressing CD62Lhi T cells no longer were detected (Fig. 5 A). In comparison, TGFβ1 transcript copy number was relatively consistent in FoxP3-expressing CD62Lhi T cells that were prepared from different aged NOR and B6 female mice (Fig. 5, B and C). At 16 wk of age, FoxP3-positive CD62Lhi T cells that were prepared from NOD female mice expressed on average 10- and 100-fold fewer TGFβ1 transcripts compared with NOR and B6 mice, respectively (P ≤ 0.009; Fig. 5).

Analogous to CD62Lhi T cells, TGFβ was detected in >80% of FoxP3-expressing CD62Llo T cells that were prepared from PLNs of NOD, NOR, and B6 female mice, regardless of age (Fig. 5). Furthermore, transcript copy number varied only minimally among the respective groups within a given mouse strain, and between NOD, NOR, and B6 mice (Fig. 5). TGFβ1 mRNA was detected in ≤4% of CD62Llo T cells which lacked FoxP3 expression (Table S2). Together, these results indicate that the percentage of TGFβ1-expressing CD62Lhi (and CD62Llo) T cells, and the level of TGFβ1 expression decreases in NOD female mice as β cell autoimmunity progresses. Furthermore, the majority of FoxP3-positive CD4+CD25+ T reg cells coexpress TGFβ1, but not IL-10.

A temporal decrease in FoxP3- and TGFβ1-coexpressing CD62Lhi T cells is not detected in NOD male mice

Differences in the extent of pancreatic inflammation between NOD versus NOR or B6 female mice may have distinct effects on the activation and/or expansion of CD4+CD25+ T reg cells, which, in turn, may explain the profiles of FoxP3- and TGFβ1-coexpressing CD62Lhi (and CD62Llo) T cells that are seen in the respective mice. To rule out this possibility and determine whether decreasing FoxP3- and TGFβ1-coexpressing CD62Lhi in NOD female mice is a consequence or cause of disease progression, NOD male mice were studied. NOD male and female mice exhibit similar insulitis with age, despite the fact that male animals develop diabetes at a markedly reduced frequency. A similar increase in the absolute number of PLN CD62Lhi and CD62Llo T cells was detected in 4- and 16-wk-old NOD male and female mice (Table I). However, the percentage of PLN CD62Lhi (and CD62Llo) FoxP3-expressing T cells in NOD male mice remained relatively constant with age (Fig. 6, A and B)

; this resembles the profiles that were seen in NOR and B6 mice (Fig. 4, C–F). Furthermore, >80% of FoxP3-expressing CD62Lhi (and CD62Llo) T cells in NOD male mice expressed TGFβ1, regardless of age (Fig. 6 C). Moreover, in contrast to NOD female mice (Fig. 5 A), TGFβ1 transcript copy number in NOD male mice also remained constant (Fig. 6 C). TGFβ1 transcript copy number was increased ∼10-fold more in PLN CD62Lhi (or CD62Llo) FoxP3+ T cells from male versus female NOD mice at 16 wk of age (Fig. 5 A and Fig. 6 C).

Next, the in vitro suppressor function of PLN CD62Lhi T cells from NOD male mice was tested. PLN CD62Lhi T cells from 4- and 16-wk-old NOD male mice suppressed CD4+CD25 T cell proliferation similarly—and with significantly greater efficacy—to CD62Lhi T cells from 16-wk-old female mice (P < 0.001; Fig. 6 D). These results demonstrate that unlike NOD female mice, the percentage of FoxP3- and TGFβ1-coexpressing CD62Lhi T cells—and associated TGFβ1transcript copy number and suppressor function—remain constant with age.

Suppression mediated by PLN CD4+CD25+ T reg cells is TGFβ dependent

The above findings suggested that a progressive decline in the percentage of TGFβ1-expressing PLN CD4+CD25+ T reg cells promoted disease progression in NOD female mice. To determine a role in CD4+CD25+ T reg cell–mediated suppression, the level of TGFβ secretion in an in vitro suppression assay was investigated. As the number of CD4+CD25+ T cells was decreased, significantly reduced amounts of TGFβ were detected in cultures that were established from 16-wk-old versus 4- and 8-wk-old NOD female mice (P ≤ 0.02; Fig. 7 

A). Similarly, a marked reduction in TGFβ secretion was detected between CD4+CD25+ T cells that were isolated from the PLNs of 16-wk-old NOD versus NOR and B6 female mice (Fig. 7 B). The declining levels of TGFβ reflected the reduced efficacy of the CD4+CD25+ T cells that were prepared from 16-wk-old NOD female mice to suppress the proliferation of CD4+CD25 T cells (Figs. 1 A and 2 A).

To address whether decreased suppressor activity largely was due to reduced TGFβ production, the function of CD62Lhi T cells that were prepared from 16-wk-old NOD female mice was examined in an in vitro suppression assay that was supplemented with recombinant human (rh)TGFβ. Here, 3 × 103 CD62Lhi T cells and 105 CD4+CD25 T cells were cultured with or without 425 pg of rhTGFβ. The amount of rhTGFβ was based on detected levels of TGFβ secretion by CD4+CD25+ T cells (2.5 × 105) from 4-wk-old NOD female mice mediating maximum suppression (Fig. 7 A). When rhTGFβ was added, CD62Lhi T cells from 16-wk-old NOD female mice suppressed proliferation of CD4+CD25 T cells similarly to 4-wk-old CD62Lhi T cells that had not been supplemented with cytokine (Fig. 7 C). CD4+CD25 T cells alone did not produce TGFβ upon activation (Fig. 7 A), nor was FoxP3 expression induced in CD4+CD25 T cells that were incubated with 425 pg of rhTGFβ. For instance, no difference was detected in the frequency (∼7.0%) of FoxP3-positive CD4+CD25 T cells that were cultured with or without rhTGFβ. Furthermore, when 5 × 104 CD62Lhi or CD62Llo T cells that were prepared from PLNs of 4-wk-old NOD female mice were stimulated individually in vitro, 345 ± 48 pg/ml and 125 ± 25 pg/ml of TGFβ was detected, respectively. These data demonstrate that the addition of TGFβ can enhance the suppressor function of CD62Lhi T cells from 16-wk-old NOD female mice, and that the source of TGFβ that was detected in in vitro suppression assays are CD62Lhi and CD62Llo T cells.

Finally, the addition of anti-TGFβ Ab, but not an isotype control Ab, effectively blocked—in a dose-dependent manner—the capacity of CD4+CD25+ T cells from female and male NOD and B6 mice to mediate suppression (Fig. 7 D). Together, these data demonstrate that TGFβ produced by the CD62Lhi (and CD62Llo) plays a key role in the observed suppression; diminishing TGFβ protein expression within the pool of PLN CD4+CD25+ T reg cells contributes to reduced effector function in an age-dependent manner.

Pancreatic islets and popliteal lymph nodes of NOD female mice exhibit a progressive loss of FoxP3- and TGFβ1-coexpressing CD62Lhi T cells

Islets also provide a key site for immunoregulation of β cell specific T cell reactivity (32). Accordingly, CD4+CD25+ T reg cells that infiltrate the islets of 8- and 16-wk-old NOD female mice were examined. As demonstrated in Table II 

, a decrease in the number of CD62Lhi T cells (162 ± 5 versus 111 ± 3; P = 0.04) and a concomitant increase in the number of CD62Llo T cells (320 ± 3 versus 731 ± 11; P = 0.02) that infiltrated the islets of 8- and 16-wk-old NOD female mice, respectively, were detected. Similar to PLN (Fig. 4 A), a twofold reduction in FoxP3-positive CD62Lhi (and CD62Llo) T cells was observed in the islet infiltrates of 8- versus 16-wk-old NOD female mice (P = 0.01; Fig. 8 

A). Again, as seen in PLN, two distinct subsets of FoxP3-expressing CD62Lhi T cells, based on TGFβ1 transcript copy number, were detected in the islets of 8-wk-old, but not 16-wk-old, NOD female mice (Fig. 8 B). Furthermore, none of the FoxP3-positive T cells expressed IL-10 only, although <10% of FoxP3- and TGFβ1-coexpressing T cells were positive for IL-10 mRNA, independent of age (data not shown). These findings demonstrate that analogous to the PLN, the percentage of islet-infiltrating CD62Lhi (and CD62Llo) T cells that coexpress FoxP3 and TGFβ1 declines as β cell autoimmunity progresses in NOD female mice.

To determine if the observed temporal decrease in the percentage of FoxP3-expressing T cells was unique to sites of inflammation, such as the PLN and islets, the popliteal lymph nodes from 4- and 16-wk-old NOD female mice were analyzed; B6 mice served as a control. A significant decrease in the percentage of CD62Lhi T cells that expressed FoxP3 was detected in NOD, but not B6, female mice (Fig. 8, C and D). Transcript copy number of FoxP3 did not vary and was similar to levels detected in PLN and islet-infiltrating CD62Lhi (and CD62Llo) T cells (Fig. S1 A). As above, >82% of FoxP3-positive T cells also expressed TGFβ1, independent of age. These data indicate that the temporal decrease in the percentage of FoxP3- and TGFβ1-coexpressing T cells in NOD female mice is not due to activation and subsequent down-regulation of expression of these genes, but rather is an intrinsic defect in the maintenance of this T cell population.

Discussion

The interplay between natural CD4+CD25+ T reg cells and pathogenic T cells in autoimmunity, including T1D, is poorly understood. In part, this is due to the difficulty of accurately distinguishing between CD4+CD25+ T reg cells and other subsets of immunoregulatory or pathogenic T effectors. In this study and others, natural CD4+CD25+ T reg cells have been defined based on a CD62Lhi surface phenotype, and expression of markers, such as CTLA-4, GITR, CD45RB, CD103 (20, 29, 30), and more recently, lymphocyte activation gene-3 (33). Furthermore, various groups have demonstrated that CD4+CD62Lhi T cells that were prepared from the thymus or spleen of young NOD mice block diabetes in adoptive transfer experiments (29, 30). Detection of FoxP3 mRNA also was used recently to define CD4+CD25+ T reg cells (2123). However, bulk populations typically are studied so that the number/frequency of, and the corresponding cytokines that are expressed by, FoxP3-positive T cells cannot be determined readily. The latter is of significance in view of the apparent heterogeneity within CD4+CD25+ T reg cells regarding expression and the respective roles of TGFβ1 and/or IL-10 in suppressor activity (2426). We have addressed these issues by using a real-time PCR assay to measure FoxP3 and cytokine expression in single cells.

Two conclusions pertaining to the general properties of CD4+CD25+ T reg cells can be drawn based on the single cell analyses that were made in NOD, NOR, and B6 mice. First, only a subset (∼40%) of single CD62Lhi T cells express significant levels of FoxP3. Furthermore, a considerable number (∼20%) of CD62Llo T cells also express FoxP3; this is consistent with work by the Salomon group (34) that showed that proliferating CD4+CD25+ T reg cells down-regulate CD62L expression. The pool of CD62Llo T cells exhibited a markedly reduced suppressor function relative to CD62Lhi T cells (Fig. 2, A and B); this suggests that CD4+CD25+ T reg cells effector function is regulated tightly. Together, however, these results further underscore the difficulty of characterizing CD4+CD25+ T reg cells based on surface phenotype alone. Second, the majority (>80%) of single, FoxP3-positive T cells express TGFβ1 but not IL-10. Noteworthy, however, is that a small percentage (<10%) of FoxP3- and TGFβ1-positive CD62Lhi T cells in the islets, but not PLNs, of NOD mice also expressed IL-10 (data not shown). FoxP3-positive T cells that express IL-10, but not TGFβ1, normally may represent a minor population of CD4+CD25+ T reg cells that undergo expansion under certain proinflammatory conditions (26). A role for TGFβ1 was demonstrated by the capacity of neutralizing Ab to block completely CD4+CD25+ T reg cell–mediated suppression in vitro (Fig. 7 D). The latter does not rule out cell–cell contact as a means by which these CD4+CD25+ T reg cells mediate suppression in vivo. In fact, cell–cell contact is likely to be the predominant mechanism of suppression of the few FoxP3-positive T cells (<20%) that lack TGFβ1 expression (Fig. 5 and Table S2). Our observation that anti-TGFβ Ab blocked the in vitro suppressor activity of NOD and B6 CD4+CD25+ T reg cells (Fig. 7 D) is consistent with findings that were reported by Nakamura et al; they used the same polyclonal anti-TGFβ Ab (1D11) to inhibit the function of BALB/c CD4+CD25+ T reg cells in vitro (35). In addition, several studies demonstrated that the suppressor activity of CD4+CD25+ T reg cells upon adoptive transfer is blocked in vivo by anti-TGFβ Ab treatment (36, 37). Nevertheless, under certain conditions, suppression of CD4+CD25+ T cells by CD4+CD25+ T reg cells can be mediated independent of TGFβ (38); this suggests that the role of this cytokine in CD4+CD25+ T reg cell function is complex. Shevach and colleagues suggested that TGFβ1, in an autocrine manner, may have a direct effect on the induction of CD4+CD25+ T reg cell function (38). Reduced TGFβ1 expression (Fig. 5 A) correlated with a decreased number of FoxP3-expressing CD4+CD25+ T cells in older NOD female mice (Fig. 4, A and B). Accordingly, lower levels of autocrine TGFβ1 may affect the maintenance/expansion and the function of FoxP3-positive CD4+CD25+ T reg cells adversely in older NOD female mice.

Another key observation that was made in this study is that the function and phenotype of CD4+CD25+ T reg cells in NOD female mice undergo marked changes in an age-dependent fashion. For example, the in vitro and in vivo suppressor function of PLN CD4+CD25+ T cells declined significantly from 4 to 8 and 16 wk of age (Fig. 1). The loss of suppressor activity was detected, despite increasing numbers of PLN CD62Lhi T cells. Therefore, the progressive loss of suppressor activity by sorted CD62Lhi (and CD62Llo) T cells (Fig. 2 A) indicated an intrinsic defect within this pool of immunoregulatory effectors. Consistent with this finding, real-time PCR analysis demonstrated that the percentage of single CD62Lhi T cells that coexpresses FoxP3 and TGFβ1 declined with age in the PLN (Fig. 4, A and B) and islets (Fig. 8 B). CD62Lhi T cells that were sorted from 4-wk-old NOD mice exhibited the most effective suppressor activity (Fig. 2 A), which, in turn, correlated with the greatest number of FoxP3- and TGFβ1-coexpressing T cells (Fig. 4, A and B). In comparison, the pool of CD62Lhi T cells from 8-wk-old NOD female mice exhibited decreased suppressor activity (Fig. 2 A). The latter was due to a reduced frequency of FoxP3- and TGFβ1-expressing CD62Lhi T cells in 8- versus 4-wk-old NOD female mice (Fig. 4 A). Notably, TGFβ1 transcript copy number was equivalent between the FoxP3-expressing CD62Lhi T cells of 4- and 8-wk-old NOD female mice. In older NOD female mice, the suppressor function of CD62Lhi T cells was reduced further (Fig. 2 A). However, in this case, not only was the frequency of FoxP3- and TGFβ1-coexpressing CD62Lhi T cells reduced but so was the level of TGFβ1 expression by these T cells (Fig. 4, A and B; Fig. 5 A). It is likely that both factors contribute to the relatively ineffective suppressor activity that is exhibited by this pool of CD62Lhi T cells. Nevertheless, at high numbers (i.e., 5 × 104), CD62Lhi T cells from 16-wk-old NOD female mice exhibited in vitro suppressor activity that was equivalent to that of 4- and 8-wk-old NOD female mice (Fig. 2 A). This result indicates that at an optimal number/frequency, CD62Lhi T cells exhibit effective immunoregulatory function, despite reduced TGFβ production. These data further demonstrate that the progressive loss of suppressor activity within the pool of CD4+CD25+ T reg cells in NOD female mice is associated with multiple defects.

The temporal profile of FoxP3- and TGFβ1-coexpressing T cells differed markedly between NOD female mice versus B6, NOR, and most notably, NOD male, mice (Figs. 4 and 6). Almost a twofold increase in FoxP3- and TGFβ1-positive CD62Lhi T cells was detected in 4-wk-old NOD female mice relative to B6, NOR, and NOD male mice of the same age. This was despite the fact that in 4-wk-old NOR and B6 mice, the number of PLN CD62Lhi T cells was approximately two- to threefold greater than that observed in NOD female mice (Table S1). However, the progressive loss of FoxP3- and TGFβ1-coexpressing cells within the pool of CD62Lhi (and CD62Llo) T cells was detected only in NOD female mice. In B6, NOR, and NOD male mice, equivalent numbers of FoxP3- and TGFβ1-positive T cells remained constant with age (Fig. 4, C–F, and Fig. 6, A and B). The fact that infiltration of the PLNs and islets is similar between NOD male and female mice (Table I) suggests that ongoing inflammation does not account for down-regulation of FoxP3 (and TGFβ1) expression as a result of the activational and/or proliferative status of CD4+CD25+ T reg cells. A similar decline in the frequency of FoxP3- and TGFβ1-coexpressing CD62Lhi and CD62Llo T cells was detected in the popliteal lymph nodes of NOD female mice (Fig. 8 C), a site that lacks obvious inflammation. These data strengthen the hypothesis that the progressive loss of FoxP3- and TGFβ1-coexpressing CD4+CD25+ T reg cells in NOD female mice directly contributes to the development of aggressive insulitis and overt diabetes.

The events that influence the frequency of FoxP3- and TGFβ1-coexpressing CD4+CD25+ T reg cells in NOD mice is not clear, although a number of mutually nonexclusive possibilities exist. Thymic selection may become limiting for CD4+CD25+ T reg cells in NOD mice, thereby indirectly affecting the size of the peripheral pool. Various groups, for instance, have documented aberrant thymocyte selection (3, 4) and abnormal thymic physiology (39) in the NOD mouse. Treatment of neonatal NOD mice with anti-TNFα Ab increases the number of thymic CD4+CD25+ T cells (40). Second, a defect in maintenance and/or expansion of FoxP3- and TGFβ1-coexpressing CD4+CD25+ T reg cells may contribute to the decline of these immunoregulatory effectors. Various molecules, including CTLA-4, TNF-related activation-induced cytokine, GITR, and lymphocyte activation gene-3, are potential candidates which affect the induction and/or expansion of CD4+CD25+ T reg cells (20, 33, 4143). The temporal decline in TGFβ1 mRNA that was detected in FoxP3-expressing CD4+CD25+CD62Lhi T cells (Fig. 5), and the corresponding decrease in TGFβ secretion (Fig. 7), may hinder the expansion and/or maintenance of CD4+CD25+ T reg cells during inflammation. Recent work by the Flavell group demonstrated that ectopic expression of TGFβ1 by β cells increased the frequency of FoxP3-expressing CD4+CD25+ T reg cells, and protected the transgenic NOD mice from diabetes (44). Finally, it has been suggested that peripheral induction of FoxP3 expression by CD4+CD25 T cells may contribute to the maintenance and/or expansion of the CD4+CD25+ T reg cells. The intrinsic properties of CD4+CD25 T cells or the lack of appropriate signals in NOD mice may limit the efficiency of such a mechanism. For instance, TGFβ1 was shown to induce FoxP3 expression in CD4+CD25 T cells in vitro (45, 46). Once again, limiting amounts of TGFβ that is produced by NOD CD4+CD25+ T reg cells would be expected to reduce the efficiency of such a mechanism.

This study and others highlight the complexity among immunoregulatory T cells that modulate the progression of T1D. Adorini and colleagues demonstrated that CD4+ CD25+ T cells that were prepared from the spleen of NOD mice also exhibited varying suppressor activity in an age-dependent manner (7). In contrast to the findings herein, immunoregulatory CD4+CD25+ T cells were characterized by expression of IL-10, although in vitro suppression was mediated in an IL-10–independent manner. Expression of FoxP3 or TGFβ1 was not assessed in this pool of CD4+ CD25+ T cells. Recent work by Benoist et al. showed that progression of insulitis in BDC2.5 mice was dependent on FoxP3-expressing CD4+CD25+CD69+ T cells which reside within the islets (32). This population of immunoregulatory T effectors also was characterized by expression of IL-10; in vivo persistence was dependent on inducible costimulatory expression. Sufficient expansion of this subset of CD4+ CD25+ T reg cells may have been facilitated by expression of the BDC2.5 clonotypic TCR in this model.

In summary, we propose that a progressive decline in the percentage of FoxP3- and TGFβ1-coexpressing CD4+CD25+ T reg cells contributes to the development of “aggressive” insulitis in NOD female mice. During the prediabetic stage, a balance between pathogenic T cells and CD4+CD25+ T reg cells that reside in the PLNs and islets is maintained. However, as FoxP3- and TGFβ1-positive–expressing T cells decline within the pool of CD4+CD25+ T reg cells, a minimum threshold is surpassed and pathogenic β cell–specific T effectors are permitted to expand (Table I) and drive the response to an overt diabetic state. However, in NOD male mice, the pool of FoxP3- and TGFβ1-coexpressing CD4+CD25+ T reg cells is relatively stable; this is reflected by a markedly reduced frequency of diabetes (∼20%) compared with NOD female mice (∼80%). Defining the mechanism that determines the frequencies of FoxP3- and TGFβ1-coexpressing CD4+CD25+ T reg cells is likely to provide key insight into events that regulate T1D and other tissue-specific autoimmune diseases.

MATERIAL AND METHODS

Mice.

NOD/LtJ, NOR/LtJ, NOD.BDC2.5, C57BL/6 (B6), and NOD.scid mice were housed under specific pathogen-free conditions. Diabetes develops in ∼80% and ∼20% of NOD female and male mice, respectively, by 1 yr of age. Mice were maintained in an American Association of Laboratory Animal Care-accredited facility, and studies were approved by the Institutional Animal Care and Use Committee of the University of North Carolina at Chapel Hill.

Flow cytometric analysis.

Abs were purchased from BD Biosciences and used for analysis of CD4+CD25+ T cells: FITC-labeled rat anti–CD3 (17A2), PerCP-labeled rat anti–CD4 (GK1.5), APC-labeled rat anti–CD25 (PC61), PE-labeled rat anti–CD62L (Mel-14), PE-labeled anti–CD103 (M290), PE-labeled anti–CTLA-4 (UC10-4F10-11), biotinylated rat anti–CD69 (H1.2F3), and biotinylated anti–CD45RB (16A). Purified rat anti–mouse GITR (108619) was purchased from R&D Systems. CD4+CD25+ and CD4+CD25 T cells were sorted using a MoFlo high-speed sorter (DakoCytomation) for single cell real-time PCR analysis, suppression assays, and adoptive transfer experiments. Sorted bulk populations, >98% purity, were stained with the aforementioned Abs. T cell preparations were stained in 5% FBS/PBS for 30 min on ice. When necessary, T cells also were incubated with secondary antibodies, PE-labeled streptavidin (BD Biosciences) or PE-labeled anti–rat IgG (Southern Biotechnology Associates Inc.) for 30 min on ice and washed. Flow cytometry data were acquired at the University of North Carolina Flow Cytometry and evaluated by Facility using a FACS Calibur (Becton Dickinson) and Summit software (DakoCytomation).

Single cell real-time PCR.

Single CD4+CD25+CD62Llo, CD4+CD25+ CD62Lhi, or CD4+CD25 T cells were sorted into individual wells of a 96-well polypropylene RNase/DNase-free plate (USA Scientific) which contained 4 μl of cell lysis buffer and stored at −80°C. cDNA was synthesized using SuperScript II (Invitrogen), oligo(dT) primers (Invitrogen), RNase Out (Invitrogen), and thermocycler (GeneAmp 9700, Applied Biosystems) conditions that consisted of 40°C for 40 min and 70°C for 15 min. Real-time PCR primers and probes that are specific for FoxP3 (5′-CCCAGGAAAGACAGCAACCTT-3′; 5′-TTCTCACAACCAGGCCACTTG-3′; 5′-fam/ATCCTACCCACTGCTGGCAAATGGAGTC/tam-3′), TGFβ1 (5′-CACTGATACGCCTGAGTG-3′; 5′-GTGAGCGCTGAATCGAAA-3′; 5′-fam/CCGTCTCCTTGGTTCAGCCACTG/tam-3′), and IL-10 (5′-CAGAGCCACATGCTCCTA-3′; 5′-GGAGTCGGTTAGCAGTATG-3′; 5′-fam/CTGCGGACTGCCTTCAGCCAG/tam-3′) were synthesized by Integrated DNA Technologies. A standard curve that was generated from serial dilutions of purified plasmid DNA that encoded the respective genes was used to measure mRNA transcript copy number. The lower limit of detection was 10 copies for all genes tested. Each gene was detected in independent real-time PCR reactions using 5 μl of a 25-μl total cDNA mixture. Copy number of mRNA encoding β-actin was analyzed in the same sample using 5′-AGAGGGAAATCGTGCGTGAC-3′; 5′-CAATAGTGATGACCTGGCCGT-3′; 5′-fam/CACTGCCGCATCCTCTTCCTCCC/tam-3′ by Integrated DNA Technologies. Data are expressed as a copy number normalized to β-actin content. The normalized mRNA copy number for a gene was determined by: [raw transcript copy number derived from standard curve] × [β-actin corrective ratio]. The β-actin corrective ratio was calculated as [lowest β-actin copy number within sample set]/[β-actin copy number for cell of interest].

In vitro CD4+CD25+ T cell suppression assay.

CD4+CD25+ T cells that were sorted from PLNs were cocultured in 96-well plates with 105 CD4+CD25 T cells that were sorted from PLNs plus 1 μg/ml of anti-CD3ε (145-2C11; eBioscience) and 105 irradiated T cell depleted splenocytes. Splenocytes were depleted of T cells by incubation with anti-Thy1.1–coated magnetic microbeads (Miltenyi Biotec) and then magnetic-activated cell separation was selected according to the manufacturer's protocol. T cells were cultured in RPMI 1640 which was supplemented with 10% FBS, 50 μM 2-ME, 1× nonessential amino acids, 1 mM sodium pyruvate, 2 mM L-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin for 72 h at 37°C, and then pulsed with 1 μCi/well of [3H]thymidine for 18 h. [3H]thymidine incorporation was measured by a scintillation counter (Beckman LS6500). Percent inhibition was determined by the following equation: 1 − {[(CD4+CD25 + CD4+CD25+) cpm]/[(CD4+CD25 alone) cpm]} × 100. Some studies included the addition of anti-TGFβ1,2,3 (1D11) or IgG1 isotype control Ab (R&D Systems) at 10 or 100 μg/ml, or 425 pg of rhTGFβ (BD Biosciences).

Measurement of IL-10 and TGFβ secretion via ELISA.

A capture ELISA was used to measure levels of IL-10 and TGFβ in supernatants that were harvested from individual wells that were prepared for the in vitro suppression assay that was described above. In brief, 96-well plates were coated with 2 μg/ml of rat anti–TGFβ or rat anti–IL-10 (BD Biosciences) prepared in carbonate buffer pH 9.0 and incubated overnight at 4°C. Plates were blocked, and 50 μl of culture supernatant (diluted 1:2) was added per well and incubated overnight at 4°C. Plates were washed (1% BSA/.05% Tween 20/PBS) and incubated for 2 h at room temperature with 1 μg/ml of biotinylated rat anti–TGFβ or rat anti–IL-10 (BD Biosciences). Plates were washed, and streptavidin-alkaline phosphatase (BD Biosciences) was added at a 1/1000 dilution and incubated for 2 h. After washing, plates were developed using 200 μl of diethenolamine buffer with p-N-PP substrate (Sigma-Aldrich). Cytokine concentrations in culture supernatants were determined using standard curves of hTGFβ and mIL-10 (BD Biosciences). The lower limit of detection for IL-10 and TGFβ was 40 pg/ml.

In vivo CD4+CD25+ T cell suppression assay.

5 × 105 CD4+CD25+ T cells that were sorted from the PLNs (>90% purity) of NOD female mice were coinjected i.v. with 5 × 106 splenic BDC2.5 CD4+ T cells into NOD.scid recipients. Splenic BDC2.5 CD4+ T cells were isolated with anti-CD4–coated magnetic microbeads (Miltenyi Biotec) followed by MACS selection. Mice were monitored daily for the development of glycosuria with diastix (Bayer Corporation).

Pancreatic islet isolation.

Pancreases from groups of eight or nine mice were perfused with 2 mg/ml collagenase P (Roche) and digested for 20 min at 37°C. Islets were purified via Ficoll gradient, handpicked, and then dissociated into a single cell suspension using enzyme-free cell dissociation solution (Sigma-Aldrich). Cells were washed and prepared for flow cytometric analysis of CD4+CD25+ T cells.

Online supplemental material

Fig. S1 provides data for FoxP3 expression in CD62Lhi and CD62Llo sorted T cells prepared from the popliteal lymph nodes of NOD and B6 female mice. Table S1 provides the absolute number and percentage of FoxP3+ and FoxP3 CD4+CD25+ T cells expressing TGFβ1 prepared from the PLNs of NOD, NOR, and B6 female mice varying in age. Table S2 provides the absolute number of CD4+, CD4+CD25+, CD62Lhi, and CD62Llo T cells prepared from the PLNs of NOR and B6 female mice varying in age.

Acknowledgments

We would like to thank Dr. J. Frelinger for critical reading of the text.

This work was supported by grants from the Juvenile Diabetes Research Foundation (no. 1-2001-838) and the National Institutes of Health (no. 1R01AI058014). C.P. Wong was supported by a fellowship from the Juvenile Diabetes Research Foundation.

The authors have no conflicting financial interests.

References

References
1
Tisch, R., and H. McDevitt.
1996
. Insulin-dependent diabetes mellitus.
Cell.
85
:
291
–297.
2
Delovitch, T.L., and B. Singh.
1997
. The nonobese diabetic mouse as a model of autoimmune diabetes: immune dysregulation gets the NOD.
Immunity.
7
:
727
–738.
3
Kishimoto, H., and J. Sprent.
2001
. A defect in central tolerance in NOD mice.
Nat. Immunol.
2
:
1025
–1031.
4
Choisy-Rossi, C.M., T.M. Holl, M.A. Pierce, H.D. Chapman, and D.V. Serreze.
2004
. Enhanced pathogenicity of diabetogenic T cells escaping a non-MHC gene controlled near death experience.
J. Immunol.
173
:
3791
–3800.
5
Rapoport, M.J., A. Jaramillo, D. Zipris, A.H. Lazarus, D.V. Serreze, E.H. Leiter, P. Cyopick, J.S. Danska, and T.L. Delovitch.
1993
. Interleukin-4 reverses T cell proliferative unresponsiveness and prevents the onset of diabetes in nonobese diabetic mice.
J. Exp. Med.
178
:
87
–99.
6
Salomon, B., D.J. Lenschow, L. Rhee, N. Ashourian, B. Singh, A. Sharpe, and J.A. Bluestone.
2000
. B7/CD28 costimulation is essential for the homeostasis of the CD4+CD25+ immunoregulatory T cells that control autoimmune diabetes.
Immunity.
12
:
431
–440.
7
Gregori, S., N. Giarratana, S. Smiroldo, and L. Adorini.
2003
. Dynamics of pathogenic and suppressor T cells in autoimmune diabetes development.
J. Immunol.
171
:
4040
–4047.
8
Andre, I., A. Gonzalez, B. Wang, J. Katz, C. Benoist, and D. Mathis.
1996
. Checkpoints in the progression of autoimmune disease: lessons from diabetes models.
Proc. Natl. Acad. Sci. USA.
93
:
2260
–2263.
9
Amrani, A., J. Verdaguer, P. Serra, S. Tafuro, R. Tan, and P. Santamaria.
2000
. Progression of autoimmune diabetes driven by avidity maturation of a T cell population.
Nature.
406
:
739
–742.
10
Tisch, R., X.-D. Yang, S.M. Singer, R.S. Liblau, L. Fugger, and H.O. McDevitt.
1993
. Immune response to glutamic acid decarboxylase correlates with insulitis in non-obese diabetic mice.
Nature.
366
:
72
–75.
11
Kaufman, D.L., M. Clare-Salzler, J. Tian, T. Forsthuber, G.S. Ting, P. Robinson, M.A. Atkinson, E.E. Sercarz, A.J. Tobin, and P.V. Lehmann.
1993
. Spontaneous loss of T cell tolerance to glutamic acid decarboxylase in murine insulin dependent diabetes.
Nature.
366
:
69
–72.
12
Muir, A., A. Peck, M. Clare-Salzler, Y.H. Song, J. Cornelius, R. Luchetta, J. Krischer, and N.K. Maclaren.
1995
. Insulin immunization of nonobese diabetic mice induces a protective insulitis characterized by diminished intraislet interferon-gamma transcription.
J. Clin. Invest.
95
:
628
–634.
13
Shevach, E.M.
2000
. Regulatory T cells in autoimmmunity.
Annu. Rev. Immunol.
18
:
423
–449.
14
Groux, H., A. O'Garra, M. Bigler, M. Rouleau, S. Antonenko, J.E. de Vries, and M.G. Roncarolo.
1997
. A CD4+ T-cell subset inhibits antigen-specific T-cell responses and prevents colitis.
Nature.
389
:
737
–742.
15
Weiner, H.L.
2001
. Induction and mechanism of action of transforming growth factor-beta-secreting Th3 regulatory cells.
Immunol. Rev.
182
:
207
–214.
16
Shevach, E.M.
2002
. CD4+CD25+ suppressor T cells: more questions than answers.
Nat. Rev. Immunol.
2
:
389
–400.
17
Maloy, K.J., and F. Powrie.
2001
. Regulatory T cells in the control of immune pathology.
Nat. Immunol.
2
:
816
–822.
18
Sakaguchi, S., N. Sakaguchi, J. Shimizu, S. Yamazaki, T. Sakihama, M. Itoh, Y. Kuniyasu, T. Nomura, M. Toda, and T. Takahashi.
2001
. Immunologic tolerance maintained by CD25+CD4+ regulatory T cells: their common role in controlling autoimmunity, tumor immunity, and transplantation tolerance.
Immunol. Rev.
182
:
18
–32.
19
Bluestone, J.A., and A.K. Abbas.
2003
. Natural versus regulatory T cells.
Nat Rev.
3
:
253
–257.
20
McHugh, R.S., M.J. Whitters, C.A. Piccirillo, D.A. Young, E.M. Shevach, M. Collins, and M.C. Byrne.
2002
. CD4+CD25+ immunoregulatory T cells: Gene expression analysis reveals a functional role for the glucocorticoid-induced TNF receptor.
Immunity.
16
:
311
–323.
21
Hori, S., T. Nomura, and S. Sakaguchi.
2003
. Control of regulatory T cell development by the transcription factor Foxp3.
Science.
299
:
1057
–1061.
22
Khattri, R., T. Cox, S.A. Yasayko, and F. Ramsdell.
2003
. An essential role for Scurfin in CD4+CD25+ T regulatory cells.
Nat. Immunol.
4
:
337
–342.
23
Fontenot, J.D., M.A. Gavin, and A.Y. Rudensky.
2003
. Foxp3 programs the development and function of CD4+CD25+ regulatory T cells.
Nat. Immunol.
4
:
330
–336.
24
Nakamura, K., A. Kitani, I. Fuss, A. Pedersen, N. Harada, H. Nawata, and W. Strober.
2004
. TGF-β1 plays an important role in the mechanism of CD4+CD25+ regulatory T cell activity in both humans and mice.
J. Immunol.
172
:
834
–842.
25
Ostroukhova, M., C. Seguin-Devaux, T.B. Oriss, B. Dixon-McCarthy, L. Yang, B.T. Ameredes, T.E. Corcoran, and A. Ray.
2004
. Tolerance induced by inhaled antigen involves CD4+ T cells expressing membrane-bound TGF-β and FOXP3.
J. Clin. Invest.
114
:
28
–38.
26
Asseman, C., S. Mauze, M.W. Leach, R.L. Coffman, and F. Powrie.
1999
. An essential role for interleukin 10 in the function of regulatory T cells that inhibit intestinal inflammation.
J. Exp. Med.
190
:
955
–1004.
27
Hoglund, P., J. Mintern, C. Waltzinger, W. Heath, C. Benoist, and D. Mathis.
1999
. Initiation of autoimmune diabetes by developmentally regulated presentation of islet cell antigens in the pancreatic lymph nodes.
J. Exp. Med.
189
:
331
–339.
28
Katz, J.D., B. Wang, K. Haskins, C. Benoist, and D. Mathis.
1993
. Following a diabetogenic T cell from genesis through pathogenesis.
Cell.
74
:
1089
–1100.
29
Szanya, V., J. Ermann, C. Taylor, C. Holness, and C.G. Fathman.
2002
. The subpopulation of CD4+CD25+ splenocytes that delays adoptive transfer of diabetes expresses L-selectin and high levels of CCR7.
J. Immunol.
169
:
2461
–2465.
30
Herbelin, A., J.M. Gombert, F. Lepault, J.F. Bach, and L. Chatenoud.
1998
. Mature mainstream TCR alpha beta+CD4+ thymocytes expressing L-selectin mediate “active tolerance” in the nonobese diabetic mouse.
J. Immunol.
161
:
2620
–2628.
31
Prochazka, M., D.V. Serreze, W.N. Frankel, and E.H. Leiter.
1992
. NOR/Lt mice: MHC-matched diabetes-resistant control strain for NOD mice.
Diabetes.
41
:
98
–106.
32
Herman, A.E., G.J. Freeman, D. Mathis, and C. Benoist.
2004
. CD4+CD25+ T regulatory cells dependent on ICOS promote regulation of effector cells in the prediabetic lesion.
J. Exp. Med.
199
:
1479
–1489.
33
Huang, C.-T., C.J. Workman, D. Flies, X. Pan, A.L. Marson, G. Zhou, E.L. Hipkiss, S. Ravi, J. Kowalski, D.M. Pardoll, et al.
2004
. Role of LAG-3 in regulatory T cells.
Immunity.
21
:
503
–513.
34
Fisson, S., G. Darrasse-Jeze, E. Litvinova, F. Septier, D. Klatzmann, R. Liblau, and B.L. Salomon.
2003
. Continuous activation of autoreactive CD4+CD25+ regulatory T cells in the steady state.
J. Exp. Med.
198
:
737
–746.
35
Nakamura, K., A. Kitani, and W. Strober.
2001
. Cell-contact-dependent immunosuppression by CD4+CD25+ regulatory T cells is mediated by cell-surface-bound transforming growth factor β.
J. Exp. Med.
194
:
629
–644.
36
Liu, H., B. Hu, D. Xu, and F.Y. Liew.
2003
. CD4+CD25+ regulatory T cells cure murine colitis: the role of IL-10, TGF-beta and CTLA-4.
J. Immunol.
171
:
5012
–5017.
37
Oida, T., X. Zhang, M. Goto, S. Hachimura, M. Totsuka, S. Kaminogawa, and H.L. Weiner.
2003
. CD4+CD25+ T cells that express latency-associated peptide on the surface suppress CD4+CD45R Bhigh-induced colitis by a TGF-beta-dependent mechanism.
J. Immunol.
170
:
2516
–2522.
38
Piccirillo, C., J.J. Letterio, A.M. Thornton, R.S. McHugh, M. Mamaura, H. Mizuhara, and E.M. Shevach.
2002
. CD4+CD25+ Regulatory T cells can mediate suppressor function in the absence of transforming growth factor b1 production and responsiveness.
J. Exp. Med.
196
:
237
–245.
39
Atlan-Gepner, C., M. Naspetti, R. Valero, M. Barad, F. Lepault, B. Vialettes, and P. Naquet.
1999
. Disorganization of the thymic medulla precedes evolution towards diabetes in female NOD mice.
Autoimmunity.
31
:
249
–260.
40
Wu, A.J., H. Hua, S.H. Munson, and H.O. McDevitt.
2002
. Tumor necrosis factor-alpha regulation of CD4+CD25+ T cell levels in NOD mice.
Proc. Natl. Acad. Sci. USA.
99
:
12287
–12292.
41
Read, S., V. Malmstrom, and F. Powrie.
2000
. Cytotoxic T lymphocyte-associated antigen 4 plays an essential role in the function of CD25+CD4+ regulatory cells that control intestinal inflammation.
J. Exp. Med
.
192
:
295
–302.
42
Shimizu, J., S. Yamazaki, T. Takahashi, Y. Ishida, and S. Sakaguchi.
2002
. Stimulation of CD25+CD4+ regulatory T cells through GITR breaks immunological self tolerance.
Nat. Immunol.
3
:
135
–142.
43
Green, E.A., Y. Choi, and R.A. Flavell.
2002
. Pancreatic lymph node-derived CD4+CD25+ Treg cells: highly potent regulators of diabetes that require TRANCE-RANK signals.
Immunity.
16
:
183
–191.
44
Peng, Y., Y. Laouar, M.O. Li, E.A. Green, and R.A. Flavell. TGF-β regulates in vivo expansion of Foxp3-expressing CD4+CD25+ regulatory T cells responsible for protection against diabetes. Proc. Natl. Acad. Sci. USA. 101:4572–4577.
45
Chen, W., W. Jin, N. Hardegen, K. Lei, L. Li, N. Marinos, G. McGrady, and S.H. Wahl.
2003
. Conversion of peripheral CD4+ CD25− naïve T cells to CD4+CD25+ regulatory T cells by TGF-β induction of transcription factor Foxp3.
J. Exp. Med.
198
:
1875
–1886.
46
Fantini, M.C., C. Becker, G. Monteleone, F. Pallone, P.R. Galle, and M.F. Neurath.
2004
. Cutting edge: TGF-β induces a regulatory phenotype in CD4+CD25- T cells through FoxP3 induction and down-regulation of Smad7.
J. Immunol.
172
:
5149
–5153.

Abbreviations used: Ab, antibody; CD62Lhi T cells, CD4+CD25+CD62Lhi T cells; CD62Llo T cells, CD4+CD25+CD62Llo T cells; CTLA-4, cytotoxic T lymphocyte–associated antigen 4; GITR, glucocorticoid-induced TNF receptor gene; NOD, nonobese diabetic; PLN, pancreatic LN; rh, recombinant human; T reg, regulatory T; T1D, type 1 diabetes.

Supplementary data