Suppression of CD4⁺ T Lymphocyte Effector Functions by CD4⁺CD25⁺ Cells In Vivo¹

The existence of a subpopulation of T cells that specialize in the suppression of immune responses was originally postulated in the early 1970s (1). However, interest in such cells was gradually lost because the molecular mechanisms responsible for such suppressive phenomena were difficult to characterize. Recently, advances in the identification of CD4⁺ T cell subpopulations, together with the use of genetically modified mice, have led to a renaissance of this field; the concept of regulatory T cells has been reborn as an additional mechanism to explain the maintenance of peripheral self-tolerance, alongside T cell deletion and T cell anergy (2, 3, 4, 5, 6, 7). The notion of regulatory T cells is conceptually attractive because it explains how tolerance can be adoptively transferred by T cells in some experimental systems, how pathological responses to self and harmless foreign Ags are prevented, and how host bystander tissue insult is avoided during normal immune responses.

Physiologically generated CD4⁺CD25⁺ cells are the most widely studied type of regulatory T cells. These cells inhibit a wide range of autoimmune and inflammatory manifestations such as gastritis, oophoritis, orchitis, thyroiditis, inflammatory bowel disease (IBD),⁴ and spontaneous autoimmune diabetes (8, 9, 10, 11). Despite numerous studies, the mechanisms by which regulatory T cells exert their function are unclear. Some studies have shown that regulation is dependent in vivo on the production of suppressive cytokines such as IL-10 and TGF-β and cell surface molecules such as CTLA-4 (12, 13, 14, 15, 16, 17, 18). In vitro experiments aimed at further dissecting the mechanisms by which T cells exert their regulatory function have given controversial results. Indeed, in contrast to in vivo studies, neither soluble cytokines nor CTLA-4 seem to be required for the suppressive effects of CD4⁺CD25⁺ cells in vitro (19, 20, 21, 22).

Despite numerous studies, it remains to be shown how regulatory T cells exert their suppressive function on pathogenic effector T cells. The conflict between in vitro and in vivo data now makes it mandatory to study regulatory functions in vivo, using experimental models. Naive CD4⁺ T cells transferred into T cell-deficient mice expand strongly and rapidly induce IBD (23). This inflammatory disorder depends on IFN-γ production by expanding CD4⁺ T cells (24). Coinjection of CD4⁺CD25⁺ regulatory T cells protects recipient mice by inhibiting IFN-γ production by pathogenic effector CD4⁺ T cells (18, 25, 26, 27), but the precise mechanism by which they act remains to be elucidated. Indeed, prevention of IBD could be due to deletion of pathogenic T cells, induction of a non reactive state (anergy) among pathogenic effector T cells, preferential induction of Th2 effectors rather than Th1 effectors or permanent action (direct or indirect (through APCs)) of regulatory T lymphocytes on effector T cells (suppression).

C57BL/6 mice (Thy1.2; H-2^b) and BALB/c mice (H-2^d) were from Centre d’élevage Janvier (Le Genest Saint Isle, France), and C57BL/6 CD3ε-deficient mice (CD3ε^−/− mice) (28) were from Centre de Développement des Techniques Avancées pour l’Experimentation Animale (Orléans, France). C57BL/Ba mice (Thy1.1, H-2^b) were maintained in our own animal facilities.

Lymph node cells were depleted of macrophages, granulocytes, and CD8⁺ T cells by incubating them first with anti-CD11b (Mac-1) Ab, anti-GR1 (8C5) Ab, and anti-CD8 (Lyt-2) Ab, and then with magnetic beads coupled to anti-rat Ig (Dynal, Great Neck, NY). B cells were removed using magnetic beads coupled to anti-mouse Ig (Dynal). Purified CD4⁺ T cells from C57BL/6 mice (Thy1.2) were labeled with biotinylated anti-CD25 (clone PC61). CD4⁺CD25⁺ T cells were then positively selected using MACS streptavidin microbeads (Miltenyi Biotec, Paris, France). CD4⁺CD25⁺ T cells were usually 90–95% pure. Purified CD4⁺ T cells from C57BL/Ba mice (Thy1.1) were labeled with biotinylated anti-CD25 (clone PC61) and PE anti-CD44 (clone 1M7). CD4⁺ CD25⁻ CD44⁻ naive T cells were then purified by sorting in a FACSVantage flow cytometer (BD Biosciences, Mountain View, CA).

Purified naive CD4⁺ T cells (1 × 10⁶) were injected i.v., with or without purified CD4⁺CD25⁺ T cells (0.2 × 10⁶), into C57BL/6 CD3ε-deficient mice. The spleen and lymph nodes of these mice were recovered, pooled for cell preparation, and analyzed at various times after CD4⁺ T cell transfer.

In some experiments (Fig. 7), naive CD4⁺ T cells (Thy1.1) were injected, alone or together with CD4⁺CD25⁺ regulatory T cells (Thy1.2), before being purified as described above. Thy1.2⁺ cells were removed by incubating purified CD4⁺ T cells first with anti-Thy1.2 Ab (clone 53-2.1) and then with magnetic beads coupled to anti-rat Ab. Purified Thy1.2⁻ CD4⁺ T cells (0.5 × 10⁶) were then retransferred into CD3ε-deficient mice.

Colons were removed from mice 4 wk after transfer and fixed in PBS containing 10% formaldehyde. Five-micrometer paraffin-embedded sections were cut and stained with H&E.

Lymph nodes and spleens were pooled; homogenized in PBS, 5% FCS, and 0.2% NaN3 with a nylon cell strainer (Falcon, Franklin Lakes, NJ); and distributed in 96-well U-bottom microplates (4 × 10⁶ cells per well). Staining was performed on ice for 30 min per step.

Abs were purchased from BD PharMingen (San Diego, CA) unless otherwise indicated. The following Ab combinations were used: for four-color analysis, PE anti-Thy1.2, anti-CD19, anti-CD44, FITC anti-TCR_β, anti-GR1, PercP anti-CD4, anti-CD8, and biotinylated anti-CD5, anti-CD25, anti-CD28, anti-CD44, anti-CD45 or anti-CD69 with allophycocyanin-streptavidin development (BD PharMingen). Four-color immunofluorescence was analyzed using a FACSCalibur cytometer (BD Biosciences). List-mode data files were analyzed using CellQuest software (BD Biosciences).

Peripheral T cells were recovered at different time-points after transfer and then stimulated 6 h with PMA and ionomycin in the presence of brefeldin A. After surface staining, cells were fixed and permeabilized, and then intracytoplasmic staining was performed using PE anti-IL-2 (JES6-5H4), anti-IFN-γ (XMG1.2) and anti-IL4 (11B11). PE rat IgG1 was used as the isotype control. Abs were purchased from BD PharMingen.

One milligram of BrdU (Sigma-Aldrich, St. Louis, MO) was injected i.p., twice at a 30-min interval. To detect and characterize DNA-synthesizing cells, spleens and lymph nodes were removed 30 min after the second injection and pooled for cell preparation. Surface-stained cells were fixed and permeabilized in PBS containing 1% paraformaldehyde plus 0.01% Tween 20 for 48 h at 4°C and then submitted to the BrdU DNase detection technique as described in Ref. 29 , using FITC-conjugated anti-BrdU Ab (BD Biosciences).

A total of 10⁵ purified CD4⁺ T cells was cultured in RPMI Glutamax 1640 medium (Life Technologies-Invitrogen Corporation, U.K.), 10% FCS, 2 mM l-glutamine and antibiotics, with 5 × 10⁴ irradiated (2000 rad) peritoneal macrophages from H-2^b (C57BL/6) or H-2^d (BALB/c) mice in 96-well U-bottom microplates (Falcon). Supernatants were collected 96 h after the beginning of culture. IFN-γ production was assessed by ELISA. [³H]Thymidine was added to the culture at the same time, and proliferation was measured 16 h later.

CD4⁺ T cells from C57BL/Ba mice (Thy1.1) were purified by depletion of B cells, macrophages, granulocytes, and CD8⁺ T cells using magnetic beads; naive CD4⁺ T cells were then electronically sorted on the basis of their nonexpression of CD25 and low or absent expression of CD44, an activation marker expressed at high densities on effector/anergic/memory/regulatory T cells (Fig. 1,A). CD4⁺CD25⁺ T cells from C57BL/6 mice (Thy1.2) were purified by depletion of B cells, macrophages, granulocytes and CD8⁺ T cells, followed by positive selection of CD25-expressing cells (Fig. 1 B).

A total of 1 × 10⁶ purified naive CD4⁺ T cells were transferred with or without 0.2 × 10⁶ purified CD4⁺CD25⁺ T cells (to respect the normal proportion of CD25⁺ cells among CD4⁺ T cells, i.e., between 10 and 20% depending on the mouse strain) to T cell-deficient mice (CD3ε-deficient mice; Fig. 1,C; groups B and C). Total CD4⁺ T cells were also injected (Fig. 1 C; group A).

Mice injected with naive CD4⁺ T cells alone lost 20–40% of their initial weight 4 wk after transfer (Fig. 1, D and E). This loss of weight was associated with poor general health, severe diarrhea, and rectal prolapse. More precisely, these mice developed IBD characterized by extensive mononuclear cell infiltrates, depletion of mucin-secreting cells, ulcers, and pronounced epithelial cell hyperplasia (Fig. 2). As already described by other groups (23), coinjection of CD4⁺CD25⁺ cells together with naive CD4⁺ T cells prevented wasting and histological signs of inflammation (Figs. 1,E and 2). Similarly, mice injected with total CD4⁺ T cells showed no signs of IBD, demonstrating that a normal proportion of CD4⁺CD25⁺ regulatory T cells was sufficient to protect mice from the disease (Fig. 1, D and E).

The use of Thy1 as an allotypic marker to discriminate between injected naive CD4⁺ T cells and CD4⁺CD25⁺ regulatory T cells allowed us to study the absolute number and cycling status of peripheral expanding “naive” CD4⁺ T cells at various times after transfer (Fig. 3). The proliferation and expansion of naive CD4⁺ cells (Thy1.1) was studied according to whether or not regulatory T cells (Thy1.2) were coinjected. To study proliferation rates, mice were pulsed with BrdU, a thymidine analog whose incorporation into cycling cells is readily detected by using a specific mAb (29, 30).

Injected naive CD4⁺ T cells expanded to a similar extent during the first 2 wk, independently of regulatory T cell coinjection (Fig. 3,A, group B, group C: Thy1.2⁻ cells). More precisely, expansion was maximum in all groups between the first and the second week. Indeed, two days after transfer, CD4⁺ T cell recovery was around 20% of the absolute number of injected cells, i.e., around 0.2 × 10⁶ CD4⁺ T cells (data not shown). Expansion was only up to 2-fold during the first week whereas absolute numbers of recovered CD4⁺ T cells increased by 50-fold between the first and the second week. The proliferation rate (BrdU incorporation) and cell size (FSC) of recovered naive CD4⁺ T cells were similar in the two groups 1 and 2 wk after transfer (Fig. 3, B and C; Thy1.2⁻ cells). Only when they were injected alone (group B) did naive T cells continue to expand, giving rise to a CD4⁺ T cell pool five times larger than that obtained when regulatory T cells were coinjected (group C: Thy1.2⁻ cells) or when total CD4⁺ T cells were injected (group A). Accordingly, naive CD4⁺ T cells exhibited a higher proliferation rate and contained a higher proportion of blast cells 3 wk after transfer when injected alone than when coinjected with CD4⁺CD25⁺ regulatory T cells (Fig. 3, B and C).

In all groups, the absolute number of granulocytes increased strongly during the first 2 wk posttransfer (Fig. 3,A). Only in mice injected with total CD4⁺ T cells (group A) did the absolute number of peripheral granulocytes then return to normal, suggesting that a subset of CD4⁺ T cells other than CD25⁺ regulatory T cells participates in the control of inflammation. The absolute numbers of B lymphocytes as determined by CD19 staining varied similarly with time in all groups (Fig. 3 A).

The expression kinetics of activation markers (CD25, CD44, CD45RB, CD62L, CD69), costimulatory molecules (CD28, CD152), adhesion molecules (CD11a, CD54), and all molecules that could play a role in modulating T cell responses to self (TCR itself, CD5, CD45) by expanding naive CD4⁺ T cells were the studied according to whether or not CD4⁺CD25⁺ regulatory T cells were coinjected (Fig. 4).

We found that the pattern of surface molecule expression by expanding naive CD4⁺ T cells started to diverge as early as 1 wk after transfer according to whether or not regulatory T cells had been coinjected. Indeed, 1 and 2 wk after transfer, CD45 expression by naive CD4⁺ T cells increased less strongly when CD4⁺CD25⁺ regulatory T cells were coinjected. Moreover, naive CD4⁺ T cells exhibited stronger CD5 expression, at all time-points after transfer, when they were coinjected with CD25⁺ regulatory T cells. No such differences were observed concerning the expression of activation markers (CD25, CD44, CD45RB, CD62L, CD69), costimulatory molecules (CD28, CD152), adhesion molecules (CD11a, CD54), or the TCR itself (Fig. 4 and data not shown).

Thus, although the extent of naive CD4⁺ T cell proliferation during the first 2 wk after transfer was independent of CD4⁺CD25⁺ regulatory T cell cotransfer, their phenotype was already modulated by CD4⁺CD25⁺ regulatory T cells.

At different time-points after transfer, peripheral T cells were recovered and their capacity to produce IL-2, IFN-γ, and IL-4 (Fig. 5) was estimated ex vivo. More precisely, peripheral T cells were submitted to a short (6 h) stimulation by PMA and ionomycin in the presence of brefeldin A followed by intracytoplasmic staining of their cytokine production. Previous data have shown that regulatory CD4⁺CD25⁺ T cells did not interfere with T cell stimulation by PMA and ionomycin (19). Therefore, such a protocol allowed us to determine the differentiation status of expanding naive CD4⁺ T cells according to whether or not regulatory T cells were coinjected.

Independently of regulatory T cell coinjection, injected naive CD4⁺ T cells (Thy1.1) exhibited a similar pattern of cytokine production at all tested time-points after their transfer (Fig. 5). Only 4 wk after transfer did naive CD4⁺ T cells injected alone produce significantly less IL-2 than naive CD4⁺ T cells coinjected together with regulatory CD4⁺CD25⁺ T cells. Such a result certainly reflected the poor general health of mice injected only with naive CD4⁺ T cells at this time-point (Figs. 1,E and 2) and therefore the effect of stress on immune cells. More importantly, in both groups B and C, most expanding naive T cells differentiated into IFN-γ-producing cells, i.e., into Th1 effector/memory T cells. No significant IL-4 production was detected in both groups (Fig. 5 B). Thus, no evidence of a shift from a Th1 to a Th2 cytokine pattern induced by regulatory CD4⁺CD25⁺ T cells can be characterized in this model. These results demonstrate that regulatory CD4⁺CD25⁺ T cells do not interfere with the differentiation of expanding naive CD4⁺ T cells into Th1 effector/memory T cells.

The use of an allotypic marker to discriminate between injected naive CD4⁺ T cells and CD4⁺CD25⁺ cells also allowed us to purify expanding naive T cells and to analyze their capacities to respond to antigenic stimulation according to whether or not regulatory T cells were coinjected. The functional capacities of effector CD4⁺ T cells were tested both in vitro (Fig. 6) and in vivo, by analyzing their ability to induce IBD upon retransfer (Fig. 7).

The ability of H-2^b transferred CD4⁺ T cells to respond to H-2^d APCs was evaluated 4 wk after transfer, when the absolute number of recovered CD4⁺ T cells reached a plateau in all groups. When injected alone, expanding naive CD4⁺ T cells proliferated and produced IFN-γ to the same extent as control CD4⁺ T cells (Fig. 6; group B). By contrast, no response was observed when total CD4⁺ T cells were injected, or when CD4⁺CD25⁺ regulatory T cells were coinjected at the same time as naive CD4⁺ T cells, suggesting that, in these two groups (A and C), following their expansion, naive CD4⁺ T cells became anergic. Nevertheless, the ability of naive CD4⁺ T cells to respond to H-2^d APCs was restored by removing CD4⁺CD25⁺ regulatory T cells before culture. Indeed, in these conditions, they proliferated and produced IFN-γ (Fig. 6; group C: Thy1.2⁻) and IL-2 (data not shown) to the same extent as control CD4⁺ T cells. Thus, naive CD4⁺ T cells coinjected with CD4⁺CD25⁺ regulatory T cells do not completely lose their ability to respond to antigenic stimulation but are rather subject to suppression as long as they are together with CD4⁺CD25⁺ regulatory T cells.

To determine whether this also applied to pathogenic CD4⁺ T cells responsible for IBD, naive CD4⁺ T cells (Thy1.2⁻), injected alone or together with CD4⁺CD25⁺ regulatory T cells, were purified 4 wk later and then retransferred into empty hosts (Fig. 7,A). Naive CD4⁺ T cells injected first alone induced rapidly and uniformly IBD upon retransfer (Fig. 7,B; group B′ vs group B). For their part, naive CD4⁺ T cells coinjected first with CD4⁺CD25⁺ regulatory T cells also induced IBD upon retransfer although less rapidly that what could be observed with naive CD4⁺ T cells injected first alone (Fig. 7,B; group C′ vs group B′). When we assessed the ability of retransferred CD4⁺ T cells to respond to H-2^d APCs 4 wk after transfer, no significant differences were observed between the two groups (Table I).

Several mechanisms might explain how CD4⁺CD25⁺ regulatory T cells prevent autoimmune and inflammatory disorders, such as induction of anergy (31) or clonal deletion (32) of pathogenic T cells, or induction of Th2 rather than Th1 effectors. In this study, we show that CD4⁺CD25⁺ regulatory T cells do not affect the initial activation/proliferation of injected naive T cells as well as their differentiation into Th1 effectors. Moreover, expanding naive T cells do not become anergic when coinjected with CD4⁺CD25⁺ regulatory T cells, as they are still able to respond in vitro to antigenic stimulation when the regulatory T cells are removed. In these conditions, they still produce as much IFN-γ as before injection and as when injected alone. Finally, when purified, naive CD4⁺ T cells are able to induce IBD on reinjection into lymphopenic hosts. Thus, pathogenic T cells are not deleted, are still able to respond to self Ags, and are still able to produce IFN-γ, a cytokine crucial for IBD induction (23, 33).

In this study we analyzed in detail, during the first weeks following their transfer into lymphopenic hosts, the absolute number and phenotype of naive T cells according to whether or not CD4⁺CD25⁺ regulatory T cells were injected simultaneously. Initial expansion of naive T cells was not altered by coinjection of a “physiological” proportion of CD4⁺CD25⁺ regulatory T cells (Fig. 2) in agreement with recent studies performed in non lymphopenic recipients showing that the expansion of transferred naive T cells in response to their specific Ag was suppressed relatively late (34, 35). However, naive T cells stopped expanding after 2 wk when injected together with CD4⁺CD25⁺ regulatory T cells, whereas they continued to expand for up to 4 wk posttransfer when injected alone. In other words, CD4⁺CD25⁺ regulatory T cells controlled the late expansion of injected naive T cells. Other groups have also shown that inhibition of naive T cell expansion by CD4⁺CD25⁺ regulatory T cells is a late event, although the kinetics were quite different from those we observed (36, 37, 38). Indeed, on injecting the same number of regulatory T cells and naive T cells, or five times more regulatory T cells than naive T cells, Annacker et al. (36, 37) and Almeida et al. (38) observed a more rapid and more marked inhibition of naive T cell expansion. However, these authors failed to show whether CD4⁺CD25⁺ regulatory T cells acted by increasing naive T cell death or by decreasing the rate of division. In this study, using a physiological proportion of CD4⁺CD25⁺ regulatory T cells and BrdU as a marker of cell proliferation, we clearly found that CD4⁺CD25⁺ regulatory T cells inhibited the late proliferation of naive T cells injected into lymphopenic mice. During this late phase of expansion, the T cell repertoire of naive cells injected alone would certainly be skewed toward T cells with the highest affinity for self. This could explain why, when retransferred into lymphopenic hosts, naive T cells that had been injected first alone induced IBD more rapidly than initially observed, and also more rapidly than retransferred naive T cells that had first been coinjected with CD4⁺CD25⁺ regulatory T cells.

Interestingly, we found that the expression pattern of surface molecules on expanding naive CD4⁺ T cells started to diverge as early as 1 wk after transfer according to whether or not regulatory T cells were coinjected. In particular, cotransferred CD4⁺CD25⁺ regulatory T cells inhibited CD45 up-regulation on expanding naive T cells. CD45 augments TCR signaling by positively regulating the activity of both Lck and Fyn (39, 40, 41, 42). Thus, CD4⁺CD25⁺ regulatory T cells probably limit the expansion of expanding naive T cells by inhibiting CD45 up-regulation. It is noteworthy that knock-in mice in which a single critical amino acid in the inhibitory structural wedge of CD45 has been mutated, resulting in inappropriate CD45 activation and inappropriate lymphocyte activation, develop lymphoproliferation and autoimmunity (43).

Naive T cells also rapidly expressed higher surface densities of CD5 when injected together with CD4⁺CD25⁺ regulatory T cells than when injected alone. Current data indicate that CD5 inhibits TCR signal transduction (43, 44, 45, 46). Up-regulated CD5 expression at the surface of naive T cells would raise their activation threshold and thereby limit their expansion. Taken together, these results point to promising lines of investigation regarding the underlying regulatory mechanisms, for example by using CD5-deficient mice as a source of naive cells.

In conclusion, we show that the behavior of naive CD4⁺ T cells transferred into lymphopenic mice differs according to whether or not regulatory T cells are injected simultaneously. First, in the presence of CD4⁺CD25⁺ regulatory T cells, CD45 up-regulation is diminished on naive T cells while CD5 up-regulation is enhanced. These phenotypic changes would raise the activation threshold of naive T cells, possibly explaining why these cells stopped proliferating earlier when injected together with CD4⁺CD25⁺ regulatory T cells. Nevertheless, pathogenic cells were still found among expanding naive T cells, and were also still functional, as they were able to induce IBD upon retransfer to lymphopenic hosts. Moreover, we found no signs of a Th1-to-Th2 cytokine shift ex vivo or in vitro. Thus, all these results suggest that CD4⁺CD25⁺ regulatory T cells act by actively suppressing effector T cell functions as long as the two subsets remained confined together. Whether CD4⁺CD25⁺ regulatory T cells act directly on effector T cells or indirectly in modulating APCs remains to be elucidated.

These results should help to elucidate the mechanisms by which CD4⁺CD25⁺ regulatory cells regulate pathogenic T cell effector functions in vivo and protect from inflammatory and autoimmune diseases. They also point to new types of immunological manipulations and/or cellular therapy protocols aimed at preventing inflammatory and autoimmune diseases.

We thank B. Faideau, F. Lepault, and C. Pénit for critical reading of the manuscript and A. Benraiss for his invaluable help in histology.