Resistance to CD4⁺CD25⁺ Regulatory T Cells and TGF-β in Cbl-b^−/− Mice

Optimal T cell activation requires signaling through both the TCR and the CD28 costimulatory receptor. CD28 costimulation is believed to set the threshold for T cell activation. Cbl-b, a member of the Cbl/Sli family of molecular adaptors and a ubiquitin ligase, has been shown to negatively regulate CD28-dependent T cell activation (1, 2). Recent studies using gene-targeted mice lacking Cbl-b (Cbl-b^−/− mice) have demonstrated the importance of this molecule in both T cell activation and the development of autoimmunity (1, 2). Cbl-b^−/− T cells display increased proliferation after TCR stimulation and produce increased amounts of IL-2, but not IFN-γ or TNF-α (1). Cbl-b^−/− mice have T cells that are independent of CD28 costimulation, in that they do not require CD28 engagement for IL-2 production and proliferation (1, 2).

Bachmaier et al. (1) have demonstrated that Cbl-b^−/− mice develop spontaneous autoimmunity characterized by autoantibody production, infiltration of activated T and B lymphocytes into multiple organs, and parenchymal damage. Chiang et al. (2) have demonstrated that Cbl-b^−/− mice are highly susceptible to experimental autoimmune encephalomyelitis. The increased autoimmunity found in Cbl-b^−/− mice is believed to be a product of the signaling abnormalities, the CD28 independence, and the resultant T cell hyper-reactivity. However, the mechanisms leading from T cell hyper-reactivity to autoimmunity remain unclear. In the present study we have asked whether the signaling abnormalities in Cbl-b^−/− T cells could play a role in the increased autoimmunity in Cbl-b^−/− mice by affecting the interactions of Cbl-b^−/− regulatory T cells (T_reg)² and Cbl-b^−/− effector T cells (T_eff). AlthoughCD4⁺CD25⁺ T_reg have been shown to suppress CD4⁺CD25⁻ T_eff and CD8⁺ T cells in vitro and to play a role in the prevention of autoimmunity in vivo (3, 4, 5, 6, 7, 8, 9), the function and interaction of T_reg and T_eff have not been examined in Cbl-b^−/−mice.

Tang et al. (10) have recently demonstrated that CD28 controls both the survival and proliferation of T_reg in vivo. In addition, T_reg derived from TCR-transgenic mice have been found to be capable of proliferating in response to specific Ags, but this proliferation is dependent on dendritic cells and B7 costimulation (11). Thus, in addition to the fact that Cbl-b^−/− mice are prone to autoimmunity, the relevance of CD28-B7 interactions in the normal physiology of T_reg suggested that CD4⁺ T cell regulatory interactions could be abnormal in CD28-independent Cbl-b^−/− mice.

In the present studies we have examined for the first time the status of CD4⁺ T cell regulatory interactions in vitro in Cbl-b^−/− mice. We now report that CD4⁺CD25⁺ T_reg from Cbl-b^−/− mice function normally in vitro in suppressing wild-type (WT) CD4⁺CD25⁻ T_eff. However, we have found that CD4⁺CD25⁻ T_eff from Cbl-b^−/− mice are resistant to regulation by both Cbl-b^−/− and WT T_reg. In addition, Cbl-b^−/− CD4⁺CD25⁻ T_eff are resistant to regulation by exogenously added TGF-β. This resistance of T_eff to inhibition by T_reg and TGF-β may help explain the increased autoimmunity seen in Cbl-b^−/− mice. Furthermore, in light of similar findings in both aged NOD mice and T_eff exposed to LPS-treated dendritic cells, our present findings suggest that resistance to regulation may be a relevant mechanism in both normal immune function and autoimmunity (12, 13).

C57BL/6 (WT) mice were obtained from The Jackson Laboratory (Bar Harbor, ME). Cbl-b^−/− mice on a C57BL/6 background (a gift from Dr. H. Gu (National Institutes of Health, Bethesda, MD) were bred in our facilities under specific pathogen-free conditions. All animals were used between the ages of 6 and 10 wk.

The following Abs/reagents were used: from Miltenyi Biotec (Auburn, CA): anti-CD25-PE (clone 7D4) and anti-PE microbeads; and from BD Pharmingen (San Diego, CA): anti-CD25-PE (clone PC61), anti-CD4-FITC (clone GK1.5), anti-CD8α (clone 5H10-1), anti-CD24 (clone J11d), anti-CD3ε (clone 145-2C11), anti-CD4 PerCP (RM4-5), anti-CD45RB, and CD62L. Affinity-purified goat anti-rat IgG (Kierkegaard & Perry, Keene, NH) was used for panning. Human recombinant TGF-β1 was purchased from R&D Systems (Minneapolis, MN).

Splenocytes from WT and Cbl-b^−/− mice were stained with anti-CD8α and anti-CD24 Abs and plated onto petri dishes (27 × 10⁶/petri dish) coated with goat anti-rat IgG Ab (25 μg/ml). After 1.5 h at 4°C, CD8α⁺ T cells and CD24⁺ cells were depleted by panning, and the CD4⁺-enriched population was separated on magnetic columns (MS MACS separation column; Miltenyi Biotec) after staining with anti-CD25-PE, followed by anti-PE-conjugated microbeads.

CD4⁺CD25⁻ T_eff (5 × 10⁴), 5 × 10⁴ irradiated (2600 rad) WT splenocytes, and various numbers of CD4^+/+CD25^+/+ T_reg were plated in round-bottom, 96-well plates in RPMI 1640 supplemented with 10% FCS and 5 × 10⁻⁵ M 2-ME. Cultures were stimulated with varying doses of soluble anti-CD3 Ab for 48 h, and [³H]thymidine was added for the last 6 h. Cultures were harvested using a semiautomated cell harvester and assayed using a beta scintillation counter. To derive T cell-depleted splenocytes, Cbl-b^−/− splenocytes were stained with anti-CD4 and anti-CD8 microbeads. This population was then negatively selected on magnetic columns, and the negatively selected cells were used as a source of T cell-depleted APCs. This population routinely consisted of <5% T cells (data not shown).

Freshly isolated CD4⁺CD25⁻ WT and Cbl-b^−/− T_eff (10 × 10⁶cells/ml) were labeled with 2.5 μM CFSE for 10 min at 37°C and subsequently washed once in 20% FCS-PBS and twice in RPMI 1640 with 10% FCS. CFSE-labeled Cbl-b^−/− T_eff and CFSE-labeled WT T_eff were cultured with irradiated T cell-depleted WT splenocytes (as APCs) and 0.5 μg/ml soluble anti-CD3 Ab with or without T_reg for the time periods indicated. After culture, cells were stained with PerCP-anti-CD4 (BD Pharmingen; RM4-5), washed, and fixed for subsequent FACS analysis.

Freshly isolated CD4⁺CD25⁻ T_eff (5 × 10⁶) were plated in 24-well plates in 1 ml of RPMI 1640 supplemented with 10% FCS and 5 × 10⁻⁵ M 2-ME. Cultures were stimulated with or without TGF-β1 (5 ng/ml) for 30 min and harvested, and lysates were prepared for Western blots.

Samples were prepared by lysing 5 × 10⁶ purified cell populations in lysis buffer (50 mM Tris, 150 mM NaCl, 5 mM EDTA, 0.5% Nonidet P-40, 0.5% octyl glucoside, and 4 mM Pefabloc SC (Sigma-Aldrich, St. Louis, MO); for the Smad3 studies, 1 mM NaF, 1 mM NaVO₄, and 50 mM β-glycerophosphate were also added), followed by centrifugation at 14,000 rpm to remove debris. The protein content of the lysates was quantified by the Bradford method using the Bio-Rad protein assay (Hercules, CA). The indicated quantities of lysates were denatured by boiling 5 min in sample buffer, resolved by reducing SDS-PAGE (10% (w/v) acrylamide), and electrophoretically transferred to an Immun-Blot polyvinylidene difluoride membrane (Bio-Rad). Membranes were blocked for 1 h in blocking buffer (PBS with 5% dehydrated milk) and probed with anti-TGF-β receptor type II (anti-TGF-βRII) Ab (sc-17799; Santa Cruz Biotechnology, Santa Cruz, CA) at a dilution of 1/500 in blocking buffer, followed by incubation with HRP-conjugated goat anti-mouse secondary Ab (170-6516; Bio-Rad) at a dilution of 1/2000 in blocking buffer. Smad3 was detected with anti-Smad3 Ab (sc-6202; Santa Cruz Biotechnology), followed by HRP-conjugated anti-goat secondary Ab. For p-Smad3, membranes were blocked for 1 h in TBS with Tween 20 with 5% dehydrated milk, then probed with anti-phospho-Smad3 Ab (gift from Dr. E. Loef, Mayo Clinic, Rochester, MN), followed by incubation with HRP-conjugated goat anti-rabbit secondary Ab (170-6515; Bio-Rad). Alternatively, membranes were stripped and probed with anti-β-actin Ab (A-2066; Sigma-Aldrich) at a dilution of 1/1000, followed by incubation with HRP-conjugated goat anti-rabbit secondary Ab (170-6515; Bio-Rad) at a dilution of 1/2000 in blocking buffer. Bands were visualized using ECL substrate (Amersham Pharmacia Biotech, Arlington Heights, IL) detected by BioMax LS film (Eastman Kodak, Rochester, NY) and developed on an M35A X-OMAT film processor (Eastman Kodak).

Our first goal in these studies was to characterize CD4⁺CD25⁺ T_reg cells derived from Cbl-b^−/− mice. We isolated CD4⁺CD25⁺ T_reg and CD4⁺CD25⁻ T_eff from the spleens of WT C57BL/6 mice and Cbl-b^−/− mice. The T_reg populations isolated were usually >95% pure, and the T_eff populations isolated were usually 88–95% pure, as assessed by FACS analysis. To more completely characterize the T_eff populations used, we analyzed the expression of CD45RB and CD62L as well as the expression of CD25 within the T_eff populations derived from WT and Cbl-b^−/− mice. Our WT and Cbl-b^−/− T_eff populations were similar not only in CD25 expression, but also in CD45RB and CD62L expression (Fig. 1, A–C). These results are consistent with those reported by Bachmaier et al. (1) and indicate that the T_eff populations derived from Cbl-b^−/− mice did not consist of a greater proportion of activated or memory cells than the T_eff populations derived from WT mice. Also shown in Fig. 1 D is the similarity of CD25 expression of the WT and Cbl-b^−/−-derived T_reg populations. The number of T_reg isolated from WT mice was ∼0.35–0.40 × 10⁶/spleen, and the number of T_reg isolated from Cbl-b^−/− mice was ∼0.40–0.55 × 10⁶/spleen.

WT and Cbl-b^−/− CD4⁺CD25⁺ T_reg and CD4⁺CD25⁻ T_eff were tested for their ability to respond and/or suppress after stimulation in vitro with soluble anti-CD3 Ab. In confirmation of previous reports, WT T_reg showed no significant proliferation in response to anti-CD3 Ab using either 0.5 or 2 μg/ml anti-CD3 Ab. As expected, the WT T_reg were capable of suppressing the anti-CD3 Ab-stimulated response of WT T_eff using either 0.5 or 2 μg/ml anti-CD3 Ab (Fig. 2, A and B). Cbl-b^−/− CD4⁺CD25⁺ T_reg also showed no significant proliferation in response to anti-CD3 Ab using either 0.5 or 2 μg/ml anti-CD3 Ab. However, in contrast to WT cultures, we found that the Cbl-b^−/− CD4⁺CD25⁺ T_reg were incapable of suppressing the response of Cbl-b^−/− CD4⁺ CD25⁻ T_eff at both 0.5 and 2 μg/ml anti-CD3 Ab stimulation (Fig. 2, C and D). On occasion, the responses of Cbl-b^−/− T_eff were actually increased in cocultures with Cbl-b^−/− CD4⁺CD25⁺ T_reg.

To further assess the lack of regulation seen in cultures ofCbl-b^−/− CD4⁺CD25⁺ T_reg and CD4⁺ CD25⁻ T_eff, we next set up mixed cultures combining WT T_eff with Cbl-b^−/− T_reg and Cbl-b^−/− T_eff with WT T_reg. We found that Cbl-b^−/− T_reg were fully capable of suppressing the anti-CD3 Ab-stimulated response of WT T_eff. This was true using either 0.5 or 2 μg/ml anti-CD3 Ab (Fig. 3, A and B). This suggests that Cbl-b^−/− T_reg are fully capable of normal regulatory function. In contrast, WT T_reg, which were capable of suppressing WT T_eff, were not capable of suppressing the response of Cbl-b^−/− T_eff using either 0.5 or 2 μg/ml anti-CD3 Ab (Fig. 3, C and D). Again, the Cbl-b^−/− T_eff response was often slightly increased when cells were cocultured with WT T_reg. These results suggest that although the in vitro regulatory function of Cbl-b^−/− T_reg is normal, the ability of Cbl-b^−/− CD4⁺CD25⁻ T_eff to be suppressed by either WT or Cbl-b^−/− T_reg is abnormal.

Having found an abnormality in the suppressibility of Cbl-b^−/− T_eff at what is usually considered an optimal or supraoptimal T_reg:T_eff ratio of 1:1, we next tested the ability of WT and Cbl-b^−/− T_eff to be suppressed at a lower and a higher ratio. At a ratio of 0.5:1, the response of WT T_eff was suppressed 62% by WT T_reg. In contrast, at this ratio, the response of Cbl-b^−/− T_eff was not suppressed by Cbl-b^−/− T_reg (Fig. 4). We then asked whether we could detect any suppression of the Cbl-b^−/− T_eff response using the very high T_reg:T_eff ratio of 2:1. At a ratio of 2:1, the response of WT T_eff was suppressed 90% by WT T_reg. In contrast, at a 2:1 ratio, the response of Cbl-b^−/− T_eff was only 34% suppressed by Cbl-b^−/− T_reg (Fig. 4). Thus, even at the very high T_reg:T_eff ratio of 2:1, there was very little suppression of Cbl-b^−/− T_eff.

All of our prior studies were performed using WT-derived, irradiated, whole splenocytes as the APC population. We next wanted to examine the possibility that there might be functional differences between the WT and Cbl-b^−/− APC populations that played a role in the resistance to regulation seen with Cbl-b^−/− T_eff. In addition, we wanted to confirm our results using an APC population that was depleted of T cells. We, therefore, repeated our studies using irradiated, T cell-depleted (Tds), Cbl-b^−/− splenocytes as the APC population. Using these APCs, we found no significant difference in results from previous experiments. At a 1:1 T_reg:T_eff ratio and using 0.5 μg/ml anti-CD3 Ab for stimulation, the responses of WT T_eff were ∼85% suppressed. This level of suppression was mediated by both WT T_reg and Cbl-b^−/− T_reg (Fig. 5, A and B). However, the responses of Cbl-b^−/− T_eff were unable to be suppressed by either WT T_reg or Cbl-b^−/− T_reg (Fig. 5, C and D). These results suggest that the resistance to regulation of Cbl-b^−/− T_eff is not related specifically to either WT or Cbl-b^−/− APCs.

To further evaluate the resistance to regulation seen in Cbl-b^−/− T_eff, we next examined the T_eff responses at the level of cell division rather than at the level of DNA synthesis (i.e., [³H]thymidine incorporation). WT and Cbl-b^−/− CD4⁺CD25⁻ T_eff were labeled with CFSE, stimulated with anti-CD3 Ab and APCs, and cultured with or without the addition of T_reg. In the FACS analyses used to evaluate these responses, we gated on CD4⁺ T cells. As such, in those cultures to which T_reg were added, the T_reg were seen as a CSFE-negative population (Fig. 6, B, D, and F). The T_eff responses were examined at 48 and 60 h. In confirmation of our [³H]thymidine incorporation studies, WT T_reg suppressed the anti-CD3 Ab-stimulated division of WT T_eff. This was seen as a decrease in the number of waves of division in the presence of T_reg. The results of a typical experiment for WT T_eff, harvested at 60 h, are shown in Fig. 6, A and B (data not shown for the 48 h WT T_eff responses). In contrast, Cbl-b^−/− T_reg did not suppress anti-CD3 Ab-stimulated division of Cbl-b^−/− T_eff. This was seen as either no decrease in the waves of division or even a slight increase in the progression through cell division of the CFSE-labeled Cbl-b^−/− T_eff. The results of typical experiments for Cbl-b^−/− T_eff are shown in Fig. 6, C and D (harvested at 48 h), and in Fig. 6, E and F (harvested at 60 h). These results suggest that the resistance to T_reg regulation of Cbl-b^−/− T_eff occurs not only at the level of DNA synthesis, but also at the level of cell division.

TGF-β has been demonstrated to be an important regulator of immune responses in vivo (14, 15). In addition, TGF-β has been postulated to be a mechanism of suppression used by T_reg (5, 16). To investigate the possibility that the resistance to regulation of Cbl-b^−/− T_eff is a more general characteristic of these cells, we next examined the effect of exogenously added TGF-β on the anti-CD3 Ab-stimulated responses of WT and Cbl-b^−/− T_eff. No T_reg were used in these assays. As shown in Fig. 7, A and B, 5 and 25 ng/ml TGF-β suppressed the anti-CD3 Ab response of WT T_eff by ∼55–65%. In contrast, neither 5 nor 25 ng/ml TGF-β was able to significantly suppress the response of Cbl-b^−/− T_eff. At 5 ng/ml TGF-β, the response of Cbl-b^−/− T_eff actually increased (Fig. 7,C). At 25 ng/ml TGF-β, the Cbl-b^−/− T_eff response decreased, but only by ∼11% (Fig. 7 D). Ab to TGF-β reversed the effects of the added cytokine, confirming the specificity of the TGF-β effect (data not shown). Thus, Cbl-b^−/− T_eff demonstrated resistance to regulation not only by both WT and Cbl-b^−/− T_reg, but also by TGF-β. Together these findings of Cbl-b^−/− T_eff resistance to both TGF-β-mediated suppression and regulation by T_reg suggest that the autoimmunity seen in Cbl-b^−/− mice might be the result of numerous regulatory abnormalities associated with Cbl-b^−/− T cells.

To further characterize the defect in suppression of Cbl-b^−/− T_eff by TGF-β, we examined the expression of TGF-βRII by WT T_eff and Cbl-b^−/− T_eff. Western blots were performed on lysatesof freshly isolated, nonstimulated WT T_eff and Cbl-b^−/−CD4⁺CD25⁻ T_eff. Using Ab specific for TGF-βRII, we examined two dilutions of lysates for both Cbl-b^−/− CD4⁺CD25⁻ T_eff (Fig. 8,A, lanes b and c) and WT CD4⁺CD25⁻ T_eff (Fig. 8,A, lanes d and e). We found that the levels of expression of TGF-βRII were not significantly different in Cbl-b^−/− T_eff vs WT T_eff (Fig. 8 A). These results suggest that the abnormality in TGF-β-mediated suppression of Cbl-b^−/− T_eff is not the result of a decrease in expression of TGF-βRII.

Finally, we examined Smad3 phosphorylation in WT T_eff and Cbl-b^−/− T_eff using Western blots. The ligation of TGF-βRII by TGF-β leads to the heterodimerization of TGF-βRII with TGF-βRI. This activated receptor complex then recruits Smad2 and Smad3, which are phosphorylated by TGF-βRI. The phosphorylated Smad2 and Smad3 moieties each then heterodimerize with Smad4, and these Smad complexes can translocate to the nucleus and activate transcriptional responses (17). We incubated WT and Cbl-b^−/− CD4⁺CD25⁻ T_eff with 5 ng/ml TGF-β for 30 min and then performed immunoblots on the cell lysates using Abs to both Smad3 and phosphorylated Smad3. We found that non-TGF-β-stimulated WT T_eff and Cbl-b^−/− T_eff expressed equal levels of nonphosphorylated Smad3 (Fig. 8,B). Furthermore, although TGF-β suppresses WT CD4⁺CD25⁻ T_eff proliferation, but does not suppress (and often enhances) Cbl-b^−/− CD4⁺CD25⁻ T_eff proliferation, we found that WT T_eff and Cbl-b^−/− T_eff demonstrated an equal increase in phosphorylated Smad3 after incubation with TGF-β (Fig. 8 B). Thus, our results suggest that the resistance of Cbl-b^−/− T_eff to TGF-β-mediated suppression is the result of neither a decrease in expression of TGF-βRII nor an abnormality in the subsequent phosphorylation of Smad3, but may relate instead to abnormalities distal to these initial signaling events.

Cbl-b, a ubiquitin ligase, acts as a type of gate-keeper in that it serves as a negative influence on T cell activation unless unleashed through CD28 stimulation (18). This unleashing occurs, at least in part, by CD28’s ligation leading to Cbl-b ubiquitination and degradation (19). Two different laboratories have generated C57BL/6 mice that do not express Cbl-b (1, 2). In these mice the Cbl-b deficiency does not affect cellular development, but does lead to T cells that demonstrate increased proliferation and produce increased amounts of IL-2 after TCR stimulation (1, 2, 18). Chiang et al. (2) have shown that in Cbl-b^−/− T cells, the main TCR signaling pathways are not affected, but Vav activation is significantly enhanced. Thus, mutation of Cbl-b uncouples T cell proliferation, IL-2 production, and phosphorylation of Vav1 from the requirement for CD28 costimulation, leading to CD28 independence of Cbl-b^−/− T cell activation (1, 2).

Cbl-b^−/− mice have been found to be prone to both spontaneous and induced autoimmune diseases (1, 2). The spontaneous autoimmunity involves the development of autoantibodies to dsDNA and massive infiltration in multiple organs (1). The induced autoimmunity involves an increased susceptibility to the induction of experimental autoimmune encephalomyelitis (2). In addition, a mutation in the Cbl-b gene has been shown to be associated with a spontaneous model of diabetes in the rat (20). Although the CD28 independence, the decreased threshold for stimulation, and the T cell hyper-reactivity are postulated to be responsible for the increased autoimmunity in Cbl-b^−/− mice, the actual mechanisms underlying the increased autoimmunity are as yet unclear.

The mechanisms of suppression via CD4⁺CD25⁺ T_reg have yet to be definitively identified (9). Unresolved issues include questions about the cytokines and stimulatory factors required for the development and function of these cells and the nature of their Ag specificity. Shevach et al. (7, 9) have demonstrated that T_reg require stimulation through their TCR to mediate suppressive function and that T_reg suppress IL-2 production by T_eff. This suppression is noted at both the IL-2 mRNA and protein levels (7, 9). These CD4⁺CD25⁺ T_reg appear to have significant in vivo functional relevance, as evidenced by their ability to prevent certain autoimmune syndromes and their regulatory role in normal immune responses (3, 4, 9, 21, 22). The suppressive function of T_reg has also been demonstrated in vitro, where they have been found to suppress both anti-CD3 Ab-stimulated and Ag-specific responses of CD4⁺ and CD8⁺ T cells (8, 9).

CD4⁺CD25⁺ T_reg have been shown to depend on CD28 for their development, homeostatic proliferation, and Ag-driven proliferation (4, 10, 11). CD28^−/− NOD mice have been demonstrated to develop exacerbated diabetes because of the lack of CD4⁺CD25⁺ T_reg development (4). Tang et al. (10) have recently shown that CD28 controls both the survival and the proliferation of T_reg in vivo. In addition, T_reg derived from TCR-transgenic mice have been found to be capable of proliferating in response to specific Ags, but this proliferation is dependent on dendritic cells and B7 costimulation (11). Thus, given the relevance of CD28-B7 interactions in the normal physiology of T_reg, the CD28 independence of Cbl-b^−/− T cells and the increased proclivity of Cbl-b^−/− mice to autoimmunity suggested that CD4⁺ T cell regulatory interactions could be abnormal in these mice.

We now report that the function of Cbl-b^−/− CD4⁺CD25⁺ T_reg is essentially identical with that of T_reg derived from WT mice. However, the ability of Cbl-b^−/− CD4⁺CD25⁻ T_eff to be regulated is significantly different from that of T_eff derived from WT mice. Specifically, in the in vitro response to soluble anti-CD3 Ab, Cbl-b^−/− T_eff, compared with WT T_eff, are significantly less able to be suppressed by either Cbl-b^−/− or WT T_reg. Using [³H]thymidine incorporation proliferation assays, this resistance to regulation of Cbl-b^−/− T_eff is seen over various ratios of T_reg:T_eff. In addition, this resistance is found when the strength of activating signal is varied by stimulation with either 0.5 or 2 μg/ml anti-CD3 Ab and when either WT or Cbl-b^−/− APCs are used in the assays. Finally, using CFSE labeling, we confirmed that the resistance to T_reg regulation of Cbl-b^−/− T_eff occurs not only at the level of DNA synthesis, but also at the level of cell division.

To date, almost all studies of CD4⁺ regulatory function have focused on the T_reg populations rather than on the regulated T_eff. However, our present results of T_eff resistance to regulation are suggestive of two other recently described findings (12, 13). In the first, a decrease in suppressibility of murine CD4⁺CD25⁻ T_eff has been found in conjunction with LPS-activated dendritic cells (13). Such LPS-activated dendritic cells have been found to mediate a decrease in CD4⁺CD25⁻ T_eff suppressibility through the secretion of IL-6 and an as yet unidentified cytokine (13). At a 1:1 ratio, T_reg-mediated suppression has been shown to decrease from ∼85% with normal dendritic cells to ∼25% when LPS-treated dendritic cells were used (13). These results are similar to ours; at a 1:1 ratio we have found 85% suppression of the WT response and no suppression of the Cbl-b^−/− T_eff response. In a second relevant study, Gregori et al. (12) have reported age-related changes in both CD4⁺CD25⁺ T_reg and CD4⁺CD25⁻ T_eff in NOD mice. These authors have found that as autoimmune diabetes progresses, the CD4⁺CD25⁻ pathogenic (effector) T cells become progressively less sensitive to immunoregulation by CD4⁺CD25⁺ T_reg (12). These results in NOD mice suggest that our present findings may be applicable to various autoimmune states in addition to that seen in Cbl-b^−/− mice.

The mechanisms underlying our findings of Cbl-b^−/− T_eff resistance to regulation are as yet unclear. There is evidence suggesting that the strength of the activating signal may play a role in determining the suppressibility of T_eff (23, 24). These studies have demonstrated that stronger activating signals are associated with a decrease in the suppressibility of CD4⁺CD25⁻ T_eff. Baecher-Allen et al. (23) have demonstrated a correlation between the in vitro strength of signal of activation and the ability of human CD4⁺CD25⁻ T_eff to be suppressed by T_reg in vitro. In these studies the stronger activating signals are associated with a decrease in suppressibility of T_eff (23). The stronger activating signals are also associated with a decrease in the suppressor function of T_reg (23). Although we did not examine this issue of T_reg function specifically, in general, our findings indicate that WT T_reg and Cbl-b^−/− T_reg function with equal efficiency. We believe that it is possible that the resistance to regulation described now for Cbl-b^−/− T_eff results from the fact that the hyper-reactivity of Cbl-b^−/− T cells is equivalent to a state of increased strength of activating signal for any given stimulus. However, the resistance to regulation ofCbl-b^−/− T_eff is not simply related to differences in absolute levels of proliferation between Cbl-b^−/− and WT T_eff. This can be seen in our results, where Cbl-b^−/− T_eff are not suppressible when proliferating at a level of ∼38,000 cpm of [³H]thymidine incorporation (Fig. 3,D), whereas WT T_eff are highly suppressible at a similar level of proliferation (Fig. 5, A and B).

Given the finding of resistance to T_reg regulation, we also examined whether Cbl-b^−/− CD4⁺CD25⁻ T_eff are resistant to other forms of immunoregulation. It has previously been reported that Cbl-b^−/− T cells are suppressed normally by CTLA-4 (2). Suppression by IL-10 has been reported to be mediated through down-regulation of B7 expression on APCs and also through alteration of the CD28 costimulation pathway in T cells (25, 26). The CD28 dependence of these mechanisms suggested that Cbl-b^−/− T_eff are unlikely to be suppressible by IL-10. We therefore examined the suppressibility of Cbl-b^−/− T_eff by TGF-β. TGF-β plays a role in immunoregulation, and there are reports suggesting that cell-bound TGF-β is important in the suppression mediated by T_reg (5, 14, 15, 16). TGF-β inhibits T cell proliferation and function at many different levels, including the production of IL-2 (27). We now report that Cbl-b^−/− T_eff are also significantly less suppressed by exogenously added TGF-β than are WT T_eff. To further understand this TGF-β resistance, we first characterized the expression of TGF-βRII by freshly isolated T_eff. We found that the expression of TGF-βRII did not significantly differ between WT and Cbl-b^−/− T_eff. We then examined Smad3 phosphorylation in WT T_eff and Cbl-b^−/− T_eff after incubation with TGF-β. Although TGF-β suppresses WT CD4⁺CD25⁻ T_eff proliferation, but does not suppress (and often enhances) Cbl-b^−/− CD4⁺CD25⁻ T_eff proliferation, we found that WT T_eff and Cbl-b^−/− T_eff demonstrated an equal increase in phosphorylated Smad3 after incubation with TGF-β (Fig. 8 B). This suggests that the resistance of Cbl-b^−/− T_eff to TGF-β-mediated suppression is the result of neither a decrease in the expression of TGF-βRII nor an abnormality in the subsequent phosphorylation of Smad3, but may relate instead to abnormalities distal to these initial signaling events. Future studies in our laboratory will involve identifying these more distal TGF-β signaling alterations in Cbl-b^−/− T cells.

The findings of resistance to TGF-β-mediated suppression together with the resistance to regulation by T_reg suggest that the autoimmunity seen in Cbl-b^−/− mice might be the result of numerous regulatory abnormalities associated with Cbl-b^−/− T cells. Given recent reports of similar resistance to T_reg regulation in aged NOD mice and in T_eff exposed to LPS-treated dendritic cells (12, 13), resistance to regulation may be a relevant mechanism in both normal immune function and autoimmunity. Future studies in our laboratory will attempt to confirm these immunoregulatory abnormalities in Cbl-b^−/− mice using in vivo approaches.

We thank Dr. Ethan Shevach, National Institutes of Health, for his critical reading of this manuscript and his helpful comments.