The mechanism and stimulatory requirements of regulatory T cell (Treg)-mediated suppression are still unclear. To assess the requirement for Treg stimulation by cognate peptide:MHC, we used T cells from OTII and AND TCR transgenic mice that are specific for and restricted by distinct, noncrossreactive peptide:MHC combinations. This allowed us to independently activate Tregs and their conventional T cell (Tconv) targets. Surprisingly, we found that suppression can occur in the absence of peptide:MHC-mediated stimulation of Tregs. This suppression was Treg dependent and not due to cold target inhibition. Using Rag1−/− TCR transgenic T cells, we show that regulation of Tconv proliferation by heterogeneous Tregs is not due to alloreactivity or crossreactivity. Finally, using anti-TCR-Vβ8-coated microbeads and Vβ8− Tregs, we show that TCR stimulation-independent suppression can occur in the absence of APCs. These data suggest that Tregs may possess constitutive regulatory activity that can be mediated in the absence of cognate peptide:MHC-TCR stimulation.
It has become increasingly clear that regulatory T cells (Treg)3 represent an important subset in the control of the immune response. Not only do they play a significant role in peripheral tolerance and protection against autoimmune disease, they also modulate immune responses to infection and tumors (1). Several studies have assessed the requirements for Treg-mediated suppression in vitro. Tregs have been shown to be nonresponsive to TCR stimulation in the absence of exogenous IL-2 (2). It is believed that suppression in vitro occurs via a cell-cell contact-dependent, cytokine-independent mechanism (3, 4); however, a role for short-acting and/or labile factors has not been ruled out (5). Furthermore, Treg-mediated suppression does not have to be mediated by Tregs that have the same Ag specificity and MHC restriction as the suppressed T cell population (2, 6). Lastly, it has been suggested that Tregs require activation through their TCR to exert their suppressive activity (2, 4, 6). This was observed for both polyclonal and transgenic populations of T cells.
A number of mechanisms of Treg suppression have been proposed using in vitro and in vivo models (7). These include suppression via contact with target cells or APCs (2, 4, 8), soluble factors (9, 10) or through cell killing (11, 12). Although it is still unclear how Tregs exert their activity, growing evidence suggests that multiple mechanisms are used depending upon context, location, and disease state (7). Studies investigating the kinetics of Treg activity in vitro suggest that Tregs exert their suppressive activity in a narrow kinetic window, within 6–10 h of culture (13, 14). These observations suggest that perhaps resting Tregs may exist in a somewhat “preactivated” state and thus raise the possibility that they possess some constitutive regulatory activity.
This prompted us to re-evaluate the requirements for Treg-mediated suppression. We used purified, conventional, naive T cells (Tconv) and Tregs from two TCR transgenic mouse models, AND and OTII, which have distinct and noncrossreactive specificities and restriction elements. In accordance with what others have seen, we show that suppression by Tregs is neither Ag- nor MHC-restriction dependent. However, surprisingly, we found that suppression by Tregs can occur in the absence of stimulation through the TCR.
Materials and Methods
Congenic C57BL/6J and B10.BR mice were purchased from The Jackson Laboratory. Foxp3gfp mice (15), OTII TCR transgenic mice (16), and AND TCR transgenic mice (17) (on wild-type, Rag1−/−, and/or Foxp3gfp crosses) were bred and maintained at St. Jude Children’s Research Hospital. All animal experiments were performed in American Association for the Accreditation of Laboratory Animal Care-accredited, specific pathogen-free, helicobacter-free facilities following national, state and institutional guidelines. Animal protocols were approved by the St. Jude Animal Care and Use Committee.
Cell purification and flow cytometry
T cells (Tconv and Treg) from spleen and/or lymph nodes were stained and purified by FACS using anti-CD4, anti-CD25 and, in some experiments, anti-CD45RB. Dendritic cells (DCs) were purified from spleen by FACS using anti-CD11c. Vβ8 Tconv and Treg were purified by FACS using anti-Vβ8 Ab (F23.1). Abs against Vβ3, CD3, and Foxp3 were additionally used for flow cytometric analysis. The anti-Foxp3 Ab was purchased from eBioscience. All other Abs used were from BD PharMingen.
Tconv (CD4+CD25−; 2.5 × 104) were cultured in 96-well, round-bottom plates with DCs (104) and various concentrations of peptide (OVA326–339) and/or pigeon cytochrome c (PCC) peptide 88–104 (PCC88–104) for 68 h (200-μl final volume). Cells were pulsed with [3H]thymidine for the last 8 h of culture and proliferation was measured.
In vitro suppression assays
Tconv (2.5 × 104) were cultured with DCs or irradiated splenocytes and peptide (Ova326–339 and/or PCC88–104) for 68 h with titrating amounts of Tregs. Cells were pulsed with [3H]thymidine for the last 8 h of culture and proliferation was measured. For CFSE dilution, Tconv were purified by FACS, washed in PBS, resuspended in PBS plus 0.1% BSA with 5 μM CFSE and incubated for 10 min at 37°C. Cells were washed, resuspended in medium, and incubated with titrating numbers of Tregs and DCs with peptide for 60 h. CFSE dilution was measured by flow cytometry. For assays using Vβ8+- and Vβ8−-purified cells, Tconv and Tregs were incubated with anti-Vβ8-coated sulfate latex microbeads (Molecular Probes) for 68 h, pulsed with [3H]thymidine for the last 8 h of culture, and and proliferation measured. Percent suppression is calculated by subtracting the cpm of Tconv plus Treg from the cpm of Tconv alone and dividing by the cpm of Tconv alone.
Statistical analysis was performed using the unpaired Student’s t tests.
Results and Discussion
Tregs do not require activation through their TCR to function
To evaluate the stimulatory requirements for Treg-mediated suppression, we cocultured T cells from OTII TCR transgenic mice (OVA326–339-specific; H-2Ab restricted) (16) and AND TCR transgenic mice (PCC88–104-specific; H-2Ek restricted) (17) with DC from B10.BR (expressing H-2Ek) and C57BL/6 (expressing H-2Ab) mice. To verify that there was no crossreactivity, we assessed the ability of purified OTII and AND T cells to proliferate in the presence of C57BL/6 or B10BR DCs with either peptide. As is clearly shown, no crossreactivity to the irrelevant peptide or MHC was observed (Fig. 1 A). In other words, OTII T cells (Tconv or Tregs) do not respond to PCC88–104 in the presence of either C57BL/6 or B10.BR DCs, nor to Ova326–339 when incubated with B10.BR DCs. Likewise, AND T cells do not respond to Ova326–339 in the presence of either C57BL/6 or B10.BR DCs nor to PCC88–104 when incubated with C57BL/6 DCs. Neither T cell proliferated when incubated with MHC class II-deficient DCs (data not shown).
Previous studies have suggested that Tregs require activation through their TCR to function, because polyclonal populations of Tregs are unable to suppress TCR transgenic T cells proliferating in response to their cognate Ag. However, this inability can be overcome with the addition of anti-CD3 (2, 4). Furthermore, it was suggested, but not shown, that Ag-specific Treg were able to suppress Tconv of a different Ag specificity but only in the presence of the Ag for which the Tregs were specific (4, 6). To re-examine this requirement, we performed coculture experiments to assess the ability of Tregs to suppress Tconv of different Ag specificities in the absence of their cognate MHC:peptide interaction. As expected, OTII Tregs were able to suppress proliferation of OTII Tconv when cultured with C57BL/6 DCs and both stimulating peptides (Fig. 1,B and supplemental Fig. 1A).4 Likewise, AND Tregs suppressed AND Tconv in the presence of peptide-pulsed B10.BR DCs. Surprisingly, AND Treg were also capable of suppressing OTII Tconv when cultured with C57BL/6 DCs, and OTII Tregs were capable of suppressing AND Tconv in the presence of B10.BR DCs. This suggested that Tregs do not need to be stimulated through their TCR to function, as in each system the Treg-stimulating DC and restriction element is absent. To confirm these results and alleviate any concerns about potential activation of Tregs by their cognate peptides, these assays were repeated with cultures containing only the Tconv-stimulating peptide (Fig. 1 C and supplemental Fig. 1B). Again, Tregs were able to suppress Tconv of different Ag specificity in the absence of any Treg-specific TCR-activation signal. On the whole, the “opposing” Tregs were slightly less efficient than Tregs with the same transgenic TCR but were nonetheless quite effective regulators in vitro. These observations were very surprising and clearly contrasted with previous observations (2, 4).
A potential concern with these experiments is that suppression of Tconv proliferation may be occurring due to cold target inhibition, in which T cell contact with stimulatory APCs is sterically limited by the third party cell population. To rule out this possibility, we cultured Tconv with their cognate peptide:DC combination and replaced the Tregs in these assays with irrelevant, nonstimulated Tconv cells (i.e., stimulation of OTII Tconv with OVA peptide plus C57BL/6 DCs in the presence or absence of increasing numbers of AND Tconv). Suppression was observed at a 1:1 ratio of stimulated Tconv to nonstimulated Tconv, demonstrating that this ratio should not be used to assess Treg activity (Fig. 1 D). However, no suppression was observed at lower ratios, suggesting that cold target inhibition cannot account for the ability of Tregs to suppress in the absence of cognate MHC:peptide stimulation.
TCR stimulation-independent suppression by transgenic T cells is not due to crossreactivity or alloreactivity
Several studies have suggested that the TCR repertoire used by Tconv and Tregs is distinct and that Tregs in TCR transgenic mice may use nonclonotypic TCR during selection (18, 19), although this has been contested by others (20). Furthermore, it has recently been shown that high affinity, autoreactive TCR are rare on both Tconv and Tregs (21). Nevertheless, the possibility remains that Tregs in these assays may be stimulated by the noncognate or allo-MHC, whereas their Tconv brethren are not. To perform these assays in the absence of any potential alloreactivity or crossreactivity due to the expression of nonclonotypic TCR chains, we crossed the OTII and AND TCR transgenic mice onto a Rag1−/− background. Whereas Rag1−/− OTII TCR transgenic mice appeared to have a negligible Foxp3+ Treg population, Rag1−/− AND TCR transgenic mice do have a clearly definable, if reduced, Treg population as determined by intracellular staining for Foxp3 (Fig. 2,A and supplemental Fig. 2A). This was further verified by crossing the Rag1−/− AND TCR transgenic mice with Foxp3gfp reporter mice (15). Flow cytometric analysis of both the spleen and thymus demonstrates that the CD4+CD25+CD45RBlow population of AND Tregs used in the suppression assays are Foxp3+ (Fig. 2,B and supplemental Fig. 2B) as well as essentially 100% Vβ3 clonotype positive (supplemental Fig. 3). Although the Rag1−/− OTII TCR transgenic mice had a minimal number of Foxp3+ T cells, there was a small but detectable CD4+CD25+CD45RBlow population that appeared to have potent regulatory activity as measured by its ability to inhibit CD4+CD25−CD45RBhigh Rag1−/− OTII Tconv proliferation in a CFSE dilution assay (Fig. 3 and supplemental Fig. 4). Strikingly, Rag1−/− AND Tregs were also capable of suppressing the proliferation of Rag1−/− OTII Tconv in the absence of the appropriate Treg stimulatory peptide:DC combination.
Finally, we used two additional experimental systems to further verify the ability of Tregs to function in the absence of TCR stimulation. First, purified OTII Tconv were incubated with Tregs from either OTII or C57BL/6 mice in the presence of the OVA peptide. The polyclonal population of C57BL/6 Tregs was able to suppress OTII Tconv proliferation almost as well as the OTII Tregs under conditions that would only activate OVA-reactive T cells (Fig. 4,A and supplemental Fig. 5A). Second, to eliminate MHC+ APCs from the experimental system, we used Vβ8+ Tconv stimulated with anti-Vβ8-coated microbeads. Purified Vβ8− Tregs were equally capable of suppressing Vβ8+ Tconv proliferation as the Vβ8+ Tregs without TCR stimulation (Fig. 4 B and supplemental Fig. 5B). Taken together, these data suggest that Tregs can mediate some degree of regulatory activity in the absence of TCR or MHC stimulation, indicating that they possess constitutive regulatory activity.
Tregs possess constitutive regulatory activity
Consistent with previous findings, our data confirm that Tregs need not be specific for, or restricted by, the same peptide:MHC combination as the target Tconv (2, 6). However, to our surprise, activation of the Tregs via their TCR does not appear to be an absolute requirement, contrary to previous suggestions (2, 4, 6). Although the reason behind these discrepancies is unclear, it could be due to differences in the Ag presenting populations and/or the transgenic/background combinations used. For instance, in most assays we used purified DCs as opposed to T cell-depleted splenocytes, which may be better suited to mediate Treg activity because DCs have been suggested to play a role in Treg-mediated suppression in vivo (8). Therefore, their “contribution” may be amplified in our assays. It is important to note that in all assays suppression in the absence of the Treg-activating peptide:MHC is seen, further validating our results.
The ability of Tregs to mediate suppression without prerequisite TCR-mediated activation may not be without precedent. Indeed, unmanipulated Tregs, directly ex vivo, appear to have a somewhat activated phenotype, suggesting that they might be “primed for action” (22). For instance, a number of molecules shown to be up-regulated on Tregs, such as CD25, CTLA-4, OX40, and LAG-3, are also found on activated Tconv (23, 24, 25). Furthermore, several studies have suggested that the kinetics of Treg activity in vitro are very rapid (6–10 h), further supporting the notion that Tregs can mediate some regulatory activity in the absence of initiating a complete activation program (13, 14). Although it seems unlikely, we cannot completely rule out the possibility that the sorting and manipulation of the Tregs led to some preactivation that might have affected their performance in the in vitro Treg assays. It has been suggested that Tregs possess a distinct TCR repertoire and have a higher reactivity to self-peptide:MHC (18, 19), although this notion has been contested (20). However, it could be argued that the Rag1−/− AND Tregs used our assays mediate suppression due to stimulation by the self-peptide:MHC expressed by the APCs. Although we cannot completely rule out this possibility, it is noteworthy that we reached similar conclusions regarding the ability of Tregs to suppress without TCR stimulation using anti-Vβ8-stimulated Vβ8+ Tconv with Vβ8− Tregs. In this system there are no DCs and, thus, no MHC by which Tregs might be stimulated. Thus, our data provide evidence that Tregs can suppress in the absence of cognate peptide:MHC-mediated stimulation, suggesting that further re-evaluation of the requirements of Treg-mediated suppression may be required. Whether Tregs need to be stimulated via their TCR in vivo to mediate their regulatory activity remains to be determined. It is conceivable that some measure of constitutive regulatory activity afforded by Tregs in vivo may be critical in mediating the timely control of immune responses.
We thank Steve Schoenberger, Bill Heath, Terry Geiger, and Alexander Rudensky for TCR transgenic and Foxp3gfp reporter mice, and Richard Cross, Jennifer Rogers, and Greig Lennon for assistance with flow cytometry. We are also very grateful to Karen Forbes and the staff in the St. Jude Animal Resource Center for mouse colony management.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by National Institutes of Health Grants AI39480 and AI058156, St. Jude Cancer Center Support Center of Research Excellence Grant CA-21765, and the American Lebanese Syrian Associated Charities.
Abbreviations used in this paper: Treg, regulatory T cell; DC, dendritic cell; PCC, pigeon cytochrome c; Tconv, conventional T cell.
The online version of this article contains supplemental material.