Characterization of Human CD39+ Th17 Cells with Suppressor Activity and Modulation in Inflammatory Bowel Disease

Induced regulatory T-cells (iT-reg) and T helper type 17 (Th17) in the mouse share common CD4 progenitor cells and exhibit overlapping phenotypic and functional features. Here, we show that human Th17 cells endowed with suppressor activity (supTh17) can be derived following exposure of iT-reg populations to Th17 polarizing conditions. In contrast to “pathogenic” Th17, supTh17 display immune suppressive function and express high levels of CD39, an ectonucleotidase that catalyzes the conversion of pro-inflammatory extracellular nucleotides ultimately generating nucleosides. Accordingly, supTh17 exhibit nucleoside triphosphate diphosphohydrolase activity, as demonstrated by the efficient generation of extracellular AMP, adenosine and other purine derivatives. In addition supTh17 cells are resistant to the effects of adenosine as result of the low expression of the A2A receptor and accelerated adenosine catalysis by adenosine deaminase (ADA). These supTh17 can be detected in the blood and in the lamina propria of healthy subjects. However, these supTh17 cells are diminished in patients with Crohn’s disease. In summary, we describe a human Th17 subpopulation with suppressor activity, which expresses high levels of CD39 and consequently produces extracellular adenosine. As these uniquely suppressive CD39+ Th17 cells are decreased in patients with inflammatory bowel disease, our findings might have implications for the development of novel anti-inflammatory therapeutic approaches in these and potentially other immune disorders.

In contrast, T helper type 17 lymphocytes (Th17) are an effector subset that develops independently of Th1 and Th2 cell lineages. Th17 cells drive inflammatory and autoimmune conditions in both mice and humans and have been linked to intestinal inflammation [15,16]. CD4 + T-cells can be differentiated into Th17 cells when exposed to TGF-b in combination with IL-6 or IL-21 in mice and to IL-6, TGF-b and IL-1b in humans, or into induced (i)T-reg under the influence of TGF-b [15,16]. Additional studies have shown that, in addition to TGF-b, other factors including IL-2 [17,18] and anti-CD3/anti-CD28 [19] play a role in iT-reg generation, even after a short stimulation period [19]. iT-reg and Th17 cells, however, may not be terminally differentiated and iTreg in particular show phenotypic and functional plasticity [20].
Using genetic lineage tracing of Foxp3 T-reg, Zhou and colleagues observed that a significant proportion of Foxp3 + cells undergo down-regulation and in some cases loss of Foxp3 expression is noted [21]. These 'ex-Foxp3' cells display an effector memory cell phenotype, produce pro-inflammatory cytokines and are numerically increased in experimental autoimmune diabetes [21]. Moreover, exposure of T-reg to IL-6 can down-regulate both Foxp3 and IL-17 expression, suggesting that T-reg may be 'subverted' to Th17-like cells [22]. In addition, it has been reported that T-reg can further acquire effector properties -i.e. IFNc production -when cultured in the presence of IL-12 [23]. These 'Th1-like' T-reg show diminished suppressive activity that can be only partially reversed by blockade of IFNc or IL-12 removal [23].
The stimulation of naïve T-cells with TGF-b and IL-6 triggers IL-17 production but it also induces the expression of IL-10, limiting the pathogenic potential of these cells [24]. Indeed, additional studies have reported that IL-17 + T-cells can limit tissue damage during inflammation [25,26]. In experimental murine tumor settings, it has been demonstrated that CD39 and CD73 expressed by 'suppressor' Th17 cells (supTh17) suppress tumorspecific immunity [27]. Whether comparable human supTh17 cells exist has been unexplored to date.
CD39 hydrolyses ATP and ADP into AMP, which is then converted into adenosine by CD73. The regulatory properties of CD39 were initially noted in studies conducted on CD39 null mice in which an enhanced production of IFNc, IL-1b, IL-6 and TNFa was found [28,29]. CD39 and CD73 expression on murine Treg is required for the suppressive function of these cells, which results from the production of adenosine [4]. Accordingly, T-reg isolated from CD39 null mice are unable to block allograft rejection in adoptive transfer studies [13].
Expression of CD39 has been reported on human T-reg in parallel to FOXP3 and low levels of CD127 [30,31]. Human Treg do not co-express high levels of CD73 with CD39 in contrast to murine counterparts. Thus AMP conversion to adenosine by human CD39 + T-reg is thought to result from paracrine mechanisms by the presence of CD73 on target or neighboring cells [31]. Regardless of the molecular mechanism involved, it has been shown that human CD39 + T-reg exert preferential suppression on CD4 target cell IL-17 production [32].
Defective numbers of CD39 + T-reg have been reported in patients with multiple sclerosis, autoimmune hepatitis [33] and CD39 polymorphisms linked to low-level CD39 expression have also been described in Crohn's disease [30,32,34]. Recent studies have shown that in addition to T-reg, CD39 is also expressed on a subset of memory cells with effector function [35]. Although this expression of CD39 by human T-reg has been reported and putative roles dissected, the demonstration and relevance of specific CD39 expression by human Th17 cells has been unexplored to date.
We describe here a population of human supTh17 cells that in contrast to prototypic pathogenic Th17 display high levels of both CD39 and FOXP3 and exhibit immune suppressive properties. Our new observations also provide mechanistic insights into the development of supTh17 and indicate the role of CD39 and purinergic immunomodulation. The pathophysiological relevance of these cells is supported by the detection of decreased frequencies of CD39 + supTh17 cells in both peripheral blood and lamina propria of patients with Crohn's disease, an illness characterized by manifestations of unfettered intestinal inflammation.

Subjects
Peripheral blood mononuclear cells (PBMCs) were isolated from platelet-depleted blood (leukofilters) obtained from 68 healthy blood donors (Blood Donor Center at Children's Hospital, Boston, MA). PBMCs were also obtained from 25 patients with Crohn's disease, recruited from the Gastroenterology Division, Beth Israel Deaconess Medical Center (BIDMC), Boston MA. Of these patients, 11 were studied during active disease (median Harvey Bradshaw Index, HBI: 8, range 2 to 25) while 14 were in remission (median HBI: 0, range 0-12). At the time of investigations, 11 patients were receiving infliximab, 2 were on steroids and 2 on immunomodulatory drugs.

Ethics Statement
The study was approved by BIDMC Institutional Review Committee. Written consent was obtained from all study participants.

Cell Purification
PBMCs were obtained by density gradient centrifugation on Ficoll-Paque (GE Healthcare, Uppsala, Sweden). Cell viability, determined by Trypan Blue exclusion, exceeded 98%. Lamina propria mononuclear cells (LPMCs) were isolated from freshly biopsied colonic tissue. The tissue was initially washed with PBS, cut into small sections and incubated in calcium and magnesiumfree HBSS containing 4 mM EDTA and 1 mM dithiothreitol at 37uC for 15 min. Epithelia were removed by discarding the supernatants. This procedure was repeated three times. The tissue was then minced, resuspended in RPMI 1640 containing 10% FCS, 400 U/ml collagenase D and 0.01 mg/ml DNase I, and then incubated at 37uC for 1.5 hour with pipetting every 30 min. The digested tissue was filtered and centrifuged at 6006g for 7 min. Collected cells were pelleted, resuspended in PBS 1% FCS and stained as indicated below.

In vitro Suppression Assay
The ability of Th17, iT-reg and supTh17 to control target cell proliferation and effector cytokine production was evaluated following 4-day co-culture with CD4 responder cells. Following 24 hour resting in cytokine and bead-free medium, Th17, iT-reg and supTh17 were added at 1/8 ratio to autologous CD4 target cells (2.5610 4 cells/well) previously exposed to IL-2 (30 U/ml) and T-cell expander (bead/cell ratio: 1:2) for 5 to 7 days. The 1:8 ratio was selected because capable of exerting a detectable regulatory function in preliminary experiments where ratios of 1:16, 1:8, 1:4 and 1:2 were compared as these putatively reflect pathophysiological proportions between suppressor and effector lymphocytes. Parallel cultures of CD4 responder cells and of Th17, iT-reg and supTh17 on their own were performed under identical conditions. All experiments were performed in duplicates. After 4 days, cultures were pulsed with 0.25 mCi/well 3 H- thymidine and harvested 18 hours later using a cell harvester (Tomtec, Hamden, CT). Incorporated thymidine was measured by liquid scintillation spectroscopy. In preliminary experiments, inhibition of CD4 target cell proliferation in the absence and presence of suppressor cells was also analyzed using carboxy fluorescein succinimidyl ester (CFSE) staining. As CFSE-and 3 Hthymidine-based assays gave comparable results, given the requirement for fewer cells, 3 H-thymidine was used to measure proliferation in subsequent experiments. The ability of Th17, iTreg and supTh17 cells to control the production of IFNc and IL-17 by target cells was determined by intracellular cytokine staining after 4-day co-culture as detailed above. The effect of adenosine on Th17, iT-reg and supTh17 ability to suppress was tested in parallel experiments.

Quantitative Real-time PCR
Expression of A1, A2A, A2B and A3 adenosine receptors, and of phosphodiesterases (PDE) 4A and PDE4B was determined by real-time PCR. Total RNA was extracted from 2-3610 5 cells using TRIzol reagent (Invitrogen) and mRNA was reverse transcribed using iScript cDNA Synthesis kit (Bio-Rad Laboratories, Hercules, CA) according to the manufacturer's instructions. Sequences of adenosine receptors were as previously described [41]. PDE primer sequences were as follows: PDE4A: Forward 59 ACACAGCAGTGACGCTAATCCAGA 39 Reverse 59 ATTCACTGGAGGAGGTGGCTCAAA 39 PDE4B: Forward 59 ACAGCCTGATGCTCAGGACATTCT 39 Reverse 59 AAACTTCTCCATCAGACCCTGGCA 39 PCR amplification conditions were as previously reported [41]. Samples were run on a Stratagene MX3005P (Agilent Technologies, Santa Clara, CA) and results were analyzed by matched software and expressed as relative quantification. Relative gene expression was determined by normalizing to human b-actin (primer sequence as previously reported [41]).
Immunoblot Analysis 5610 5 cells were lysed in ice-cold RIPA buffer, containing 1% NP-40, 0.25% sodium deoxycolate, 50 mM Tris-HCl and 150 mM NaCl and supplemented with Complete Proteinase Inhibitor Cocktails (Roche Diagnostics, Indianapolis, IN) and Phosphatase Inhibitor Cocktails (Sigma-Aldrich). Following 30 minutes incubation on ice, samples were spun at 14,0006g for 30 minutes. Supernatants (containing total cell lysates) were collected and total protein concentration determined using Bio-Rad Dc protein assay reagent (Bio-Rad Laboratories) using bovine serum albumin as standard. Following protein denaturation with SDS, cell lysates were separated on a 4-12% Criterion XT Bis-Tris SDS-PAGE (Bio-Rad Laboratories). 10 mg of protein were loaded per lane. Gels were run for 20 minutes at 80 V and then at 110 V for additional 80 minutes. Proteins were then transferred onto PVDF membranes (Immobilon-P, Millipore, Billerica, MA) by semi-dry electroblotting and subsequently incubated in blocking buffer containing 2.5% skimmed milk. Following 60 minutes, mouse anti-adenosine deaminase (ADA) antibody (Abcam, Cambridge, MA) was applied at 1 mg/ml. Following overnight incubation membranes were incubated for 60 minutes with HRP-labeled goat anti-mouse (Thermo Scientific, Rockford, IL) at 1/50,000. Bands were visualized using SuperSignal West Femto Maximum Sensitivity Substrate (Thermo Scientific) according to the manufacturer's instructions. For immunoblot normalization, the same membranes were stripped (using a buffer containing 15 g glycine, 1 g SDS and 10 ml Tween20) and re-probed with mouse anti-human b-actin (Abcam) at 1/20,000 and subsequently with a HRP-labeled goat anti-mouse polyclonal antibody at 1/20,000. ADA band density was determined using Scion Image Processing Program (Release Beta 4.0.2).

Statistical Analysis
Results are expressed as mean6SEM (obtained from at least 5 subjects per group and from at least 3 independent in vitro experiments). Smirnov goodness of fit test was performed to test the normality of variable distribution. Paired and unpaired Student's t test were used for comparing normally distributed data; Wilcoxon's rank sum test and Mann Whitney test were used for non-normally distributed data. ANOVA repeated measures or one-way ANOVA, followed by Tukey's multiple comparisons test, was used to compare the means of multiple samples. For all

supTh17 are Phenotypically Different from Prototypic Th17 Cells and Display Regulatory Function
In order to first investigate whether human Th17 cells can acquire regulatory functions per se, we activated CD4 + CD45RO + memory (CD4 mem ) and CD4 + CD45RA + naïve (CD4 naive ) T-cells under Th17 polarizing conditions. Next, we exposed these cells to iT-reg polarizing conditions. Finally, to evaluate the stability of the polarized T-cells, we re-activated them in the presence of Th17 skewing conditions, as detailed in Methods (see also Figure S1). In these studies, Th17 polarizing conditions consisted of 3-day exposure to IL-6, IL-1band TGF-b, a cytokine cocktail previously shown to result in the efficient differentiation of IL-17 producing cells in humans [36][37][38]44], and to low dose anti-CD3/anti-CD28. Further, iT-reg polarizing conditions consisted of 4-day stimulation in the presence of high concentration IL-2 and anti-CD3/ anti-CD28, shown to be particularly effective at inducing high numbers of effective iT-reg [39,40,44].
We found that iT-reg obtained from CD4 mem -derived Th17 cells had persistent and stable suppressor activity following ''reactivation'' in the setting of Th17 polarizing conditions (Figure 1). In contrast, iT-reg obtained from CD4 naive -derived Th17 cells, had lost most of their suppressive ability once re-activated in the presence of Th17 polarizing conditions (Figure 1). Therefore, we focused consequent studies on iT-reg derived from CD4 mem . Figures 2 and S2 illustrate the phenotype of CD4 mem cells at baseline; after 3-day exposure to Th17 polarizing conditions; after further 4-day stimulation in the presence of iT-reg polarizing conditions; and then after 3-day re-exposure to Th17 driving cytokines. CD4 mem cells at baseline contained low frequencies of IL-17-producing, CD25 + and FOXP3 + lymphocytes (Figures 2A  and S2A-C). Following 3-day exposure to IL-6, IL-1b and TGF-b, CD4 mem cells displayed higher numbers of IL-17-producing cells, while maintaining low frequencies of CD25 + and FOXP3 + lymphocytes (Figures 2A and S2A-C).
Cells obtained following Th17 exposure to iT-reg polarizing conditions displayed a decrease in the number of IL-17 + lymphocytes and an increase in the frequency of CD25 + and FOXP3 + cells (Figures 2A and S2A-C). These cells contained minimal proportions of effector cytokines like IFNc or IL-2 ( Figure  S3). After iT-reg exposure to Th17 polarizing conditions, we noted marked increases in the number of cells producing IL-17, decreases in lymphocytes positive for CD25 and frequencies of FOXP3 + lymphocytes that were similar to iT-reg although higher than Th17 cells (Figures 2A and S2A-C). When contrasted to prototypic Th17, these supTh17 cells displayed higher expression of RORC, higher numbers of IL-22 + lymphocytes and similar proportions of cells positive for CCR6 and IL-23 receptor (IL-23R) ( Figure 2B). When next considering suppressive functions ( Figures 3A and  S4A), we observed that supTh17 controlled CD4 target cell proliferation in a comparable manner to iT-reg, and more effectively than did prototypic Th17 cells. With regard to suppression of pro-inflammatory cytokine production (Figures 3B and S4B-C), supTh17 effectively controlled IL-17 and IFNc cytokine production by CD4 effector cells. In contrast, iT-reg, while effectively inhibiting production of IL-17, exerted only weak control over CD4 T-cell IFNc production.
In summary, supTh17 can be obtained following exposure of CD4 mem -derived iT-reg to Th17 polarizing conditions. In contrast to prototypic Th17, these cells contain higher frequencies of IL-17 In a manner distinct from iT-reg which are anergic, however, supTh17 express low levels of A2A receptor and exhibit nucleoside scavenging ecto-enzymatic activity. These properties confer on supTh17 an important intrinsic resistance to suppressive effects of adenosine, which may develop in parallel with prolonged cellular activation in accordance with memory T-cell status. These differences suggest that supTh17 might undergo conversion and be recruited as suppressor-type cells in the later evolution of immune responses where these cells may persist at sites of resolving injury. doi:10.1371/journal.pone.0087956.g008 producing and FOXP3 + lymphocytes and furthermore display effective and stable suppressive function.

The supTh17 Cells Express both CD39 and CD73 thereby Generating Adenosine and other Nucleoside Derivatives
Given the regulatory properties displayed by supTh17 and the association between CD39 and immunoregulation [13,32], we determined the expression of CD39 in supTh17 and compared it with that of CD4 mem cells at baseline, Th17 and iT-reg. As shown in Figure 4A and 4B, supTh17 contained the highest frequencies of CD39 + cells and displayed the highest CD39 MFI, being therefore clearly distinguishable from prototypic Th17 cells that displayed low numbers of CD39 + lymphocytes and low CD39 MFI.
To evaluate whether different Th17 polarizing conditions influence CD39 expression, we obtained Th17 and supTh17 cells upon exposure to IL-6, IL-1b and IL-23 or IL-6, IL-1b, IL-23 and TGF-b. As depicted in Figure S5A, no differences were observed in the frequency of CD39 + cells in the presence of different Th17inducing cytokine cocktails.
We next evaluated the phenotypic properties of CD39 + cells within supTh17 and compared with those of CD39 + cells within CD4 mem at baseline, Th17 and iT-reg ( Figures 4C, S5B and S6A-C). supTh17 cells contained proportions of cells positive for CD73 -the ectonucleotidase working in tandem with CD39 to generate adenosine -and for FOXP3 comparable to iT-reg and higher than Th17 cells and CD4 mem cells at baseline (Figures 4C, S5B and S6A). No significant differences in the frequencies of IL-10 + and RORC + cells were noted between supTh17 and the other cell subsets ( Figure S6B-C).
Given the concomitant expression of CD39 and CD73 by both supTh17 and iT-reg, we determined the ability of these cells to generate adenosine. Cell ectoenzymatic activity was assessed by thin layer chromatography (TLC) following cell incubation with [C 14 ] radiolabeled ADP. As depicted in Figure 4D, supTh17 and iT-reg were both able to generate adenosine that supTh17 cells further effectively degraded into inosine. In contrast, Th17 cells were capable of hydrolyzing ADP into AMP but did not generate extracellular adenosine, in accordance with low levels of CD39 and CD73 expression. In keeping with concomitant CD39 and CD73 expression, supTh17 are therefore competent in generating adenosine, which is then effectively degraded into inosine.

Effects of Adenosinergic Signaling on Cell Phenotype and Function
We then tested the effect of adenosine exposure on supTh17 and compared with that in iT-reg and Th17 cells. Adenosine increased the frequency of CD39 + and CD73 + cells among iT-reg while not having any effect on the frequency of these cells among Th17 and supTh17 cells ( Figures 5A and S7 and data not shown). Exposure to exogenous adenosine did not affect the proportion of FOXP3 + and IL-17 + cells in any of the three cell subsets ( Figure 5A). With regard to suppressive function, adenosine enhanced the ability of iT-reg and, though to a lesser extent, Th17 cells to control CD4 target cell proliferation while not having any effect on the suppression exerted by supTh17 ( Figure 5B). The above data show that adenosine boosts the phenotypic and functional properties of iT-reg while not having any effect on supTh17.
We then examined possible mechanisms that could further account for resistance of supTh17 to exogenous adenosine. We considered that resistance to adenosine may result from low expression levels of adenosine receptors, from high levels of adenosine deaminase (ADA), which degrades adenosine into inosine, and/or high expression of phosphodiesterases (PDE)the enzymes degrading the phosphodiester bond of cAMP. Thus, we determined the expression of A1, A2A, A2B and A3 adenosine receptors by quantitative real-time PCR. The expression of A2A receptor, known to be involved in down-regulation of inflammation and protection from tissue damage [45], was decreased at mRNA levels in supTh17, when compared to Th17 and iT-reg ( Figure 6A).
In order to test whether adenosine resistance of supTh17 was the result of enhanced adenosine clearance, we first assessed the expression of ADA. We observed expression of ADA in Th17 and supTh17 ( Figure 6B), indicating that both these cell types have the ability to deaminate adenosine. In contrast, ADA was only weakly expressed in iT-reg ( Figure 6B). ADA is completely functional at the cell surface (known as ecto-ADA), where it directly interacts with the dipeptidylpeptidase IV (CD26) and regulates adenosine receptors. ADA activity depends on CD26, the expression of which has been recently reported to be increased on human Th17 [46]. We therefore assessed the expression of CD26 and found that the CD26 MFI was higher in iT-reg and supTh17 compared to Th17 ( Figure 6C). These data indicate that the effective degradation of adenosine into inosine displayed by supTh17 relies on the co-expression of both ADA and CD26. In contrast Th17 and iT-reg, which express either ADA (Th17) or CD26 (iT-reg), do not display effective deamination activity ( Figure 6C).
We next determined the expression of PDE4A and PDE4B. We found that both enzymes are expressed by Th17, iT-reg and supTh17 ( Figure 6D). The supTh17, however, did not overexpress any of the PDE, ruling out the possibility that the adenosine resistance noted in these cells result from high levels of cAMP clearance. We therefore conclude that supTh17 resistance to exogenous adenosine is associated with low A2A adenosine receptor expression and enhanced scavenging of nucleosides by ecto-adenosine deaminase.

Demonstration of supTh17 in Healthy Subjects and in Patients with Crohn's Disease
To investigate the biological relevance of supTh17, the frequency of CD4 + IL-17 + and that of supTh17 was determined in PBMCs and LPMCs obtained from healthy subjects and Crohn's patients. These supTh17 were identified by initially gating CD4 + CD45RO + cells within PBMCs or LPMCs and then by determining the proportion of CD39 + IL-17 + and FOXP3 + within this population.
While the proportion of CD4 + IL-17 + in PBMCs cells was similar in the two groups, that of CD4 + IL-17 + lymphocytes obtained from the lamina propria was higher in Crohn's patients than in healthy subjects ( Figure 7A). In patients, the frequency of CD4 + IL-17 + cells was markedly higher in the lamina propria compared to the circulation ( Figure 7A). We then determined the frequency of supTh17 in both compartments. These supTh17 were decreased in Crohn's patients, when compared to healthy subjects, both within PBMC and LPMC populations ( Figures 7B  and S8). In both groups, supTh17 were increased in the lamina propria compared to the circulation (Figures 7B and S8). When analyzed for expression of Stat-3 known to modulate Th17 immunosuppressive activity through up-regulation of CD39 [27]circulating supTh17 from Crohn's patients displayed higher proportion of cells positive for this marker than did comparably prepared cells from healthy subjects ( Figure 7C) [27]. In both groups, supTh17 from the lamina propria contained higher proportions of lymphocytes that were positive for Stat-3 ( Figure 7C). Analysis of cytokine profiles show that supTh17 from Crohn's disease patients had higher frequencies of TNF-a + and IL-2 + cells -pro-inflammatory cytokines previously reported to be decreased in populations of Th17 cells with suppressive properties isolated from the small intestine [47], than the respective supTh17 from healthy subjects in the circulation (albeit not in the lamina propria; Figure 7D). In both groups, supTh17 in the lamina propria contained higher proportions of TNF-a + and IL-2 + cells than did the counterparts in the circulation ( Figure 7D). No differences in the frequencies of supTh17 and of TNF-a + and IL-2 + cells within them were noted between patients with either active or inactive disease. The above data indicate that supTh17 are highly represented in the lamina propria and that the frequencies of these cells are lower in the periphery in Crohn's disease, where in contrast these cells also express increased levels of proinflammatory cytokines.

Discussion
We have shown that a population of human supTh17 cells can be derived following the exposure of iT-reg to Th17 polarizing conditions in vitro. These putative supTh17 display phenotypic features of both effector Th17 (i.e. production of IL-17 and expression of CCR6, IL-22 and IL-23R) and iT-reg (expression of FOXP3) and importantly control effector cell function by inhibiting CD4 cell proliferation as well as production of IFNc and IL-17. It is not clear whether these cells might be representative of a late stage in Th17 differentiation, in which the effector potential of prototypic Th17 cells is attenuated or rather constitute unique cell subsets in which overlapping regulatory and effector features coexist.
The robust in vitro system used in the present study enabled us to observe changes in T-cell phenotype and function upon stimulation in the presence of Th17 and iT-reg polarizing conditions. Previous studies have documented differentiation of CD4 cells into Th17 or iT-reg following short-and medium-term cell culture in vitro [17,18,48,49]. These and our studies may be of particular relevance to disease settings, in which antigen-primed CD4 memory cells may be sequentially exposed to different cytokine milieus and undergo modulation accordingly, during either inflammatory or remission phases.
Given the putative importance of CD39 in immunoregulation, particularly concerning purinergic mechanisms governing the suppressive function of iT-reg, we studied whether supTh17 expressed this ectonucleotidase. Our data show that, in contrast to prototypic Th17 cells, supTh17 display high levels of CD39. Furthermore, supTh17 cells also co-express ecto-59-ectonucleotidase CD73, which is pivotal in the generation of adenosine from AMP. These supTh17, in contrast to prototypic Th17, have the potential to generate adenosine in a manner comparable to iT-reg, which can be noted by standard biochemical tests ( Figure 4D). However, in a manner distinct from iT-reg, the extracellular adenosine that is generated by supTh17 undergoes further degradation, given the concomitant expression of adenosine deaminase and CD26 by these cells. In accordance with the low CD73 levels expressed, prototypic Th17 cells were unable to generate adenosine.
When we examined the effect of exogenous adenosine on supTh17 phenotypic and functional properties, we could observe that these cells were resistant to the effect of this mediator. Curiously, these cells did not undergo upregulation of CD39 expression nor did these cells exhibit amelioration of suppressive function, in the manner observed in anergic type iT-reg.
Adenosine resistance in supTh17 cells is likely to be conferred by low levels of A2A adenosine receptor and by higher levels of adenosine catalysis. The A2A adenosine receptor is primarily known to mediate anti-inflammatory effects: lymphocytes from A2A receptor (2/2) mice show higher rates of cell proliferation and produce high IFNc levels upon stimulation [50]. A2A receptor stimulation has established inhibitory effects on Th1 and Th17 effector cell generation and, in contrast, favors generation of FOXP3 + and LAG-3 + regulatory T-cells [51]. Our data suggest that the most likely mechanisms for supTh17 resistance to adenosine are linked to low A2A receptor levels and enhanced levels of adenosine catalysis, enabled by ADA and CD26 co-expression. We have observed that iT-reg display marked decreases in mRNA levels of the A2B adenosine receptor. This observation might have relevance for the differential effects of A2B versus A2A signaling by these cells. Furthermore, Moriyama and Sitkovsky have demonstrated in studies of A2AR versus A2BR expression in transfected cells that substantive proportions of A2BR are preferentially degraded by the proteasome, a mechanism that might be also operative here in differentiating Th17 cells from iT-reg [52].
The observation that supTh17 are resistant to adenosinergic modulation implies that these cells are not conventional suppressors, nor are these cells anergic. The supTh17 cells might adapt their own intrinsic ability to regulate or inflict damage according to the immunological context in which they operate. Given that in these cells, both regulatory (i.e. adenosine generation, suppressive function) and pro-inflammatory (i.e. low levels of A2A and inosine generation) features co-exist, it is plausible to suggest that supTh17 might reside in a form of 'purinergic limbo' and unresponsiveness until a crucial time where this balance may be perturbed. This temporal 'status' would enable supTh17 cells to influence extrinsic homeostatic properties of target cells via suppression mediated through generation of adenosine while maintaining intrinsic resistance to this immune suppressive molecule. In contrast, iTreg exert suppression via production of adenosine, while being also wholly susceptible to the nucleoside modulatory effects that might stabilize their immune suppressive phenotype (Figure 8; [4,31]). At variance with iT-reg, supTh17 are not subjected to this autocrine loop, suggesting that these cells, in contrast, may play roles as both regulators of late stage immune responses and in the maintenance of T-cell memory functionality.
Another important finding of this investigation is that supTh17 could be enumerated in both circulation and lamina propria of healthy subjects and patients with Crohn's disease. These cells appear to preferentially home to the intestine, as demonstrated by their high percentages in the lamina propria, suggesting that the intestine may be the compartment where Th17 cells undergo regulation. Our data indicate there are higher percentages of supTh17 cells expressing Stat-3 in the lamina propria indicating that this transcription factor may have a role in the expression of CD39 and induction of supTh17 in the colon [27]. In agreement with these data, in mouse models of colitis, pathogenic Th17 cells are also considered to undergo regulation in the intestine where these cells acquire phenotypic and functional T-reg-like properties [47].
Importantly, supTh17 numbers are markedly decreased in Crohn's patients. This decrease might theoretically result in disease exacerbation and perpetuation because of the decreased ability of effector Th17 to undergo regulation. Previous clinical studies demonstrated impaired immunoregulation and particularly numerically defective and dysfunctional T-reg in these same disease settings [53,54].
Hence, the increased numbers of effector Th17 cells, also shown here may originate from defective control usually operated by active immune suppression -i.e. primarily defective T-reg, or alternatively be the result of decreased Th17 autoregulation. Interestingly, supTh17 from Crohn's patients appear skewed towards a pro-inflammatory phenotype as these cells contain higher frequencies of TNF-a and IL-2 pro-inflammatory cytokines than those noted in healthy controls.
In conclusion, we have shown that human supTh17 can be obtained upon exposure of iT-reg to Th17 driving conditions in vitro.
High levels of CD39 expression distinguish these immune suppressive cells from effector pathogenic Th17 cells. We propose that these fundamental alterations in purinergic signaling might control tissue damage while limiting cellular pathogenicity in local and systemic inflammatory illnesses, such as in Crohn's disease. Promoting the local expansion of supTh17 cells and the maintenance of these should boost local immune suppressive activities and augment diminished T-reg functionality, as previously noted in IBD [34,55]. Indeed, these studies and development of modalities to boost CD39 expression have implications for the development of novel therapeutic strategies in Crohn's disease. Figure S1 Experimental protocol for T-cell activation. CD4 mem and CD4 naive T-cells, purified as CD4 + CD45RO + and CD4 + CD45RA + cells, were initially activated under Th17 polarizing conditions. This comprised of IL-6+IL-1b+rTGF-b+ anti-CD3/anti-CD28 T-cell expander (bead/cell ratio: 1:50) for 3 days. Cells were then exposed to iT-reg skewing conditions with high concentration IL-2 and anti-CD3/anti-CD28 T-cell expander (bead/cell ratio: 1:2) for 4 days, and then were re-activated under Th17 polarizing conditions for 3 days.