The incidence of allergy and asthma in developed countries is on the increase and this trend looks likely to continue. CD4+ T helper 2 (Th2) cells are major drivers of these diseases and their commitment is controlled by cytokines such as interleukin 4, which are in turn regulated by the suppressor of cytokine signaling (SOCS) proteins. We report that SOCS2−/− CD4+ T cells show markedly enhanced Th2 differentiation. SOCS2−/− mice, as well as RAG-1−/− mice transferred with SOCS2−/− CD4+ T cells, exhibit elevated type 2 responses after helminth antigen challenge. Moreover, in in vivo models of atopic dermatitis and allergen-induced airway inflammation, SOCS2−/− mice show significantly elevated IgE, eosinophilia, type 2 responses, and inflammatory pathology relative to wild-type mice. Finally, after T cell activation, markedly enhanced STAT6 and STAT5 phosphorylation is observed in SOCS2−/− T cells, whereas STAT3 phosphorylation is blunted. Thus, we provide the first evidence that SOCS2 plays an important role in regulating Th2 cell expansion and development of the type 2 allergic responses.
Allergic diseases such as atopic dermatitis (AD) and asthma are orchestrated and enhanced by CD4+ Th2 lymphocytes and the cytokines they produce, namely IL-4, IL-5, and IL-13. These type two cytokines drive elevated serum IgE and infiltration of Th2 cells, eosinophils, and basophils, which have been observed in patients with asthma and AD (Venge et al., 1996; Sudowe et al., 2000; Walter et al., 2001; Nakajima and Takatsu, 2007). IL-4 and IL-13 mediate their responses by activating JAK kinases and STAT6, a transcription factor responsible for activating many allergy-associated genes (Kato and Schleimer, 2007).
The suppressor of cytokine signaling (SOCS) family of proteins is a key controller of cytokine responses, which can down-regulate specific cytokine signals and consequently modify the immune response (Alexander, 2002; Croker et al., 2008; Dimitriou et al., 2008). SOCS1 and SOCS3 have been shown to influence the Th1/Th2 balance (Kubo and Inoue, 2006; Yoshimura et al., 2007; Palmer and Restifo, 2009). In particular, SOCS3 expression correlates with the severity of asthma and AD, as well as serum IgE levels in patients with allergy (Seki et al., 2003). Although SOCS2 is expressed in activated CD4+ T cells and its expression can regulate IL-2 and IL-3 signals (Tannahill et al., 2005), no role for SOCS2 in CD4+ T cell differentiation has yet been defined.
In this paper, we demonstrate that activated SOCS2−/− CD4+ T cells show significantly higher IL-4, IL-5, and IL-13 production and Th2 differentiation and that they had elevated SOCS3- and IL-4–induced STAT6 but blunted IL-6–induced STAT3 phosphorylation. Moreover, SOCS2−/− mice were highly susceptible to AD with a marked eosinophilic infiltrate and, after helminth antigen challenge, produced significantly more type 2 cytokines relative to WT animals. In addition, OVA-sensitized SOCS2−/− mice had significantly increased airway resistance, specific IgE, and airway inflammation after OVA challenge. Our data suggest a novel role for SOCS2 in controlling CD4+ lineage commitment and type 2 allergic responses.
SOCS2 deficiency is associated with enhanced IL-4, IL-5, and IL-13 secretion and Th2 differentiation both in vitro and in vivo
SOCS2−/− mice present with gigantism but with no reported obvious immune defect (Metcalf et al., 2000). Phenotyping of thymus, spleen, and LN of SOCS2−/− and WT mice showed similar frequency and numbers of CD4+, CD8+, double-positive, and double-negative T cells (Fig. S1, A and B), suggesting that T cell development was normal. However, as other SOCS family members have important roles in determining the cytokine responsiveness of different Th subsets (Palmer and Restifo, 2009), we examined the cytokine profile of CD4+ T cells from SOCS2−/− mice after TCR stimulation in vitro. IL-2 and IFN-γ secretion were similar in CD4+ T cells from SOCS2−/− and WT mice but, surprisingly, IL-4, IL-5, IL-10, and IL-13 levels were markedly higher in the absence of SOCS2 (Fig. 1 A). Moreover mRNA levels of these cytokines were also enhanced when examined by real-time PCR (Fig. S1 C). Importantly the level of GATA-3 mRNA was significantly higher in SOCS2−/− CD4+ T cells (Fig. S1 C), suggesting spontaneous enhancement of Th2 polarization. To explore this more closely, we sorted naive CD4+CD25−CD44low T cells and drove de novo Th2 differentiation in vitro. As shown in Fig. 1 B and Fig. S1 D, in naive CD4+ cells the absence of SOCS2 resulted in a significantly higher frequency of Th2 cells (CD4+ IL-4+ cells; P < 0.001), thus indicating that SOCS2 must control polarization toward the Th2 lineage.
Because parasitic helminth Schistosoma mansoni eggs induce a potent Th2 response (Fallon and Mangan, 2007), we used this model to compare Th2 responses in vivo in SOCS2−/− mice relative to WT animals. SOCS2−/− and WT mice were injected intraperitoneally with 10,000 S. mansoni eggs and, 10 d later, splenocytes were restimulated with anti-CD3/CD28 or with schistosome soluble egg antigen (SEA). Both approaches induced marked type 2 responses in WT mice relative to PBS-challenged mice (Fig. 1 C). Moreover, there was significantly (P < 0.05) greater production of the type 2 cytokines IL-4, IL-5, and IL-13 by splenic T cells from egg-injected SOCS2−/− mice relative to WT mice after both nonspecific and antigen-specific TCR stimulation (Fig. 1 C). Interestingly, splenocytes from egg-injected SOCS2−/− mice also produced significantly lower amounts of IL-17 compared with WT mice (Fig. 1 C). To specifically enumerate the in vivo expansion of Th2 cells, S. mansoni eggs were injected into the footpads of SOCS2−/− and WT mice. After 10 d, the frequency of Th2 (CD4+ IL-4+) cells in the popliteal LN was significantly higher (P < 0.001) in SOCS2−/− mice than in WT animals (Fig. 1 D). To confirm that SOCS2 deficiency in CD4+ T cells was directly responsible for the observed Th2 bias, naive SOCS2−/− or WT CD4+ T cells were transferred into RAG-1−/− recipients and challenged with S. mansoni eggs. 10 d after injection, the number of Th2 (CD4+ T1/ST2+) cells in popliteal LN was significantly (P < 0.05) higher in mice reconstituted with SOCS2-deficient T cells when compared with those reconstituted with WT cells (Fig. 1 E), showing a requirement for SOCS2 expression in CD4+ T cells in controlling in vivo differentiation of Th2 cells. These observations indicate that SOCS2 deficiency unexpectedly favors Th2 polarization in vitro and in vivo and, therefore, that SOCS2 is a key regulator of Th2 differentiation.
SOCS2−/− mice are more susceptible to AD
Atopic disorders are driven by an elevated Th2 response; therefore, we used an in vivo model of AD, which was previously described by Li et al., (2006), to assess whether SOCS2−/− mice are more susceptible to atopy. Mice were treated daily with an application of the vitamin D3 analogue calcipotriol (MC903). Topical application of MC903 triggered an AD-like response—increase in skin thickness—in all mice, but SOCS2−/− mice exhibited earlier onset and more severe disease (Fig. 2 A). Histological examination showed red scaly lesioned skin, accompanied by epidermal hyperplasia with dermal lymphocyte infiltration that was more abundant in SOCS2−/− mice (Fig. 2 B). Furthermore, the dermal inflammation was associated with a marked eosinophilic infiltrate that was significantly exaggerated in the SOCS2−/− mice (P < 0.05; Fig. 2 C). Finally, analysis of cytokine secretion in restimulated auricular LN and spleen cells showed enhanced IL-4, IL-5, and IL-13, but not IFN-γ, levels from MC903-treated SOCS2−/− compared with treated WT animals (P < 0.05; Fig. 2 D). All together, the data demonstrates that SOCS2−/− mice were more susceptible to AD in this model, confirming an inhibitory effect of SOCS2 on the allergic response in vivo.
SOCS2−/− mice have increased susceptibility to allergic airway inflammation
Given the marked predisposition of SOCS2−/− mice to AD, we investigated whether SOCS2 deficiency resulted in an increased response to OVA-induced allergic airway inflammation. SOCS2−/− and WT mice were OVA challenged systemically and by aerosol to evoke airway inflammation and airway hyper-responsiveness (AHR). In the absence of allergen challenge, AHR, determined by lung resistance (RL), was not induced in PBS-treated SOCS2−/− mice, with these mice having RL values comparable to those of PBS-treated WT mice in response to methacholine challenge (Fig. 3 A). In comparison, OVA-sensitized SOCS2−/− mice had a striking dose-dependent increase in RL, relative to OVA-sensitized WT animals, that was statistically significant (P < 0.01) at higher methacholine doses (Fig. 3 A). Histological analysis of lungs of OVA-sensitized mice showed more marked inflammation in SOCS2−/− mice relative to WT, with peribronchial infiltrating eosinophils and goblet cell hyperplasia (Fig. 3 B). Consistent with a greater allergen-specific type 2 response, OVA-sensitized SOCS2−/− mice had significantly enhanced levels of both total IgE (P < 0.05) and OVA-specific IgE (P < 0.001) compared with OVA-treated WT animals (Fig. 3 C). Furthermore, cells from the lung-draining mediastinal LN from OVA-sensitized SOCS2−/− mice had significantly (P < 0.05) greater production of type 2, but not type 1, cytokines relative to allergen-sensitized WT animals (Fig. 3 D). Again, we observed a significant reduction (P < 0.05) in IL-17 production in response to antigen challenge in T cells from SOCS2−/− mice (Fig. 3 D). To formally determine that the enhanced allergen-induced lung inflammation in SOCS2-deficient mice (Fig. 3, A–D) was attributed to specific alterations in CD4+ cell function, we generated SOCS2−/−xOT/II mice. RAG-1−/− mice were reconstituted either with CD4+ T cells from SOCS2−/−xOT/II or OT/II mice. Recipients were then sensitized and challenged as previously described with OVA and AHR evaluated. RAG-1−/− animals that received the SOCS2−/−xOT/II cells had significantly higher RL responses than RAG-1−/− mice reconstituted with OT/II CD4+ T cells (Fig. 3 E). Moreover, this was associated with enhanced peribronchial cell infiltration and goblet cell hyperplasia in recipients of CD4+ from SOCS2−/−xOT/II mice relative to cells from OT/II mice (Fig. S2 A) and increased bronchoalveolar cell infiltration (Fig. S2 B). Recipients of CD4+ from SOCS2−/−xOT/II mice had significantly increased secretion of Th2-associated cytokines, lower IL-17, and unaffected IFN-γ both in the lungs (Fig. 3 F) and in BAL fluid (Fig. S2 C). Collectively, the data demonstrate that SOCS2−/− mice are more susceptible to OVA-induced airway inflammation and this can be a direct result of increased Th2 phenotype in SOCS2-deficient CD4+ T cells.
Enhanced STAT5 and STAT6 phosphorylation and SOCS1 and SOCS3 expression after T cell activation in the absence of SOCS2
In the different allergic models used in this study, SOCS2−/− mice consistently had enhanced IL-4, IL-5, and IL-13 and reduced IL-17 expression. It was important, therefore, to determine which cytokines involved in T helper cell polarization actually induced SOCS2 expression. Consequently, we next investigated which cytokine signals induced SOCS2 in purified WT CD4+ T cells. Both IL-4 and IL-6 strongly induced SOCS2 with expression being clearly up-regulated after 48 h (Fig. 4 A). Moreover, IL-4 was the most potent inducer of SOCS2, suggesting a possible role for SOCS2 in limiting Th2 differentiation. Interestingly, TGF-β induced no detectable SOCS2 expression and IL-12 induced limited amounts in CD4+ T cells (Fig. 4 A). SOCSs are known to inhibit STAT activation, and STATs, in turn, play a key role in the differentiation and commitment of helper T cell subsets. STAT6 is associated with Th2 development, whereas STAT4 and STAT3 reciprocally control Th1 and Th17 development, respectively (O’Shea et al., 2011). To examine the mechanism of preferential Th2 differentiation in SOCS2−/− mice, we investigated the profile of STAT activation in T cells. SOCS2−/− CD4+ T cells displayed constitutive and significantly enhanced STAT6 phosphorylation compared with WT CD4+ T cells (Fig. 4 B and Fig. S3 A), suggesting that the SOCS2-deficient CD4+ T cells are hyper-responsive to IL-4 in vivo. Moreover, IL-2–induced STAT5 activation is thought to promote early Th2 cell differentiation (Zhu et al., 2003; Cote-Sierra et al., 2004; Paul and Zhu, 2010) and we found that IL-2–induced STAT5 phosphorylation is enhanced in SOCS2−/− CD4+ T cells. (Fig. 4 C and Fig. S3 B). Importantly, IL-2-mediated STAT5 activation was shown to induce and maintain IL-4Rα expression (Liao et al., 2008). However, we found that IL-4Rα expression was unchanged (Fig. S3 D). Finally, although IL-12–mediated STAT4 activation was unaffected by the absence of SOCS2 (Fig. 4 D), we observed consistent blunting of IL-6–induced STAT3 phosphorylation (Fig. 4 E and Fig. S3 C). Interestingly, SOCS2 expression was induced after IL-6 and TGF-β stimulation (Fig. S3 E) and, when stimulated under Th17 polarization conditions, SOCS2-deficient T cells express less RORγt (Fig. S3 F) and secrete less IL-17 (Fig. S3 G). These results are consistent with the decreased IL-17 levels observed in vivo (Fig. 1 C and Fig. 3 D), including in the CD4+ cell reconstitution experiments (Fig. 3 F and Fig. S2 C), and they suggest that SOCS2 is required for normal Th17 differentiation. STAT5 is known to inhibit Th17 differentiation (Laurence et al., 2007; Yang et al., 2011) and, therefore, our data suggest that the reduced STAT3 activation and increased IL-2–mediated STAT5 phosphorylation in the SOCS2−/− T cells primes Th2 development, probably at the expense of Th17.
Interestingly, SOCS3 is a strong inhibitor of STAT3 signaling (Elliott and Johnston, 2004) and constitutive expression of SOCS3 in T cells was shown to favor Th2 polarization (Seki et al., 2003). Moreover, several studies suggest that SOCS2 can control the expression of both SOCS1 and SOCS3 (Tannahill et al., 2005; Piessevaux et al., 2006). We therefore hypothesized that SOCS1 and SOCS3 levels may be altered in SOCS2-deficient T cells. When splenocytes from SOCS2−/− mice were stimulated with anti-CD3/CD28 and IL-2 for periods of up to 24 h, we consistently observed significantly elevated SOCS1 and SOCS3 expression (Fig. S4 A). Interestingly, in WT cells, SOCS2 expression after T cell activation correlated with loss of detectable SOCS3 protein (Fig. S4 A). To assess whether SOCS3 levels were elevated in Th2 cells from SOCS2−/− mice, we purified and cultured CD4+ T cells for 7 d under Th2 conditions. Basal levels of SOCS3 were higher in SOCS2-deficient T cells relative to controls (Fig. S4 B), with the difference between WT and SOCS2−/− cells being amplified under polarizing conditions (Fig. S4 C). Finally, we also found that SOCS1 and SOCS3 mRNA levels are elevated in SOCS2-deficient CD4+ T cells stimulated for 3 and 5 d with anti-CD3/CD28 antibodies (Fig. S4 D). Because SOCS1 and SOCS3 are important regulators of Th differentiation (Palmer and Restifo, 2009), the elevated SOCS3 expression in SOCS2−/− CD4+ T cells would be consistent with the observation that SOCS3 blocks Th17 development by inhibiting IL-6 and IL-23 responses (Chen et al., 2006) while favoring Th2 differentiation (Seki et al., 2003). It is therefore plausible that the increased levels of SOCS1 and SOCS3 may also affect polarization toward other subsets, indirectly favoring Th2 differentiation.
Collectively, our findings show that SOCS2−/− mice are predisposed to AD and allergic lung inflammation and developed enhanced type 2 immune responses to helminth antigen. Because the SOCS2 mice are ubiquitous knockouts, other immune cells also lacking SOCS2 could contribute to this phenotype. Of note, we made the observation that the SOCS2-deficient B cells do not spontaneously class-switch toward IgE (unpublished data), which could have explained the predisposition of the SOCS2−/− mice for atopy. Of course, this should be further investigated, together with the effect of SOCS2 deficiency on APC, eosinophils, and mast cells in particular. However, as RAG-1−/− mice reconstituted with SOCS2-deficient CD4+ T cells developed enhanced pathology compared with those reconstituted with WT (Fig. 1 E; and Fig. 3, E and F), these findings suggest that SOCS2 exerts an intrinsic inhibition on CD4+ T cells, which prevents polarization toward Th2 both in vitro and in vivo. Interestingly, and consistent with the normal IFN-γ production we observe, SOCS2−/− mice mount a normal Th1 response to L. major infection (Bullen et al., 2003).
SOCS2 has primarily been studied in relation to the regulation of growth hormone and cytokine signaling and has not previously been thought to influence T cell lineage commitment. CD4+ T cell lineage maturation is governed by the expression of master regulator transcription factors that drive differentiation, but recent evidence suggests that their expression is not as stable as initially thought; indeed, there are now many examples of plasticity in Th cell subsets (Murphy and Stockinger, 2010). T cell plasticity and lineage fate are also governed by the cytokines they encounter after antigen recognition. Therefore it is plausible that the SOCS proteins might play a significant role in Th cell polarization. Indeed, constitutive expression of SOCS3 during CD4+ T cell differentiation facilitates Th2 expansion (Seki et al., 2003; Ozaki et al., 2005), whereas selective deletion of SOCS3 in T cells facilitated enhanced STAT3 activation and elevated IL-17 production (Chen et al., 2006). Moreover, recent studies have shown that SOCS3 may play a role in controlling regulatory T cell responses because SOCS3-deficient T cells demonstrate a Th3-like phenotype with increased TGF-β and IL-10 and increased STAT3 activation (Kinjyo et al., 2006; Taleb et al., 2009).
In conclusion, therefore, we propose that SOCS2 inhibits Th2 development, and that the enhanced STAT5 and STAT6 phosphorylation in the absence of SOCS2 favors Th2 polarization. Finally, we have defined new functions for SOCS2 in the regulation of allergic inflammation and, potentially, CD4+ T cell differentiation, providing new insights on how the influence of SOCS proteins on T cell plasticity may be exploited for the development of therapies for atopic disorders.
MATERIALS AND METHODS
All experiments used 12–16-wk-old sex- and age-matched C57BL/6 and SOCS2−/− mice. SOCS2−/− mice were a gift from D.J. Hilton (Royal Melbourne Hospital, Victoria, Australia) and were generated on a C57BL/6 background as previously described (Metcalf et al., 2000). C57BL/6 mice were either purchased from Harlan Laboratories or were WT littermates. RAG-1−/− and OT/II strain mice on a C57BL/6 background were purchased from The Jackson Laboratory. OT/II mice were subsequently crossed to SOCS2−/− mice in house to generate SOCS2−/−xOT/II mice. Mice were housed under specific pathogen-free conditions. All animal experiments were performed in compliance with either the UK Home Office or the Irish Department of Health and Children regulations and approved either by the Queen’s University Ethical Review Committee or by the Trinity College Bioresources Ethical Review Board.
T cell culture, purification, and differentiation.
All in vitro assays were performed by using between two and three mice per group. After extraction, splenocytes and lymphocytes were depleted of erythrocytes by lysis with ammonium chloride solution. Cells were cultured at 106/ml in complete RPMI medium. Cells were activated with 2 µg/ml (eBioscience) or 0.5 µg/ml (BD) of plate-bound anti-CD3 (145-2C11) and 1 µg/ml (eBioscience) or 4 µg/ml (BD) of plate-bound anti-CD28 (37–51). CD4+ T cells were isolated using positive or negative selection kits (Miltenyi Biotec). Naive CD4+CD25−CD44low T cells were preenriched using CD4 negative selection kit and sorted to >99% purity with a FACSAria II (BD) using anti-CD4 (BD), anti-CD25 (eBioscience), and anti-CD44 (BD) antibodies. Naive cells were polarized toward Th2 with 5 µg/ml of plate-bound anti-CD3 and 5 µg/ml of plate-bound anti-CD28 in the presence of 20 ng/ml of recombinant IL-4 (R&D Systems) and 20 µg/ml of blocking IFN-γ (R&D Systems) as previously described (Watford et al., 2010). Th17 differentiation was driven by stimulating CD4+CD25− T cells with 2 µg/ml of plate-bound anti-CD3 and 1 µg/ml of plate-bound anti-CD28 antibodies in the presence of 1 ng/ml of recombinant human (rh) TGF-β (R&D Systems), 200 U/ml rhIL-6 (R&D Systems), and 10 µg/ml of blocking IL-4 and IFN-γ antibodies (R&D Systems) in complete IMDM medium (Invitrogen).
For real-time PCR, mRNA was extracted with the RNeasy kit (QIAGEN) and converted as first-strand cDNA with the MMLV reverse transcription (Invitrogen). Each PCR was performed on an MX3000P qPCR machine (Agilent Technologies) from 50 ng cDNA using SYBR green technology (QIAGEN), and genes were amplified with the following primers (Invitrogen): IL-4 (forward, 5′-ACAGGAGAAGGGACGCCAT-3′; reverse, 5′-GAAGCCCTACAGACGAGCTCA-3′), IL-5 (forward, 5′-AGCACAGTGGTGAAAGAGACCTT-3′; reverse, 5′-TTCAATGCATAGCTGGTGATTT-3′), IL-13 (forward, 5′-AGACCAGACTCCCCTGTGCA-3′; reverse, 5′-TGGGTCCTGTAGATGGCATTG-3′), GATA3 (forward, 5′-AGGCAAGATGAGAAAGAGTGCCTC-3′; reverse, 5′-CTCGACTTACATCCGAACCCGGTA-3′), RORγt (forward, 5′-CCGCTGAGAGGGCTTCAC-3′; reverse, 5′-TGCAGGAGTAGGCCACATTACA-3′), SOCS1 (forward, 5′-ACCTTCTTGGTGCGCGAC-3′; reverse, 5′-AAGCCATCTTCACGCTGAGC-3′), SOCS2 (forward, 5′-GGTTGCCGGAGGAACAGTC-3′; reverse, 5′-GAGCCTCTTTTAATTTCTCTTTGGC-3′), and SOCS3 (forward, 5′-CCTTCAGCTCCAAAAGCGAG-3′; reverse, 5′-GCTCTCCTGCAGCTTGCG-3′). Cytokine secretion in cell culture supernatants was analyzed by ELISA (BD or R&D Systems). Total and antigen-specific serum IgE were measured by ELISA, as described previously (Mangan et al., 2007).
Surface marker expression was assessed by flow cytometry using anti-CD3 (BD), anti-CD4 (BD), or anti-CD8 (eBioscience). For cytokine staining, cells were restimulated in complete medium at 37°C for 4 h with 50 ng/ml PMA (Sigma-Aldrich), 500 ng/ml ionomycin (Sigma-Aldrich), and GolgiPlug (BD) or for 6 h with 2.5 µg/ml Concanavalin A (Sigma-Aldrich) and 10 µg/ml Brefeldin A (Sigma-Aldrich). Cells were fixed and permeabilized using the Cytofix/Cytoperm kit (BD), followed by intracellular anti–IL-4, anti–IFN-γ, or anti–IL-17 (BD or eBioscience) staining. For phospho-STAT analysis, the cells were fixed after stimulation, permeabilized with the Perm III buffer (BD), and stained with antibodies specific for phospho-STAT5, phospho-STAT3, and phospho-STAT6 according to the Phosflow (BD) instructions. Data were collected on a CyAn flow cytometer (Beckman Coulter) or on BD FACSCanto II (BD) and analyzed using FlowJo software (Tree Star).
Helminth egg injections.
Eggs from S. mansoni–infected mice were obtained as described previously (Smith et al., 2009). SEA at <1 endotoxin unit/mg (Chromogenic LAL; BioWhittaker) was prepared as described previously (Smith et al., 2009). Mice were given an i.p injection of either 10,000 S. mansoni eggs or PBS. For the adoptive transfer experiments, RAG-1−/− mice were reconstituted with CD4+ sorted from WT or SOCS2−/− mice before injection. In separate experiments, 5,000 S. mansoni eggs or PBS were injected into the footpad of mice to examine local cellular responses to SEA. Mice were culled and spleens or popliteal LNs removed 10 d after injection.
Calcipotriol (MC903; Sigma-Aldrich) was resuspended in ethanol at 50 µM. A total of 1 nM was applied daily to the outer and inner surfaces of the right ear (20 µl/ear) as described previously (Li et al., 2006; Elentner et al., 2009). Ethanol (20 µl/ear) was applied as a vehicle control. Ear thickness was monitored daily using digital calipers (Kroeplin). Spleens and auricular draining LNs of the MC903 or vehicle-treated mice were extracted at day 5 and at the end of the study. The cells were restimulated with anti-CD3 and anti-CD28 and the supernatants harvested after 48 h. MC903 or vehicle-treated ear samples were formalin fixed, embedded in paraffin, and stained with hematoxylin and eosin. 10 fields were randomly defined throughout the tissue and counted blindly by two people using a graticule at 400×. The number of eosinophils per square millimeter of tissue was then calculated.
Experimental allergic airway inflammation.
Mice were sensitized with OVA to induce allergic airway inflammation in a protocol optimized for mice on C57BL/6J strain background, as described previously (Fallon et al., 2009). Mice were administered 50 µg OVA in alum (Thermo Fisher Scientific) i.p. on days 0 and 14 and subjected to airway exposure with 1% OVA in PBS for 20 min on days 27–31. 24 h after the last OVA aerosol challenge, lung function was analyzed by an invasive method where mice were tracheostomized and ventilated using a whole-body plethysmography with a pneumotachograph linked to a transducer (EMMS) to determine changes in RL in response to increasing doses of inhaled methacholine (10, 30, 60, and 100 mg/ml; Sigma-Aldrich). Lungs from allergen-sensitized and challenged mice were taken and fixed in 10% formaldehyde-saline for histology (Mangan et al., 2007). Sections of lungs were stained with hematoxylin and eosin for assessment of lung inflammation and Periodic acid-Schiff stained for the enumeration of goblet cells and mucus production.
OT/II CD4+ cell transfer model of allergic airway inflammation.
Spleen and LNs were harvested from OT/II and SOCS2−/−xOT/II mice. CD4+ T cells were enriched from single cell suspension by negative selection using the EasySep mouse CD4+ T cell enrichment kit (STEMCELL Technologies). Enriched CD4+ T cells were subsequently sorted to >99% purity with a MoFlo (Beckman Coulter) using anti-CD4 (BD) antibodies. A total of 1.5 × 107 purified CD4+ cells isolated from OT/II or SOCS2−/−xOT/II mice were injected i.v. into RAG-1−/− recipient mice. Control mice received PBS. RAG-1−/− mice were administered 50 µg OVA adsorbed onto alum (Thermo Fisher Scientific) i.p. on days 1 and 8 and were subjected to airway exposure with 1% OVA aerosol in PBS for 20 min on days 9–13. 24 h after the last OVA aerosol challenge, lung function and immunological analysis was performed as described in the previous section. Lavage was performed and cytokines were detected in BAL fluid as in the previous section. Lungs were homogenized and cytokines determined (expressed as picogram of cytokine per milligram of lung protein) as previously described (Mangan et al., 2007).
Immunoblotting and immunoprecipitation.
CD4+ cells were prestimulated with anti-CD3 and anti-CD28 for 48 h before activation with rIL-4 (R&D Systems), rIL-12 (PeproTech), or rIL-6 (PeproTech) as indicated. Cells lysates were analyzed by Western blotting as previously described (Tannahill et al., 2005). Phosphorylated STAT proteins were detected using the following antibodies: phospho-STAT6 (Imgenex), phospho-STAT3 (Abcam), and phospho-STAT4 (Invitrogen). Blots were subsequently reprobed with total STAT6, 3, or 4 (Santa Cruz Biotechnology, Inc.). To analyze SOCS2 and SOCS3 expression, total splenocytes from WT or SOCS2−/− mice were prestimulated with anti-CD3, anti-CD28, and rIL-2 (R&D Systems) as shown. Cell lysates were immunoblotted with antibodies against SOCS3 (IBL International), SOCS2 (Cell Signaling Technology), and γ-tubulin (Sigma-Aldrich). For SOCS3 analysis in CD4+ cells, equal amounts of total protein were immunoprecipitated with anti-SOCS3 (IBL International) that had been precoupled to protein A Sepharose beads. SOCS3 protein was subsequently analyzed by Western blotting as previously described (Tannahill et al., 2005). Densitometry was performed using ImageJ software (National Institutes of Health).
Data were analyzed with Prism4 (GraphPad Software). Statistical differences were determined by two-tailed unpaired or paired Student’s t test as appropriate.
Online supplemental material.
Fig. S1 shows CD4+ and CD8+ numbers in thymus, spleen, and LN of WT and SOCS2−/− mice, increased IL-4, IL-5, IL-13, and GATA-3 mRNA expression in SOCS2-deficient CD4+ T cells by real-time PCR, and dot plot showing enhanced Th2 differentiation in the absence of SOCS2. Fig. S2 shows lung histology, cell infiltration in BAL, and IL-4, IFN-γ, and IL-17 concentration in BAL of RAG-1−/− mice reconstituted with PBS, WT, or SOCS2−/− CD4+ T cells after OVA challenge. Fig. S3 shows increased STAT6 and STAT5 phosphorylation and reduced STAT3 phosphorylation in SOCS2-deficient CD4+ T cells by flow cytometry, unaffected IL-4Rα levels, SOCS2 induction in CD4+ T cells after IL-6 and TGF-β stimulation, and reduced RORγt and IL-17 levels in SOCS2 deficient CD4+ T cells polarized toward Th17 lineage. Fig. S4 shows that in the absence of SOCS2, SOCS3 and SOCS1 levels are elevated in splenocytes, SOCS3 expression is higher in freshly isolated and Th2-polarized CD4+ T cells, and SOCS1 and SOCS3 mRNA expression is higher in CD4+ cells stimulated with anti-CD3 and anti-CD28.
We would like to thank D. Hilton for providing the SOCS2−/− mice (Royal Melbourne Hospital, Victoria 3050, Australia), S. Dubrac (Innsbruck Medical University, Austria) and L. Hanna (BRU, Queen’s University Belfast) for helping to design AD studies, C. Fulton (Queen’s University Belfast) for assistance with cell sorting, M. Walsh (City Hospital, Belfast) for helping with eosinophil counting, and O. Lyubomska and M. McCaigue, (Queen’s University Belfast) for assistance.
P.G. Fallon was supported by Science Foundation Ireland (07/IN1/B902) and National Children’s Research Centre. J.A. Johnston, A. Kissenpfennig, and J. Elliott were supported by the Department for Employment and Learning of Northern Ireland, The Welcome Trust (WT008532) and Health Research Board (Ireland; PD12008/01), and the MRC (UK; G1001795).
The authors have no conflicting or competing financial interests.