Sprouty2 positively regulates T cell function and airway inflammation through regulation of CSK and LCK kinases

The function of Sprouty2 (Spry2) in T cells is unknown. Using 2 different (inducible and T cell–targeted) knockout mouse strains, we found that Spry2 positively regulated extracellular signal-regulated kinase 1/2 (ERK1/2) signaling by modulating the activity of LCK. Spry2−/− CD4+ T cells were unable to activate LCK, proliferate, differentiate into T helper cells, or produce cytokines. Spry2 deficiency abrogated type 2 inflammation and airway hyperreactivity in a murine model of asthma. Spry2 expression was higher in blood and airway CD4+ T cells from patients with asthma, and Spry2 knockdown impaired human T cell proliferation and cytokine production. Spry2 deficiency up-regulated the lipid raft protein caveolin-1, enhanced its interaction with CSK, and increased CSK interaction with LCK, culminating in augmented inhibitory phosphorylation of LCK. Knockdown of CSK or dislodgment of caveolin-1–bound CSK restored ERK1/2 activation in Spry2−/− T cells, suggesting an essential role for Spry2 in LCK activation and T cell function.


Introduction
Since its discovery as an inhibitor of fibroblast growth factor (FGF)-stimulated tubular morphogenesis [1], Sprouty (Spry) proteins have been implicated in numerous biological processes including limb development [2], lens morphogenesis [3], inner ear development [4], and kidney development [5] (reviewed in [6,7]). To date, 4 mammalian Spry orthologs (Spry1 to 4) have been identified [1,8], and their inhibitory function is both cell and ligand specific. While the expression of Spry3 is restricted to brain and testis, the rest of the Spry proteins are ubiquitously expressed and have been shown to inhibit the Ras/ERK1/2 signaling pathway [9,10]. This general inhibitory function of Spry is due to its ability to interact with key signaling molecules in the Ras/ERK1/2 signaling pathway such as Grb2, Shp2, Raf, Sos1, TESK, and FRS2 [11][12][13]. However, a role for Spry2 in enhancing epidermal growth factor (EGF)-induced extracellular signal-regulated kinase 1/2 (ERK1/2) activation by preventing c-Cbl-mediated degradation of EGF receptor has also been described [14]. We have shown a positive role for Spry2 in inducing sustained phosphorylation of ERK1/2 leading to its bistability in airway epithelial cells upon repetitive stimulation [15]. In contrast, overexpression of Spry2 in B cells compromised B cell receptor (BCR)-triggered ERK1/2 activation by engaging and antagonizing Syk, RAF, and BRAF [16]. Similarly, Spry1-deficient T cells [17] or Spry1/2-deficient antigen-specific memory CD8 + T cells [18] showed increased proliferation in response to T cell receptor (TCR) stimulation, suggesting a negative regulatory effect. While ectopically expressed Spry1 has a positive effect on TCR signaling in naïve T cells, it has an opposite effect in differentiated Th1 clones [19]. Collectively, these results demonstrate the context-dependent ability of Spry to function as either a positive or a negative regulator of ERK1/2 activation. ERK1/2 activation is directly implicated in T cell maturation [20], activation [21], and Th2 differentiation [22,23]. TCR ligation triggers the assembly of a multimolecular complex with concomitant phosphorylation of the z-chain of the CD3 complex by the kinase LCK [24,25]. This dynamic and complex process links activation of cell surface receptors to nuclear effectors and subsequent transcription of cytokines [25]. Although several molecular mechanisms of ERK1/2 activation have been reported [20,21,23,26], our understanding of ERK1/2 regulatory mechanisms is incomplete. In this study, we for the first time show that Spry2 positively regulated LCK and ERK1/2 activation and CD4 + T cell function by modulating the recruitment and activity of CSK.

Effect of Spry2 ablation on immune cells
To study the role of Spry2, we generated ERT 2 -Cre:Spry2 flox/flox mice, which allows conditional deletion of Spry2 upon Tamoxifen administration (Spry2 −/− ). Tamoxifen treatment eliminated Spry2 protein ( Fig 1A) and mRNA ( Fig 1B) expression in T cells. Tamoxifen-treated Spry2 f/f and ERT 2 -Cre had a normal level of Spry2, and we used Tamoxifen-treated Spry2 f/f as controls (Spry2 +/+ ). Flow cytometric analysis revealed a significantly reduced CD4 + but not CD8 + T cells in the spleens of Spry2 −/− mice (Fig 1C and 1D). The numbers of double-positive (DP), single-positive (SP) CD4 + , and SP CD8 + were similar in the thymus from Spry2 −/− and Spry2 +/ + mice (Fig 1E and 1F), indicating that Spry2 was dispensable for transition from double-negative to DP and then SP T cell selection. To assess if reduced T cell numbers in Spry2 −/− mice was due to Tamoxifen toxicity, we generated mice with CD4-targeted Spry2 deletion by crossing Spry2 f/f mice with CD4-Cre mice (S1A Fig). Similar to ERT 2 -Cre:Spry2 flox/flox mice, CD4-Cre:Spry2 f/f mice had reduced peripheral CD4 + T cell numbers compared to Spry2 f/f littermate controls (S1B Fig).
Reduced T cell numbers in peripheral organs could be due to survival defects. Hence, splenic T cells from Spry2 +/+ and Spry2 −/− mice were assayed for B cell lymphoma 2 (Bcl-2), a survival protein. Bcl-2 expression in CD4 + T cells from Spry2 −/− mice was compromised ( Fig  1G), suggesting that Spry2 promotes survival. Cleaved Caspase-3 levels ( Fig 1H) and Annexin-V staining (Fig 1I and 1J) in Spry2 −/− CD4 + T cells was increased upon anti-CD3/CD28 stimulation, suggesting increased apoptosis. CD4 + T cells from Spry2 −/− mice showed defective proliferation as shown by the carboxyfluorescein succinimidyl ester (CFSE) dilution assay, especially in response to anti-CD3/CD28+IL2 stimulation (Fig 1K, inducible knockout; S1C Fig, CD4-targeted knockout). The expression of TCRβ in Spry2 −/− was comparable to Spry2 +/+ mice ( Fig 1L). Collectively, these results suggest that Spry2 is critical for survival and proliferation of CD4 + T lymphocytes. The differential effect of Spry2 deletion on splenic and thymic T cell populations prompted us to check Spry2 expression in the splenocyte and thymocyte compartments under homeostatic conditions. Spry2 expression was higher in splenocytes as compared to thymocytes in B6 mice (Fig 2A). Further, Spry2 was induced in murine CD4 + T cells upon TCR stimulation in a timedependent manner (Fig 2B). To confirm the inducible nature of Spry2, human CD4 + T cells were stimulated and immunoblotted for Spry2 and p-ERK. We observed progressive induction of Spry2 and a positive correlation between Spry2 induction and ERK1/2 activation in a timeand stimulatory strength-dependent manner ( Fig 2C). This positive correlation was established in mouse CD4 + T cells wherein Spry2 deficiency substantially abrogated ERK1/2 activation in CD4 + T cells and modestly in CD8 + T cells ( Fig 2D). Further, under homeostatic conditions, Spry2 expression was higher in splenic naïve CD4 + T cells relative to memory CD4 + T cells (Fig 2E).
To assess the role of Th2 cells independent of other type-2 innate immune cells, we adoptively transferred CD4 + T cells from Spry2 +/+ and Spry2 −/− mice to Rag2 −/-γC −/− mice lacking ILC2s. Recipient mice were sensitized and IN challenged either with Sal or Asp ( Fig 4A) and assessed for AHR, BAL inflammatory cell influx, and lung inflammation. Consistent with our previous results, mice receiving Spry2 −/− CD4 + T cells showed reduced AHR, decreased eosinophils and lymphocytes in the BAL, attenuated lung inflammation, mucus production, and collagen deposition upon Asp exposure (Fig 4B-4G). Similarly, secretion of IL5 and IL13 was severely impaired in allergen-challenged Rag2 −/-γC −/− mice that received CD4 + T cells from Spry2 −/− mice, as compared to the recipient mice which received CD4 + T cells from Spry2 +/+ mice (Fig 4H and 4I). These data suggest that Spry2 positively regulates T cell-driven asthma by up-regulating Th2 cytokines.

Spry2 expression is increased in asthma
To further evaluate its contribution to human asthma, we assayed Spry2 expression in CD4 + T cells from blood and BAL from patients with asthma and disease controls by flow cytometry. Demographic and clinical characteristics of patients with asthma and disease control are in S1 Table. The expression level of Spry2 in CD4 + T cells from blood and BAL was significantly higher in patients with asthma ( Fig 5A).

Enhanced interaction between CSK and LCK in Spry2 −/− CD4 + T cells
LCK is negatively regulated by CSK. The latter is predominantly cytosolic and requires membrane adapters in the vicinity of LCK-containing lipid rafts to exert its inhibitory effect [29]. In addition, phosphorylation of CSK at Ser-364 results in increased kinase activity [30]. Thus, the membrane-bound active CSK phosphorylates LCK at Y505 to suppress TCR signaling. We determined CSK subcellular localization in CD4 + T cells. CSK was diffusely distributed in the cytosol of Spry2 +/+ CD4 + T cells, whereas a major fraction of CSK was localized to the plasma membrane in Spry2 −/− CD4 + T cells (Fig 7A and 7B). However, the total levels of CSK remained the same in Spry2 +/+ and Spry2 −/− CD4 + T cells (Fig 7C), suggesting that it is the altered membrane localization of CSK that is responsible for increased LCK(Y505) phosphorylation in Spry2 −/− CD4 + T cells. Subcellular fractionation and immunoblotting showed increased levels of p-CSK and CSK in the membrane fractions of Spry2 −/− CD4 + T cells ( Fig  7D). Immunoprecipitation studies showed that higher levels of CSK coprecipitated with LCK from Spry2 −/− CD4 + T cells and that enhanced LCK:CSK interaction persisted for 10 min after anti-CD3/CD28 stimulation ( Fig 7E). We examined the phosphorylation status of CSK in the immunoprecipitates and found higher level of p-CSK in Spry2 −/− CD4 + T cells relative to controls ( Fig 7E). Immunostaining of LCK and CSK under unstimulated and stimulated conditions (Fig 7F and 7G) showed enhanced colocalization of LCK and CSK in Spry2 −/− CD4 + T cells. Enhanced colocalization with CSK may increase phosphorylation (Y505) of LCK. We therefore examined in situ LCK(Y505) phosphorylation in CD4 + T cells from the lung sections obtained from our mouse asthma model and observed increased phospho-LCK (Y505) staining in CD4 + T cells from Asp-challenged Spry2 −/− mice (Fig 7H and 7I). Thus, Spry2 deletion alters the localization and activity of CSK, enhanced its interaction with LCK, and augmented LCK Y505 phosphorylation.

Enhanced interaction between CSK and Cav-1 leads to increased CSK activity in Spry2 −/− CD4 + T cells
Differences in CSK activity prompted us to determine the role of Spry2 in the regulation of CSK. To test whether Spry2 is targeted to the lipid rafts, anti-CD3/CD28-stimulated CD4 + T cells were stained for Spry2, CSK, and Cholera Toxin-B (CTX-B, a lipid raft marker). Colocalization of Spry2 with CTX-B was minimal prior to TCR triggering (S4A Fig); however, 5 min post-TCR triggering, Spry2 was associated with CSK in the lipid rafts, which completely diminished 30 min post-TCR stimulation. Hence, we hypothesized that Spry2 regulated LCK activity directly by interfering with the complex formation between LCK and CSK. To test this in a reductionist model, we pulled down LCK and subjected the immunoprecipitates to a kinase reaction in the presence or absence of recombinant CSK (r-CSK) and Spry2 (r-Spry2) (S4B Fig). Addition of r-CSK to the immunoprecipitates enhanced LCK-Y505 phosphorylation (S4B Fig, lane 4). However, r-CSK in the presence of increasing concentrations of r-Spry2 (S4B Fig, lanes 5, 6, and 7) failed to block the kinase activity of recombinant CSK, thereby suggesting that Spry2 may function by altering CSK recruitment but not its inherent activity.
The N-terminal domain of Cav-1 contains a caveolin scaffolding domain (CSD, residues 82 to 101), which binds to and serves as a scaffold for several caveolin-interacting proteins including CSK [34,35]. Recombinant CSD peptide fused to Antennapedia homeodomain (AP-CSD) act as a "dominant negative inhibitor" by discharging the interacting proteins from Cav-1 [36]. We hypothesized that disrupting the Cav-1 complex would inhibit CSK activity and enhance ERK1/2 activity in response to TCR stimulation. Treatment of Spry2 −/− CD4 + T cells with AP-CSD showed compromised CSK activity (reduced p-CSK levels) and elevated ERK1/2 phosphorylation compared with scrambled peptide treatment (Fig 8H). Similarly, Spry2 −/− CD4 + T cells treated with AP-CSD showed significantly elevated levels of both IL2 and IFNγ as compared to Spry2 −/− CD4 + T cells treated with control peptide (Fig 8I and 8J). These results suggest that in the absence of Spry2, Cav-1-bound CSK inhibits LCK and diminishes TCRinduced responses in CD4 + T cells.

Discussion
While the role of Spry proteins as negative feedback regulators of receptor tyrosine kinase (RTK) signaling in stromal cells is well established [9,37], very little is known about their role in lymphoid cells. In this study, we show that Spry2 functions as a positive regulator of CD4 + T cell function and differentiation by modulating the activity of non-RTKs. Consistent with earlier studies [18,19], we observed that expression of Spry2 is induced upon TCR stimulation, suggesting that Spry2 expression is regulated by TCR at transcriptional level. Our data indicate that Spry2 is an important determinant of effector T cell differentiation as deletion of Spry2 resulted in reduced expression of Th-specific transcription factors and cytokines under Th1-, Th2-, and Th17-skewing conditions. The robust induction of Spry2 in differentiated T effector cells as opposed to metabolically quiescent naïve or memory T cells indicate that Spry2 plays a fundamental role in orchestrating pathways involved in signaling reprogramming that is displayed by T effector cells.
Recently, Spry2 was identified as a negative regulator of BCR-mediated ERK1/2 signaling in B cells [16]. In contrast, we show that Spry2 functions as a positive regulator of ERK1/2 signaling in response to TCR stimulation. This differential effect on ERK1/2 signaling by Spry2 suggests diverse upstream mechanisms of activation in different cell types.
LCK associates with CD4 and plays a central role in initiation of TCR signaling [25]. LCK activation requires an alteration in its conformation. LCK phosphorylation at the C-terminal Y505 residue by CSK results in a closed, inactive conformation, while phosphorylation of Y394 facilitates open conformation and leads to full activation [41]. Spry2 deficiency decreases the activating phosphorylation of Y394 and increases the inhibitory phosphorylation of Y505, suggesting Spry2 antagonizes CSK and prevents Y505 phosphorylation to allow LCK activation. The inability to activate multiple downstream signaling pathways-ERK1/2, and nuclear factor of activated T cells (NFAT)-and produce cytokines belonging to multiple T helper lineages in Spry2-deficient mice is consistent with a failure of the TCR to activate receptor-proximal signaling molecules. An earlier report suggests that ERK1/2 positively regulates the induction of Spry2 [42]. Here, we show a positive correlation between Spry2 induction and ERK1/2 activation in activated CD4 + T cells. Thus, we speculate that a dynamic interplay exists between Spry2 induction and ERK1/2 activation in CD4 + T cells that leads to optimal TCRgenerated signals.
CSK is predominantly cytosolic [29]. Through its SH2 domain, CSK interacts with a number of adapter molecules (e.g., Cbp/PAG1, Caveolin-1, Paxillin, and LIME) which serve as membrane anchors [31][32][33]43,44]. The association with these adaptor molecules increases CSK activity, thereby enabling inhibitory (Y505) phosphorylation of LCK. Cbp/PAG1 is a major regulator of CSK [31,32]. However, recent studies indicate that Cbp is dispensable for raft localization of CSK in T cells [45]. We show that Spry2 deletion leads to a marginal decrease in Cbp/PAG1 but a robust increase in Cav-1, which results in elevated CSK activity and, consequently, maintenance of LCK in an inhibited state. Functional redundancies exist among Csk-binding proteins as knockdown of either Cav-1 or Cbp results in a compensatory increase in expression of the other to inhibit Src activity [46].
The lipid raft protein, Cav-1, plays an important role in cell migration, endocytosis, and vesicular transport, and regulates a number of signaling pathways via its scaffolding domain [34,47]. Cav-1 also plays an important role in TCR synaptic polarity, signaling strength, T cell proliferation, and cytokine production [48][49][50]. Cav-1, through its C-terminal and scaffolding domain, interacts with Spry2 and CSK, respectively [51,52]. One of the perturbations in Spry2 −/− T cells was an overabundance of Cav-1 in the membrane, which positively correlated with CSK abundance and activity. Our findings reveal, for the first time, that Spry2 regulates the Cav-1 protein turnover in CD4 + T cells by promoting ubiquitination and subsequent proteosomal degradation. Cav-1 is ubiquitinated [53,54], and Spry2, acting as an adaptor, could bring ubiqutin E3 ligases in close proximity to Cav-1 to promote its ubiquitination and degradation. Spry2 interacts with the E3 ubiquitin ligase Cbl [55]. Whether Spry2 directs Cbl to promote Cav-1 ubiquitination is unknown. While it needs to be investigated as to how Spry2 regulates Cav-1 levels, our work does support the notion that the Cav-1-bound CSK plays a major role in defective LCK activation in Spry2 −/− CD4 + T cells. A recent study [56] demonstrated that Cav-1 negatively regulated Janus kinase (JAK)/STAT5 signaling pathway. The present study in CD4 + T cells from Spry2−/− mice showing an up-regulation of Cav-1 and corresponding defective phosphorylation of STAT5 and STAT6, in response to IL-2 and TSLP, suggest that a similar mechanism may exist by which Cav-1, via its scaffolding domain, negatively regulates upstream JAK tyrosine kinases.
This study suggests a novel role for Spry2 as a positive regulator of ERK1/2 signaling and T cell function by modulating LCK activity via CSK. Spry2 −/− mice fail to develop airway hyperreactivity and type 2 inflammation in T cell model of asthma. We employed adoptive transfer of Spry2 −/− T cells to Rag2 −/-γC −/− mice to eliminate the possibility that this defect was due to global Spry2 deletion. Our results demonstrated that Spry2 in CD4 + T cells is indispensable for the asthma phenotype. In human T cells, Spry2 knockdown preferentially inhibited type 2 cytokines but not type 1 cytokines. Augmented TNFα production in Spry2 knockdown cells that we observed suggests a differential regulation of human T helper cells and their cytokines by Spry2. Our result is in agreement with a paper by Chiu and colleagues [57], who observed increased TNFα expression in Spry2 knockdown HIV-specific T cells. This was associated with increased polyfunctionality of T cells.
Our study identifies Spry2 as a positive regulator of CD4 + T cell function, and type 2 immunity, improving the understanding of Spry2 regulation of TCR signaling. Our results also uncover a novel mechanism of induction of type 2 inflammation in disorders such as asthma.

Airway hyperreactivity (AHR) measurement
Three days after the last IN challenge, mice were anesthetized with ketamine (180 mg/kg), xylazine (9 mg/kg), and acepromazine (4 mg/kg), underwent tracheotomy, and were attached via an 18-gauge cannula to small-animal ventilator with computer-controlled piston (Flexivent; Scireq, Montreal, Quebec, Canada). AHR was measured in response to methacholine (Mch) as described previously [27]. Group averages were expressed as the fold increase over baseline resistance (mean ± SEM). Statistical significance was determined using ANOVA and unpaired Student t test.

Histology and morphometric analysis
Formalin-fixed and paraffin-embedded lung tissue sections (5 μm) were stained with hematoxylin and eosin (H&E) for morphometric analysis, periodic acid-Schiff (PAS) for mucin staining, and Masson trichrome for collagen deposition and mounted using permount medium (Thermo Fisher Scientific). Images were acquired on a Nikon Eclipse TE2000-U microscope using ×20 dry lenses. Inflammation, PAS, and Trichrome staining intensity was quantified with Metamorph image acquisition and analysis software (Molecular Devices, Eugene, Oregon). Airway inflammation was measured as the ratio of the total area of peribronchial and perivascular inflammatory infiltrates over the perimeter of the associated basement membrane obtained from a minimum of 5 airways per mouse and 5 mice per group.

Aspergillus-specific IgE
Briefly, Aspergillus extract-coated ELISA plates (10 μg/mL) were washed with PBS-Tween (0.05%) and blocked with 10% FBS in PBS for 1 h at room temperature (RT). Triplicate 50 μL volumes of serum from each mouse were then incubated for 2 h at RT. Plates were washed 5 times in PBS and incubated with 1:1,000 biotinylated anti-IgE (BD PharMingen, San Jose, California, United States of America). Plates were then washed and incubated with streptavidinhorseradish peroxidase (BD PharMingen) for 1 h at RT. Plates were washed 7 times and developed with TMB substrate (Pierce) for 20 min. 2N H 2 SO 4 was added to stop the reaction and the optical density (OD) read at 450 nm. Results are reported as raw OD values.

Flow cytometry
One million cells were incubated with 1 μg of purified anti-mouse CD16/32 (clone: 93, Biolegend, San Diego, California) in FACS buffer prior to staining. Cell staining was performed with fluorescently labeled anti-mouse monoclonal antibodies (described below) for 30 min at 4˚C. Cells were washed with FACS buffer, fixed with 4% paraformaldehyde for 15 min at 4˚C, and data acquired using either LSRII or LSRFortessa (BD Biosciences, San Jose, California) flow cytometer. For intracellular staining, cells were incubated for 4 h with monensin (2 μmol/ L). Surface staining was performed as above. Cells were then washed in FACS buffer, fixed, and permeabilized (Foxp3 kit, eBioscience, San Diego, California; #00-5523-00) for 1 h at RT. Cell were washed and stained with antibodies against intracellular proteins (cytokines, transcription factors) for 30 min at 4˚C. Cells were washed and data acquired using either LSRII or LSRFortessa. All data were analyzed using FlowJo v10 software. Anti-mouse antibodies used for surface and intracellular staining were given in S2 Table.
Equal amounts of protein were loaded on SDS-PAGE, transferred on to Nitrocellulose membrane (GE, Amersham). Blots were subsequently blocked with 5% Blotto in TBS-T for 1 h at RT, followed by overnight incubation with primary antibodies. After 3 washes with TBS-T (5 min each), blots were incubated in HRP-conjugated secondary Abs for 1 h at RT, followed by 3 washes with TBS-T. Pierce ECL western blotting substrate was used to visualize bands, and the bands were quantified using the ImageJ software. The primary antibodies used for western blotting were given in S2 Table. Immunoprecipitation Coimmunoprecipitation was done as described previously [59]. Briefly, CD4 + T cells were washed with ice-cold PBS and lysed in buffer containing 25 mM Tris-HCl (pH7.4), 150 mM NaCl, 1.0% TritonX-100, 5 mM EDTA, 0.1% BSA, and protease and phosphatase inhibitor cocktail (Sigma) at 4˚C for 20 min. Lysates were centrifuged at 10,000g for 10 min, and supernatants were incubated either with 4 μg of LCK antibody or appropriate control Ig for 12 h at 4˚C on a Roto-torque followed by 2 h of incubation with protein A/G plus agarose beads to immunoprecipitate LCK-bound proteins. The bound fraction was washed, and immunoprecipitates were lysed in SDS containing Laemmli sample buffer. The samples were then subjected to western blotting analysis as described above.
For in vitro kinase assay, CD4 + T cells (5 × 10 7 ) were lysed in Pierce IP lysis buffer (# 87787, Thermo Fisher) containing protease and phosphatase inhibitor cocktail (Sigma) at 4˚C for 20 min. Lysates were centrifuged at 10,000g for 10 min, and supernatants were incubated either with 4 μg of LCK antibody or appropriate control Ig for 12 h at 4˚C on a Roto-torque followed by 2 h of incubation with Protein G Magnetic Beads (#70024, CST, Danvers, Massachusetts). Magnetic bead-antibody complex were washed 4 times with IP lysis buffer and twice with 500 μl 1× kinase buffer (#9802, CST) using a magnetic separation rack. Washed pellets were suspended in 40 μl of 1× kinase buffer supplemented with 200 μM ATP (#9804, CST) and 1 μg of recombinant mouse CSK (#50893-M20B, Sino Biological, Beijing, China) or 1 μg of recombinant mouse CSK with increasing concentrations (0.5 μg or 1 μg or 2 μg) of recombinant mouse Spry2 (MyBiosource, San Diego, California; #MBS1362850). Pellets were incubated for 30 min on a heating block at 30˚C, and the kinase reaction was terminated using 4× Laemmli sample buffer. Immunoblot analysis was done as described above.

Immunohistochemistry (IHC) and immunocytochemistry (ICC)
Formalin-fixed paraffin-embedded lung tissue sections from Aspergillus-treated Spry2 +/+ and Spry2 −/− (n = 3) mice were deparaffinized with Citrisolv, rehydrated in graded alcohol series, and further rehydrated in PBS. To retrieve epitopes for efficient antibody binding, sections were microwaved for 15 min in 10 mM citrate buffer (pH 6.0). Sections were blocked with blocking buffer (3% BSA in PBS) for 1 h at RT and were incubated overnight in a humid chamber at 4˚C in primary antibodies CD4 and p-LCKY505 diluted in blocking buffer; followed by 3 washes with PBS. Appropriate fluorescent conjugated secondary antibodies (Molecular Probes, Eugene, Oregon) diluted in blocking buffer (1:200) were added to the sections for 1 h at RT in dark, followed by 3 PBS washes, and mounted using Prolong antifade reagent containing DAPI.
For ICC, CD4 + T cells were fixed (2% paraformaldehyde) and cytospun onto charged slides. We followed the same staining procedure as described above, except that slides were not subjected to deparaffinization, rehydration, or heat-induced antigen retrieval steps. However, cells were permeabilized in PBS containing 0.5% TritonX-100 and 0.05% Tween-20 for 5 min before the staining procedure. Images were acquired using Zeiss LSM 700 Confocal microscope with 63× oil immersion objective lens. Images were analyzed and processed using Zen Blue software (Carl Zeiss, Germany) and further processed using Adobe photoshop. Pearson correlation coefficients for colocalization were calculated by Zen Blue software. The details of primary antibodies and the dilutions at which they were used for ICC were given in S2 Table. Proximity ligation assay (PLA) Unstimulated and anti-CD3/CD28-stimulated CD4 + T cells were fixed (10 min in 3.7% formaldehyde); cytospun onto charged slides and washed in PBS (2× 5 min each) before permeabilization in PBS containing 0.5% TritonX-100 and 0.05% Tween-20. PLA was done essentially according to manufacturer's instructions provided along with the kit (Duolink In Situ Red Starter Kit Mouse/Rabbit; Cat# DUO92101). Briefly, slides were incubated in blocking buffer for 1 h at RT and further incubated overnight in primary antibodies (mouse CSK and rabbit Cav-1) diluted in appropriate buffer at 4˚C. Slides were then rinsed 2× 5 min in wash buffer, incubated in PLA probes for 1 h at 37˚C, washed again (2× 5 min), and incubated in ligase for 30 min at 37˚C before proceeding to amplification step. Finally, slides were washed and mounted using Duolink In Situ Mounting Medium with DAPI; images were acquired using Zeiss LSM 700 Confocal microscope and analyzed by Zen Blue software.

Human CD4 + T cell isolation and ELISA
Peripheral blood mononuclear cells (PBMCs) were isolated by density gradient centrifugation using Histopaque (Sigma). CD4 + T cells were purified from PBMCs by negative selection using CD4 + T cell isolation kit (Miltenyi Biotech, Germany). Sort-purified T cells were transduced with Lentiviral vectors harboring either Control shRNA (shLuc) or SPRY2 shRNA (transOMIC Technologies, Huntsville, Alabama, USA) as described below. Cells were cultured under T helper differentiation conditions for 6 days, supernatants were collected and analyzed by ELISA for human TNF-α (Biolegend), IL-4 (R&D Systems), IL-10, IL-17, and IFN-γ (eBioscience) according to manufacturer's instructions. BALF collected from saline or Asptreated Spry2 +/+ and Spry2 −/− mice were analyzed for IL5 and IL13 using Duoset ELISA kits according to manufacturer's instructions.

Peptides
Peptides corresponding to either CSD (N-terminal cytoplasmic region of Cav-1 with amino acids 82-101; DGIWKASFTTFTVTKYWFYR) or the scrambled control (WGID-KASFTTFTVTKYWFRY; fused at the N-terminus to the cell-permeable Antennapedia internalization sequence (43-58) RQIKIWFQNRRMKWKK) were purchased from Sigma (Cat# 219482 and 219483). Lyophilized peptides were resuspended in DMSO at 10 mM, then diluted to 1 mM in distilled water. CD4 + T cells from Spry2 −/− mice (10 6 cells/group) were treated with CSD or scrambled control peptides (10 μM) for 8 h and subsequently remained unstimulated or were stimulated with plate-bound anti-CD3/CD28 for 10 min. Cells were then analyzed by western blotting.

Statistics
All statistical analyses were performed using GraphPad Prism 7 software. Statistical significance in human studies was calculated by the nonparametric Mann-Whitney U test. Statistical analyses in mouse studies were done by Student t test. Results are shown as mean +/− SEM. All in vitro experiments were repeated independently at least 3 times. For all analyses, significance was determined at � : p < 0.05, �� : p < 0.005, ��� : p < 0.0001.

Ethics statement
All experiments involving mice were carried out in accordance with protocols approved by National Jewish Health Institutional Animal Care and Use Committee (Protocol # AS2614).