Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Asthma is characterized by both acute and chronic airway inflammation (1). In particular, affected individuals often show peripheral and pulmonary eosinophilia. Experimental allergen challenge of atopic individuals causes marked airway eosinophilia (2). These data strongly suggest that eosinophils are critical components of pulmonary inflammation and thus involved in the pathogenesis of asthma.

Interleukin (IL)-5 and granulocyte macrophage colony-stimulating factor (GM-CSF) are potent regulators of eosin-ophil production, chemotaxis, effector functions, and survival (2, 3). Elevated levels of IL-5 and GM-CSF have been detected in the sera (4) and bronchoalveolar lavage (BAL) fluid of asthmatics (5). IL-5 and GM-CSF are produced by activated T helper 2 lymphocytes (6), pulmonary epithelial cells (7), and eosinophils themselves (8).

The signal transducers and activators of transcription (STAT) family of transcription factors often mediate cytokine-induced responses. Receptor-bound cytokines trigger the phosphorylation and activation of receptor-associated Janus family tyrosine kinases (JAKs). Enzymatically active JAKs phosphorylate latent cytoplasmic STATs and in this process convert them into active factors. The tyrosine-phosphorylated STATs form homo- or heterodimers and translocate into the nucleus where they bind to their specific target DNA sequences and alter transcription (9). Studies on the involvement and function of different STAT proteins in cytokine signaling events have documented cell- type specificities. In human neutrophils, for example, GM-CSF induced the tyrosine phosphorylation of JAK2, followed by activation of STAT1, STAT3, and STAT5B, whereas only STAT5 was phosphorylated in the OTT1 cell line (10). Similarly, STAT5 but not STAT1, was activated in human monocytes in response to GM-CSF (13, 14). We and others have previously shown that the JAK/STAT pathway is activated in peripheral blood eosinophils (PBEs) by IL-5 family members (15, 16). However, little is known about STAT5-dependent gene transcription initiated by IL-5 and GM-CSF in PBEs. Neither the kinetics of STAT5 activation in response to IL-5 and GM-CSF nor a comparison of the downstream signaling events induced by these two cytokines has been performed in PBEs. We show that in response to both IL-5 and GM-CSF, STAT5 was rapidly tyrosine phosphorylated and bound the palindromic interferon gamma activation site (GAS) DNA motif (17) in extracts of cytokine-stimulated PBEs. Transcription of the STAT5 responsive gene cytokine-inducible, SH2-containing protein (CIS1) was upregulated in PBEs within 1 h of cytokine stimulation. These data suggest that IL-5 and GM-CSF activated STAT5 as well as downstream transcriptional events in human eosinophils.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents

-Glycerophosphate, sodium vanadate, and other laboratory reagents were obtained from Sigma (St. Louis, MO) and Fisher Scientific (Itasca, IL). All the protease inhibitors were from Boeh-ringer Mannheim (Indianapolis, IN). Anti-STAT5 antibody was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-phosphotyrosine STAT5 antibody, which detects Tyr (694)-phosphorylated STAT5, was procured from New England Biolabs (Beverly, MA). Anti-phosphotyrosine monoclonal antibody 4G10 from UBI (Lake Placid, NY) was used for the immunoprecipitation study (Figure 1). From R&D Systems (Minneapolis, MN), we obtained human recombinant IL-5 and GM-CSF. H-7, dihydrochloride [1-(5-isoquinolinesulfonyl)-2-methylpiperazine, HCl], was obtained from Calbiochem (La Jolla, CA).

View larger version (35K): [in this window] [in a new window]   Figure 1.   IL-5 and GM-CSF induce STAT5 phosphorylation in eosinophils. Freshly isolated eosinophils were treated with control buffer, IL-5 (1 nM), or GM-CSF (250 pM), immunoprecipitated either with anti-STAT5 antibody or with nonspecific rabbit IgG, and immunoblotted either with anti-phosphotyrosine antibody (upper panel) or with anti-STAT5 antibody (lower panel).

Isolation of PBEs

Eosinophils were purified from the heparinized peripheral blood of both atopic and normal volunteer donors, with eosinophils composing 2 to 10% of the peripheral blood leukocytes (18). A granulocyte mixture was obtained from the leukocyte buffy coat after centrifugation through a Percoll monolayer (1,090 g/ml) and lysis of erythrocytes by hypotonic shocks. The suspension was depleted of neutrophils by incubation with anti-CD16-conjugated paramagnetic microbeads. The recovered eosinophils were resuspended in Hanks' balanced salt solution supplemented with 2% newborn calf serum. These cell preparations were at least 95% eosinophils, based on morphologic examination.

Eosinophil Stimulation

Freshly isolated eosinophils were stimulated with control buffer, IL-5, and GM-CSF, and incubated at 37°C for various times, as indicated in each experiment. After the reactions were stopped by addition of ice-cold buffer, the cells were collected by centrifugation and cell fractions were prepared. A range of concentrations of IL-5 and GM-CSF (10 to 1,000 pM) was tested and the represented figures are shown with the concentrations stated in individual figure legends.

Preparation of Cell Lysates for Electrophoretic Mobility Shift Assays

Cytoplasmic cell extracts (Figure 3) were prepared as described previously (19). Alternatively, a protocol by van der Bruggen and coworkers (20) was used with modifications (Figure 4). Briefly, eosinophils were resuspended in 400 µl of a hypotonic buffer containing 10 mM N-2-hydroxyethylpiperazine-N'-ethane sulfonic acid (Hepes), pH 7.9, 1.5 mM MgCl₂, 10 mM KCl, 1 mM Na₃VO₄, 20 mM -glycerophosphate, 0.5 mM dithiothreitol (DTT), 0.2 mM phenylmethylsulfonyl fluoride (PMSF), and 10 µg/ml aprotinin. After incubation on ice for 15 min, Nonidet (NP)-40 was added to a final concentration of 0.6% and the mix was vortexed full speed for 10 s. The cytoplasmic fraction was collected after centrifugation at 4°C for 10 s at 7,200 × g. The pellet was resuspended in hypertonic buffer containing 20 mM Hepes, pH 7.9, 420 mM NaCl, 1.5 mM MgCl₂, 0.2 mM ethylenediaminetetraacetic acid (EDTA), 0.5 mM DTT, 1 mM Na₃VO₄, 0.2 mM PMSF, and 10 µg/ml aprotinin and incubated on ice for 20 min. After centrifugation for 2 min at 13,500 × g, the soluble nuclear extract (supernatant) was collected. Both fractions were stored at 80°C until used. Bradford assay (Pierce, Rockford, IL) was performed to determine the protein concentration.

View larger version (54K): [in this window] [in a new window]   Figure 3.   Kinetics of induction of GAS binding activity in IL-5- and GM-CSF-treated eosinophils. After stimulation with IL-5 (1 nM) or GM-CSF (250 pM) for the indicated time intervals, soluble cytoplasmic fractions were prepared, and 10 µg of these fractions were subjected to EMSA with radiolabeled duplex oligonucleotides from the -casein promoter region harboring the GAS element.

View larger version (63K): [in this window] [in a new window]   Figure 4.   Anti-STAT5 antibody supershifts the STAT5-GAS complex. EMSA reactions were prepared with 5 µg of either cytoplasmic (upper panel) or nuclear extracts (lower panel) of eosinophils that had been treated with IL-5 (300 pM) or GM-CSF (100 pM) for 15 min. Parallel reaction mixtures contained either anti-STAT5 antibody or nonspecific rabbit IgG.

Immunoprecipitation and Immunoblotting

Stimulated eosinophils were diluted in ice-cold buffer (20 mM Tris, pH 7.4, 137 mM NaCl, 1 mM EDTA, 1 mM PMSF, 1 µg/ml leupeptin, 1 mM Na₃VO₄, 2 µg/ml aprotinin, 20 mM -glycerophosphate, 10 mM NaF), pelleted by centrifugation, and the supernates discarded. The cell pellets were resuspended in the same buffer with detergents (1% NP-40, 0.25% deoxycholate, and 0.1% sodium dodecyl sulfate [SDS]) at a concentration of 2.5 × 10⁷ cells/ml and kept on ice for 10 min and centrifuged at 14,000 rpm for 10 min. The lysates were precleared for 4 h by incubation with rabbit immunoglobulin (Ig) G and agarose-conjugated protein A before incubation with 1.5 µg rabbit anti-STAT5 antisera at 4°C and precipitated by agarose-conjugated protein A. Control immunoprecipitates contained 1.5 µg of preimmune rabbit IgG instead of specific antisera. The agarose beads were washed with six changes of buffer with detergents, resuspended in SDS-polyacrylamide gel electrophoresis (PAGE) loading buffer, and immunoblotted as previously described (18).

Electrophoretic Mobility Shift Assay and Preparation of the Probe

Complimentary oligonucleotides (5'-GATCAGATTTCTAGGA-ATTCAAATC-3' and 5'-GATCGATTTGAATTCCTAGAAA-TCT-3') from the -casein promoter region with four base overhangs were combined, boiled in 0.5 M NaCl, and cooled to room temperature (21). Duplexes were end-labeled with [-³²P] deoxycytidine triphosphate (> 3,000 Ci/mmol) with terminal transferase (Promega, Madison, WI) following the manufacturer's recommendations. Aliquots of each lysate were incubated with 10 µl of 2.5× electrophoretic mobility shift assay (EMSA) buffer containing 25 mM Hepes, pH 7.8, 125 mM KCl, 2.5 mM EDTA, 12.5 mM MgCl₂, 25% glycerol, 0.25 µg/µl herring sperm DNA, 10 mg/ml bovine serum albumin, in a final volume of 24 µl, and the mix was incubated at room temperature for 15 min. A radiolabeled probe (1 × 10⁵ cpm) was then added and the incubation continued for another 20 min before electrophoresis on 5% native polyacrylamide gels. For the supershift experiments, 1 µg of the respective antibody was added to the lysate and incubated on ice for 1 h before the addition of the probe.

Reverse Transcription and Polymerase Chain Reaction

After stimulation with GM-CSF or IL-5, cells were centrifuged, washed, and immediately solubilized in TRI-reagent (Molecular Research Center Inc., Cincinnati, OH). RNA was isolated according to the manufacturer's instructions. RNA preparations were treated with RNase-free DNAse (Ambion Message Machine Kit) and were used for reverse transcription (RT) with oligo dT primers followed by polymerase chain reaction (PCR) with actin (5'-TCACCAACTGGGACGACATG-3' and 5'-GTACAGGGATAGCACAGCCTT-3'), glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (5'-TCAAGGT CGGAGTCAACGATTTGGT-3' and 5'-CATGTGGGCCATGAGGTCCA CCAC-3'), or CIS1 (5'-TCCTCTGCGTTCAGGGACCT-3' and 5'-ACACTAGGC-GCATCCTCCTT-3') specific primers. PCR products were separated by electrophoresis through an agarose gel and visualized with ethidium bromide. Molecular masses of expected PCR products were 0.2 kb (actin), 0.98 kb (GAPDH), and 0.57 kb (CIS1).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

STAT5 Is Phosphorylated at Tyrosine Residues after IL-5 or GM-CSF Treatment

According to the current model of cytokine-mediated JAK-STAT signaling, the phosphorylation and enzymatic activation of JAK promote the tyrosine phosphorylation of STAT factors, which is a necessary event for STAT dimerization, nuclear translocation, and DNA binding. To determine if STAT5 participates in IL-5 or GM-CSF signaling in PBEs, we examined the tyrosine phosphorylation status of STAT5 after cytokine treatment of eosinophils. Eosinophils were stimulated with IL-5 or GM-CSF for 15 min and cell lysates were immunoprecipitated either with anti-STAT5 antibody or with nonspecific rabbit IgG and immunoblotted with anti-phosphotyrosine and anti-STAT5 antibodies (Figure 1). Anti-phosphotyrosine antibodies detected a protein of molecular weight 92 to 97 kD from cytokine-stimulated eosinophils. No signal was detected in unstimulated cells or in the control immunoprecipitates where nonspecific rabbit IgG was used (Figure 1, upper panel). The same membrane reprobed with anti-STAT5 antibody demonstrated that equal amounts of target protein were present in immunoprecipitates from resting and stimulated cells (Figure 1, lower panel). The migration of STAT5 was slightly slower in stimulated cells, consistent with its phosphorylation.

Phosphorylation at tyrosine residue 694 of STAT5 is obligatory for activation of STAT5 (22). So we examined the kinetics of STAT5 activation by immunoblotting lysates of IL-5- or GM-CSF-treated eosinophils with an antibody that detects STAT5 only when activated by phosphorylation at Tyr 694 (Figure 2). A 92- to 97-kD protein band was discernibly labeled by this antiserum after 1 min and was maintained after 30 min of stimulation. These results demonstrate that STAT5 was rapidly tyrosine phosphorylated in GM-CSF- or IL-5-treated PBEs.

View larger version (23K): [in this window] [in a new window]   Figure 2.   Kinetics of STAT5 tyrosine phosphorylation in IL-5- and GM-CSF-treated eosinophils. Eosinophils were stimulated with IL-5 (300 pM), GM-CSF (1,000 pM), or control for the time indicated. Resulting whole cell lysates (50 µg) were separated by SDS-PAGE and immunoblotted with anti-phospho-Tyr (694)-STAT5 antibody (upper panel) and anti-STAT5 antibody (lower panel).

Median effective dose values of IL-5 and GM-CSF that primed the eosinophil effector mechanism are 50 and 5 pM, respectively (20). Our dose-response studies revealed that IL-5 or GM-CSF concentrations from 10 to 1,000 pM (the highest concentration tested) promoted detectable tyrosine phosphorylation of STAT5 (data not shown).

Both IL-5 and GM-CSF Induce a GAS Binding Activity in PBEs

As phosphorylated STAT5 is capable of binding specific DNA sequences, we examined eosinophil extracts for GAS binding activity by EMSA. Consistent with the rapid onset of STAT5 tyrosine phosphorylation, a dramatic induction of GAS binding activity was observed in cell extracts within a few minutes of stimulation with either GM-CSF or IL-5 (Figure 3). The binding activity remained elevated for at least 1 h, suggesting it was maintained as long as cytokines were present. The kinetics of activation and the electrophoretic migration of protein-DNA complex were indistinguishable for both cytokines, and identical results were obtained from three different donors. These data are consistent with studies suggesting that initial signaling induced by GM-CSF and IL-5 significantly overlap (23).

Characterization of the GAS Binding Factor

To demonstrate that the GAS binding complex induced by GM-CSF or IL-5 contained STAT5, supershift experiments were performed with cytoplasmic and nuclear lysates from stimulated cells. As shown in Figure 4, the inclusion of anti-STAT5 antibody in the EMSA reactions supershifted the DNA binding complex, whereas nonspecific antirabbit IgG had no effect. Similar results were obtained from five different donors with concentrations of IL-5 or GM-CSF ranging from 100 to 1,000 pM. These results show the presence of active STAT5 in the cytoplasmic and the nuclear fractions in response to GM-CSF or IL-5.

Transcriptional Upregulation of the STAT5-Dependent CIS1 Gene in Response to Stimulation

The presence of STAT5 in the nuclear fraction (Figure 4) suggested it may interact with and modulate transcription from promoters containing a GAS element. We examined the transcriptional induction of a STAT5-dependent gene, CIS1. Its promoter contains several STAT5 binding sites (24), and the protein has been implicated as a regulatory molecule in IL-3- and GM-CSF-mediated signal transduction (25). Total RNA was isolated from resting, IL-5-treated (Figure 5, upper panel ), or GM-CSF-treated (Figure 5, lower panel ) eosinophils, and subjected to RT-PCR with CIS1 gene-specific oligonucleotides. Although CIS1 messenger RNA (mRNA) was barely detectable in resting cells, a strong signal was present in the activated PBEs. Parallel PCRs with actin or GAPDH-specific oligonucleotides showed comparable levels of message in resting or stimulated cells. CIS1-specific PCR products peaked at 1 h of GM-CSF treatment and diminished by 2 h. Similar kinetics of induction of CIS1 mRNA have been observed in the GM-CSF-dependent cell line UT7 (26). These data suggest that transcriptional upregulation of CIS1 is most likely due to translocation of active STAT5 from the cytoplasm to the nucleus after cytokine activation of PBEs.

View larger version (48K): [in this window] [in a new window]   Figure 5.   GM-CSF and IL-5 induce transcriptional upregulation of CIS1. RNA was isolated from IL-5- treated (300 pM, for 1 h; upper panel) and GM-CSF- treated (250 pM; lower panel) cells, and subjected to reverse transcription followed by PCRs with CIS1-,  actin-, and GAPDH-specific oligonucleotides.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Eosinophils are important effector cells of allergy and asthma. Late-phase allergic reactions are characterized by a substantial influx of eosinophils into the circulatory system and airways. Cytokines produced from eosinophils may prolong airway inflammation by recruiting and activating additional inflammatory cells. GM-CSF- and IL-5-mediated signaling likely plays a critical role in this process (27).

GM-CSF and IL-5 bind to heterodimeric membrane receptors that contain identical  subunits. This common receptor subunit with its large cytoplasmic domain plays the major role in intracellular signaling and increases the binding affinity of the cytokine-specific  chain (23). Because neither of these subunits has discernible intrinsic enzymatic activity, signaling by the receptor is mediated by activation of a number of cytosolic protein tyrosine kinases, including the JAK family. Members of this family of tyrosine kinases are noncovalently associated with the receptor subunits (28). Ligand occupancy triggers phosphorylation and activation of JAK activity with initiation of downstream events, including the phosphorylation and activation of STAT proteins (29). Activated STATs then form homo- or heterodimers by virtue of their SH2 interaction domains, translocate to the nucleus, bind to specific target sequences identical to or closely related to the GAS element, and participate in alteration of gene expression of cognate genes (17).

Given the importance of phenotypic context in the outcome of different signaling processes, we undertook the present study of STAT5 in PBEs to provide a comprehensive examination of the involvement of STAT5 in the mediation of responses to IL-5 or GM-CSF. Studies in a variety of cell lines have shown STAT5 to be an important mediator of signaling through cytokine receptors (10, 11). Bone marrow-derived macrophages from STAT5A-deficient mice showed delayed proliferation in response to GM-CSF (30). STAT5 and related molecules also play a role in myeloid cell differentiation and/or activation, suggesting the crucial role of this factor for both gene expression and cell proliferation (31). Because the best described function of STAT factors is in transcriptional regulation, we evaluated tyrosine phosphorylation and DNA binding activity of STAT5 in stimulated PBEs. After IL-5 and GM-CSF stimulation, tyrosine-phosphorylated STAT5 was capable of binding DNA. These data strongly suggest that STAT5 can be activated by either cytokine, migrate into the nucleus, and affect gene transcription.

To confirm the relevance of eosinophil STAT5 activation to transcriptional regulation, we examined an additional indicator of STAT5-dependent signaling, namely CIS1 gene expression. A number of lines of evidence suggest that CIS1 is a STAT5-dependent gene. First, the CIS1 gene was shown to be a target of STAT5 in IL-3- and erythropoietin (EPO)-dependent hematopoietic murine cells (26). STAT5 binds to its specific recognition element in the promoter region of CIS1, which if deleted, abolished the response to EPO. Second, mutant IL-2 receptor  chains, which cannot activate STAT5, failed to induce CIS1 expression (26). Third, expression of CIS1 was markedly reduced in GM-CSF-stimulated bone marrow cells derived from STAT5 knockout mice (30). In this present study, we observed that within 1 h of stimulation with IL-5 or GM-CSF, CIS1 message was markedly enhanced in PBEs. To our knowledge, this is the first report of IL-5 inducing the transcription of CIS1. While it remains possible a non-STAT pathway may also regulate CIS1, these data strongly suggest that active nuclear STAT5 contributed to enhanced CIS1 gene transcription. Serine/threonine phosphorylation may also contribute to the transcriptional activation of STAT proteins. Inhibition of phosphorylation of these residues abrogates the transcriptional upregulation of STAT5 responsive genes irrespective of the phosphorylation status of the tyrosine residue (32). We assessed the sensitivity of cytokine-mediated CIS1 induction to 100 µM H7, a pharmocologic agent known to inhibit the serine/threonine phosphorylation of STATs. We observed that H7 inhibited CIS1 upregulation in cytokine-treated PBEs, although it did not antagonize the tyrosine phosphorylation of STAT5 (data not shown).

The function of CIS1 likely includes participation in negative feedback regulation of the JAK2-STAT5 signal transduction pathway. For example, CIS1 protein was physically associated with the IL-3 tyrosine-phosphorylated  receptor, and under these conditions, the growth of an IL-3-dependent hematopoietic cell line was slowed (25). Similarly, CIS1 overexpression blocked EPO-dependent activation of STAT5 and downstream signaling (24). These data suggest that CIS1 binds to phosphorylated sites on the  receptor, preventing STAT5 interactions and subsequent phosphorylation (24). As ubiquitinated CIS1 was stably associated with the tyrosine-phosphorylated receptor of a thrombopoietin-dependent cell line, proteosome-mediated degradation of ubiquitinated CIS1 together with the associated receptor may also downregulate cytokine-mediated signaling (33). Furthermore, the phenotype of transgenic mice in which CIS1 was overexpressed was strikingly similar to that of STAT5A and/or STAT5B knockout mice (34). Given these reports and our observation that IL-5 as well as GM-CSF induce the transcription of CIS1, it is likely that CIS1 induction may suppress both IL-5 and GM-CSF regulation of eosinophil function.

Our data demonstrate that GM-CSF and IL-5 activate STAT5, induce CIS1 transcription, and reinforce previous in vivo and in vitro findings (33) that these cytokines have considerable functional overlap. Indeed, both inhibit eosin-ophil apoptosis, participate in the recruitment of eosinophils from the periphery to the lung, and are elevated in the BAL fluid of active asthmatics (35). Thus, the molecular mechanisms underlying this similarity likely involves STAT5 activation (35) through the shared receptor  chain. Pharmacologic intervention at this point in the signaling cascade could block both GM-CSF- and IL-5-mediated signaling and eventually be a useful therapeutic for the treatment of bronchial asthma.

In summary, we have presented evidence that IL-5 and GM-CSF rapidly induce the tyrosine phosphorylation of STAT5, thereby allowing it to bind to GAS elements, and that this is associated with an increase in CIS1 gene transcription in human PBEs.