| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 282, Issue 5, 3014-3026, February 2, 2007
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
11
23
From the
Departments of Molecular and Cellular Biology and Cancer Genetics, Roswell Park Cancer Institute, Buffalo, New York 14263 and the ¶Department of Biochemistry, Christian-Albrechts-Universität zu Kiel, D-24098 Kiel, Germany
Received for publication, October 13, 2006 , and in revised form, November 16, 2006.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
subunit, but does not engage gp130, the common subunit of all other IL-6 cytokine receptors. Because the response of epithelial cells to IL-31 is unknown, the action of IL-31 was characterized in the human alveolar epithelial cell line A549 in which the expression of the ligand-binding IL-31R
subunit was increased. IL-31 initiated signaling that differed from other IL-6 cytokines by the particularly strong recruitment of the STAT3, ERK, JNK, and Akt pathways. IL-31 was highly effective in suppressing proliferation by altering expression of cell cycle proteins, including up-regulation of p27Kip1 and down-regulation of cyclin B1, CDC2, CDK6, MCM4, and retinoblastoma. A single STAT3 recruitment site (Tyr-721) in the cytoplasmic domain of IL-31R exerts a dominant function in the entire receptor complex and is critical for gene induction, morphological changes, and growth inhibition. The data suggest that inflammatory and immune reactions involving activated T-cells regulate functions of epithelial cells by IL-6 cytokines through receptor-defined signaling reactions. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
Unexplained thus far is the basis for the divergent outcome of IL-6 cytokine action that, in certain epithelial cell types, yields a suppression of proliferation (11), whereas in other cell types it yields a growth promotion (12, 13). Because the subunits for IL-6 cytokine receptors are encoded by single-copy genes, comparable structures and biochemical actions are expected for the receptor proteins expressed in the various cell types, regardless of the ultimate outcome of cytokine treatment.
An independent assessment of receptor signaling specificity among IL-6 cytokine receptors became possible with the identification of IL-31R (gp130-like protein; see Ref. 14) as a member of the IL-6R group that does not engage gp130, but OSMR, to form a signaling receptor complex (15, 16). Based on the effects of ectopic expression in transgenic mice, IL-31, a CD4+ T-cell-derived cytokine, has been associated with pruritis and dermatitis affecting dermal keratinocytes and dorsal root ganglia (17–19). Because antigenic challenge increased levels of IL-31 in a mouse model of airway hyper-responsiveness, this cytokine has also been proposed to play a role in airway hypersensitivity (15). The fact that keratinocytes and lung epithelial cells abundantly express OSMR and gp130 and respond to OSM by profound changes in gene transcription and proliferation (20–23), the question arises whether IL-31 and OSM cooperate in cell regulation. At issue is whether the signaling by IL-31 through the dimeric complex of IL-31R·OSMR is equivalent to signaling by OSM through OSMR·gp130 and whether signaling by these two receptor combinations equals signaling by the homodimer of the gp130·gp130 complex as part of stimulation by IL-6 or hyper-IL-6, a fusion protein of IL-6 and the soluble IL-6R, which stimulates a gp130 dimer independently of the membrane-bound IL-6R (24).
Transient transfection experiments with native and chimeric IL-31R constructs indicated the recruitment of those signaling pathways known to be common to IL-6 cytokine receptors, including STAT1, -3, and -5, phosphoinositide 3-kinase, and ERK (16, 25). The studies demonstrated the requirement of the cytoplasmic domain of IL-31R, as present in the full-length receptor subunit, to enable signaling and determined tyrosine residues 652 and 721 (Tyr-652 and Tyr-721) to direct activation of STAT1/STAT5 and STAT1/STAT3, respectively. Considering the limitations of transient expression systems, the precise function of the IL-31R complex beyond the immediate signaling reactions and quantitative comparison to the action by the other IL-6 cytokine receptors remains to be determined. Despite the prediction of IL-31-responsive epidermal keratinocytes and other epithelial cells (15, 17–19), the study of the receptor signaling in these cells proved to be challenging because the level of IL-31R is low relative to that of OSMR.
We hypothesize that sequence elements, which include the kinase-modified tyrosine residues within the cytoplasmic domains of the IL-6 family of receptors, define not only the recruitment of signal-transducing pathways (16, 25) but also the specificity of the cell response to the cytokines. This hypothesis was tested for IL-31R in lung epithelial cells. We identified that normal human lung epithelial cells and the transformed line A549 express low levels of functional IL-31R. To better visualize IL-31 action, expression of IL-31R was enhanced in A549 cells by stable integration of a retroviral expression vector. These cells permitted the functional analysis of the receptor combinations IL-31R·OSMR, OSMR·gp130, and gp130·gp130 and indicated a striking specificity of the signaling by IL-31 as compared with OSM and hyper-IL-6.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
; ATCC) were cultured in serum-free hormonally defined keratinocyte medium (Invitrogen). COS-1 cells were used to express recombinant human cytokines and Hep3B cells (26) for functional reconstitution of the IL-31R complex by transient transfection of receptor subunits.
Cytokine and Cytokine Receptor Constructs—The protein-coding region for human IL-31 and the long form (745 amino acids) IL-31R was identified in the genome data bank based on the published sequence (14, 15). The IL-31 cDNA (encoding amino acid residues 24–164, NCBI accession number NP001014358.1) was generated by reverse transcription-PCR using RNA from mitogen-activated human peripheral T-cells. Because the first exon could not be unambiguously identified in the human genomic data base, the cDNA encoding the signal peptide of human IL-6 (amino acid residues 1–29) was added to the 5' end of the cloned IL-31 cDNA under inclusion of an extra triplet encoding glycine. The resulting construct was inserted into the pcDNA3.1 vector (Invitrogen) and verified by sequencing. The IL-31R cDNA was amplified from RNA extracted from human glioma cell line U87 and cloned into pCMV-Tag 4A (Stratagene, La Jolla, CA) vector between the HindIII and XhoI sites upstream and in-frame with the DYKDDDDK (FLAG) sequence. The IL-31R-FLAG sequence was transferred into pLHCX retroviral delivery and expression vector (Clontech). The three tyrosine residues in the intracellular domain, Tyr-652, Tyr-683, and Tyr-721, were mutated to phenylalanine (YF), singly and in combinations, by QuickChange® site-directed mutagenesis kit (Stratagene, La Jolla, CA). IL-31R forms with the mutations of both Tyr-652 and Tyr-683, or Tyr-652 and Tyr-721 are labeled Y652F, Y683F and Y652F, Y721F, respectively. IL-31R with all three tyrosine mutations is labeled 3YF. Retroviruses were generated by a 2-day transient transfection procedure in PT67 packaging cells. The chimeric receptor GM-CSFR-IL-31R was constructed by inserting the EcoRI-NotI fragment encoding the IL-31R (amino acid residues 515–745) into the EcoRI-NotI restriction-digested pLNCX2-GM-CSFR, encoding the extracellular domain of GM-CSFR (amino acid residues 1–437). The chimeric receptor GM-CSFR-OSMR was constructed by inserting the PCR-generated EcoRI-NotI fragment encoding the OSMR (amino acid residues 723–979) into the EcoRI-NotI restriction-digested pSVSPORT-GM-CSFR, encoding the extracellular domain of GM-CSFR (amino acid residues 1–315) (27). The chimeric receptor cDNA was excised using HindIII and PmeI restriction enzymes, a NotI site added 3' to the PmeI site, and re-cloned into the HindIII and NotI sites of the retroviral pLPCX vector. The tyrosine residues 917 and 945 in the cytoplasmic domain of OSMR were mutated to phenylalanine and verified by sequencing. Expression vectors for human OSMR, OSM (Amgen Corp., Seattle (28)), hyper-IL-6 (24), IL-6 and LIF (Wyeth Pharmaceuticals, Cambridge, MA), IL-22 and IL-24 (Origene, Rockville, MD), and the CAT-reporter gene construct pCRP (219)-CAT (29) were applied as described (5).
Transfection and Transduction of Cells—The production of IL-6, IL-31, OSM, hyper-IL-6, IL-22, and IL-24 for cell treatments was carried out as follows. COS-1 cells in 10-cm diameter dishes (3 x 106 cells) were transfected with 4 µg of cytokine expression vector and 0.2 µg of pEGFP using Mirus TransIT-LT1 transfection reagent (Mirus Bio Corp., Madison, WI). After 24 h, the cultures with >50% transfected (GFP+) cells were changed to fresh Dulbecco's modified Eagle's medium containing 10% fetal calf serum and 4.5 mg/ml glucose. The presence of 10% fetal calf serum was important to preserve the bioactivity of the synthesized cytokines. After 72 h of incubation, during which the cultures reached 100% confluence, the conditioned medium was collected and filtered through a 0.2-µm membrane. The cytokine concentrations were determined by Luminex multiplex assay (30, 31) using commercial antibodies and recombinant cytokines as standards (R&D Systems, Minneapolis, MN). The bioactivity of every cytokine preparation was independently determined by an in-house bioassay in H-35 cells (32), which were stably transduced with human OSMR (5) and human IL-31R. These cells have no measurable production of endogenous IL-6 cytokines and permit the measurement of the bioactivity of all human IL-6 cytokines by the dose-dependent induction of type II acute phase proteins, such as fibrinogen and thiostatin (5, 33). Half-maximal expression of the plasma proteins defines one hepatocyte-stimulating factor unit and is obtained with 
1 ng/ml of IL-6 cytokines. The accuracy of the bioassay was confirmed by using purified recombinant IL-6 cytokines (R&D Systems). Generally, the cytokine concentrations in the COS-1 media ranged from 10 to 50 µg/ml. From these stocks, an initial dilution to 100 ng/ml cytokines was prepared that has been demonstrated to produce in all cell types optimal signaling reactions and cell responses. These preparations, along with further 10-fold serial dilutions, were applied to the experimental cell cultures (see "Results"). Conditioned medium from COS-1 cells transfected with control vector served as background reference (contained <0.1 ng/ml IL-6 equivalent bioactivity). This medium was diluted like the cytokine stocks and was included in the experimental applications as a no-cytokine control.
Receptor expression vectors along with the reporter gene constructs pCRP(219)CAT and pEGFP (Invitrogen) were transfected into Hep3B cells, and cytokine-regulated CAT activity was determined and normalized as described (5). A549 cells were transduced with receptor-encoding retroviruses, cloned by limited dilution within 4 days following transduction, and selected in hygromycin (1 mg/ml). Positive clones were identified by the expression of a FLAG epitope. All selected clones (at least three independently derived clones per receptor construct) were subjected to a second round of subcloning. Homogeneity of the selected clonal lines was verified by immunostaining for the FLAG epitope.
Cell Treatments—The cell culture conditions and duration of cytokine treatments depended on the type of analysis. Generally, cells were plated into 6-well or 24-well culture plates. When the cultures reached a density of 50–100% (100% = 2 x 105 cells/cm2 culture area), the cells were incubated first in serum-free medium for 2 h and then changed to medium containing cytokines (0–100 ng/ml). For determining receptor-mediated signaling, treatments were limited to 15 min; for measuring induction of genes, altered cell morphology, or change in cell cycle markers, the treatments ranged from 24 to 48 h. To determine cell proliferation, the cells were seeded at 1% density into 24-well culture plates. After 24 h, the medium was replaced by fresh, serum-containing medium that also included 10-fold serially diluted cytokines or dexamethasone (34). After 3 days, the culture medium was replaced, and another 3 days later, the number of viable cells in each well was counted in a hematocytometer. Values from quadruplicate cultures were expressed as mean ± S.D. Incorporation of [3H]thymidine was determined as described previously (22). Morphology of the cell cultures was recorded by inverted phase microscopy at x10 magnification (TE2000; Nikon, Melville, NY).
Immunoblot Analysis—Cells were washed with phosphate-buffered saline and lysed within the culture plate (10 µl/1 x 105 cells) with RIPA buffer (50 mM Tris-HCl (pH 7.4), 1% Nonidet P-40, 0.25% sodium deoxycholate, 1 mM NaF, 1 mM sodium orthovanadate, 1 µg of leupeptin, 1 µg of aprotinin/ml, 1 mM phenylmethylsulfonyl fluoride, 10% glycerol). Replicate aliquots of lysates containing 1–50 µg of proteins were separated on 6–10% SDS-polyacrylamide gels (Bio-Rad) with series of 15 samples per gel. Proteins were transferred to nitrocellulose membrane (Schleicher & Schuell), and equal loading and transfer were confirmed in all cases by staining with Ponceau S red. Subsections of the membrane containing the proteins of interest were reacted with antibodies. The following phospho-specific antibodies were used: phospho-STAT1-Tyr-701 (9171), phospho-STAT3-Tyr-705 (9131), phospho-STAT5-Tyr-694 (9351), phospho-p44/42 MAPK-Thr-202/Tyr-204 (phospho-ERK1/2; 9106), phospho-SAPK/JNK-Thr-183/Tyr-185 (9251), phospho-Akt-Ser-473 (9271), or phospho-Hsp27-Ser-82 (2401) (Cell Signaling Technology Inc., Beverly, MA). The antibodies against total proteins included the following: STAT3 (C-20), ERK (K-23), JUNB (C-11), p27Kip1 (C-19), p53 (DO-1), cyclin B1 (H-433), cyclin D1 (M-20), cyclin E (HE-12), CDK2 (M-2), CDK4 (sc-260), CDK6 (C21), CDC2 (sc-54), MCM2 (H-126), MCM4 (H-300), MCL-1 (S-19), and SOCS3 (M-20) from Santa Cruz Biotechnology (Santa Cruz, CA). The antibody against FLAG (026K4848) was obtained from Sigma; antibody against Rb (554136) was from Pharmingen; antibody against human IL-31R was from R&D Systems; and antibody against human 1-antichymotrypsin, 1-antitrypsin, and fibrinogen were from Dako Immunoglobins, Glostrup, Denmark. The membranes were incubated with the appropriate peroxidase-conjugated secondary antibodies (ICN Biomedical, Aurora, OH), and the antibody binding was visualized by enhanced chemiluminescence reaction (Pierce). In each experimental series, immunoblots were exposed to x-ray films for various lengths of time (1 s to 2 h) to obtain images that are in the linear range of signal detection (22). For pictorial presentation of immunoblot data, the signals for either total STAT3 or ERK1/2 served as loading controls. Immunoblots were digitalized and quantified with ImageQuant TL software (Amersham Biosciences). The net pixel value for each protein band that lies within the linear range of detection was compared with co-analyzed control cultures.
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
. A survey of various glioma cell lines indicated that IL-31R mRNA and protein level and response to IL-31 treatment varied considerably among lines. When tested with serially diluted cytokines, one clonal derivative of the Roswell Park glioma line, ROS (ROS-9), was identified that showed the highest IL-31 responsiveness, which, based on phosphorylation of signaling proteins at the optimal treatment dose of 100 ng/ml cytokine, was approximately half of OSM (Fig. 1A). IL-31 response of glioma cells was characterized by high level phosphorylation of STAT3 and ERK and relatively low level phosphorylation of STAT1. OSM produced a similarly high phosphorylation of STAT3 and ERK but also a higher relative activation of STAT1. LIF and IL-6 had a minor stimulatory activity on all glioma lines due to low LIFR and IL-6R expression. The function of gp130 per se became evident by treatment with hyper-IL-6 that yielded the highest relative activation of STAT1. Some of the signaling functions of IL-6 cytokines were comparable to the effect of factors known to be highly effective in activating ERK (by epidermal growth factor) and STAT1 (by interferon-
). Further analysis of the IL-31 response indicated a dose- and time-dependent transcriptional induction of known IL-6 target genes, such as JUNB and SOCS3 (data not shown). However, most likely because of the highly transformed phenotype of the glioma cells, no measurable consequences of any IL-6 cytokine treatments were evident on proliferation or cell morphology. Hence, to identify a cell line suitable to study the full range of proposed cellular action of IL-31/OSM, the survey for IL-31-sensitive cells was extended to the other cell types that had not been characterized previously (15, 16). For grading the IL-31 responsiveness among the cell lines, IL-31-stimulated STAT3 phosphorylation in ROS-9 cells served as reference.
Normal primary, immortalized, and transformed cultures of various human cell types were tested for expression of IL-31R mRNA by reverse transcription-PCR and regulatory effects of IL-31 by STAT activation (details will be described elsewhere). No appreciable IL-31 response (<1% of ROS-9) was confirmed for hepatic, lymphoid, nonactivated myeloid, melanoma, fibroblastic, endothelial, and neuroblastoma cells. Low IL-31-dependent signaling (1–5% of ROS-9) was found in primary (Fig. 1B) and immortalized dermal keratinocytes. The IL-31 response of epithelial cells from prostate and lung ranged from <1 to 16% of ROS-9. Among lung cells, primary cultures of normal alveolar and bronchial epithelial cells, although displaying a highly consistent response pattern to OSM, IL-6, and hyper-IL-6 (Fig. 1C) (22), showed an IL-31 response that varied substantially from donor to donor. Functional analyses of more that 10 independent preparations of either epithelial cell type indicated that in half of all cases the IL-31-stimulated STAT3 phosphorylation was barely detectable (<1% of ROS-9), whereas it could range from 1 to 10% of ROS-9 in the other half (Fig. 1C). Although in cells with the highest IL-31 response, the level of STAT3 activation was still minor when compared with that of OSM, it was sufficient to mediate a low but detectable induction of type 2 acute phase genes, such 1-antichymotrypsin and l-antitrypsin (Fig. 1, D and E). The variable expression of IL-31R in primary epithelial cells may relate to the health status of the donor's lung, and the underlying mechanism is under investigation.
|
hyper-IL-6 > IL-6 > IL-22> IL-31 > LIF > IL-24. The immune detectable IL-31R protein in A549 cells was 1% of ROS-9 cells (Fig. 1G). The fact that normal and some of the transformed lung epithelial cell lines, including A549, respond to OSM by a strong inhibition of proliferation, morphological changes, and prominent transcriptional induction of several acute phase proteins, including serpins and fibrinogen (Figs. 4A and 5) (21–23, 35–38), we chose A549 cells as an experimental model to define the signaling specificity of IL-31. To do so, the IL-31 responsiveness of A549 cells needed to be increased to achieve a more prominent manifestation of the receptor signals and cellular reactions.
IL-31R-transduced A549 Cells as Model to Define IL-31 Action—IL-31R cDNA was cloned from U87 human glioma cells, and a FLAG epitope was added to the 3' end of the open reading frame and inserted into a retroviral expression vector. To confirm the function of the cloned receptor subunit, Hep3B cells were transfected with the IL-31R expression vector together with a vector encoding human OSMR and the OSM-responsive reporter construct, pCRP (219)-CAT. Hep3B cells do not express OSMR, LIFR, and IL-31R, but they have an effective endogenous IL-6R/gp130 system (28). Western blotting of extracts from receptor-transfected cells indicated the expression of IL-31R protein (data not shown).
Cytokine-regulated expression of the reporter gene construct indicated, by a gain-of-function approach, a subunit-specific reconstitution of receptor action. Transfected IL-31R and OSMR yielded an IL-31 response and OSMR and endogenous gp130 yielded an OSM response (Fig. 2A). The dependence of OSMR, but not of IL-31R, on endogenous gp130 was evident by the sensitivity of the cytokine treatment to inhibition by pan-neutralizing gp130 antibodies (Fig. 2B) (16). The regulation of the CAT reporter construct also indicated differences in the gene regulatory signals derived from transfected OSMR and IL-31R and from endogenous IL-6R and gp130. Maximal IL-31 response consistently exceeded that of OSM by about 2-fold. Although regulatory processes, which rely on transfected receptors, have merit in identifying qualitative reactions (5, 16, 25), the heterogeneous receptor protein expression in the transfected cell cultures limits the interpretation of quantitative differences. Therefore, the quantitative assessment of cytokine action and comparison to endogenous IL-6 cytokine receptors demanded a cell system with uniform and stable expression of IL-31R.
A549 cells were transduced with a retroviral expression system for IL-31R-FLAG. A clonal line (termed A549-IL-31R-wt) was selected that expressed IL-31R at an 10-fold higher level than ROS-9 cells (Fig. 1C). Of note is that IL-31R synthesized by glioma cells, such as U87 or ROS-9 cells, migrated on an SDS gel with a 120-kDa apparent molecular mass, whereas the same protein made by A549 cells was 6 kDa smaller. This size difference suggested a cell type-specific modification of the receptor protein. The predominant plasma membrane localization of IL-31R was verified by immunofluorescent confocal microscopy (Fig. 2C, panel b).
This cell line offered the opportunity to compare the signaling function of three distinct compositions of IL-6 cytokine receptor subunits, IL-31R·OSMR, OSMR·gp130, and gp130·gp130. The signaling specificity by each receptor form was defined by the profile of representative reactions mediated by the cytokine treatments. Three temporally sequential but interrelated regulatory categories are chosen to define receptor action as follows: (i) immediate activation of signal-transducing pathways; (ii) induced expression of differentiation genes and changed cell morphology; and (iii) altered expression of proteins associated with cell cycle and replication and suppression of cell proliferation.
Activation of STAT and MAPK Pathways Indicates IL-6 Receptor-specific Signal-transducing Reactions—First we confirmed the cytokine dose dependence of receptor action (Fig. 3A). As already demonstrated for OSMR and gp130 in other cell systems (5, 39), maximal initiation of signaling through STATs and MAPKs to the induction gene (i.e. fibrinogen) was achieved with 10 ng/ml (4 nM) of agonist. Half-maximal phosphorylation occurred at a concentration of 1 ng/ml. The characterization of the cell response was thus performed at the cytokine concentration of 100 ng/ml to ensure a receptor action that was not limited by the cytokines. Time course of cytokine treatments revealed the following consistent signaling features (Fig. 3B; confirmed in three or more independent experimental series).
The specificity of the IL-31R/OSMR/gp130 signaling in A549 cells showed traits that were already detected in ROS-9 cells (Fig. 1A). IL-31 has a preference to activate STAT3 > STAT5 > STAT1. In contrast, OSM produced maximal phosphorylation of STAT5 and hyper-IL-6 of STAT1. As evident from the phosphorylated ERK and JNK, both IL-31 and OSM activated MAPK pathways to a higher level than hyper-IL-6. The activation of the JNK pathway led to a proportionally higher activation of downstream targets such as Akt (Fig. 3, A and B) and HSP27 (Fig. 3A and Fig. 6, below) and eventually induction of JUNB expression (Fig. 3B).
|
to an 30% level by 6 h of treatment and a return to pretreatment levels by 24 h (Fig. 3B).
A change in cell morphology was a visual consequence of long term treatment of A549 cells with IL-6 cytokines. This change, which occurred over a 24–48-h time period, entailed a reduction of cell-cell interaction but maintained focal adhesions whereby the spread cell shape with prominent lamellipodia altered to a retracted morphology with podosomal extension and diffracting appearance in phase microscopy (Fig. 2D). The efficacy to induce this morphological change follows the order IL-31 > OSM > hyper-IL-6. The IL-6 cytokine-induced change differed, however, from that mediated by transforming growth factor-, which was characterized by the prominent stress fiber formation, enlarged and elongated morphology, and enhanced number of focal adhesion sites (Fig. 2D) (41).
|
-transduced A549 cells indicated that IL-31 strongly suppressed proliferation with an IC50 of 0.2 ng/ml. OSM was less potent (IC50 of 1 ng/ml); hyper-IL-6 did not reduce proliferation below 50%, and IL-6 was ineffective. The reduced cell proliferation correlated in part with lower DNA synthesis (Fig. 4C). An accumulation of cells in G1 state, but without an appreciable increase of apoptosis, was confirmed by flow cytometry (data not shown). The same cytokines were also effective in dose-dependent induction of STAT3-responsive genes, including fibrinogen (Fig. 4A). Of note is that IL-31 displayed a dose-dependent stimulation of fibrinogen expression that differed from suppression of proliferation and also from OSM action (Figs. 3A and 4A). IL-31 was 10 times less effective than OSM in stimulating fibrinogen expression (half-maximal stimulation at 20 ng/ml) and only at high concentrations produced a maximal level of fibrinogen that was equal to that of OSM. To identify potential mechanisms leading to the cytokine-mediated inhibition of proliferation, the changes in the level of cell cycle-controlling proteins were determined (Fig. 5). Six-hour cytokine treatment increased the expression of the CDK inhibitor p27Kip1. The elevated level was maintained for at least 48 h and was proportional to the level of inhibited proliferation. However, p27 induction alone appeared insufficient to account for cell cycle arrest because dexamethasone did not alter the p27 level. Temporally delayed relative to p27 induction was the cytokine- and dexamethasone-mediated reduction of cyclin B1, CDC2, CDK2, and RB, with a maximal reduction reached by 24–48 h of treatment. Loss of MCM4 and dephosporylation of MCM2 suggested impairment of DNA replication and arrest at G0, respectively (42–44). The 2–3-fold induced expression of the anti-apoptotic MCL1, together with the cytokine-specific reduction of p53, could explain in part the survival of the treated cell cultures. Also consistently observed was a slow 2-fold increase in STAT3 (but not STAT1 or STAT5; not shown) in cytokine- and dexamethasone-treated cells (Fig. 5, bottom). Maximal level was attained after 48–72 h of treatment. The cytokine-restricted induction of SOCS3 was maximal within 1 h and remained elevated for at least 48 h. The time course of SOCS3 expression did not coincide with down-regulation of STAT3 phosphorylation suggesting that SOCS3 was not a major component in the negative feedback regulation of cytokine receptor signaling (45, 46). An additional marker of cytokine-enhanced gene regulation in long term treated A549 cells was fibrinogen (Fig. 5), which reached maximal levels by 24 h of treatment. The magnitude of fibrinogen induction was proportional to the receptor-activated STAT3.
|
Dominant Role of Tyrosine Residues in IL-31R Signaling—Signaling by the heteromeric IL-31·OSMR receptor complex is controlled by distinct cytoplasmic domains. These domains, through separate tyrosine-based elements, determine the engagement of signal-transducing pathways. The common OSMR subunit contributes a cytoplasmic domain with nine tyrosine residues of which Tyr-917 and Tyr-945 recruit STAT1 and -3, and Tyr-861 recruits SHC to link to the MAPK pathway (47). The cytoplasmic domain of IL-31R contains three tyrosine residues (Tyr-652, Tyr-683, and Tyr-721) of which the first and the last are suggested to recruit STAT1 and -5 and STAT1 and -3, respectively (16, 25). To determine the functional contribution of the three tyrosine residues in IL-31R to the overall signaling function by the IL-31 receptor (paired with the endogenous wild-type OSMR), the tyrosine residues were mutated to phenylalanine (YF), either individually or in combinations. Each mutant form was transduced into A549, and clonal lines were generated. Based on quantification by immunoblotting, IL-31R expression in these lines varied within a 5-fold range (see Figs. 7E and 8B, below), but even the line with the lowest level (Y652F, Y721F) still exceeded the ROS-9 standard by 2-fold. Each receptor form properly localized to the plasma membrane (e.g. Fig. 2C, panels c and d). The effects of the mutations on receptor functions were determined by profiling the signaling reactions (Fig. 6), STAT3-regulated genes and proliferation (Fig. 7), and altered expression of cell cycle proteins (Fig. 8). For each analysis, the response to OSM served as internal reference.
The mutation-dependent changes in STAT signaling by IL-31 indicated that each of the tyrosine residues of IL-31R contributed to the STAT1 and STAT5 signaling, yet even with all tyrosine residues mutated (3YF), 50% of STAT1 and STAT5 activation is still retained. STAT3 signaling was strongly attenuated by the Y721F mutation, the predicted STAT3 interaction site (Fig. 7C). With all three tyrosine residues mutated (3YF), STAT3 activation reached its lowest level, with 20% retained. This signaling activity was attributed to the remnant function of the transduced mutant receptor and the low level of wild-type IL-31R normally expressed by A459 cells (Fig. 1, F and G). Individually, none of the IL-31R mutations appreciably affected the regulation of MAPK, resulting in a consistently strong ERK and JNK phosphorylation.
|
forms indicated the relevance of the regulated signal pathways in directing the downstream cellular responses. The mutation of individual tyrosine residues in IL-31R lowered the transcriptional regulation of the CRP promoter in transfected Hep3B cells (Fig. 7A) and fibrinogen gene expression in stably transduced A549 cells (Figs. 7B and Fig. 8B). The Y721F had a particularly prominent impact in that it not only strongly attenuated activation of STAT3 (Figs. 6 and Fig. 7C) but essentially eliminated induction of CRP and fibrinogen expression (Fig. 7, A and B). The mutations Y652F and Y683F only marginally impaired STAT3 activation, yet these produced an 2-fold reduced expression of CRP and fibrinogen. The prominent effect of the Y721F mutation in IL-31R in combination with the wild-type OSMR suggested that the cytoplasmic domain of IL-31R exerted a dominant function on both STAT3 activation and gene regulation. This conclusion was supported by the finding that reciprocal combination of receptor mutants, wild-type IL-31R and OSMR with the double mutation Y917F, Y945F (elimination of both STAT3 recruitment sites), when tested as part of the chimeric GM-CSFR/GM-CSFR constructs, still retained half-maximal CRP-inducing activity (Fig. 7A, right side lane).
The ability of IL-31R mutants to alter cell proliferation indicated that the most critical contribution was provided by Tyr-721 despite the presence of all OSMR-dependent regulatory activities (Fig. 7D). All IL-31R forms that included Y721F lost the IL-31-dependent suppression of the proliferation. Moreover, none on the A549 cell lines expressing mutant IL-31R displayed an IL-31-stimulated proliferation that exceeded the proliferation of the control culture, indicating that IL-31R did not contain a mutation-sensitive negative regulatory element.
Like the suppression of proliferation, the contraction of cell shape induced by IL-31 was also primarily dependent on the Tyr-721-derived signal (Fig. 8A). Cells expressing IL-31R containing the Y721F mutation largely maintained the morphology of untreated cells.
The proliferative response of A549 to IL-31 through the transduced IL-31R forms permitted a definition of the accompanying regulatory components. By profiling the expression pattern of those cell cycle-associated proteins that were found to be affected by cytokine treatment (Fig. 5), we could delineate the effects determined by individual receptor signaling sites (Fig. 8B). The characteristic regulatory features for IL-31R, including increase of p27Kip1 and reduction of cyclin B1, CDC2, CDK2, CDK6, MCM4, RB, and p53, were all greatly diminished by the Y721F mutation. The same mutation had a major effect on induction of fibrinogen (Fig. 7B). The gene-inducing action reflected more the quantitative activation of STAT3 by IL-31, whereas the altered proliferation and morphology demonstrate the threshold of signaling effect.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
subunit that, in part, specifies STAT3 activation by the entire receptor complex. The study also defines differences in signal-transducing activities among various IL-6 cytokine receptor complexes indicating the functional relevance of cytoplasmic domains of the contributing subunits. The IL-31-mediated activities are expected to be effective at sites in the lung and skin tissue where inflammatory processes occur, which include activated T-cells, the producers of IL-31 and OSM. This study concerns two issues regarding IL-6 cytokine receptors. The first issue is the identification of those signaling functions that specify the action of the complementary receptor pair for IL-31 and OSM. The second issue is the potential biological relevance of the identified receptor-specific functions.
The first goal of this study was to establish the signaling specificity of IL-31 and OSM, two cytokines predicted to be relevant for regulating epithelial cells (8, 15, 22, 23, 35). Assessment of signaling specificity was in part prompted by the prediction that the receptors for these two cytokines, like the other members of the IL-6 cytokine receptors, exert redundant functions due to shared receptor subunits (25). In principle, receptor functions are determined at two levels. The first level is the magnitude of signaling that is dependent on the amount of receptor proteins and downstream signaling components. The second level is the specificity by which the subunits engage signaling reactions. Based on shared usage of OSMR and knowing that all cells express gp130, one predicts that every cell type that is responsive to IL-31 is also responsive to OSM and to trans-signaling, i.e. responsive to hyper-IL-6.
|
relative to OSMR in keratinocytic and epithelial cells (Fig. 1), the cell types that were predicted to be relevant IL-31 targets (15). Inasmuch as glioma cells have appreciable expression of IL-31R (Fig. 1C) and hence are suitable references for IL-31R signaling reactions, IL-31-controlled responses of these cells amenable to experimental analysis were limited. Therefore, the IL-6 cytokine responsiveness of A549 cells was used to define IL-31R action. To better visualize the IL-31-stimulated function, IL-31R expression in these cells was enhanced to a level exceeding that of OSMR. The OSM response of A549 cells, like that of other cell types (48), is limited by the expression of OSMR and not by gp130. Based on the model that OSMR and IL-31R function as a heterodimer, the limiting OSMR subunit determines maximal signaling by both OSM and IL-31, even in cells with excess IL-31R, as in IL-31R-transduced A549 cells. Qualitative differences in signaling are then attributed to the action of the OSMR partners, IL-31R or gp130. Signaling by gp130, through homodimerization by hyper-IL-6, is expected to be of the order of OSMR, because a comparable number of functional receptor complexes are mobilized (Fig. 1, A–C and F). Quantitative differences in signaling among the receptor complexes, in particular for STATs, are predicted based on the number of STAT recruitment sites contributed by the individual subunits (eight in gp130·gp130, six in OSMR·gp130, and four in IL-31R·OSMR).
The results of the cell analyses demonstrate that a simple arithmetic prediction of STAT signaling did not apply (Figs. 3 and 5). The differences in the actually observed signaling reactions, which are manifested in the relative levels of the phosphorylated STATs, are interpreted to represent subunit- as well as site-specific efficacies in mediating STAT activation (16, 25). This action, however, is in part dependent on the context in which the receptor subunits operate. We conclude that, while all receptor subunits engage STAT3, IL-31R is most effective, and the site Tyr-721 exerts a dominant action. In contrast, OSMR, when combined with gp130, provides the highest STAT5, and gp130 provides the highest STAT1 activation. The signaling in epithelial cells through the JAK/STAT pathway is similar to that determined in glioma cells and primary lung epithelial cells (Figs. 1 and 3), suggesting that the transduced expression of IL-31R did not introduce an altered receptor specificity. The time course of STAT signaling is comparable among the cytokine receptors with maximal effects on STAT1 and STAT5 by 15 min and a return close to basal levels by 1 h and maintained elevated levels of activated STAT3 for days. The similar kinetics suggests that there is no appreciable difference in regulatory feedback mechanisms among the receptor systems that include receptor turnover (40) and kinase regulation through SOCS (45, 46).
|
(Tyr-861) were determined to link to MAPK pathways (47, 49, 50). The combination of OSMR and gp130 has been recognized previously to be particularly effective in recruitment of MAPK pathways that activate ERK and JNK (5). This study indicated that IL-31R is even more effective than OSMR. Because the mutational analysis of the tyrosine residues failed to identify a functional element within IL-31R necessary for the IL-31-specfic MAPK activation (Fig. 6) (16, 25), we suggest that the link to MAPK is not through a tyrosine-containing motif but may be mediated by receptor-activated kinases without involvement of additional receptor elements. The mutational analysis also ruled out a tyrosine site within the cytoplasmic IL-31R domain that will recruit a protein-tyrosine phosphatase and thereby attenuate receptor signaling as found for other hematopoietic cytokine receptors (49–53). The second goal of the study was to assess the biological response of epithelial cells to IL-31 and to compare this to OSM. The most prominent cellular response to IL-31, as already determined for OSM, is the STAT3-dependent induction of genes. This property has been used to assess the function of cloned receptor subunits in transiently transfected cells by co-introducing STAT3-responsive reporter gene constructs (Fig. 2) (5, 25). More precise qualitative and quantitative assessment of regulated expression of endogenous genes was possible in receptor-transduced cells. We demonstrate the induction of fibrinogen, JUNB and SOCS3 (Figs. 3 and 5). Induction was generally proportional to the cytokine-stimulated level of STAT3. This fact became relevant not only for the characterization of the tyrosine mutant receptor forms (Fig. 8) but also when re-assessing the IL-31 response in parental A549 and normal lung epithelial cells. In those cells, a low level of cytokine-dependent phosphorylation of STAT3 was detectable (Fig. 1C) and a correspondingly reduced induction of the prominent STAT3 target genes was measured (Fig. 1, D and E).
IL-6 cytokines are part of homeostatic processes, including inflammation and tissue repair, and generally act on differentiated cells, i.e. epithelial cells in the lung. IL-6 is one of the abundant paracrine factors produced by macrophages (22), dendritic cells, mast cells, and granulocytes and which initiate the local inflammatory process. Moreover, it is predicted that soluble IL-6R is released by leukocytes and stromal cells within the tissue undergoing an inflammatory reaction (54). This soluble receptor, in complex with IL-6, promotes trans-signaling (55). Whereas lung epithelial cells are responsive to IL-6 (Fig. 1, C and F) (22), the maximal response is only one-third that of OSM. A more prominent action is predicted for IL-6 trans-signaling, provided the local formation of sIL-6R·IL-6 complexes reaches effective concentrations. That indeed trans-signaling, under condition of maximal stimulation, can exceed the action of IL-6 alone is demonstrated by the effects of hyper-IL-6 (Figs. 1, A–C and F, Fig. 2, A and B, and Fig. 4B).
OSM production by macrophages is temporally delayed relative to that of pro-inflammatory cytokines (56). Additional contribution of local OSM, as well as delivery of IL-31, is attributed to tissue-immigrating and activated CD4+ and CD8+ T-cells (15, 57). Thus, OSM and IL-31 are predicted to act on target cells at the later stage of the inflammatory process. Under these conditions, the cytokine-regulated expression of genes in epithelial cells appears to be the most relevant consequence. The biological functions of these genes include proteases, protease inhibitors, mucins, surfactants, and extracellular matrix proteins. Because most epithelial cells are terminally differentiated, proliferation control by OSM and IL-31 in those cells is not an issue.
However, in cases of lung tissue injury, the activation of the inflammatory response is invariably coupled to tissue repair with emphasis to reestablish epithelialization. This process, as part of the wound healing, demands mobilization, spreading, proliferation, and differentiation of epithelial cells that involves recruitment from the local progenitor population (58). Proliferative signals are attributed the paracrine growth factors derived from leukocytes and stroma cells and include members of the epidermal growth factor, hepatocyte growth factor, fibroblast growth factor, and insulin-like growth factor families. Based on our findings, we predict that the delivery of IL-6 cytokines, in particular OSM and IL-31, in the context of other inflammatory mediators will attenuate proliferation of epithelial cells that are involved in regeneration and may also promote differentiation of these by prompting G0 arrest. The mode of cytokine action appears to involve stimulated loss of cell cycle proteins (Figs. 5 and 8). Because this growth control of lung epithelial cells is considered to be protective, an apoptotic outcome of the cytokine action would be counterproductive.
The proliferative arrest correlates with the activation of STAT3 in IL-31R-transduced cells (Fig. 7). Although activated STAT3 has been suggested to promote proliferation in certain cell types (12, 13, 59), in epithelial cells it appears to favor suppression (11, 22, 60). The growth-suppressing activity was observed in our cell system by maintaining normal cellular level of wild-type STAT3 but eliminating STAT3 activation by the IL-31R. A comparable approach had been used by us to determine the role of STAT activated by gp130 for growth regulation in hepatoma cells (61). By deleting all STAT recruitment sites in gp130 and thus loss of STAT3 activation, growth suppression by gp130 signaling was lost. Our approach to test IL-6 cytokine-regulated STAT3 activity on proliferation differs from approaches used by others who either created an altered cellular STAT content by introducing STAT3 deficiency (62) or overexpressing constitutively active (63) or dominant negative STAT3 (64) or by using a JAK inhibitor (13). By altering the composition or the structure of STAT3, deviations from the normal STAT3-dependent signaling mechanisms may occur. The application of kinase inhibitors inevitably affects a broader range of receptor-dependent reactions than STAT activation. While STAT3 dependence of p27Kip1 induction (65, 66) can explain the lack of p27 increase in IL-31-treated A549 cells that express Y721F IL-31R, the biochemical link from STAT3 to the down-regulated cell cycle proteins remains to be elucidated.
|
FOOTNOTES |
|---|
1 Both authors contributed equally to this work.
2 Present address: Centre Hospitalier de l'Université de Montréal Research Centre, Notre Dame Hospital, 1560 Sherbrooke East, Montreal, Quebec H2L 4M1, Canada.
3 To whom correspondence should be addressed: Dept. of Molecular and Cellular Biology, Elm and Carlton St., Buffalo, NY 14263. Tel.: 716-845-4587; Fax: 716-845-5908; E-mail: heinz.baumann{at}roswellpark.org.
4 The abbreviations used are: IL, interleukin; CAT, chloramphenicol acetyltransferase; CRP, C-reactive protein; ERK, extracellular signal-regulated kinase; JAK, Janus kinase; JNK, Jun N-terminal kinase; GM-CSF, granulocyte macrophage-colony stimulating factor; LIF, leukemia inhibitory factor; OSM, oncostatin M; OSMR, oncostatin M receptor; SOCS, suppressor of cytokine signaling; STAT, signal transducer and activator of transcription; RB, retinoblastoma; MAPK, mitogen-activated protein kinase.
|
ACKNOWLEDGMENTS |
|---|
, gp130, and GM-CSFR; Drs. John Cowell and Robert A. Fenstermaker for glioma cell lines; Dr. Satrajit Sinha for cultures of primary human keratinocytes; Dr. Lesleyann Hawthorn for communicating Affymetrix data on lung analyses; Dr. Andrei Bakin for providing transforming growth factor-1 and advice on cytoskeletal structure; Dr. William Burhans for immunoreagents for cell cycle proteins and advice on cell cycle control; and Edward L. Hurley for carrying out confocal microscopy. |
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
