Journal of Leukocyte Biology eBioscience full spectrum cell analysis
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published online as doi:10.1189/jlb.0706465 on January 16, 2007

Published online before print January 16, 2007
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
jlb.0706465v1
81/4/1137    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Martinez-Moczygemba, M.
Right arrow Articles by Lei, J. T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Martinez-Moczygemba, M.
Right arrow Articles by Lei, J. T.
(Journal of Leukocyte Biology. 2007;81:1137-1148.)
© 2007 by Society for Leukocyte Biology

JAK kinases control IL-5 receptor ubiquitination, degradation, and internalization

Margarita Martinez-Moczygemba1, David P. Huston and Jonathan T. Lei

Biology of Inflammation Center and Immunology, Allergy and Rheumatology Section, Departments of Medicine and Immunology, Baylor College of Medicine, Houston, Texas, USA

1 Correspondence: Baylor College of Medicine, One Baylor Plaza, BCM 285, Houston, TX 77030-3411, USA. E-mail: mxm{at}bcm.tmc.edu


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
IL-5, IL-3, and GM-CSF are related hematopoietic cytokines, which regulate the function of myeloid cells and are mediators of the allergic inflammatory response. These cytokines signal through heteromeric receptors containing a specific {alpha} chain and a shared signaling chain, ßc. Previous studies demonstrated that the ubiquitin (Ub) proteasome degradation pathway was involved in signal termination of the ßc-sharing receptors. In this study, the upstream molecular events leading to proteasome degradation of the IL-5 receptor (IL-5R) were examined. By using biochemical and flow cytometric methods, we show that JAK kinase activity is required for ßc ubiquitination and proteasome degradation but only partially required for IL-5R internalization. Furthermore, we demonstrate the direct ubiquitination of the ßc cytoplasmic domain and identify lysine residues 566 and 603 as sites of ßc ubiquitination. Lastly, we show that ubiquitination of the ßc cytoplasmic domain begins at the plasma membrane, increases after receptor internalization, and is degraded by the proteasome after IL-5R internalization. We propose an updated working model of IL-5R down-regulation, whereby IL-5 ligation of its receptor activates JAK2/1 kinases, resulting in ßc tyrosine phosphorylation, ubiquitination, and IL-5R internalization. Once inside the cell, proteasomes degrade the ßc cytoplasmic domain, and the truncated receptor complex is terminally degraded in the lysosomes. These data establish a critical role for JAK kinases and the Ub/proteasome degradation pathway in IL-5R down-regulation.

Key Words: signal transduction • endocytosis • ßc-sharing receptors • hematopoietic cytokines • eosinophils


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
IL-5, IL-3, and GM-CSF are hematopoietic cytokines that are potent mediators of inflammatory responses by myeloid cells [1 ]. Although known for their important roles in hematopoiesis, these cytokines are of particular importance in allergic inflammation, asthma, and parasite immunity [2 3 4 5 6 7 8 9 ].

The specific cytokine signal for each of these cytokines is transmitted through cognate heteromeric receptors comprised of a ligand-specific {alpha} subunit [IL-5 receptor {alpha} (IL-5R{alpha}), IL-3R{alpha}, or GM-CSFR{alpha}] and a shared signaling subunit, ßc [10 ]. Binding of IL-5, IL-3, and GM-CSF to their respective receptors results in the activation of three main signaling pathways: JAK-STAT, Ras/Raf/MAPK (ERK, JNK, p38), and PI-3K [6 , 11 12 13 14 15 ]. Activation of JAK2 and JAK1 by these cytokines results in tyrosine phosphorylation of ßc on six critical tyrosine residues: Y577, Y612, Y695, Y750, Y806, and Y866. Of these phosphorylated residues, Y612, Y695, and Y750 serve as docking sites for the Src homology 2 (SH2) domains of three members of the STAT family of transcription factors, STAT1, STAT5, and STAT3 [6 , 11 12 13 14 15 16 17 ]. Together, these STATs regulate gene expression that controls cytokine-induced proliferation and differentiation for these three cytokines. However, members of the Src family of kinases such as Lyn, Fyn, and Hck are also reported to be activated by these three cytokines [6 , 11 12 13 14 15 16 17 ]. Recently, Lyn was shown to mediate IL-5-stimulated eosinophil survival and differentiation from bone marrow cells [18 ]. Moreover, Lyn immune complexes from eosinophils were able to phosphorylate IL-5R{alpha} and ßc immune complexes in in vitro kinase assays [18 ]. Thus, for the ßc-sharing receptors, JAK2/JAK1 and Lyn kinases regulate IL-5-mediated signal transduction, yet it is not known whether these signaling pathways also control ßc ubiquitination, proteasome degradation, or even receptor internalization.

Our previous studies demonstrated that following cytokine ligation, ßc signaling is terminated partially by ubiquitination and proteasome degradation of its cytoplasmic domain, resulting in the generation of truncated ßc products, termed ßc intracytoplasmic proteolysis (ßIP) [19 ]. Moreover, inhibition of ßc proteasome degradation resulted in prolonged activation of ßc, JAK2, STAT5, and SH2-containing tyrosine phosphatase 2. Following proteasome degradation of the ßc cytoplasmic tail, the remaining, truncated IL-5R complex (IL-5R{alpha} and ßIP) was degraded in the lysosomes [19 ]. This down-regulatory process resulted in homotypic and heterotypic desensitization of cells to further activation by ßc-engaging cytokines. However, the molecular mechanism governing IL-5R ubiquitination and proteasome degradation remains elusive.

In general, the post-translational modification of proteins by covalent attachment of ubiquitin (Ub) selectively targets these proteins for degradation by the proteasome [20 21 22 23 24 25 ]. This process of selective proteolysis basically consists of three steps: identification of the protein to be degraded; tagging of that protein for degradation by attachment of Ub to lysine (K) residues; and delivering it to the proteasome, a multienzyme protease complex, which will degrade it and recycle Ub. Protein ubiquitination requires the concerted enzymatic activities of the Ub conjugation machinery comprised of E1, the Ub activator; E2, the Ub conjugator; and E3, the Ub protein ligase [20 21 22 23 24 25 ]. PolyUb chains are formed through three different types of isopeptide linkages between the {epsilon}-amino group of one Ub molecule and a carboxy-terminal glycine of the newly added Ub molecule to the chain. These lysine residues include Lys 29, Lys 48, and Lys 63 [20 21 22 23 24 25 26 ]. In general, polyUb chains formed through Lys 48 linkages are attached to substrates destined for proteasome degradation. In contrast, polyUb chains formed through Lys 29 or Lys 63 linkages have other nonproteolytic functions in cells, such as transcriptional regulation and membrane transport [26 ]. Other types of protein ubiquitination, such as monoUb and multimonoUb, have also been reported [22 ]. MonoUb is involved in at least three distinct cellular functions: histone regulation, endocytosis, and the budding of retroviruses from the plasma membrane [22 ].

In this study, the molecular events regulating IL-5R down-regulation were investigated. Specifically, two major questions were asked: Do JAK and Lyn kinases regulate key steps in IL-5R down-regulation such as ßc ubiquitination, generation of ßIP, and IL-5R internalization? and Do proteasomes generate ßIP prior to or after IL-5R endocytosis? Our results demonstrate the direct ubiquitination of ßc on lysine residues 566 and 603. The data further show that JAK kinase activity is the dominant activity required for ßc ubiquitination, proteasome degradation, and removal of ligated IL-5Rs from the cell surface, and we propose an updated model of IL-5R down-regulation.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Cell culture, materials, and inhibitors
The human erythroleukemic cell line, TF1, was cloned as described previously [19 , 27 ]. Human eosinophils were freshly isolated from the peripheral blood of healthy donors by Ficoll gradient centrifugation, followed by negative selection with anti-CD16+ beads on an AutoMACS system. Recombinant human IL-5 was baculovirus-expressed and affinity-purified [28 ]. For all kinetic analyses in the presence or absence of inhibitors, TF1 cells were depleted of IL-5 for 24 h in RPMI 1640 containing 10% FBS (cytokine-starvation). Cells were then stimulated with 10 ng/ml IL-5 (R&D Systems, Minneapolis, MN, USA) for the indicated times. The human embryonic kidney (HEK) cell line, HEK293 (purchased from American Type Culture Collection, Manassas, VA, USA), was maintained in DMEM, supplemented with 10% FBS and 10 µg/ml gentamicin.

JAK inhibitor I, AG490, cytochalasin D, U0126 (MEK inhibitor), LY294002 (PI-3K inhibitor), SB203580 (p38 inhibitor), and brefeldin A were purchased from Calbiochem (San Diego, CA, USA). The Src family kinase inhibitor, PP1, was purchased from BioMol (Plymouth Meeting, PA, USA). Filipin was purchased from Sigma Chemical Co. (St. Louis, MO, USA). All inhibitors except AG490 were dissolved in 100% DMSO and used at the following final concentrations for 1 h prior to cytokine stimulation: JAK inhibitor I (50 µM), AG490 (100 µM), PP1 (20 µM), U0126 (10 µM), LY294002 (10 µM), SB203580 (10 µM), cytochalasin D (10 µM), brefeldin A (10 µg/ml), and filipin (5 µg/ml). TF1 cells were pretreated with AG490 (in DMSO) overnight (16 h) in the dark.

Contruction of {Delta}ßc mutants
Wild-type (WT) ßc cDNA was isolated from TF1 cells by RT-PCR (Stratagene, La Jolla, CA, USA) followed by PCR amplification of two fragments with Pfu polymerase. The two fragments consisted of an amino terminal fragment beginning at 11 bp and ending at the BglII site (1710 bp) and a carboxyl terminal fragment beginning at the BglII site and ending at 2722 bp (Table 1 ). After PCR amplification, the two fragments were restricted with BglII and ligated to form full-length ßc cDNA (2.7 Kb). The cDNA was ligated into the EcoRI- and HindIII-restricted pCMV-Script (Stratagene) mammalian expression vector and fully sequenced (sequence corresponding to GenBank Accession #AAA18171). IL-5R{alpha} cDNA was isolated from TF1 cells by RT-PCR and PCR amplification using the primers listed in Table 1 with Pfu polymerase. The full-length cDNA was cloned into the BamHI- and EcoRI-restricted pCMV-Script and fully sequenced (corresponding to GenBank Accession #NM000564).


View this table:
[in this window]
[in a new window]

  		Table 1. Primers Used for Cloning WT ßc, IL-5R{alpha}, {Delta}ßc Deletions and {Delta}ßc KtoR Mutants

 
Construction of the {Delta}ßc cytoplasmic deletion mutants, {Delta}ßc 473, 505, 560, 590, and 644, was performed by Pfu polymerase PCR amplification of WT ßc cDNA using the 5' 11 bp EcoRI primer (Table 1) and individual 3' primers designed with a stop codon inserted immediately after the last designated amino acid, followed by a HindIII restriction site (Table 1) . Each fragment was restricted with EcoRI and HindIII and cloned into the corresponding sites in the pCMV-Script vector. Site-directed mutagenesis of ßc K543, K566, and K603 to arginine was performed with the multisite, site-directed mutagenesis kit (Stratagene) using specific mutagenesis primers listed in Table 1 . The specific K to R mutations for {Delta}ßc 560, 590, and 644 are listed (see Fig. 1B ). All ßc mutants were fully sequenced to confirm sequence modifications.


Figure 1
View larger version (22K):
[in this window]
[in a new window]

  		Figure 1. Illustration of WT ßc and {Delta}ßc mutants. (A) Shown are WT and mutant ßc constructs used in this study. Indicated on the WT ßc receptor are the six critical cytoplasmic tyrosine residues (Y577, Y612, Y695, Y750, Y806, and Y866) and serine (S) 585, which become phosphorylated. Names of the {Delta}ßc constructs indicate the number of amino acids present in each mutant; numbers of lysine residues are also indicated for all constructs. The black boxes represent the transmembrane domain, and the open boxes represent Box 1 and Box 2. (B) Shown are the three {Delta}ßc constructs whose lysine residues were mutated to arginine by site-directed mutagenesis. Lysine residues mutated to arginine in {Delta}ßc 644 K to R mutant are K543, 566, and 603; mutated residues in {Delta}ßc 590 K to R mutant are K543 and 566; mutated residue in {Delta}ßc 560 K to R mutant is K543. The first three lysine residues in the ßc cytoplasmic domain (K473, K477, K483) do not appear to play a major role in ßc ubiquitination (see Fig. 2 ) and thus, were not mutated at this time.

 
Transient and stable transfections
The HEK293 cell line was transiently transfected for 48 h with plasmids encoding WT ßc, IL-5R{alpha}, or ßc mutants with GeneJammer reagent (Stratagene) for various analyses. Stably transfected TF1 cells with JAK2 cDNA or empty vector were generated by nucleofection (Amaxa) using recommended Kit V (Amaxa) and selected and cultured in 200 µg/ml G418. JAK2 overexpression was confirmed by immunoblot (IB) and flow cytometry using anti-JAK2 antibodies (Santa Cruz Biotechnology, Santa Cruz, CA, USA).

Immunoprecipitation (IP) and IB assays
All IP/IB assays were done as described previously [19 ]. Briefly, whole cell lysates (WCL) were standardized by the Bradford method (BioRad, Hercules, CA, USA), and equal amounts (usually 2–3 mg protein) were added to each IP tube. After IP with anti-ßc (S-16) antibodies and anti-actin (for some assays), proteins were detected by incubating the blots with anti-ßc polyclonal antibodies (N-20, Santa Cruz Biotechnology); anti-ßc (amino terminal, R&D Systems); anti-Ub mAb (P4D1), anti-JAK2, and anti-Lyn kinase (Santa Cruz Biotechnology); anti-actin (Sigma Chemical Co.); anti-IL-5R{alpha} (R&D Systems); anti-pSTAT5 (Upstate Biotechnology, Lake Placid, NY, USA); and anti-pJAK2, anti-pAKT, anti-pLyn kinase, antiphospho-p38, and anti-pMAPK (Cell Signaling, Beverly, MA, USA). Proteins were visualized by incubation with enhanced chemiluminescence Plus reagents (Amersham, Little Chalfont, UK), and images were captured with a FluorChem 8000 imaging system (Alpha Innotech, San Leandro, CA, USA).

ßc cell surface IP
Cell surface IP analysis of ßc was done as described by Ragimbeau et al. [29 ]. Briefly, 5 x 106 unstimulated and IL-5-stimulated TF1 cells were incubated with 2 µg anti-ßc (S-16) mAb for 2 h at 4°C, followed by three washes with cold PBS to remove unbound antibodies. Cells were lysed as described previously [19 ] with the exception of 1 mM DTT in the lysis buffer; immune complexes were then collected by precipitation with Protein G agarose beads and analyzed by SDS-PAGE.

Flow cytometry
ßc and IL-5R{alpha} cell surface expression was measured by incubating 1 x 106 unstimulated or IL-5-stimulated (30 min) TF1 cells and 1 x 105 eosinophils in PBS + 2% FBS with PE-labeled anti-ßc (BD PharMingen, San Diego, CA, USA) or PE-labeled anti-IL-5R{alpha} antibodies (R&D Systems) for 30 min on ice according to standard protocols. Labeled proteins were analyzed on a Beckman-Coulter XL flow cytometer. The hatched line in Figs. 6 , 7 and 9 represents cells labeled with an isotype-matched control antibody. The ßc and 5R{alpha} cell surface fold reduction in the absence or presence of inhibitors was calculated by dividing the mean fluorescence intensity (MFI) at 0 min by the MFI at 30 min. The flow data were analyzed using WinMDI software and graphed with Excel software.


Figure 6
View larger version (19K):
[in this window]
[in a new window]

  		Figure 6. IL-5R internalization precedes the generation of ßIP. (A) ßIP is generated intracellularly. Cytokine-starved TF1 cells (5x106 per lane) were used to IP ßc from the cell surface (Lanes 3 and 4) or from total WCL (Lanes 1 and 2) with anti-ßc (S-16) antibodies as described in Materials and Methods. Immune complexes were collected with Protein G agarose beads, separated by SDS-PAGE, and analyzed by IB with the indicated antibodies. Note how ßIP is detected only in the WCL IP and how ßc ubiquitination is increased in this lane (Lane 2). (B) Cytochalasin D does not block IL-5R internalization significantly. Cytokine-starved TF1 cells were left untreated (shaded histograms, both panels) or pretreated with 10 µM cytochalasin D (left panel, open histogram) or 5 µg/ml filipin (right panel, open histogram) for 1 h prior to 30 min of IL-5 stimulation (10 ng/ml). ßc cell surface expression was measured by labeling cells with a PE-conjugated anti-ßc mAb, followed by flow cytometry analysis. The hatched line represents cells labeled with an isotype-matched control antibody. Note how the drop and shift to the left of the open histogram are inhibited strongly in the presence of filipin (right panel), compared with cytochalasin D-treated cells (left panel). (C) Cytokine-starved TF1 cells were left untreated, treated with 5 µg/ml filipin, or treated with DMSO (DM; vehicle, Lane 7) for 1 h, followed by IP/IB analysis as described in Figure 5 . Note how ßIP generation is inhibited in the presence of filipin (upper panel, bottom arrow, Lanes 3–6) and how ubiquitinated forms of ßc accumulate in these lanes (both panels, Lanes 3–6).

 

Figure 7
View larger version (40K):
[in this window]
[in a new window]

  		Figure 7. JAK kinase activity is partially required for IL-5-induced ßc cell surface reduction. (A) Cytokine-starved TF1 cells were left untreated (left panel) or pretreated with 50 µM JAK inhibitor I (middle panel) or 20 µM PP1 (right panel) for 1 h. ßc cell surface expression was measured by labeling unstimulated (–IL-5, solid histogram) or 30 min IL-5 stimulated cells (+IL-5, open histogram) with a PE-conjugated anti-ßc mAb, followed by flow cytometry analysis. The hatched line represents cells labeled with an isotype-matched control antibody (labeled C). Note how the shift to the left of the open histogram is inhibited in the presence of JAK inhibitor I (middle panel), as compared with untreated cells. Data are also shown as means ± SEM of raw ßc MFI in the absence or presence of inhibitors (lower left panel) and as ßc fold reduction (lower right panel), which was calculated by dividing the MFI of unstimulated (0 min) cells by the IL-5-stimulated (30 min) MFI. A fold reduction of 1 means the cell surface protein level was unchanged following IL-5 stimulation; ßc untreated, n = 9; ßc (+JAK inhibitor I), n = 4; ßc (+PP1), n = 3. (B) WCL from stably transfected vector control or JAK2-overexpressing TF1 cell lines were analyzed by IB (upper panel) with anti-JAK2 antibodies. ßIP generation and ßc tyrosine phosphorylation were examined by IP/IB in vector or JAK2-overexpressing TF1 cell lines as described in Figure 4 A and 4B ). (C) JAK2 overexpression increases IL-5-induced ßc cell surface reduction. ßc cell surface reduction was analyzed and quantified as described in A from the stably transfected vector control or JAK2-overexpressing cells (n=3).

 

Figure 9
View larger version (28K):
[in this window]
[in a new window]

  		Figure 9. JAK kinase activity is required for IL-5-induced IL-5R endocytosis in freshly isolated human eosinophils (EOS). Freshly isolated EDS from peripheral blood, were left untreated or pretreated with 50 µM JAK inhibitor I and 20 µM PP1 for 1 h, followed by flow cytometry analysis of ßc cell surface expression as described in Figure 7 . Data are shown as means ± SEM of ßc MFI (lower left, n=3) and ßc fold reduction (lower right, n=3). Note how JAK inhibitor I blocks the shift to the left of the IL-5-stimulated histogram.

 

   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
The ßc cytoplasmic domain is ubiquitinated directly
As ßc and JAK2 have similar molecular weights (130 and 120 kDa, respectively) and are reported to be modified by ubiquitination, the possibility existed that the ubiquitinated bands we observed in our IP/IB ßc ubiquitination assays from TF1 cells were not ubiquitinated forms of ßc itself but rather other ubiquitinated proteins coprecipitating with ßc immune complexes, such as JAK2 [30 ]. To rule out this possibility and to determine if ßc itself were ubiquitinated directly, we used a molecular approach to generate five faster-migrating {Delta}ßc constructs with various cytoplasmic deletions, each containing a different number of lysine residues (sites of Ub attachment, 20–25): {Delta}ßc 644, {Delta}ßc 590, {Delta}ßc 560, {Delta}ßc 505, and {Delta}ßc 473 (Fig. 1 ; number of lysines present in each construct are indicated). These cytoplasmically deleted ßc constructs migrate faster than the 130-kDa full-length ßc; therefore, we hypothesized that if ßc were in fact ubiquitinated, then the migration of the ubiquitinated smears observed for full-length ßc (above 130 kDa) would shift downward in an IP/IB assay. Furthermore, if any of these smaller ßc constructs were ubiquitinated, they would provide an opportunity for partially defining potential ßc ubiquitination sites.

The HEK293 cell line was chosen for our studies, as it does not express endogenous IL-5Rs and is easily transfected. To confirm the feasibility of the HEK293 cell model system, plasmids encoding WT ßc and IL-5R{alpha} were cotransfected and analyzed for IL-5R activation by IP with anti-ßc antibodies followed by IB with antiphosphotyrosine ({alpha}-pY) and anti-Ub antibodies in a standard IL-5 time-course assay (Fig. 2A ). IL-5 stimulation of IL-5R-transfected HEK293 cells resulted in inducible and detectable ßc tyrosine phosphorylation (Fig. 2A , top panel), as well as ßc ubiquitination (Fig. 2A , middle panel), indicating that this cell line was appropriate for our studies.


Figure 2
View larger version (29K):
[in this window]
[in a new window]

  		Figure 2. The ßc cytoplasmic domain is ubiquitinated directly. (A) HEK293 cells were transiently cotransfected with plasmids encoding WT ßc and IL-5R{alpha} for 48 h and stimulated with 10 ng/ml IL-5 for the indicated time-points. WCL were prepared, and 1.5 mg total protein was IP with anti-ßc mAb (S-16), followed by IB with the indicated antibodies. (B) HEK293 cells were transiently cotransfected with plasmids encoding IL-5R{alpha} (Lanes 1–6), WT ßc (Lane 1), {Delta}ßc 473 (Lane 2), {Delta}ßc 505 (Lane 3), {Delta}ßc 560 (Lane 4), {Delta}ßc 590 (Lane 5), and {Delta}ßc 644 (Lane 6) or empty vector (Lane 7) for 48 h. WCL were prepared from IL-5-stimulated cells (30 min), IP with anti-ßc mAb (2 µg, S-16) and antiactin (1 µg) antibodies (for sample standardization), and serially IB with antibodies listed on the left side. Note the corresponding shift in the ubiquitination migration of {Delta}ßc 560, ßc 590, and ßc 644, as compared with WT ßc (second panel from top).

 
To test the hypothesis that ßc was ubiquitinated directly, plasmids encoding WT ßc and each of the {Delta}ßc constructs were transiently cotransfected with 5R{alpha} into HEK293 cells, stimulated with IL-5 for 30 min, and assayed for ßc ubiquitination in a similar manner (Fig. 2B) . HEK293 cells express low levels of JAK2 protein (unpublished observations); therefore, any Ub smears observed in the 75- to 120-kDa range should reflect ubiquitinated ßc or {Delta}ßc truncations. First, expression and migration of the new ßc constructs were confirmed by IB with anti-ßc antibodies. As expected, WT ßc migrated much slower (120 kDa) than the {Delta}ßc constructs (below 120 kDa), and all were expressed equivalently (Fig. 2B , top panel). Moreover, when the membrane was stripped and reprobed with anti-Ub antibodies, efficient ubiquitination was detected in the WT ßc lane (Fig. 2B , Lane 1). Conversely, no ßc ubiquitination was detected in lanes expressing {Delta}ßc 473 and {Delta}ßc 505 but was readily detected in the lanes expressing {Delta}ßc 560, {Delta}ßc 590, and {Delta}ßc 644. Moreover, the degree of ßc ubiquitination correspondingly increased with an increasing number of lysine residues present in each construct (compare Fig. 2B , Lanes 4–6). It is most important, however, that the ubiquitinated smears of the {Delta}ßc constructs concomitantly shifted down in migration, as compared with full-length WT ßc, thus confirming that the ßc cytoplasmic domain is ubiquitinated directly (Fig. 2B , second panel from top).

To demonstrate that our truncated ßc receptors were functional, the membrane was stripped and reprobed with antiphosphotyrosine and anti-5R{alpha} antibodies (Fig. 2B , third and fourth panels from top). Like WT ßc, tyrosine phosphorylation of {Delta}ßc 644 was readily detected as a result of the presence of two critical tyrosine residues, Y577 and Y612. It is interesting that 5R{alpha} coprecipitated with all of the {Delta}ßc constructs except for ßc 505, whose cytoplasmic domain contains the Box 1 motif but not Box 2. Last, anti-actin IP antibodies were added simultaneously to the anti-ßc IP incubation to demonstrate that equal amounts of WCL were included in each IP lane (Fig. 2B , bottom panel). In sum, these data clearly show that the ßc cytoplasmic domain is ubiquitinated directly and that ßc lysine residues 543, 566, and 603 are potential sites of ßc ubiquitination.

ßc lysine residues 566 and 603 are ubiquitination sites
Site-directed mutagenesis was performed to investigate whether ßc lysine residues 543, 566, and 603 were sites of ßc ubiquitination by mutating each lysine residue to another positively charged amino acid, arginine (illustrated in Fig. 1B ). The corresponding constructs were transiently cotransfected with IL-5R{alpha} into HEK293 cells and evaluated for their ability to support ßc ubiquitination in a similar manner as described above (Fig. 3 ). Compared with their {Delta}ßc WT counterparts (Fig. 3 , Lanes 3 and 4), IB analysis with anti-Ub antibodies revealed a significant decrease in the relative levels of IL-5-induced ubiquitination of {Delta}ßc 590 K543,566R and {Delta}ßc 644 K543,566,603R (Fig. 3 , middle panel, Lanes 6 and 7). In contrast, the degree of {Delta}ßc 560 K543R ubiquitination (Fig. 3 , Lane 5) did not differ significantly from WT {Delta}ßc 560 (Fig. 3 , Lane 2), suggesting that K543 might not be a major ßc ubiquitination site. Last, we confirmed the functionality of the {Delta}ßc K to R mutants by showing their capacity to co-IP 5R{alpha} (bottom panel). Therefore, we conclude that ßc lysine residues 566 and 603 are sites of ßc ubiquitination and that K543 might play a lesser role in ßc ubiquitination. However, it is important to note that this study characterized only six cytoplasmic lysines in ßc. Further analysis of the remaining nine lysine residues is required for complete mapping of all ßc ubiquitination sites.


Figure 3
View larger version (34K):
[in this window]
[in a new window]

  		Figure 3. ßc lysine residues 566 and 603 are sites of Ub attachment. HEK293 cells were transiently cotransfected with plasmids encoding IL-5R{alpha} (Lanes 1–7), WT ßc (Lane 1), {Delta}ßc 560 (Lane 2), {Delta}ßc 590 (Lane 3), {Delta}ßc 644 (Lane 4), {Delta}ßc 560 K543R, (Lane 5), {Delta}ßc 590 K543,566R (Lane 6), and {Delta}ßc 644 K543,566,603R (Lane 7) or empty vector (Lane 8) for 48 h. WCL were prepared from IL-5-stimulated cells (30 min), but this time, only 750 µg total protein was IP with anti-ßc mAb (1 µg, S-16) to detect the difference in ßc ubiquitination. The membrane was serially IB with indicated antibodies. Note how mutation of lysine residues 566 (Lane 6) and 603 (Lane 7) to arginine significantly reduces ßc ubiquitination.

 
JAK kinase activity is required for ßc ubiquitination and proteasome-mediated ßIP generation
For our mechanistic studies, we used the subcloned TF1-F11 cell line as a model system for two main reasons: They endogenously express the IL-5R (as well as IL-3R and GM-CSFR), and they permit the study of IL-5R regulation in a hematopoietic cell environment. To determine which IL-5-activated signaling pathway was involved in ßc ubiquitination and proteasome degradation, a specific JAK2 inhibitor, AG490 [31 ], was used to treat TF1 cells and was assayed for its ability to inhibit these molecular events. Figure 4A demonstrates that in the presence of AG490, ßc proteasome degradation was almost completely inhibited (Fig. 4A , top panel, compare Lanes 3 and 4), thus blocking the generation of ßIP (top panel, lower arrow). In addition, when the membrane was stripped and reprobed with anti-Ub antibodies, ßc ubiquitination was correspondingly inhibited (Fig. 4A , middle panel, compare Fig. 4A , Lanes 3 and 4). It is interesting that when the membrane was stripped and reblotted with an antiphosphotyrosine antibody, 4G10 (Fig. 4A , bottom panel), ßc tyrosine phosphorylation was inhibited by only 50% (Fig. 4A , Lane 4), as compared with untreated cells. Together, these data suggest that JAK2 kinase activity is required for ßc ubiquitination and proteasome degradation but only partially required for ßc tyrosine phosphorylation. Furthermore, incomplete inhibition of ßc tyrosine phosphorylation in the presence of AG490 is most likely a result of residual JAK1 activity, which is unaffected by this inhibitor and is concomitantly activated by IL-5 [31 , 11 –].


Figure 4
View larger version (22K):
[in this window]
[in a new window]

  		Figure 4. JAK kinase activity is required for ßc ubiquitination, tyrosine phosphorylation, and ßIP generation. (A) Cytokine-starved TF1 cells (24 h) were left untreated (Lanes 1 and 3) or pretreated with 100 µM AG490 (Lanes 2 and 4) overnight (in the dark) prior to IL-5 stimulation (10 ng/ml) for the indicated times. WCL were prepared and IP with anti-ßc mAb (S-16) and IB with anti-ßc polyclonal antibodies (top panel). The upper arrow indicates full-length ßc receptors, and the lower arrow corresponds to ßIP. The membrane was stripped and serially reprobed with anti-Ub (P4D1) and antiphosphotyrosine 4G10 mAb (middle and bottom panels). (B) Cytokine-starved TF1 cells (24 h) were left untreated (–) or pretreated (+) with 50 µM JAK inhibitor I for 1 h prior to IL-5 stimulation (10 ng/ml) for the indicated times. WCL were prepared and IP/IB, as described in A. As controls, WCL from this experiment were assayed by IB analysis with antiphospho-Lyn (as a non-JAK target) and antiphospho-STAT5 mAb as a JAK kinase target (C). Experiments with both inhibitors were repeated at least four times with reproducible results.

 
To minimize residual JAK1 activity, TF1 cells were treated with a pan-JAK inhibitor, JAK inhibitor I, an inhibitor that blocks all JAK kinase activity [32 ], and assayed for inhibition of these molecular events by IP/IB analysis. Figure 4B , top panel, demonstrates that in cells treated with the JAK inhibitor, proteasome degradation was almost completely inhibited, thus blocking the generation of ßIP (Fig. 4B , top panel, compare 30 min and 60 min stimulation, –/+ JAK inhibitor I). In addition, when the membrane was stripped and reprobed with anti-Ub antibodies, IL-5 stimulation induced a marked increase in the relative levels of ßc ubiquitination (~50%, top panel, compare Fig. 4B , Lane 1 with Lane 3). In contrast, in the presence of JAK inhibitor I, relative levels of ßc ubiquitination did not change appreciably in response to IL-5 stimulation (Fig. 4B , middle panel, Lanes 2, 4, and 6). To determine if this inhibitor blocked ßc tyrosine phosphorylation, the membrane was stripped and reblotted with 4G10 (Fig. 4B , bottom panel). As compared with no inhibitor (Fig. 4B , Lanes 1, 3, and 5) or to AG490 alone (Fig. 4A , bottom panel), JAK inhibitor I treatment completely blocked IL-5-induced ßc tyrosine phosphorylation (compare 30 min and 60 min stimulation, –/+ JAK inhibitor I).

To confirm nonspecific inhibition of other kinases such as Lyn, WCL from JAK inhibitor-treated and untreated cells were analyzed by IB with antiphospho-Lyn antibodies. The data demonstrate that JAK inhibitor I did not block IL-5-induced Lyn phosphorylation (Fig. 4C , top panel) but as compared with untreated cells, did result in slight accumulation of phospho-Lyn at 30 min, which returned back to basal levels at 60 min (Fig. 4C , top panel, compare Lanes 3 and 4). This phospho-Lyn accumulation is probably a result of the lack of ßc proteasome degradation seen at the same time-point in Figure 4B , middle panel. Last, to confirm that JAK kinase activity was indeed blocked, the downstream substrate STAT5 was analyzed by IB with antiphospho-STAT5 antibodies, and as predicted, the inhibitor completely blocked IL-5-induced STAT5 phosphorylation (Fig. 4C , middle panel). In sum, these data demonstrate that JAK kinase activity is required for IL-5-stimulated ßc ubiquitination, tyrosine phosphorylation, and proteasome-mediated generation of ßIP.

Role of Lyn kinase in ßc ubiquitination and proteasome degradation
To date, Lyn kinase is the only Src family kinase reportedly activated by IL-5 in TF1 cells and eosinophils [18 ]. To investigate the role of Lyn kinase in ßc ubiquitination and proteasome degradation, TF1 cells were pretreated with the Src kinase inhibitor PP1 and analyzed in a similar manner as described in Figure 4 . Compared with untreated TF1 cells (Fig. 5A , Lanes 1, 3, and 5), IB analysis with anti-ßc antibodies revealed that PP1 treatment delayed the kinetics of ßIP generation (Fig. 5A , top panel). In addition, PP1 treatment did not affect IL-5-induced ßc ubiquitination significantly (Fig. 5A , second panel from top, compare odd- and even-numbered lanes). It is surprising that inhibition of Src kinase activity did not decrease IL-5-stimulated ßc tyrosine phosphorylation (Fig. 5A , third panel from top) and did not affect the amount of Lyn protein coprecipitating with ßc immune complexes (Fig. 5A , fourth panel from top).


Figure 5
View larger version (21K):
[in this window]
[in a new window]

  		Figure 5. Lyn kinase activity is not required for ßIP generation or ßc ubiquitination. (A) Same as in Figure 4B , except TF1 cells were treated with 20 µM PP1 (Src kinase inhibitor) for 1 h prior to IL-5 stimulation. All IB antibodies are listed to the left of the panels. (B) To demonstrate reduction of Lyn kinase activity in the presence of PP1, IL-5-stimulated (30 min) WCL (Lanes 3 and 4, 5A) were analyzed by IB using antiphospho-Lyn polyclonal antibodies. The membrane was stripped and serially IB with anti-Lyn and antiactin antibodies. (C) Same as A, except TF1 cells were pretreated with 10 µM of the indicated signaling pathway inhibitors for 1 h prior to 30 min IL-5 stimulation (10 ng/ml). Inhibition of signaling pathways by each inhibitor was confirmed by IB analysis of WCL from the same experiment with phospho-specific antibodies against specific downstream targets.

 
To confirm that PP1 treatment inhibited tyrosine phosphorylation of Lyn kinase itself in TF1 cells, WCL from 30 min, IL-stimulated TF1 cells were analyzed by IB with antiphospho-Lyn antibodies (Fig. 5B) . PP1 treatment inhibited IL-5-induced Lyn tyrosine phosphorylation, compared with untreated cells, thus confirming the effectiveness of the Src kinase inhibitor on Lyn activation (Fig. 5B , top panel).

Furthermore, to determine if IL-5-activated signaling pathways downstream of JAKs and Lyn kinases were involved in ßc proteasome degradation or ubiquitination, specific inhibitors of the MEK/MAPK (U0126), PI-3K (LY294002), and p38 MAPK (SB203580) signaling pathways were analyzed by IP/IB in TF1 cells as described above (Fig. 5C) [33 , 34 ]. Compared with untreated cells (Fig. 5C , all panels, Lane 1), ßIP generation, as well as ßc ubiquitination and tyrosine phosphorylation, was not blocked significantly by treatment with these inhibitors (Fig. 5C , top three panels, Lanes 2–4). To confirm the effectiveness of each inhibitor, IB analysis of inhibitor-treated WCL with phospho-specific antibodies against downstream targets for each signaling pathway was performed (Fig. 5C , bottom three panels). As predicted, the three inhibitors blocked their respective signaling pathways effectively, suggesting that these signaling pathways do not contribute to ßc ubiquitination and proteasome degradation. However, it is interesting to note that the p38 MAPK inhibitor, SB203580, also blocked the PI-3K pathway (anti-pAKT blot, Lane 4). Whether the p38 MAPK pathway regulates the PI-3K is currently unknown.

In aggregate, these findings indicate that as compared with JAK kinases, Lyn kinase as well as other potential Src family members are not major regulators of ßc ubiquitination and proteasome degradation. However, it is worth mentioning that we have observed an additive effect on ßc ubiquitination and ßIP generation with the cotreatment of JAK inhibitor I and PP1. These data also confirm that JAK1 and JAK2, but not Lyn, are the main ßc tyrosine phosphorylating kinases (Fig. 4B , bottom panel, and Fig. 5A , third panel from top) [35 , 36 ]. Last, the data demonstrate that the molecular signals, which initiate ßc ubiquitination and proteasome degradation, are early events mediated by JAK kinases and not downstream events mediated by the MAPK, PI-3K, and the p38 MAPK signaling pathways.

ßIP is generated intracellularly
As shown in the previous data, IL-5 stimulation results in ßc ubiquitination and proteasome-mediated generation of ßIP. However, it is currently not known whether these ßc-targeted events occur at the plasma membrane or inside the cell following IL-5R endocytosis. To distinguish between these two locations, anti-ßc antibodies were used to IP ßc from the cell surface or from total cell lysates (containing cell surface and intracellular ßc) as described in Materials and Methods (Fig. 6 ) [29 ]. IB analysis with anti-ßc antibodies revealed that full-length ßc expression was detected in both locations, although lower ßc levels were present on the cell surface, especially after IL-5 stimulation (Fig. 6A , top panel). This observation is consistent with receptor down-regulation (shown in Figs. 7 8 9 ). It is interesting that the presence of ßIP was detected only in the IPs from the IL-5-stimulated, total cell lysates (Fig. 6A , Lane 2, upper panel), clearly indicating that its generation occurs inside the cell and not on the cell surface (Fig. 6A , compare Lanes 2 and 4).


Figure 8
View larger version (17K):
[in this window]
[in a new window]

  		Figure 8. Reduced ßc internalization in the presence of JAK inhibitor is not a result of increased receptor recycling. Same as in Figure 7A , except cells were treated with 10 µg/ml brefeldin alone or cotreated with 50 µM JAK inhibitor I plus 10 µg/ml brefeldin for 1 h before IL-5 stimulation. Raw ßc MFIs are expressed in the upper panel. The isotype-match control antibody is labeled –C. ßc fold reductions for all treated cells were calculated as described above; brefeldin only, n = 4; JAK kinase inhibitor + brefeldin, n = 4.

 
When the membrane was stripped and IB with anti-Ub antibodies, basal ßc ubiquitination was detected intracellularly (Fig. 6A , second panel from top, Lane 1) and on the cell surface (Fig. 6A , second panel from top, Lane 3); however, IL-5 stimulation resulted in a marked increase in ßc ubiquitination in the IP from the total cell lysates (Fig. 6A , second panel from top, Lane 2) but not in the IP from the cell surface (Fig. 6A , second panel from top, Lane 4). Last, IB analysis with anti-IL-5R{alpha} and anti-JAK2 revealed that co-IP of these molecules with ßc could be detected at the plasma membrane (Fig. 6A , third and fourth panels from top, Lane 4) but was much stronger after receptor internalization (Fig. 6A , two bottom panels). An antiactin IP control was included with the anti-ßc IP incubation to confirm that the same cell number was used in each lane (Fig. 6A , bottom panel). Together, these data indicate that ßc is ubiquitinated basally at the plasma membrane, as was shown for the epidermal growth factor receptor [37 ]. However, following IL-5 stimulation, ßc associates with IL-5R{alpha} and JAK2 and then moves inside the cell, where it becomes ubiquitinated further and is targeted by the proteasome to generate ßIP. These data suggest that ßIP is generated after receptor internalization and not prior, as previously predicted.

IL-5R internalization is required for the generation of ßIP
Our previous studies showed that cells treated with cytochalasin D still have the capacity to generate ßIP but fail to degrade the truncated IL-5R (ßIP/IL-5R{alpha}) in the lysosomes [19 ]. This observation led us to speculate that ßIP was generated prior to IL-5R endocytosis; however, data from Figure 6A demonstrate that ßIP is generated intracellularly. To clarify whether IL-5R internalization was a prerequisite for ßIP generation, we tested whether cytochalasin D, an inhibitor of actin filament function, could block the removal of IL-5Rs from the cell surface [38 ]. TF1 cells were pretreated with or without 10 µM cytochalasin D for 30 min, followed by flow cytometry analysis of ßc cell surface expression before and after IL-5 stimulation (Fig. 6B , left panel). It is surprising that cytochalasin D treatment did not block IL-5-induced ßc internalization significantly, evidenced by the overlap of the open histogram (+cytochalasin D) with the shaded histogram representing untreated cells (–cytochalasin D), as well as the isotype control (Fig. 6B , left panel). This result led us to search for alternative inhibitors, which blocked IL-5-induced IL-5R internalization.

To this end, we screened commonly used clathrin- and lipid raft-mediated endocytosis inhibitors and discovered that both classes of inhibitors blocked this process effectively (unpublished observations) [39 40 41 42 43 44 ]. We decided to use filipin, a cholesterol-binding drug, which is commonly used to block lipid raft-mediated endocytosis, as a representative endocytosis inhibitor for these studies [39 40 41 42 43 44 ]. We confirmed inhibition of IL-5-induced ßc internalization by pretreating cells with (open) or without (shaded) filipin for 30 min and assaying by flow cytometry (Fig. 6B , right panel). In contrast to Cyto D D-treated cells, IL-5-stimulated ßc internalization was blocked completely in the presence of filipin (Fig. 6B , right panel, open histogram).

We next examined whether ßIP generation was inhibited under these conditions. We predicted that if ßIP generation occurred after IL-5R internalization, then filipin treatment would decrease ßIP levels. In contrast, if ßIP generation occurred before IL-5R internalization, then blocking endocytosis with filipin would not affect its protein level. IP/IB analysis with anti-ßc antibodies revealed a dramatic reduction of ßIP protein in the filipin-treated cells as compared with untreated cells (Fig. 6C , upper panel, bottom arrow). Moreover, IB analysis with anti-Ub antibodies demonstrated that blocking receptor endocytosis caused a marked accumulation of ubiquitinated ßc receptors on the cell surface (Fig. 6C , lower panel, Lanes 4 and 6), confirming data in Figure 6A showing that ßc ubiquitination begins at the plasma membrane. Thus, these results demonstrate that degradation of the ßc cytoplasmic domain to yield ßIP occurs after the IL-5R has been internalized.

Inefficient removal of activated ßc cell surface receptors in the presence of JAK kinase inhibitor I
As the generation of ßIP required JAK kinase activity and IL-5R internalization, we hypothesized that JAK kinase activity would also be required for the latter. To test our hypothesis, flow cytometry analysis was used to evaluate ßc cell surface expression before and 30 min after IL-5 stimulation in untreated, JAK inhibitor-treated, or PP1-treated TF1 cells by labeling cells with PE-conjugated anti-ßc antibodies (Fig. 7A) . In untreated cells, ßc cell surface receptors decreased 60% following 30 min of IL-5 stimulation (Fig. 7A , upper left panel, +IL-5; MFIs quantified in lower left panel). In contrast, only a 30% reduction in ßc cell surface receptors was seen in the IL-5-stimulated cells treated with JAK inhibitor I (Fig. 7A , upper middle panel; quantified in both lower panels). It is interesting that treatment of TF1 cells with PP1 did not significantly affect IL-5-stimulated ßc cell surface reduction, as the data resembled that of untreated cells (Fig. 7A , upper right panel; quantified in bottom left panel).

As JAK kinase activity was required for IL-5-induced ßc internalization using pharmacological inhibitors (Fig. 7A) , we hypothesized that JAK2 overexpression would promote ßc internalization. To this end, we generated a stably transfected JAK2-overexpressing TF1 cell line, as well as a negative control cell line stably transfected with an empty vector (Fig. 7 B and 7C ). We confirmed JAK2 overexpression by anti-JAK2 IB analysis (Fig. 7B , upper panel). We next tested the hypothesis that JAK2-overexpressing cells would concomitantly have enhanced signaling and accelerated proteasome degradation. Indeed, IP/IB analysis revealed that relative levels of ßIP and ßc tyrosine phosphorylation were greater in the JAK2-overexpressing cells than in control cells (Fig. 7B) .

Last, we compared IL-5-induced ßc internalization in each of the stable cell lines using flow cytometry (Fig. 7C) . JAK2-overexpressing cells had an increased, IL-5-induced ßc cell surface fold reduction of 3.4 compared with the 2.3 ßc fold reduction seen in the vector control cell line, further supporting the role of JAK2 in IL-5-induced IL-5R internalization. In aggregate, these data demonstrate that JAK2 not only controls IL-5-mediated signaling and ßIP generation but also contributes to the regulation of IL-5R internalization in TF1 cells.

Reduced ßc internalization in the presence of JAK inhibitor is not a result of increased receptor recycling
As many receptors are recycled after ligand stimulation [45 ], we speculated that perhaps the decrease in IL-5R internalization in the presence of the JAK kinase inhibitor was a result of increased recycling of the receptors, as this effect would result in high ßc and 5R{alpha} MFIs after IL-5 stimulation. To test our hypothesis, we pretreated TF1 cells with the recycling inhibitor brefeldin A, alone or in combination with JAK inhibitor I. We predicted that if treatment with the JAK kinase inhibitor resulted in increased recycling of ßc back to the cell surface then blocking the recycling pathway with brefeldin in the presence of the JAK inhibitor would result in a lower ßc MFI after IL-5 stimulation (similar to that of untreated cells). Treatment of unstimulated cells with brefeldin alone (Fig. 8 , upper panel) resulted in a 5.9 ± 0.45 ßc MFI (mean MFI±SEM) compared with the 8.5 ± 0.18 ßc MFI seen in untreated cells (Fig. 7A , lower left panel), demonstrating a 30% decrease of basal ßc cell surface receptors. These data demonstrate that in the absence of IL-5, ~30% of ßc receptors are recycled back to the cell surface (Fig. 8) . However, following 30 min of IL-5 stimulation, the remaining ßc cell surface receptors internalized almost as efficiently as that of untreated cells (Fig. 8 , 30 min IL-5, +Bref., 50% reduction).

Similarly, cotreatment of unstimulated cells with JAK kinase inhibitor and brefeldin resulted in the same 30% decrease of recycled ßc cell surface receptors; however, following IL-5 stimulation, only 28% of ßc receptors were removed from the cell surface, compared with the 50% reduction seen with brefeldin alone (Fig. 8 , upper panel; ßc fold reduction is indicated in lower panel). These data demonstrate that the higher levels of ßc cell surface expression seen in JAK kinase inhibitor-treated cells are not a result of increased recycling but rather decreased receptor internalization. Therefore, JAK kinases contribute to the regulation of IL-5R internalization but are not the only regulators.

JAK kinase activity is required for IL-5R endocytosis in human eosinophils
To confirm that JAK kinase activity was required for IL-5R internalization in vivo, we repeated the ßc internalization assay described in Figure 7 with freshly isolated human eosinophils (Fig. 9) . Although ßc MFIs were lower in eosinophils than those in TF1 cells, IL-5 stimulation resulted in reduced levels of ßc cell surface receptors in the absence of inhibitor (Untreated), with an average fold reduction of 1.5 ± 0.3 (Fig. 9 , upper left panel and quantified in two lower panels). In contrast, ßc cell surface expression did not decrease in IL-5-stimulated eosinophils, which were pretreated with the JAK kinase inhibitor, resulting in an average fold reduction of 1.0 ± 0.3 (Fig. 9 , upper middle panel and quantified in two lower panels; P=0.013). It is interesting that in contrast to TF1 cells, PP1 had a slight inhibitory effect on IL-5R internalization in freshly isolated human eosinophils (Fig. 9 , upper right panel and quantified in two lower panels; fold induction, 1.1±0.14; P=0.148). Collectively, these data demonstrate that JAK and possibly Lyn kinases regulate IL-5-induced IL-5R internalization in human eosinophils and confirm our data with TF1 cells.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Receptor down-regulation is an evolutionarily conserved mechanism in eukaryotes necessary for controlling the magnitude and duration of extracellular signals. Unrestrained signaling, resulting from signal transduction pathways that are not terminated properly or from receptors not desensitized sufficiently, could potentially result in various inflammatory disorders associated with hypereosinophilia [46 , 47 ]. Therefore, understanding how to limit receptor signaling is critically important for preventing a protective response from causing injury to a host. We showed previously that the hematopoietic receptor, ßc, which is shared by IL-5, IL-3, and GM-CSF, is ubiquitinated and partially degraded by the proteasomes in response to ligand stimulation by each of these three cytokines [19 ]. Here, we extend our understanding of ßc down-regulation by showing that JAK kinases play a much broader role in IL-5R down-regulation than was appreciated originally. JAK kinase activity was required for IL-5-induced ßc ubiquitination and proteasome degradation and partially required for IL-5R internalization. Moreover, in addition to demonstrating the direct ubiquitination of ßc on lysine residues 566 and 603 (and possibly K543), we show that ßc is basally ubiquitinated at the plasma membrane and after IL-5-induced internalization, becomes ubiquitinated further intracellularly, where it is targeted by the proteasome to yield ßIP.

By using a lipid raft endocytosis inhibitor, we clearly showed that ßIP generation occurs after IL-5R internalization and not before, as predicted previously. Based on this new information, we propose an updated, working model of JAK and proteasome-dependent IL-5R down-regulation (Fig. 10 ): Step 1, IL-5 binding to the IL-5R leads to JAK2/1 and Lyn kinase activation, which results in ßc ubiquitination and tyrosine phosphorylation by the JAKs. Step 2, JAK kinases facilitate the entry of the full-length IL-5R into the lipid raft endocytic pathway (which may not be the only endocytic pathway regulating IL-5R internalization). Step 3, while in the endocytic pathway, proteasomes degrade the signaling portion of the ßc cytoplasmic domain to generate ßIP, this event contributes to signal termination. Step 4, Once the truncated IL-5R is generated, ßIP and 5R{alpha} are degraded in the lysosomes. Thus, for the IL-5R, JAK kinase activity is required for ßc activation and signal termination.


Figure 10
View larger version (76K):
[in this window]
[in a new window]

  		Figure 10. Working model of IL-5 receptor down-regulation. Step 1, IL-5 binding to the IL-5R leads to JAK2/1 and Lyn kinase activation, which results in ßc ubiquitination and tyrosine phosphorylation (P) by the JAKs. Step 2, JAK kinases and the presence of cholesterol at the plasma membrane (Fig. 6C) facilitate the entry of the full-length IL-5R into the lipid raft endocytic pathway (which is not the only endocytic pathway regulating IL-5R internalization). Step 3, Although in the endocytic pathway, proteasomes degrade the signaling portion of the ßc cytoplasmic domain to generate ßIP, this event contributes to signal termination. Step 4, Once the truncated IL-5R is generated, ßIP and 5R{alpha} are degraded in the lysosomes.

 
For most proteasome-targeted proteins, the covalent attachment of lysine-48-linked (K-48) polyUb chains is necessary for their recognition and degradation by the 26S-ATP-powered proteasome complex [20 21 22 23 24 25 26 ]. Our data show that the ßc cytoplasmic domain is ubiquitinated directly and basally at the cell surface but becomes increasingly ubiquitinated after it leaves the cell surface and moves into the intracellular compartment. Our biochemical and flow cytometry data with the JAK kinase inhibitor confirm this observation, as blocking IL-5R endocytosis with this inhibitor (Figs. 7 8 9) concomitantly blocked the IL-5-induced increase in ßc ubiquitination (Fig. 4) . One possible scenario to explain why ßc is ubiquitinated basally at the cell surface is that perhaps ßc is modified initially by K-29- or K-63-linked polyUb chains (or even multimonoubiquitination) to regulate its constitutive internalization in the absence of IL-5. However, following IL-5-induced internalization, perhaps ßc becomes modified by K-48 polyUb chains, which are the substrates for the 26S proteasome. This ßc modification could then serve as the recruitment signal forthe proteasomes to the IL-5R complex as it is being routed through the endocytic pathway and allow for degradation of the ßc cytoplasmic domain to generate ßIP, as was proposed in Figure 10 .

The exact location of proteasome-mediated ßIP generation as it is routed through the endocytic pathway is currently unknown. However, as ßIP is generated after receptor internalization but before lysosomal degradation, we predict that this event occurs in the early or late endosomes. It is tempting to speculate that proteasomes degrade the ßc cytoplasmic domain in the late endosomes (also referred to as multivesicular bodies) and that the actual removal of the cytoplasmic domain is the molecular signal for delivery to the lysosome for terminal degradation of the receptor. Consistent with this hypothesis was a study by van Kerkhof et al. [48], reporting that proteasomal inhibitors block a late step in lysosomal transport of ligands, which remain associated to their receptors within the endocytic pathway. Moreover, unpublished data in our laboratory have demonstrated that TF1 cells stimulated with fluorescently labeled IL-5 show colocalization of the ligand with ßc receptors inside the cells. Together, these data indicate that IL-5 remains bound to the IL-5R as it is routed through the endocytic pathway and that proteasomes may regulate delivery of the internalized IL-5-bound receptors to the lysosomes. Evidence proving this hypothesis and the nature of ßc ubiquitination is currently lacking but is under investigation in our laboratory.

Limiting the amount of ßc signaling induced by the activated IL-5R is important for controlling potentially damaging inflammatory signals by eosinophils. Our studies clearly show that IL-5R endocytosis and ßIP generation are essential for extinguishing IL-5R signals by the proteasomes and lysosomes. The data further suggest that if activated ßc receptors fail to internalize, or their routing through the endocytic pathway is altered then their signaling could possibly not be terminated properly. As IL-5R endocytosis precedes ßIP generation, we speculate that internalization of activated IL-5Rs could potentially be a first check-point for signal termination, whereas ßIP generation could possibly be second. Whether patients with hypereosinophilic syndromes or patients with severe asthma have defects in IL-5R signal termination by this down-regulatory pathway as a result of genetic polymorphisms remains to be investigated.


   ACKNOWLEDGEMENTS
 
This work was supported in part by grants from the National Institutes of Health AI 50686 (M. M-M.), AI 063178-01 (M. M-M.), and AI 36936 (D. P. H.), American Heart Association (M. M-M.), American Lung Association (M. M-M.), and Baylor College of Medicine, Department of Medicine Development Grant (M. M-M.). We appreciate the excellent technical assistance of Dale S. Smith, C. Jeanny Laurent, and Wei Zhang. We also thank Dr. Ryan Shanks for eosinophil isolation and the helpful comments of Dr. N. Tony Eissa. The Biology of Inflammation Center is a Federation of Clinical Immunology Societies (FOCIS) Center of Excellence.

Received July 21, 2006; revised November 14, 2006; accepted December 11, 2006.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Arai, K. I., Lee, F., Miyajima, A., Miyatake, S., Arai, N., Yokota, T. (1990) Cytokines: coordinators of immune and inflammatory responses Annu. Rev. Biochem. 59,783-836[CrossRef][Medline]
  2. Foster, P. S., Hogan, S. P., Ramsay, A. J., Matthaei, K. I., Young, I. G. (1996) Interleukin 5 deficiency abolishes eosinophilia, airways hyperreactivity, and lung damage in a mouse asthma model J. Exp. Med. 183,195-201[Abstract/Free Full Text]
  3. Lee, J. J., McGarry, M. P., Farmer, S. C., Denzler, K. L., Larson, K. A., Carrigan, P. E., Brenneise, I. E., Horton, M. A., Haczku, A., Gelfand, E. W., Leikauf, G. D., Lee, N. A. (1997) Interleukin-5 expression in the lung epithelium of transgenic mice leads to pulmonary changes pathognomonic of asthma J. Exp. Med. 185,2143-2156[Abstract/Free Full Text]
  4. Liu, L. Y., Sedgwick, J. B., Bates, M. E., Vrtis, R. F., Gern, J. E., Kita, H., Jarjour, N. N., Busse, W. W., Kelly, E. A. (2002) Decreased expression of membrane IL-5 receptor {alpha} on human eosinophils: I. Loss of membrane IL-5 receptor {alpha} on airway eosinophils and increased soluble IL-5 receptor {alpha} in the airway after allergen challenge J. Immunol. 169,6452-6458[Abstract/Free Full Text]
  5. Huang, J. C., Martinez-Moczygemba, M., Nguyen, A. P., Huston, D. P. (2003) IL-5 and eosinophils in asthma and atopy: time for a new paradigm Huston, D. P. Eissa, N. T. eds. Therapeutic Targets of Airway Inflammation ,495-518 Marcel-Dekker, Inc. New York, NY.
  6. Martinez-Moczygemba, M., Huston, D. P. (2003) Biology of common ß receptor-signaling cytokines: IL-3, IL-5, and GM-CSF J. Allergy Clin. Immunol. 112,003 653-665
  7. Lee, J. J., Dimina, D., Macias, M. P., Ochkur, S. I., McGarry, M. P., O’Neill, K. R., Proteroe, C., Pero, R., Nguyen, T., Cormier, S. A., Lenkiewicz, E., Colbert, D., Rinaldi, L., Acerman, S. J., Irvin, C. J., Lee, N. A. (2004) Defining a link with asthma in mice congenitally deficient in eosinophils Science 305,1773-1776[Abstract/Free Full Text]
  8. Humbles, A. A., Lloyd, C. M., McMillan, S. J., Friend, D. S., Xanthou, G., McKenna, E. E., Ghiran, S., Gerard, N. P., Yu, C., Orkin, S. H., Gerard, C. (2004) A critical role for eosinophils in allergic airways remodeling Science 305,1776-1779[Abstract/Free Full Text]
  9. Klion, A. D., Nutman, T. B. (2004) The role of eosinophils in host defense against helminth parasites J. Allergy Clin. Immunol. 113,30-37[CrossRef][Medline]
  10. Miyajima, A., Kitamura, T., Harada, N., Yokota, T., Arai, K. (1992) Cytokine receptors and signal transduction Annu. Rev. Immunol. 10,295-331[CrossRef][Medline]
  11. Geijsen, N., Koenderman, L., Coffer, P. J. (2001) Specificity in cytokine signal transduction: lessons learned from the IL-3/IL-5 GM-CSF receptor family Cytokine Growth Factor Rev. 12,19-25[CrossRef][Medline]
  12. Scott, C. L., Begley, C. G. (1999) The ß common chain (ß c) of the granulocyte macrophage-colony stimulating factor, interleukin-3 and interleukin-5 receptors Int. J. Biochem. Cell Biol. 31,1011-1015[CrossRef][Medline]
  13. Guthridge, M. A., Stomski, F. C., Thomas, D., Woodcock, J. M., Bagley, C. J., Berndt, M. C., Lopez, A. F. (1998) Mechanism of activation of the GM-CSF, IL-3, and IL-5 family of receptors Stem Cells 16,301-313[Abstract/Free Full Text]
  14. De Groot, R. P., Coffer, P. J., Koenderman, L. (1998) Regulation of proliferation, differentiation and survival by the IL-3/IL-5/GM-CSF receptor family Cell. Signal. 10,619-628[CrossRef][Medline]
  15. Adachi, T., Alam, R. (1998) The mechanism of IL-5 signal transduction Am. J. Physiol. 275,C623-C633
  16. Caldenhoven, E., van Dijk, T., Raaijmakers, J. A., Lammers, J. W., Koenderman, L., De Groot, R. P. (1995) Activation of the STAT3/acute phase response factor transcription factor by interleukin-5 J. Biol. Chem. 270,25778-25784[Abstract/Free Full Text]
  17. Piu, F., Magnani, M., Ader, M. E. (2002) Dissection of the cytoplasmic domains of cytokine receptors involved in STAT and Ras dependent proliferation Oncogene 21,3579-3591[CrossRef][Medline]
  18. Stafford, S., Lowell, C., Sur, S., Alam, R. (2002) Lyn tyrosine kinase is important for IL-5-stimulated eosinophil differentiation J. Immunol. 168,1978-1983[Abstract/Free Full Text]
  19. Martinez-Moczygemba, M., Huston, D. P. (2001) Proteasome regulation of ßc signaling reveals a novel mechanism for heterotypic desensitization J. Clin. Invest. 108,1797-1806[CrossRef][Medline]
  20. Pickart, C. M. (2004) Back to the future with ubiquitin Cell 116,181-190[CrossRef][Medline]
  21. Glickman, M. H., Ciechanover, A. (2002) The ubiquitin-proteasome proteolytic pathway: destruction for the sake of construction Physiol. Rev. 82,373-428[Abstract/Free Full Text]
  22. Hicke, L. (2001) Protein regulation by monoubiquitin Nat. Rev. Mol. Cell Biol. 2,195-201[CrossRef][Medline]
  23. Wilkinson, K. D. (2000) Ubiquitination and deubiquitination: targeting of proteins for degradation by the proteasome Semin. Cell Dev. Biol. 11,141-148[CrossRef][Medline]
  24. Hicke, L. (1999) Gettin’ down with ubiquitin: turning off cell-surface receptors, transporters and channels Trends Cell Biol. 9,107-112[CrossRef][Medline]
  25. Hershko, A., Ciechanover, A. (1998) The ubiquitin system Annu. Rev. Biochem. 67,425-479[CrossRef][Medline]
  26. Arnason, T., Ellison, M. J. (1994) Stress resistance in Saccharomyces cerevisiae is strongly correlated with assembly of a novel type of multiubiquitin chain Mol. Cell. Biol. 14,7876-7883[Abstract/Free Full Text]
  27. Kitamura, T., Tange, T., Terasawa, T., Chiba, S., Kuwaki, T., Miyagawa, K., Piao, Y. F., Miyazono, K., Urabe, A., Takaku, F. (1989) Establishment and characterization of a unique human cell line that proliferates dependently on GM-CSF, IL-3, or erythropoietin J. Cell. Physiol. 140,323-334[CrossRef][Medline]
  28. Dickason, R. R., Huston, D. P. (1996) Creation of a biologically active interleukin-5 monomer Nature 379,652-655[CrossRef][Medline]
  29. Ragimbeau, J., Dondi, E., Alcover, A., Eid, P., Uze, G., Pellegrini, S. (2003) The tyrosine kinase Tyk2 controls IFNAR1 cell surface expression EMBO J. 22,537-547[CrossRef][Medline]
  30. Ungureanu, D., Saharinen, P., Junttila, I., Hilton, D. J., Silvennoinen, O. (2002) Regulation of Jak2 through the ubiquitin-proteasome pathway involves phosphorylation of Jak2 on Y1007 and interaction with SOCS-1 Mol. Cell. Biol. 22,3316-3326[Abstract/Free Full Text]
  31. Samanta, A. K., Lin, H., Sun, T., Kantarjian, H., Arlinghaus, R. B. (2006) Janus kinase 2: a critical target in chronic myelogenous leukemia Cancer Res. 66,6468-6472[Abstract/Free Full Text]
  32. Thompson, J. E., Cubbon, R. M., Cummings, R. T., Wicker, L. S., Frankshun, R., Cunningham, B. R., Cameron, P. M., Meinke, P. T., Liverton, N., Weng, Y., DeMartino, J. A. (2002) Photochemical preparation of a pyridone containing tetracycle: a Jak protein kinase inhibitor Bioorg. Med. Chem. Lett. 12,1219-1223[CrossRef][Medline]
  33. Birkenkamp, K. U., Tuyt, L. M. L., Lummen, C., Wierenga, A. T. J., Kruijer, W., Vellenga, E. (2000) The p38 MAP kinase inhibitor SB203580 enhances nuclear factor-{kappa} B transcriptional activity by a non-specific effect upon the ERK pathway Br. J. Pharmacol. 131,99-107[CrossRef]
  34. Medina, M. G., Ledesma, M. D., Dominguez, J. E., Medina, M., Zafra, D., Alameda, F., Dotti, C. G., Navarro, P. (2005) Tissue plasminogen activator mediates amyloid-induced neurotoxicity via Erk1/2 activation EMBO J. 24,1706-1716