| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Flanders Interuniversity Institute for Biotechnology, Department of Medical Protein Research, Ghent University, Faculty of Medicine and Health Sciences, 9000 Ghent, Belgium
Address all correspondence and requests for reprints to: Jan Tavernier, Flanders Interuniversity Institute for Biotechnology, Department of Medical Protein Research, Ghent University, Faculty of Medicine and Health Sciences, A. Baertsoenkaai 3, B-9000 Ghent, Belgium. E-mail: Jan.Tavernier{at}Ugent.be.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Known downstream components of the JAK-STAT pathway in GH signaling are STAT5a and STAT5b, and several members of the suppressor of cytokine signaling (SOCS) protein family, including SOCS2 and cytokine-inducible Src-homology 2 (SH2) containing protein (CIS). STAT5a and -b are latent cytoplasmic transcription factors that are recruited through their SH2 domains to phosphotyrosine (pY) motifs in the activated GHR. They become phosphorylated by the activated JAK2s and subsequently translocate as dimers to the nucleus where they bind to specific promoter elements and regulate transcription of their target genes (6). These include SOCS2 and CIS that are rapidly induced upon GHR activation and that function in a negative feedback loop at the receptor level. SOCS proteins typically contain an N-terminal pre-SH2 domain followed by a SH2 domain and a C-terminal SOCS box (7).
The physiological role of STAT5 and SOCS proteins in GH signaling has been studied in detail in knockout mice. STAT5b-deficient (but not STAT5a-deficient) mice exhibit loss of the male-specific longitudinal growth rate and hepatic gene expression pattern, due to impaired GHR signaling upon pulsatile stimulation (8, 9). STAT5b mutations have also been found in humans and result in severe growth deficiency, comparable to that seen in patients with GHR-inactivating mutations (10). SOCS2–/– mice are 30–40% larger than their littermates, resembling animals overexpressing GH (11, 12). SOCS2–/– x STAT5b–/– knockout mice were not significantly different from STAT5b–/– mice, indicating that the excess growth in SOCS2–/– mice mostly depends on STAT5b (13). SOCS2 transgenic mice also show gigantism (14), which may be explained by cross-inhibition of the negative regulators SOCS1 and -3 (15). CIS–/– mice show no obvious altered growth phenotype, but overexpression of CIS mimics the phenotype of STAT5b–/– mice (1, 16).
The precise mechanism by which SOCS2 and CIS inhibit STAT5 activation by GH remains unclear. In one model, SOCS2 or CIS binding to the GHR leads to proteasome-dependent degradation of the GHR complex. SOCS proteins can bind with their SH2 domain to the activated receptor complex and recruit via their SOCS box the elongin B and C proteins, which together with cullin 5 and Rbx-2 form an E3 ubiquitin ligase complex. This complex targets SOCS-associated proteins for proteasomal degradation (17, 18, 19, 20, 21). CIS also appears to play an important role in GHR internalization, a critical step preceding termination of receptor signaling (22). In an alternative model, SOCS2 or CIS compete with STAT5a and -b for shared pY binding sites in the GHR (13, 17, 18). This is underscored by the finding that a more effective inhibition by CIS is seen in cells transfected with a lower amount of STAT5b. Moreover, CIS displayed more extensive inhibition in cells stimulated with lower concentrations of GH (18). The precise identity of all STAT5 recruitment sites in the GHR and their overlap with those for SOCS2 and CIS is, however, still a matter of debate (3, 19, 23, 24, 25, 26).
In this paper, we used mammalian protein-protein interaction trap (MAPPIT), a mammalian two-hybrid method based on cytokine signaling, to determine the tyrosine motifs of the GHR involved in STAT5, SOCS2, and CIS binding. We also investigated the binding of CIS and SOCS2 in greater detail and demonstrate functional divergence of their recruitment motifs.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
and described in detail in Eyckerman et al. (27). Briefly, MAPPIT bait constructs consist of a chimeric receptor composed of the extracellular part of the erythropoietin receptor (EpoR) and the transmembrane and intracellular domains of a STAT3 recruitment-deficient leptin receptor (LR) fused to a bait protein. MAPPIT prey constructs are composed of a prey protein C-terminally fused to a part of the gp130 chain carrying STAT recruitment sites. Both STAT3 and STAT5 can be recruited to these sites, but because the endogenous STAT5 level is low in the Hek293T cell line, STAT3 predominantly transmits the signal. An interaction between bait and prey leads to functional complementation of STAT signaling that can be measured with a luciferase reporter containing the rat pancreatitis-associated protein I (rPAP) promoter responsive for STAT3 and STAT5 (27).
|
). To be able to detect interactions with the different tyrosines, eight sections of the cytosolic domain of the GHR, each of them containing one tyrosine (or two in case of Y332-Y337), were separately cloned as bait. Importantly, because MAPPIT uses the JAK-STAT pathway, phosphorylation of the attached Y-based bait motifs can occur, allowing the detection of modification-dependent interactions (27). As preys, we used STAT5a and -b, SOCS2, and CIS and mutants of the latter two proteins. In the STAT5a and -b prey constructs, a conserved tyrosine in the C-terminal part was mutated to abolish dimerization and induction of promoter activity. All preys were fused to a gp130 polypeptide with an N-terminal FLAG tag. Expression of the different preys was checked by Western blot analysis using an anti-FLAG antibody (data not shown). Expression and activation of the chimeric bait receptor was monitored using the SH2-Bß prey (data not shown), which is known to be a JAK2-interacting protein (28). As a negative control, an irrelevant prey construct encoding amino acids 261–708 of simian virus 40 large T-antigen (SVT) that lacks its nuclear localization signal was used. Furthermore, to verify whether the interaction is bait specific, we included either the chimeric receptor without bait or a tyrosine motif previously shown to be noninteracting.
Interaction Sites of STAT5a and -b with the GHR
Because there is no consensus about the interaction positions of STAT5a and -b with the GHR, we first applied MAPPIT to further clarify this issue. Each GHR bait construct together with the STAT5a or -b prey was transfected in Hek293T cells in combination with the rPAP luciferase reporter. After 24 h, cells were stimulated with Epo, and interactions were monitored 24 h later by measuring luciferase activity. Significant signals were obtained with the GHR motifs Y534, Y566, and Y627, indicating these motifs as the most favorable sites for STAT5a and -b prey binding (Fig. 2). A conserved C-terminal tyrosine of the STAT5a and -b preys was mutated to phenylalanine to reduce the background. This mutation did not change the binding pattern because preys containing wild-type STAT5a or -b (wtSTAT5a or -b) gave the strongest signals with the same GHR bait constructs (data not shown). No activity was observed in case of expression of an irrelevant SVT prey, except with the same Y534, Y566, and Y627 baits. This weak effect is caused by endogenous STAT5 that is recruited to these baits and subsequently activates the reporter. This is supported by the observation that the background was completely diminished when using stable cell lines expressing STAT5-specific small interfering RNA (siRNA) (data not shown). Similarly, coexpression of a dominant-negative STAT5 also completely suppressed this background signal (data not shown).
|
).
|
). This weak induction was completely lost in two cell lines stably expressing a STAT5-specific siRNA. In clones B3 and B5, the endogenous STAT5 protein levels were reduced by 70 and 60%, respectively (data not shown). This indicates that both motifs are capable of binding and activating STAT5 but to a much lesser extent than the Y534, Y566, and Y627 motifs.
|
, MAPPIT experiments confirmed the recruitment sites of SOCS2. In addition, the interaction sites of CIS perfectly coincide with the SOCS2 binding sites.
|
Nuclear magnetic resonance and crystal structures of SOCS3 with a bound gp130 pY peptide showed the importance of the residue at position Y+3 (29, 30). Although residues at positions Y+1 and Y+2 are solvent exposed, the hydrophobic residue at position Y+3 is completely buried in a hydrophobic pocket of SOCS3. The hydrophobic character of the pocket residues is conserved in all SOCS proteins, suggesting a similar hydrophobic pocket. Figure 6A shows a model for the binding of SOCS2 to the pY peptide of the GHR Y487 motif, in which Y+3 residue V490 lies in a hydrophobic pocket formed by the four conserved hydrophobic residues. The hydrophobic binding site is conserved in CIS, and the Y595 motif binds in a similar way (data not shown). The pY peptides of the GHR pY motifs are listed in Table 1. Besides the SOCS2/CIS recruitment sites, only the Y436 motif has a hydrophobic residue at the Y+3 position. In contrast with the Y487 and Y595 motif, the Y+2 residue is a negatively charged amino acid (D438) in the Y436 motif. The models show that the Y+2 residue of the pY peptides interacts with D107 in SOCS2 or D141 in CIS (Fig. 6A). An interaction between D438 and D107 or D141 is very unfavorable, due to repulsion between the negative charges, probably explaining why the Y436 motif is not a SOCS2/CIS recruitment site.
|
|
). Only three GHR tyrosine motifs have a hydrophobic residue at position Y+1, and only these three lead to a strong STAT5 activation (Table 1).
Many interactions of pY peptides with SH2 domains, often mediated by bridging water molecules, contribute to binding specificity, whereas different binding modes are sometimes possible for the same SH2 domain. The aforementioned principles for pY peptide binding are therefore clearly an oversimplification. Nevertheless, they seem to correspond nicely with the observed non-overlapping binding of STAT5 and SOCS2/CIS to the GHR motifs. The SOCS2 and CIS binding motifs (Y487 and Y595) do not have a hydrophobic residue in the Y+1 position, whereas the strongest STAT5 activating motifs (Y534, Y566, and Y627) do not have a hydrophobic residue in the Y+3 position (Table 1).
Effect of Mutating the Y+1 Amino Acids in the Y487 and Y595 Motifs of the GHR on STAT5 Binding
We tested whether changing the Y+1 amino acid residue in the GHR motifs Y487 and Y595 to a leucine would turn these motifs into a STAT5 activating motif. Mutating the Y+1 residue to leucine in the Y487 and Y595 motifs did not considerably affect their binding to SOCS2, CIS, or to the CIS Y253F mutant (data not shown). On the other hand, as shown in Fig. 7, the mutated Y487 motif became a functional STAT5 recruitment site, whereas the mutated Y595 motif still behaved as wild type.
|
Effect of Mutations in the SOCS Box of SOCS2 and CIS on Their Interaction with the GHR
The interaction of SOCS proteins with their receptor targets depends on their SH2 domains (33, 34). Recently, Lavens et al. (35) demonstrated that in case of CIS, the C-terminal part of the SOCS box is also required for functional interaction with the EpoR and with the LR. This CIS SOCS box domain determines pY peptide binding specificity, because its deletion did not affect interaction with the N-terminal death domain of the Toll-like receptor adaptor MyD88. Detailed mutagenesis revealed that one single tyrosine in the CIS SOCS box at position 253 is a critical binding determinant. Much in contrast, no evidence was obtained for a role in pY peptide binding by the SOCS box of the highly related SOCS2 protein (35).
To examine whether the SOCS boxes of SOCS2 and CIS play a role in their interaction with the GHR, we tested the following mutant SOCS2 and CIS preys: SOCS2 with the conserved tyrosines in the SOCS box being replaced by phenylalanines (Y190F or Y194F mutations), or lacking its entire SOCS box, and similar CIS variants with Y249F or Y253F mutations, or lacking the SOCS box. As baits, the known Y487 and Y595 interaction motifs of SOCS2 and CIS were used, as well as the Y332-Y627 segment of the intracellular domain of the GHR containing all tyrosines and the Y332-Y337 motif. None of the different mutations in SOCS2 affected its binding to the GHR (Fig. 8A). This is in contrast with CIS, where the Y253F but not the Y249F mutation clearly caused reduced binding with the GHR Y595 motif. This effect was binding site specific, because little effect was seen with the GHR Y487 motif. The same effect on interaction patterns was observed when the SOCS box of CIS was deleted (Fig. 8B).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
We showed that the Y534, Y566, and Y627 motifs are the most important sites for STAT5a and -b recruitment. These MAPPIT data were confirmed by EMSA and using a STAT5-dependent ß-casein reporter assay, based on exogenously expressed and endogenous STAT5, respectively. Our observed STAT5 binding sites are in agreement with a study by Hansen et al. (24). They showed that by expressing mutated GHR in CHO cells, which were cotransfected with a CAT reporter containing three copies of the GH-responsive element from the Spi 2.1 promoter, any of Y534, Y566, or Y627, when present as the only cytoplasmic tyrosine in the GHR, can generate a functional GHR. In contrast, after glutathione-S-transferase fusion protein experiments, Sotiropoulos and colleagues (23) proposed Y487 and Y534 as the most important STAT5 recruitment motifs. In another study by Smit et al. (25, 26), it was suggested that Y332 and Y337 are required for maximal STAT5 phosphorylation. However, using conditional knock-in mice with deletions of specific domains of the GHR, Rowland et al. (3) showed that mouse GHR tyrosines 341 and 346, corresponding to rabbit Y332 and Y337, did not play a role in generating active STAT5. Mouse tyrosines 403, 447, and 498 (rabbit Y390, Y436, and Y487) can generate 30% of active STAT5 in the liver, whereas the 80 distal residues including mouse tyrosines 577, 606, and 639 (rabbit Y566, Y595, and Y627) account for 70% of STAT5 generation (3). Although only the Y534, Y566, and Y627 motifs consistently generate strong signals using MAPPIT, EMSA, and reporter assays, we do observe discrepancies between the assays. This difference, however, disappears when the STAT5 level is reduced. This suggests that the aforementioned inconsistencies are likely due to the different sensitivities and STAT5 levels of the assays used.
MAPPIT experiments revealed that Y487 and Y595 are the main docking sites for SOCS2 and CIS. These data confirm the SOCS2 binding sites that were previously identified by Greenhalgh et al. (14, 19). These authors also used mutational analysis to show that both tyrosines are required for SOCS2 function and that the SOCS box of SOCS2 is essential for its inhibitory role (19). In line with our findings, the C-terminal 80 amino acids of the GHR were previously shown to be sufficient for CIS recruitment (17). We here demonstrate that CIS uses exactly the same tyrosine motifs of the GHR as SOCS2. This overlapping binding pattern of the highly related CIS and SOCS2 is also seen in other cytokine receptors such as the EpoR, where SOCS2 and CIS each bind to Y402 and Y344, and the LR, where both proteins bind to Y1077. This is, however, not a strict feature because the EpoR Y480 motif is a binding motif for SOCS2 but not for CIS, and the opposite is true for the Y985 motif of the LR (36, 37).
Competition for shared binding sites is one possible inhibitory mechanism for STAT5 activation by SOCS proteins as was first suggested by Ram and Waxman (17) for GHR signaling. This inhibitory mechanism seems to be used also by other members of the class I superfamily of cytokine receptors such as the EpoR, where pY binding sites are shared by STAT5, SOCS2, and CIS (36). It was shown that CIS can abrogate the EpoR Y402-dependent STAT5 recruitment and activation (35) and that SOCS2 could compete with STAT5a association at the Y1077 motif of the LR (37). In contrast, for the prolactin receptor Endo et al. (38) suggest that suppression of STAT5 activation by CIS is not due to a simple competition with STAT5 but rather to a steric hindrance or modification of the receptor by CIS binding. Our observation of differential binding with the GHR of SOCS2 and CIS vs. STAT5a and -b now rules out that SOCS2 and CIS can inhibit STAT5 signaling via competition for binding to shared pY motifs. This is in agreement with the observation that rapid and maximal STAT5b activation by GH does not differ in hepatocytes from SOCS2–/– and SOCS2+/+ mice (13). Because the SOCS box of SOCS2 seems to play a central role in the negative regulation of GH signaling, its action appears to depend more on ubiquitination and degradation to terminate GH signaling (19). Of note, it was recently shown that SOCS2 can bind to other members of the SOCS family, e.g. SOCS1 and SOCS3, thereby suppressing their activity through proteasomal degradation. That way, SOCS2 can restore and potentiate GH signaling by antagonizing SOCS1 and SOCS3 in a SOCS box-dependent manner (15). How CIS exerts its inhibitory function remains largely unclear, but in a recent report, CIS was implicated in internalization of the activated GHR complex, a critical step preceding signal termination (22).
Homology modeling of pY peptide binding revealed a hydrophobic binding site in STAT5 that binds the residue at positions +1 with respect to the pY in a pY peptide. SOCS2 and CIS contain a conserved hydrophobic pocket that binds the Y+3 residue of a pY peptide. This is in line with the segregation between the STAT5 and SOCS2/CIS binding sites seen in the GHR: SOCS2 and CIS binding motifs of the GHR do not have a hydrophobic residue in the Y+1 position, whereas the STAT5 binding motifs do not have a hydrophobic residue in the Y+3 position. May et al. (32) already suggested that hydrophobic amino acids, preferentially leucine, at the Y+1 position are probably required for efficient STAT5 recruitment. We mutated the amino acids at the Y+1 position to leucine in the SOCS2/CIS Y487 and Y595 recruitment motifs and observed efficient STAT5a and -b binding in the former but not in the latter case.
We previously reported that the SOCS box of CIS, but not of SOCS2 can contribute to pY peptide binding. Moreover, a CIS Y253F mutant was deficient in binding to any of the CIS recruitment sites in the EpoR or the LR (35). Here we extend these findings and show that Y253 is also involved in recognition of the GHR Y595 motif. In contrast, this is not the case for the Y487 motif, demonstrating functional divergence between both motifs. Such variable effect of mutating the Y253 residue on pY peptide recognition is not unique to the GHR but was observed before for the interaction of CIS with MyD88 (35). CIS homology models were built based upon the structures of SOCS2, SOCS3, and SOCS4 domains recently determined by nuclear magnetic resonance and crystallography/x-ray diffraction studies (29, 30, 39) (Debreczeni, J. E., A. Bullock, E. Papagrigoriou, A. Turnbull, A. C. W. Pike, F. Gorrec, F. Von Delft, M. Sundstrom, C. Arrowsmith, J. Weigelt, A. Edwards, and S. Knapp, submitted for publication). These models exclude a direct interaction of the SOCS box or Y253 with the pY peptide but suggest allosteric effects resulting from the loss of the Y253 hydroxyl group. More studies are necessary to address the physiological role of this phenomenon in more detail.
In conclusion, we have shown the differential interaction pattern of STAT5a and -b (Y534, Y566, and Y627) vs. SOCS2 and CIS (Y487 and Y595) with the GHR, which makes the inhibition mechanism by competition for shared pY binding sites very unlikely. Furthermore, we have demonstrated that the SOCS box of CIS is important for the interaction with GHR Y595 but not with the Y487 motif, suggesting a functional divergence. The different binding modus has also been observed when the amino acid at the Y+1 position is mutated to leucine, which turns the Y487 motif into a functional STAT5 recruitment site, whereas the corresponding mutation in the Y595 motif has no effect. Additional studies will have to elucidate whether there is a physiological meaning for this functional divergence.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
. Each forward primer contained a SacI site and each reverse primer a stop codon and a NotI site, allowing an in-frame coupling C-terminal to the EpoR-LRF3 in pSEL1(+2L). In the pSEL1(+2L)-rbGHR-Y487bait construct, the amino acid at position +1 with respect to the tyrosine was mutated to leucine by site-directed mutagenesis using primers 5'-CACTGGCAAACATCGATTTTTACCTCCAGGTTAGTGACATTACGC-3' and 5'-GCGTAATGTCACTAACCTGGAGGTAAAAATCGATGTTTGCCAGTG-3'. This construct is further referred to as pSEL1(+2L)-rbGHR-Y487-L+1bait. The pSEL1(+2L)-rbGHR-Y595-L+1bait construct was generated by site-directed mutagenesis of the pSEL1(+2L)-rbGHR-Y595bait construct using primers 5'-GTTCTGAGATGCCTGTTCCGGACTATCTCTCCATTCATTTAGTAC-3' and 5'-GTACTAAATGAATGGAGAGATAGTCCGGAACAGGCATCTCAGAAC-3'.
|
The mutants of pMG2-SOCS2 and pMG2-CIS: the pMG2-SOCS2Y190F, pMG2-SOCS2Y194F, pMG2-SOCS2dbox (amino acids 1–159), pMG2-CISY249F, pMG2-CISY253F, and pMG2-CISdbox (amino acids 1–221) prey constructs were generated as described by Lavens et al. (35). pMG2-SH2-Bß was used as a positive control. Mouse full-length SH2-Bß was amplified using primers 5'-AGAGGTAACCATGAATGGTGCCCCTTCCCCAG-3' and 5'-GCGGCCGCCTATGGCCTCTTCCAATTCAACC-3' on the RZPD clone IRAVp968F05112D.
Generation of a luciferase reporter construct containing the STAT3- and STAT5-responsive rPAP1 promoter (pXP2d2-rPAP1-luci) was described previously (41). The pGL3-ß-casein-luci reporter construct containing five repeats of the STAT5-responsive motif of the ß-casein promoter was a gift from Dr. Ivo Touw.
Cell Lines, Transfection, Reporter Assay, and Expression Control
The STAT5(1668) siRNA sequence specifically targeting STAT5 mRNA was described by Scheeren et al. (42). This sequence was inserted in the pSUPER-puro vector (OligoEngine) via BglII-HindIII sites. Stable clones were obtained by transfecting Hek293T cells with the pSUPER-puro-STAT5 siRNA construct and a plasmid containing a hygromycin resistance gene in a 10:1 ratio. After a first selection on hygromycin, the clones were selected on puromycin. Two clones, named B3 and B5, were selected for further use.
Culture conditions for Hek293T cells and the STAT5 siRNA-expressing Hek293T clones B3 and B5 and transfection by the calcium phosphate precipitation procedure were as described elsewhere (41). One day after transfection, cells were resuspended with cell dissociation agent (Life Technologies, Rockville, MD), seeded in a black well plate, and stimulated with 5 ng/ml Epo or left unstimulated. Twenty-four hours later, luciferase activity from triplicate samples was determined using a Topcount chemiluminescence counter [PerkinElmer (Packard), Waltham, MA]. Human Epo was obtained from R&D Systems (Minneapolis, MN). Expression of the different prey constructs was detected as described by Montoye et al. (36).
For statistical analysis, the Z-score was calculated. To consider the obtained fold induction value as relevant, it should be at least three times higher than the value obtained with the corresponding no-bait transfection, except for Fig. 5 where the difference should be at least a factor of 10, because these preys give rise to a higher background. For comparison with the corresponding SVT prey transfection, no factor was taken into account. From each Z-score, the P value was derived.
EMSA
The procedure for an EMSA was as described in Lavens et al. (35). In brief, Hek293T cells were transiently transfected with the desired bait construct and pMG2-wtSTAT5b, starved for 4 h in serum-free medium, and stimulated with 5 ng/ml Epo for 20 min. Nuclear extracts (10 µg) of these cells were incubated with a radiolabeled probe (10 fmol) based on the ß-casein promoter for 10 min and separated on a 4.5% polyacrylamide gel containing 7.5% glycerol, fixed in water/methanol/acetic acid (80:10:10 by volume), dried, and autoradiographed.
For the supershift analysis, an anti-STAT5 antibody (Abcam, Cambridge, MA) was incubated with the nuclear extract for 10 min before addition of the radiolabeled ß-casein probe.
Homology Modeling of the Binding of STAT5a, SOCS2, and CIS to the GHR pY Peptides
Homology models for mouse CIS and SOCS2 were built using molecular operating environment (MOE; Chemical Computing Group, Montreal, Quebec, Canada). Structures of SOCS2 (2C9W), SOCS3 (2BBU and 2HMH), and SOCS4 (2IZV) were superposed in MOE, and the corresponding sequence alignment was aligned with the sequence of CIS. Models for the mouse SOCS2 and CIS SH2 domain were built using the coordinates of SOCS2 as template. A homology model of CIS with the SOCS box was created using the entire SOCS2 structure with its SOCS box as template. An alternative model was built by superposing the CIS SH2 model (based on SOCS2) on the corresponding domain in the SOCS4 structure (2IZV), followed by homology modeling of the CIS SOCS box (residues 221–257) using the SOCS4 SOCS box as template. Both CIS domains were connected by automatic loop modeling of residues 214–220 in MOE. In the models, pY peptides corresponding to the GHR motifs were added, using the gp130 pY peptide in the SOCS3 solution structure 2BBU as template (30). The CIS and SOCS2 SH2 models and SOCS3 were superposed, and the gp130 pY peptide was mutated to the GHR pY peptide. The resulting GHR model peptides and surrounding SOCS2 or CIS residues were subjected to a limited energy minimization, followed by 100 psec molecular dynamics simulation, using MOE and the charmm27 force field (43). A similar strategy was used for modeling of pY peptide binding in STAT5a. The STAT5a crystal structure (1Y1U) was superposed onto the STAT3 crystal structure (1BG1), which contains a C-terminal STAT3 pY peptide bound to its SH2 domain (44, 45). The STAT3 pY peptide was mutated in silico to a GHR pY peptide.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Disclosure Statement: The authors have nothing to disclose.
First Published Online July 31, 2007
1 I.U. and I.L. contributed equally to this work. 
Abbreviations: CIS, Cytokine-inducible SH2-containing protein; EpoR, erythropoietin receptor; GHR, GH receptor; JAK, Janus kinase; LR, leptin receptor; MAPPIT, mammalian protein-protein interaction trap; MOE, molecular operating environment; pY, phosphotyrosine; rPAP, rat pancreatitis-associated protein I; SH2, Src-homology 2; siRNA, small interfering RNA; SOCS, suppressor of cytokine signaling; STAT, signal transducer and activator of transcription; SVT, simian virus 40 large T-antigen; wt, wild type.
Received for publication December 18, 2006. Accepted for publication July 27, 2007.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  R. R Almon, D. C DuBois, W. Lai, B. Xue, J. Nie, and W. J Jusko Gene expression analysis of hepatic roles in cause and development of diabetes in Goto-Kakizaki rats J. Endocrinol., March 1, 2009; 200(3): 331 - 346. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. G Miquet, L. Gonzalez, M. N Matos, C. E Hansen, A. Louis, A. Bartke, D. Turyn, and A. I Sotelo Transgenic mice overexpressing GH exhibit hepatic upregulation of GH-signaling mediators involved in cell proliferation J. Endocrinol., August 1, 2008; 198(2): 317 - 330. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Piessevaux, L. De Ceuninck, D. Catteeuw, F. Peelman, and J. Tavernier Elongin B/C Recruitment Regulates Substrate Binding by CIS J. Biol. Chem., August 1, 2008; 283(31): 21334 - 21346. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Kenth, J. A M. Mergelas, and C. G. Goodyer Developmental changes in the human GH receptor and its signal transduction pathways J. Endocrinol., July 1, 2008; 198(1): 71 - 82. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Wauman, A.-S. De Smet, D. Catteeuw, D. Belsham, and J. Tavernier Insulin Receptor Substrate 4 Couples the Leptin Receptor to Multiple Signaling Pathways Mol. Endocrinol., April 1, 2008; 22(4): 965 - 977. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
