HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 28, 21596-21604, July 14, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/28/21596    most recent M910087199v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Gregorieff, A. Articles by Veillette, A. Search for Related Content PubMed PubMed Citation Articles by Gregorieff, A. Articles by Veillette, A. Social Bookmarking                      What's this?

Regulation of SOCS-1 Expression by Translational Repression^*

Alexander Gregorieff^§¶, Stéphane Pyronnet^§, Nahum Sonenberg^§, and André Veillette^§**§§

From the  McGill Cancer Centre and the Departments of ^§ Biochemistry, ^** Oncology, and  Medicine, McGill University, Montréal, Québec H3G 1Y6, Canada

Received for publication, December 17, 1999, and in revised form, April 7, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Accumulating evidence demonstrates that cytokine receptor signaling is negatively regulated by a family of Src homology 2 domain-containing adaptor molecules termed SOCS (suppressor of cytokine signaling). Previous studies have indicated that the expression of SOCS-related molecules is tightly controlled at the level of transcription. Furthermore, it has been reported that SOCS polypeptides are relatively unstable in cells, unless they are associated with elongins B and C. Herein, we document the existence of a third mechanism of regulation of SOCS function. Our data showed that expression of SOCS-1, a member of the SOCS family, is strongly repressed at the level of translation initiation. Structure-function analyses indicated that this effect is mediated by the 5' untranslated region of socs-1 and that it relates to the presence of two upstream AUGs in this region. Further studies revealed that socs-1 translation is cap-dependent and that it is modulated by eIF4E-binding proteins. In combination, these results uncover a novel level of regulation of SOCS-related molecules. Moreover, coupled with previous findings, they suggest that SOCS expression is tightly regulated through multiple mechanisms, in order to avoid inappropriate interference with cytokine-mediated effects.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cytokines such as interleukins (ILs)¹ and interferons (IFNs) play pivotal roles in a wide spectrum of processes (1-3). For example, they are required for normal hemopoietic cell development and immunity. Moreover, they participate in the pathophysiology of various autoimmune disorders like rheumatoid arthritis and systemic lupus erythematosus. Cytokines mediate their biological effects by binding to specific transmembrane receptors on target cells. As is the case for many other receptors, engagement of these receptors triggers the tyrosine phosphorylation of cellular proteins, including the cytokine receptors themselves, members of the Jak family of protein-tyrosine kinases, and the STAT transcription factors. Since cytokine receptors are devoid of intrinsic protein-tyrosine kinase activity, these phosphorylation events necessitate the participation of cytoplasmic protein-tyrosine kinases. Accumulating data show that receptor-associated Jak kinases are activated upon ligand stimulation, thus allowing tyrosine phosphorylation of the receptor and subsequent recruitment and activation of Src homology 2 domain-containing effectors such as the STATs.

Given the deleterious consequences of excessive cytokine stimulation, there has been significant interest in identifying mechanisms restricting their effects. The protein-tyrosine phosphatase SHP-1 and possibly its relative SHP-2 have been shown to play important roles in this process, by dephosphorylating the cytokine receptors and the Jak kinases (4-6). Additionally, a new group of molecules termed PIAS (protein inhibitor of activated STAT) was found to inhibit STAT function by interfering with their ability to bind DNA (7, 8). Finally, members of the SOCS (suppressor of cytokine signaling) family of adaptor molecules (also referred to as the CIS or SSI family) were reported to be potent negative regulators of cytokine receptor signaling, in part through their capacity to bind and inactivate Jak kinases (9-13).

The SOCS family is comprised of at least eight distinct members (13). These molecules share a common structure including a variable amino-terminal region, a central Src homology 2 domain, and a conserved carboxyl-terminal motif termed the SOCS box. Several observations support the idea that SOCS are negative regulators of cytokine receptor signaling. First, SOCS polypeptides can bind through their Src homology 2 domain to Jak kinases and can inhibit their activity in vivo and in vitro (11, 12, 14-16). Second, enforced expression of several SOCS family members such as SOCS-1, CIS-1, and SOCS-3 can repress cytokine-induced effects, including cellular proliferation and differentiation (10-12, 17-22). And finally, mice lacking the prototype of the family, SOCS-1, exhibit monocytic and polymorphonuclear infiltration of several organs, fatty degeneration of the liver, and early neonatal lethality (23, 24). Moreover, their hemopoietic cell progenitors have an increased sensitivity to cytokines like granulocyte-macrophage colony-stimulating factor and IFN- (25). These animals also possess markedly diminished cellularity in the thymus, seemingly as a consequence of enhanced apoptosis (38).

In light of their potent inhibitory impact, it appears likely that the expression and function of SOCS molecules needs to be tightly controlled. In keeping with the idea, the accumulation of most, if not all, SOCS family members was shown to be dependent on prior exposure to cytokines (9-12). Moreover, there are data indicating that certain SOCS proteins are unstable in cells (26-28) and that their half-life may be influenced by binding of elongins B and C to their SOCS box (27, 28). In this report, we have characterized an additional mechanism of inhibition of SOCS function. Our results show that the expression of SOCS-1 is strongly modulated at the level of translation initiation.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Antibodies-- Rabbit polyclonal antibodies against the carboxyl-terminal portion of SOCS-1 were raised against a bacterial (TrpE) fusion protein encompassing amino acids 168-212 of mouse SOCS-1. Antibodies against the amino-terminal segment of SOCS-1 were also produced in rabbits, using a TrpE fusion protein containing residues 2-78 of mouse SOCS-1. Anti-hemagglutinin (HA) monoclonal antibody 12CA5 was described previously (29).

Cells and Reagents-- Whole thymus was extracted from 3-week-old C57BL/6 mice. CTLL-2 is an IL-2-dependent mouse T-cell line and was propagated in RPMI 1640 medium supplemented with 10% fetal calf serum and recombinant IL-2. BI-141 is an antigen-specific mouse T-cell hybridoma (30). Derivatives expressing the neomycin phosphotransferase alone (Neo) were described elsewhere (31). All BI-141 cells were grown in RPMI 1640 medium containing 10% fetal calf serum and, if appropriate, G418 (0.6 mg/ml). COS-1 cells were propagated in -minimal essential medium containing 10% fetal calf serum, while HeLa cells were grown in Dulbecco's minimal essential medium plus 10% fetal calf serum. Recombinant murine IFN- was purchased from Cedarlane Laboratories Ltd.

cDNAs-- The full-length socs-1 cDNA depicted in Fig. 2B was reconstituted from two mouse EST clones (GenBank^TM accession numbers AA212316 and AA024269). cDNA variants lacking the 5'- and 3'-untranslated regions (UTRs) were generated in two ways. For transfection in BI-141 cells, a partial socs-1 cDNA already lacking most of the 3'-UTR was excised with XhoI (site located just upstream of the ATG; see Fig. 2B) and SacI (in the polylinker at the 3'-end). For COS-1 cell transfections, a socs-1 cDNA bearing an EcoRI site immediately downstream of the stop codon was created by overlap extension polymerase chain reaction. This cDNA was subsequently digested with XhoI and EcoRI. cDNAs lacking the 5'-UTR or the 3'-UTR were generated by digesting the modified full-length cDNA carrying the artificial EcoRI site with either XhoI or EcoRI, respectively. All inserts were blunt-ended and cloned in the mammalian expression vector pNT-Neo, which contains the origin of replication of SV40, the SR promoter, the SV40 polyadenylation sequence, and the neo gene. The upstream AUGs of socs-1 were mutated using the QuickChange mutagenesis kit (Stratagene). To isolate the 5'-UTR of socs-1, an additional XhoI site was first created at the 5'-end of the full-length cDNA. The 5'-UTR was subsequently released with XhoI and used to generate the plasmids pSKII-5'UTR SOCS-1-luc (for in vitro transcription/translation and HeLa cell transfection) and pcDNA3-rluc-5'-UTR SOCS-1-fluc (for transfection in HeLa cells). All constructs were verified by sequencing to ensure that no unwanted mutation had been introduced during their creation (data not shown). pcDNA3-rluc-POLIRES-fluc, pACTAG-2, pACTAG-2-4E-BP1, pACTAG-2-BP1-4E were reported previously (32, 33). pSKII-5'UTR ODC-luc will be reported elsewhere.^2,3

RNA Analysis-- Total cellular RNA was prepared by the acid isothiocyanate/phenol/chloroform method (34). The integrity of the RNA was confirmed by electrophoresis in 1% formaldehyde-agarose gels. RNase protection assays were performed according to a previously published protocol (35), using serial dilutions of total cellular RNA. The construct used to produce socs-1 riboprobes was generated by cloning a fragment corresponding to nucleotides 651-773 of the sequence shown in Fig. 2B in pBSKII. After linearization with NotI, ~190-nucleotide antisense riboprobes were synthesized using T7 RNA polymerase, according to the manufacturer's instructions (Promega). Full protection of this riboprobe by cellular socs-1 transcripts was expected to yield a 123-nucleotide fragment. Cellular transcripts generated by cDNAs with a truncated 3'-UTR resulted in a larger protected fragment, due to hybridization of the riboprobe with additional sequences at the 3'-end of the cDNA. The luciferase riboprobe was generated by cloning an EcoRV-EcoNI fragment from the luciferase cDNA into pBSKII. After linearization with HindIII, antisense riboprobes of ~390 nucleotides were synthesized using T3 RNA polymerase. Full protection of these riboprobes by cellular luciferase transcripts resulted in a 282-nucleotide fragment. Size markers were prepared by end-labeling MspI-digested pBR322 fragments with Klenow DNA polymerase and [-³²P]dCTP.

Transfections-- BI-141 cells were stably transfected by electroporation (260 V, 960 microfarads) and selected in the presence of G418 (0.75 mg/ml), according to a protocol described elsewhere (36). Monoclonal cell lines expressing the transfected cDNAs were identified by a combination of RNase protection assays and immunoprecipitations (data not shown). COS-1 cells were transiently transfected by the DEAE-dextran method using 4 µg of plasmid DNA. After 12 h, cells were treated with chloroquine (60 mg/ml) for 5 h (37). Following an additional growth period of 48 h, cells were harvested for immunoprecipitation. HeLa cells were first infected for 1 h with recombinant vaccinia virus vTF7-3, which allows the production of T7 RNA polymerase (38, 39). They were then transfected with the indicated DNAs using the Lipofectin reagent, according to the manufacturer's instructions (Life Technologies, Inc.). Cells were harvested approximately 18 h later for luciferase assays, RNase protection assays, and/or immunoblots.

Immunoprecipitations and Immunoblots-- T-cells, thymus cells, and COS-1 cells were lysed in 1× TNE buffer (50 mM Tris, pH 8.0, 1% Nonidet P-40, 2 mM EDTA) supplemented with 10 µg/ml each of the protease inhibitors leupeptin, aprotinin, N-tosyl-L-phenylalanine chloromethyl ketone, N-tosyl-L-lysine-chloromethyl ketone, and phenylmethylsulfonyl fluoride as well as the phosphatase inhibitors sodium fluoride (50 mM) and sodium orthovanadate (1 mM). SOCS-1 polypeptides were recovered from various amounts of total cellular proteins using the antiserum raised against the carboxyl-terminal portion of SOCS-1. Immune complexes were subsequently collected with Staphylococcus aureus protein A (Pansorbin; Calbiochem). After several washes with lysis buffer, proteins were eluted in sample buffer, boiled, and subjected to 8 or 12% SDS-polyacrylamide gel electrophoresis. Immunoblots were performed according to a previously described protocol (40) using enhanced chemiluminescence reagents (Amersham Pharmacia Biotech) (for SOCS-1) or ¹²⁵I-labeled goat anti-mouse IgG (ICN Pharmaceuticals Canada Ltd.) (for HA tag).

In Vitro Transcription and Translation-- Luciferase RNAs were synthesized in vitro from linearized pSKII-luc, pSKII-5'UTR SOCS-1-luc, or pSKII-5'UTR ODC-luc using T7 RNA polymerase, in the presence of m⁷GpppG and GTP to allow capping of the RNAs, as described (41). Various quantities of transcripts were then translated in vitro for the indicated periods of times at 30 °C using rabbit reticulocyte lysates, as outlined by the manufacturer (Promega).

Luciferase Assays-- Luciferase activity was determined using either the luciferase reporter assay system or the dual luciferase reporter assay system from Promega. Values were measured in a luminometer (EG & G Berthold).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Identification of the SOCS-1 Protein in Mouse T-lymphocytes-- Previous studies have indicated that socs-1 is expressed highly in thymus (10). Therefore, in order to understand better the regulation of SOCS-1, we first wished to identify the SOCS-1 protein in T-cells. To this end, lysates from mouse thymus and IL-2-treated CTLL-2 T-cells were immunoprecipitated with a polyclonal rabbit antiserum directed against the carboxyl-terminal portion of SOCS-1, and the presence of SOCS-1 in these immunoprecipitates was revealed by immunoblotting with another antibody raised against the amino-terminal segment of SOCS-1 (Fig. 1A). Ribonuclease protection assays were performed in parallel to quantitate the levels of socs-1 transcripts, using serial dilutions of total cellular RNA (Fig. 1B). Although socs-1 transcripts could be readily documented in these samples (Fig. 1B), we found that the SOCS-1 protein was very difficult to detect under standard immunoprecipitation conditions. Nonetheless, when large amounts of cell lysates (9 mg of total cellular proteins) were utilized for immunoprecipitation (Fig. 1A), we were capable of detecting a ~24-kDa immunoreactive polypeptide consistent with SOCS-1 in anti-SOCS-1 immunoprecipitates (lanes 2 and 4). This product was absent in immunoprecipitates generated with normal rabbit serum (lanes 1 and 3).

View larger version (27K): [in this window] [in a new window]   Fig. 1.   Expression of endogenous SOCS-1 transcripts and protein in mouse T-lymphocytes. A, detection of SOCS-1 protein in mouse T-cells. Lysates were prepared from mouse thymus and IL-2-propagated CTLL-2 T-cells as specified under "Materials and Methods." Nine mg of total cellular proteins were immunoprecipitated with either normal rabbit serum (lanes 1 and 3) or an antiserum produced against the carboxyl-terminal portion of SOCS-1 (lanes 2 and 4). The presence of SOCS-1 in these immunoprecipitates was determined by subsequent immunoblotting with an antiserum directed against the amino-terminal portion of SOCS-1. The positions of prestained molecular weight markers are shown on the right, whereas those of the heavy chain of immunoglobulin (Ig(H)) and SOCS-1 are indicated on the left. Exposure was for 30 s. B, RNase protection assay. Total RNA was extracted from the samples described in A. The abundance of socs-1 RNA was then determined by RNase protection assay, using a socs-1-specific riboprobe and serial dilutions of cellular RNA (lanes 3 and 6, 30 µg; lanes 4 and 7, 10 µg; lanes 5 and 8, 3.3 µg). The migrations of the undigested probe and of the protected riboprobe fragment are indicated on the right, while those of radiolabeled size markers are shown on the left. Exposure was for 36 h.

Translational Repression of SOCS-1 Expression in Mouse T-cells-- Our difficulty in detecting endogenous SOCS-1 polypeptides, but not socs-1 transcripts, raised at least two possibilities. First, it is plausible that the SOCS-1 protein was highly unstable in T-cells. However, this scenario seems unlikely, since we estimated that the half-life of SOCS-1 was ~4 h in these cells,⁴ in keeping with previous reports (28). Second, it suggested that the SOCS-1 protein may be translationally regulated. To address this possibility, we first wanted to identify the noncoding elements in mouse socs-1 RNAs. Characterization of these untranslated sequences was performed through a combination of database searches, cDNA cloning, EST clone sequencing, and comparison with the reported genomic sequence of the mouse socs-1 gene. The socs-1 cDNA sequence defined by these analyses is depicted in Fig. 2A. It exhibits a length of 1177 nucleotides, in agreement with the reported size of mouse socs-1 transcripts (10, 12). This sequence includes 134 nucleotides of 5'-UTR, a coding portion of 639 bases and a 3'-UTR of 404 nucleotides. It has an additional 5 nucleotides at the 5' end compared with the cDNA reported by DeSepulveda et al. (GenBank^TM accession no. AF120490) and 16 nucleotides more than the cDNA submitted by Starr et al. (GenBank^TM accession no. U88325). Based on the sequence of the socs-1 gene, it was determined that socs-1 RNAs are encoded by two exons (Fig. 2B). While exon 1 contributes only to the 5'-UTR, exon 2 participates in the formation of part of the 5'-UTR, all of the coding region, and the 3'-UTR.

View larger version (51K): [in this window] [in a new window]   Fig. 2.   Structure and sequence of full-length mouse socs-1 cDNA. A, sequence of full-length mouse socs-1 cDNA. The cDNA sequence was reconstituted from mouse EST clones and verified by comparison with the genomic sequence (see "Materials and Methods"). The 5'-UTR and 3'-UTR are underlined, whereas the initiating codon (ATG), the stop codon (TGA), and the putative polyadenylation signal are boxed. B, schematic representation of the mouse socs-1 gene based on the available cDNA (GenBankTM accession no. AF180302) and genomic sequence (GenBankTM accession no. Z47352).

To examine the impact of the UTRs on the translation of socs-1, BI-141 mouse T-cells were stably transfected by electroporation with either a full-length socs-1 cDNA or a cDNA containing only the coding region of socs-1. Levels of SOCS-1 protein and RNA in representative derivatives were subsequently compared with those observed in control BI-141 cells that were either unstimulated or treated with IFN- (Fig. 3). As reported for other cell types (17, 20), IFN- provoked an increase (~5-fold) in the abundance of socs-1 transcripts in control BI-141 cells (Fig. 3B, compare lanes 6-8 with lanes 3-5). It also induced the appearance of detectable amounts of SOCS-1 protein in these cells (Fig. 3A, compare lanes 1 and 2). The levels of endogenous SOCS-1 protein and RNA observed in IFN--treated BI-141 cells were essentially similar to those found in the cell line transfected with the full-length socs-1 cDNA (Fig. 3, A, lane 6, and B, lanes 12-14). However, the amount of SOCS-1 protein observed in these cells was significantly lower than that found in cells transfected with the cDNA devoid of UTRs (Fig. 3A, lanes 3-5), even if comparable quantities of RNA existed (Fig. 3B, compare lanes 6-8 with lanes 9-11). Taking into account the differences in the quantities of cell lysates used for these immunoprecipitations, it was estimated that socs-1 transcripts possessing intact UTRs produced at least 30 times lower levels of SOCS-1 protein in comparison with full-length RNAs. Similar results were obtained with other SOCS-1-expressing BI-141 clones (data not shown).

View larger version (34K): [in this window] [in a new window]   Fig. 3.   Differential expression efficiency of endogenous and exogenous socs-1 RNAs. A, anti-SOCS-1 immunoblot. The accumulation of SOCS-1 protein in control BI-141 cells treated or not treated with IFN- (30 ng/ml for 2 h) or in transfectants expressing either a socs-1 cDNA lacking both the 5'-UTR and the 3'-UTR (lanes 3-5) or a full-length socs-1 cDNA (lane 6) was determined by immunoblotting as outlined in the legend of Fig. 1A. The amounts of cellular proteins used for immunoprecipitation are indicated. The positions of prestained molecular weight markers are shown on the right; those of the heavy chain of Ig and SOCS-1 are indicated on the left. Exposure was for 30 s. B, RNase protection assay. The abundance of socs-1 RNA in the cells described in A was determined as specified in the legend of Fig. 1B. The migrations of the undigested probe and of the protected riboprobe fragment are shown on the right; those of radiolabeled size markers are indicated on the left. Exposure was for 16 h.

Inhibition of SOCS-1 Protein Expression Is Mediated by the 5'-Untranslated Region of socs-1 Transcripts-- Next, we wanted to determine which noncoding sequence(s) was responsible for inhibiting the translational efficiency of socs-1 RNAs. To this end, socs-1 cDNAs lacking either the 5'-UTR or the 3'-UTR or both were constructed as detailed under "Materials and Methods." All cDNAs were cloned in the expression vector pNT-Neo, flanked by an SR promoter and the polyadenylation sequence from SV40. After transient expression in COS-1 cells, the abundance of SOCS-1 protein was measured by anti-SOCS-1 immunoblotting of SOCS-1 immunoprecipitates (Fig. 4A), whereas the accumulation of socs-1 transcripts was monitored by RNase protection assay (Fig. 4B).

View larger version (33K): [in this window] [in a new window]   Fig. 4.   Inhibition of socs-1 translation by the 5'-untranslated region. A, anti-SOCS-1 immunoblot. COS-1 cells were transiently transfected with the indicated cDNAs. Three-fold dilutions of cellular lysate (lanes 1, 4, 7, and 10, 1.5 mg; lanes 2, 5, 8, and 11, 500 mg; lanes 3, 6, 9, and 12, 167 mg) were then utilized for immunoprecipitation and immunoblot, as outlined in the legend of Fig. 1A. The positions of prestained molecular weight markers are indicated on the right, whereas that of SOCS-1 is shown on the left. Exposure was for 50 s. B, RNase protection assay. The abundance of socs-1 RNA was determined as detailed in the legend of Fig. 1B, except that 5-fold dilutions of RNA were used (lanes 4, 7, 10, and 13, 0.5 µg; lanes 5, 8, 11, and 14, 0.1 µg; lanes 6, 9, 12, and 15, 0.02 µg). For lane 3, 30 µg of RNA from thymus was examined as a positive control. The migrations of the undigested probe and of the protected riboprobe fragments are shown on the right, while those of radiolabeled size markers are indicated on the left. Exposure was for 14 h.

First, this experiment showed that the levels of SOCS-1 protein generated from the full-length cDNA in COS-1 cells (Fig. 4A, lanes 10-12) were lower (~7-fold) than those synthesized from the cDNA lacking both UTRs (5'/3'; lanes 1-3), in keeping with the observation made above in BI-141 T-cells. This difference was not due to variations in RNA expression, as documented by the parallel RNase protection assay (Fig. 4B, compare lanes 13-15 with lanes 4-6). It should be mentioned, however, that the protected riboprobe fragment observed with the 5'/3' and 3' cDNAs was 5-10 nucleotides larger than that seen with the full-length cDNA. This alteration was due to the fact that polylinker sequences present in the riboprobe were identical to sequences found at the 3'-end of the 5'/3' and 3' cDNAs. Second, this study demonstrated that removal of the 5'-UTR (Fig. 4A, lanes 4-6) was sufficient to relieve the suppressive effect of the UTRs on SOCS-1 protein expression. Deletion of the 3'-UTR (lanes 7-9) had no consequence.

In light of these results, we wished to ensure that the 5'-UTR was not only responsible, but also sufficient, to suppress translation of SOCS-1. For this purpose, the 5'-UTR of socs-1 was cloned upstream of the reporter gene luciferase in a vector suitable for in vitro transcription (pSKII) (Fig. 5A). A construct containing the inhibitory 5'-UTR of ornithine decarboxylase mRNA (42-44) was used as control. After in vitro transcription of the cDNAs with T7 RNA polymerase, increasing amounts of capped RNA (Fig. 5B) were translated in vitro for a fixed period of time under linear assay conditions (17 min) using rabbit reticulocyte lysate, and the accumulation of luciferase protein was monitored using an enzymatic reaction (Fig. 5C). This experiment clearly indicated that the 5'-UTR of socs-1 was sufficient to suppress (~17-fold at 0.25 µg of RNA) translation of luciferase RNAs in vitro. Its inhibitory impact was actually stronger than that of the 5'-UTR of ornithine decarboxylase (~4-fold inhibition at 0.25 µg of RNA).

View larger version (27K): [in this window] [in a new window]   Fig. 5.   The 5'-UTR of socs-1 is sufficient to inhibit translation. A, schematic representation of the constructs used in this Fig. The 5'-UTR of socs-1 or ornithine decarboxylase was cloned immediately upstream of the luciferase ATG. Luciferase RNA expression was initiated at the T7 promoter in pBSK. B, in vitro transcription. Ethidium bromide staining of a formaldehyde-agarose gel containing RNA samples (0.5 µg) from in vitro transcription. C, in vitro translation. Various concentrations of capped RNAs were translated in rabbit reticulocyte lysates for 17 min. Luciferase activity (expressed in arbitrary units) was measured as outlined under "Materials and Methods." This assay was repeated under various conditions at least five times. The results shown here are representative.

The Inhibitory Effect of the 5'-UTR of socs-1 Is Probably Due to Upstream AUGs-- These results prompted us to examine more closely the characteristics of the 5'-UTR of socs-1. As represented in Fig. 6A, this sequence is highly conserved (~75% identity) between mouse and human socs-1. In mouse socs-1, it contains an upstream open reading frame, including two AUGs, as well as a stop codon (UAG) positioned immediately upstream of the bona fide AUG (Fig. 6, A and B). A similar, albeit not exactly identical, open reading frame is also present in human socs-1 (see legend of Fig. 6B). This feature could be significant, since upstream AUGs have been shown to inhibit translation in other RNAs, presumably by interfering with ribosomal scanning to downstream AUGs (42, 43, 45, 46). In addition, the socs-1 5'-UTR possesses a very high content (~80%) of G and C nucleotides. Based on computer predictions of secondary structure (data not shown), this feature could allow the formation of multiple internal hairpin loops that could contribute to the suppression of SOCS-1 protein synthesis (47). Last, despite the similarities between the mouse and human sequences, human socs-1 bears an additional GC-rich 19-nucleotide insertion in the body of exon 1. The basis and significance of this alteration are unclear.

View larger version (41K): [in this window] [in a new window]   Fig. 6.   Sequence of the 5'-untranslated region of socs-1 transcripts. A, alignment of the 5'-UTR of mouse and human socs-1. The positions of the upstream AUGs and the initiating methionine are highlighted. The stop codons (UAGs) for the putative upstream open reading frames are underlined. Identical nucleotides between mouse and human socs-1 are boxed. B, upstream open reading frame in the 5'-UTR of mouse socs-1. The location of the potential upstream open reading frame in the 5'-UTR of mouse socs-1 is shown. It should be noted that a related open reading frame initiating at the first AUG may also exist in human socs-1. However, in that case, the second AUG is in a different reading frame and the stop codon is located immediately 3' of the bona fide AUG (see A).

The possible role of the two upstream AUGs was examined (Fig. 7). Each AUG was individually replaced through polymerase chain reaction by a UUG, a codon that is inefficiently recognized by the ribosome as a site of translation initiation. When expressed in COS-1 cells, the resulting cDNAs, termed AUG1  UUG1 (Fig. 7A, lanes 7-9) and AUG2  UUG2 (lanes 10-12), also provoked low levels of SOCS-1 polypeptide expression, in a manner roughly comparable with the wild-type cDNA (lanes 4-6). However, it should be pointed out that we consistently observed that mutation of the first AUG caused a small (less than 2-fold) increase in SOCS-1 protein expression (Fig. 7A; data not shown). Since upstream AUGs are known to be functionally redundant in their ability to suppress translation, we also tested the impact of replacing both AUGs by UUGs (Fig. 7, C and D). Interestingly, mutation of the two upstream AUGs (Fig. 7C, lanes 7-9) was found to alleviate fully the inhibitory effect of the 5'-UTR of socs-1. Based on this finding, we concluded that the inhibitory effect of the 5'-UTR of socs-1 was caused in large part by the upstream AUGs.

View larger version (38K): [in this window] [in a new window]   Fig. 7.   Impact of upstream AUGs of the 5'-untranslated region on translation efficiency. A and B, effects of individual mutations of upstream AUGs. The indicated constructs were transiently transfected in COS-1 cells. A, anti-SOCS-1 immunoblot. The levels of expression of SOCS-1 protein were analyzed after immunoprecipitation using 3-fold dilutions of cell lysates (lanes 1, 4, 7, and 10, 1.5 mg; lanes 2, 5, 8, and 11, 500 µg; lanes 3, 6, 9, and 12, 167 µg). The positions of prestained molecular weight markers are shown on the right, whereas that of SOCS-1 is indicated on the left. Exposure was for 15 s. B, RNase protection assay. The abundance of socs-1 transcripts was determined using 5-fold dilutions of RNA (lanes 4, 7, 10, and 13, 0.5 µg; lanes 5, 8, 11, and 14, 0.1 µg; lanes 6, 9, 12, and 15, 0.02 µg). For lane 3, 30 µg of RNA from thymus was examined as a positive control. The migrations of the undigested probe and of the protected riboprobe fragments are shown on the right, while those of radiolabeled size markers are indicated on the left. Exposure was for 20 h. C and D, impact of mutation of both upstream AUGs. The indicated constructs were transiently transfected in COS-1 cells, and the expression of SOCS-1 protein and RNA was monitored as described for A and B. Exposures were for 10 s (C) and 36 h (D).

Translation of SOCS-1 Is Cap-dependent-- In most cases, the initiation of translation occurs by interaction of the ribosomal 40 S subunit with the 5' cap structure of the messenger via the cap-binding proteins (cap-dependent translation) (48). However, in some situations, the ribosomal 40 S subunit binds to an internal ribosomal entry site (IRES) in the 5'-UTR (cap-independent translation). These two mechanisms are not mutually exclusive. In order to determine which one(s) was responsible for translation of socs-1 mRNAs, we took advantage of a bicistronic reporter system (rluc-fluc), which was shown in earlier studies to distinguish successfully between cap-dependent and cap-independent translation (33). As shown in Fig. 8A, this vector produces a T7 RNA polymerase-driven bicistronic transcript that codes for the Renilla luciferase as the first cistron and the firefly luciferase as the second cistron. Whereas translation of the Renilla luciferase is cap-dependent, the firefly luciferase can be synthesized only if it is preceded by an IRES. Hence, to assess whether translation of SOCS-1 occurred through an IRES-mediated mechanism, its 5'-UTR was inserted upstream of the firefly luciferase sequence (rluc-5'-UTR SOCS-1-fluc). A plasmid in which the 5'-UTR of poliovirus, which possesses a proven IRES (49), was cloned upstream of the firefly luciferase (rluc-POLIRES-fluc) was used as positive control.

View larger version (40K): [in this window] [in a new window]   Fig. 8.   Translation initiation of SOCS-1 is cap-dependent. A, schematic representations of the bicistronic reporter constructs used (see "Results" for details). B, luciferase assays. HeLa cells were infected with recombinant vaccinia virus vTF7-3 and transfected with 0.75 µg of the indicated plasmids. Renilla and firefly luciferase activities were measured according to the manufacturer's instructions. Luciferase activity is expressed in arbitrary units. This assay was repeated at least three times. The results shown here are representative.

When transfected in HeLa cells in the presence of vaccinia virus-encoded T7 RNA polymerase, all constructs yielded similar amounts of Renilla luciferase activity (Fig. 8B, bottom panel), indicating that they were capable of comparable degrees of cap-dependent translation. However, only the construct encompassing the 5'-UTR of poliovirus produced the firefly luciferase (top panel). On this basis, we concluded that translation of socs-1 was unlikely to occur via a cap-independent mechanism and, therefore, was most probably cap-dependent. Further support for this notion will be presented below (Fig. 9).

View larger version (27K): [in this window] [in a new window]   Fig. 9.   Impact of 4E-BP1 overexpression on socs-1 translation. A, luciferase assays. HeLa cells were infected in triplicate with recombinant vaccinia virus vTF7-3 and transfected with the indicated reporter plasmids in the presence or absence of HA-tagged versions of 4E-BP1. Luciferase activity was measured using equivalent amounts (5 µg) of whole cell extracts. Each value represents the average of three independent transfections. S.D. values are shown. B, quantitation of data from Fig. 9A. The extent of translational inhibition of luciferase alone and 5'-UTR socs-1-luciferase by the two forms of 4E-BP1 is compared. C, anti-HA immunoblot. Expression of wild-type 4E-BP1 and a variant unable to associate with eIF4E (BP1-4E) was determined by immunoblotting of total cell lysates (15 µg) with anti-HA monoclonal antibody 12CA5. The position of the HA-tagged variants of 4E-BP1 is indicated on the right. It is of note that the deletion mutant (lanes 5 and 6) migrated more slowly than the wild-type protein (lanes 3 and 4). This alteration was also noted in other studies, and its basis was not determined. Exposure was for 36 h. D, RNase protection assay. For each transfection in A, total RNA was extracted and probed by RNase protection assay using a luciferase-specific riboprobe. Three-fold dilutions of RNA (1, 0.33, and 0.11 µg) were used. The migrations of the undigested probe and of the protected riboprobe fragment are indicated on the right, while those of radiolabeled size markers are shown on the left. Exposure was for 10 h.

Cap-dependent translation requires the participation of several proteins forming a large complex that binds the 5' cap structure of the mRNA (50). This complex, termed eIF4F, binds to the cap and is thought to unwind mRNA secondary structure, thus facilitating the binding of ribosomal RNA. eIF4F consists of three subunits: eIF4A (an RNA helicase), eIF4G (a scaffolding molecule) and eIF4E (the "cap-binding" protein). The latter is thought to be the rate-limiting component for translation. Importantly, in quiescent cells, eIF4E is functionally inactivated by binding to at least three eIF4E-binding proteins (4E-BP1, -2, and -3), which prevent its interaction with eIF4G and thus inhibit translation (50-53). Upon mitogenic stimulation, 4E-BPs become phosphorylated on several serine and threonine residues, causing the release of eIF4E, full assembly of eIF4F, and initiation of protein synthesis.

To examine whether expression of the SOCS-1 protein can be regulated by modulation of the translation machinery, we assessed the impact of overexpression of one of the eIF4E-binding proteins (4E-BP1) on the inhibitory effect of the socs-1 5'-UTR. To facilitate our analyses, a construct in which the 5'-UTR was fused to the luciferase gene was utilized (Fig. 9). HeLa cells were transfected and assayed for luciferase activity and RNA as detailed under "Materials and Methods," using plasmids under the control of the T7 promoter. This experiment showed that, as expected, luciferase translation was diminished (~4-fold) by the 5'-UTR of socs-1 (Fig. 9A, compare 5'-UTR SOCS-1-luc with luc). It was further suppressed (for a total reduction of ~20-fold) by overexpression of wild-type 4E-BP1. No inhibition occurred when a 4E-BP1 mutant unable to associate with eIF4E (4E-BP1-4E) was used, indicating that the effect of 4E-BP1 was due to inhibition of eIF-4E activity. Importantly, all luciferase constructs resulted in similar amounts of luciferase RNA, as judged by a parallel RNase protection assay (Fig. 9D). In addition to supporting further the notion that translation of socs-1 is cap-dependent, this finding implied that changes in the abundance and/or activity of eIF4E-binding proteins may modulate the efficiency of this process in mammalian cells. It is noteworthy, though, that the sensitivity of 5'-UTR SOCS-1-luciferase to the effect of 4E-BP1 was comparable with or perhaps slightly lower than that of luciferase alone (Fig. 9B), suggesting that socs-1 may not be more dependent on eIF4E than other cap-dependent RNAs.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Herein, we have provided evidence that the expression of at least one SOCS family member, i.e. SOCS-1, can be modulated at the level of translational initiation. This observation was prompted by our finding that protein expression from endogenous socs-1 RNAs in T-cells was inefficient. Further studies demonstrated that this feature was due to repression of translation by the 5'-UTR of socs-1. Our analyses indicated that the 5'-UTR strongly suppressed translation both in transfected mammalian cells and in in vitro transcription-translation assays. The inhibitory effect of the 5'-UTR was self-autonomous, since it did not require the presence of the 3'-UTR or the coding region of socs-1. Interestingly, its impact was stronger than that of the 5'-UTR of ornithine decarboxylase, a well known repressor of translation initiation (42-44).

5'-UTRs can inhibit translation through several mechanisms (48, 56). After ribosome binding to the cap (cap-dependent translation) or to an IRES (cap-independent translation), ribosomal scanning along the 5'-UTR can been inhibited by several types of elements. First, the presence of upstream AUGs can interfere with translation at the bona fide AUG, by allowing translation of an alternative open reading frame. The finding that mutation of the two upstream AUGs of socs-1 essentially abolished the inhibitory effect on translation strongly suggested that this mechanism was largely responsible for translation repression. Presumably, translation was prone to initiate at one or both of the upstream AUGs, thus terminating at a stop codon (UAG) located just 5' of the bona fide AUG (Fig. 6). Since the stop codon is positioned too close to the bona fide AUG, this feature presumably prevented efficient reinitiation of translation, resulting in a dramatic reduction of SOCS-1 protein expression. Second, 5'-UTRs can down-regulate translation by forming stable secondary structures that block ribosomal binding and movement. While insufficient to repress translation, the GC-rich content of the 5'-UTR of socs-1 may facilitate the inhibitory effect of the upstream AUGs, perhaps by augmenting translation initiation at these sites. Future studies will be required to address this possibility. Finally, 5'-UTRs can possess binding sites for cellular proteins that block the interaction of the ribosome with the cap. The best characterized example of this mechanism is provided by ferritin mRNAs, in which two proteins termed iron regulatory protein 1 and 2 bind to the 5'-UTR in the absence of iron, thereby repressing ferritin synthesis (57, 58). Since these proteins recognize primary nucleotide sequences in the 5'-UTR, they would not be expected to bind the RNA when the 5'-UTR is inverted. In light of our observation that the 5'-UTR of socs-1 still inhibited translation when inserted in the antisense orientation,⁴ we believe that it is unlikely that its influence was mediated via this mechanism. It is of note that the inversion of the 5'-UTR also created a novel upstream AUG, which presumably explained the persisting inhibitory effect of the antisense 5'-UTR on translation.

The inability of the 5'-UTR of socs-1 to function as an IRES is consistent with the idea that translation of SOCS-1 is cap-dependent. This conclusion is also supported by the finding that an inhibitor of the cap-binding protein eIF4E, i.e. 4E-BP1, enhanced the inhibitory impact of the 5'-UTR on translation. Given that the functions of eIF4E and 4E-BPs can be regulated by a variety of physiological stimuli (50, 53, 59, 60), it is plausible that the translation of socs-1 is also dynamically controlled in mammalian cells. Changes in the abundance and/or phosphorylation of eIF4E or 4E-BPs could either alleviate or accentuate the inhibitory effect of the 5'-UTR. The results of our experiments with 4E-BP1 overexpression are certainly in agreement with this idea. Another possibility is that changes in RNA helicase activity may regulate socs-1 translation by unfolding secondary structures in the 5'-UTR. This argument is substantiated by the observation that the repressive effect of the 5'-UTR of socs-1 is augmented by expression of a dominant-negative eIF4A (an RNA helicase) in HeLa cells.⁴ Since there is little known of the physiological mechanisms controlling helicase function, additional experiments will be necessary to examine this intriguing possibility.

At this time, the exact purpose on translational repression of socs-1 is not known. However, it is reasonable to presume that this phenomenon helps minimize the levels of SOCS-1 protein in the cell, thus avoiding undue interference with cytokine receptor signaling. Under conditions of active cell signaling leading, for example, to phosphorylation of eIF4E-BPs, induction of helicase activity, or other effects, translational inhibition of SOCS-1 may be acutely relieved, permitting a rise in SOCS-1 protein expression and inhibition of cytokine-induced responses. Transcriptional activation of the socs-1 gene may further contribute to this effect. Although perhaps less appealing, it is also plausible that the repression of socs-1 translation is constitutive rather than regulated. In this context, it could help maintain SOCS-1 protein levels at a minimum in the cell, as has been postulated with other proteins playing critical roles in cellular proliferation and differentiation such as p53, Myc, Lck, and ornithine decarboxylase (61-63). The presence of proportionally larger amounts of socs-1 RNA may simply permit a more diffuse distribution of SOCS-1 polypeptides in the cytoplasm. Obviously, these two possibilities are not mutually exclusive.

Our observations also raise interesting questions regarding the interpretation of previous studies on enforced SOCS-1 expression. It is noteworthy that, in these reports, the socs-1 cDNAs used for overexpression always lacked the 5'-UTR. Thereby, they would be expected to provoke a very marked enhancement of SOCS-1 protein expression, probably much greater than that achieved under any physiological circumstance. It is possible that the impact of SOCS-1 on cytokine receptor signaling was not only quantitatively accentuated, but perhaps qualitatively altered, as a result of such a degree of overexpression. Some experimental support for this idea has been adduced for SOCS-2, another member of the SOCS family (64). In that situation, low amounts of SOCS-2 were noted to inhibit growth hormone receptor signal transduction, whereas higher quantities restored signaling through this receptor.

Together, these results show that SOCS-1 is regulated not only at the levels of transcription and protein stability but also by a translational mechanism. Since other socs family members such as socs-3 also bear upstream AUGs, this notion could apply broadly within this group of molecules.

	ACKNOWLEDGEMENTS

We thank members of our laboratories for useful discussions. We also acknowledge A.-C. Gingras, F. Poulin, and H. Imataka for gifts of reagents.

	FOOTNOTES

^* This work was supported by grants from the Medical Research Council of Canada and the National Cancer Institute of Canada (to A. V. and N. S).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBank^TM/EMBL Data Bank with accession number(s) AF180302.

^¶ Supported by a studentship from the Medical Research Council of Canada.

Distinguished Scientist of the Medical Research Council of Canada.

^§§ Senior Scientist of the Medical Research Council of Canada. To whom correspondence should be addressed: IRCM, 110 Pine Ave. W., Montréal, Québec H2W 1R7, Canada. Tel.: 514-987-5561; Fax: 514-987-5562; E-mail: veillea@ircm.qc.ca.

Published, JBC Papers in Press, April 10, 2000, DOI 10.1074/jbc.M910087199

² S. Pyronnet and N. Sonenberg, unpublished results.

³ H. Imataka and N. Sonenberg, unpublished results.

⁴ A. Gregorieff, S. Pyronnet, N. Sonenberg, and A. Veillette, unpublished results.

	ABBREVIATIONS

The abbreviations used are: IL, interleukin; IFN, interferon; STAT, signal transducers and activators of transcription; HA, hemagglutinin; UTR, untranslated region; IRES, internal ribosomal entry site.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				M. C. Haffner, A. Jurgeit, C. Berlato, S. Geley, N. Parajuli, A. Yoshimura, and W. Doppler Interaction and Functional Interference of Glucocorticoid Receptor and SOCS1 J. Biol. Chem., August 8, 2008; 283(32): 22089 - 22096. [Abstract] [Full Text] [PDF]

				P. Isomaki, T. Alanara, P. Isohanni, A. Lagerstedt, M. Korpela, T. Moilanen, T. Visakorpi, and O. Silvennoinen The expression of SOCS is altered in rheumatoid arthritis Rheumatology, October 1, 2007; 46(10): 1538 - 1546. [Abstract] [Full Text] [PDF]

				J. M. Zimmerer, G. B. Lesinski, S. V. Kondadasula, V. I. Karpa, A. Lehman, A. RayChaudhury, B. Becknell, and W. E. Carson III IFN-{alpha}-Induced Signal Transduction, Gene Expression, and Antitumor Activity of Immune Effector Cells Are Negatively Regulated by Suppressor of Cytokine Signaling Proteins J. Immunol., April 15, 2007; 178(8): 4832 - 4845. [Abstract] [Full Text] [PDF]

				J. Travagli, M. Letourneur, J. Bertoglio, and J. Pierre STAT6 and Ets-1 Form a Stable Complex That Modulates Socs-1 Expression by Interleukin-4 in Keratinocytes J. Biol. Chem., August 20, 2004; 279(34): 35183 - 35192. [Abstract] [Full Text] [PDF]