Chronic septic abscess formation causes an inhibition of protein synthesis in gastrocnemius not observed in rats with a sterile abscess. Inhibition is associated with an impaired mRNA translation initiation that can be ameliorated by elevating IGF-I but not insulin. The present study investigated the ability of IGF-I signaling to stimulate protein synthesis in gastrocnemius by accelerating mRNA translation initiation. Experiments were performed in perfused hindlimb preparations from rats 5 days after induction of a septic abscess. Protein synthesis in gastrocnemius from septic rats was accelerated twofold by the addition of IGF-I (10 nM) to perfusate. IGF-I increased the phosphorylation of translation repressor 4E-binding protein-1 (4E-BP1). Hyperphosphorylation of 4E-BP1 in response to IGF-I resulted in its dissociation from the inactive eukaryotic initiation factor (eIF) 4E·4E-BP1 complex. Assembly of the active eIF4F complex (as assessed by the association eIF4G with eIF4E) was increased twofold by IGF-I in the perfusate. In addition, phosphorylation of eIF4G and ribosomal protein S6 kinase-1 (S6K1) was also enhanced by IGF-I. Activation of mammalian target of rapamycin, an upstream kinase implicated in phosphorylating both 4E-BP1 and S6K1, was also observed. Thus the ability of IGF-I to accelerate protein synthesis during sepsis may be related to a stimulation of signaling to multiple steps in translation initiation with an ensuing increased phosphorylation of eIF4G, eIF4E availability, and S6K1 phosphorylation.
- messenger ribonuclease
- translation initiation
- eukaryotic initiation factors
- eukaryotic initiation factor 4G
- 4E-binding protein-1
- eukaryotic initiation factor 4E
- ribosomal protein S6 kinase-1
sepsis, the systemic manifestation of bacterial infection (1), induces profound alterations in whole body protein metabolism (64). Nitrogen losses as high as 17% of total body protein may be observed in septic patients despite aggressive nutritional support (for example see Ref. 32). Because skeletal muscle comprises ∼45% of body weight, whole body nitrogen balance reflects changes in protein turnover in muscle during sepsis. As such, ∼70% of the septic-induced whole body protein loss comes from erosion of skeletal muscle (32).
Protein wasting in skeletal muscle results from an imbalance between the rate of protein synthesis and protein degradation during sepsis (for review see Refs. 6, 46, and 47). Synthesis of proteins is a complex process in mammalian cells (for review see Ref. 14). The process involves the association of the 40S and 60S ribosomal subunits, mRNA, initiator methionyl-tRNA (met-tRNAimet), other amino acyl-tRNAs, cofactors (i.e., GTP; ATP), and protein factors, collectively known as eukaryotic initiation factors (eIF), eukaryotic elongation factors (eEF), and releasing factors (RF), through a series of discrete reactions that lead to translation of mRNA into proteins. Translation of mRNA is comprised of three phases: 1) initiation, whereby met-tRNAimet and mRNA bind to 40S ribosomal subunits, and subsequent binding of the 40S ribosomal subunit to the 60S subunit to form a ribosome complex capable of translation; 2) elongation, by which tRNA-bound amino acids are incorporated into growing polypeptide chains according to the mRNA template; and 3) termination, where the completed protein is released from the ribosome.
Sepsis causes a preferential slowing in mRNA translation initiation in the protein synthetic pathway (55). Two steps in the process of mRNA translation initiation appear to control protein synthesis, namely the binding of met-tRNAimet to the 40S ribosomal subunit to form the 43S preinitiation complex and the binding of mRNA to the 43S preinitiation complex, which is mediated by eIF4F. Furthermore, our reports indicate that sepsis inhibits both of these regulatory steps involved in mRNA translation initiation (5, 16, 43, 53, 55, 58, 59).
Binding of mRNA to the 43S preinitiation complex is catalyzed by a multisubunit complex of eukaryotic factors referred to as eIF4F (38, 39, 42). The eIF4F complex serves to recognize, unfold, and guide the mRNA to the 43S preinitiation complex. eIF4F is composed of 1) eIF4A (a RNA helicase that functions with eIF4B to unwind secondary structure in the 5′ untranslated region of mRNA), 2) eIF4E (a protein that binds directly to the m7GTP cap structure present at the 5′ end of most eukaryotic mRNAs), and 3) eIF4G (a protein that functions as a scaffold among eIF4E, eIF4A, mRNA, and the ribosome). eIF4G appears to be the nucleus around which the initiation complex forms, because it has binding sites not only for eIF4E but also for eIF4A and eIF3 (21).
Assembly of an active eIF4E·eIF4G complex may limit mRNA translation initiation in skeletal muscle. We were the first to report a positive linear relationship between rates of protein synthesis and amount of eIF4G associated with eIF4E in muscle in vivo (51). Although this correlation does not prove cause and effect, the relationship between protein synthesis and amount of eIF4G associated with eIF4E is consistent with the proposed role of the eIF4G·eIF4E complex in the overall regulation of protein synthesis. We were the first to show that the assembly of the eIF4E·eIF4G complex is significantly diminished in skeletal muscle from septic rats (51, 54). Diminished cellular content of either eIF4E or eIF4G is not the reason for deceased assembly of the eIF4E·eIF4G complex. Instead, reduced amounts of eIF4E associated with eIF4G after chronic sepsis would be expected to attenuate the association of mRNA with the ribosome and hence limit protein synthesis.
The interaction between eIF4E and eIF4G can be regulated, at least in part, by several mechanisms including the availability of eIF4E and/or phosphorylation of eIF4G. Availability of eIF4E is controlled through binding to the translation initiation repressor 4E-binding protein-1 (4E-BP1) to form an inactive eIF4E·4E-BP1 complex. Ability of eIF4E to be dissociated from this inactive complex is dependent on phosphorylation of 4E-BP1. In addition, phosphorylation of the eIF4G on Ser1108 may modulate the ability of eIF4G to bind with eIF4E (28, 37). Increased phosphorylation of eIF4G correlates with conditions known to stimulate protein synthesis (37).
Previous reports from this laboratory (43, 45) established that IGF-I stimulates protein synthesis in muscles from septic rats through accelerating mRNA translation initiation. However, the potential mechanisms responsible for the anabolic effects of IGF-I in skeletal muscle during sepsis remain unresolved. We hypothesize that IGF-I increases assembly of eIF4G·eIF4E complex in gastrocnemius of septic rats through mammalian target of rapamycin (mTOR)-dependent regulation of eIF4E availability and/or phosphorylation of eIF4G, thereby stimulating mRNA translation initiation and, hence, protein synthesis. We performed these investigations using the perfused hindlimb preparation because it allows a defined medium to be circulated through the vasculature of the gastrocnemius, thereby alleviating confounding variables associated with changes in other metabolites and hormones inherent in studies using septic animals in vivo.
MATERIALS AND METHODS
Adult male Sprague-Dawley rats were maintained on a 12:12-h light-dark cycle with water and standard rat chow provided ad libitum for at least 1 wk before the experimental protocol. With the animals anesthetized, sepsis was induced by implanting a fecal-agar pellet inoculated with 105 colony-forming units (CFU) of Escherichia coli and 109 CFU of Bacteroides fragilis into the abdominal cavity (15, 16, 51, 54, 55). The intra-abdominal abscess was allowed to develop for 5 days. Septic rats develop an abdominal abscess and show profound leukocytosis (57) with a hyperdynamic, hypermetabolic septic condition (29).
Hindlimb perfusions were carried out according to the methods described by Bylund-Fellenius et al. (4) as modified in our laboratory (15, 17, 45, 50–52). Five days after implantation of the fecal-agar pellet, rats were anesthetized with pentobarbital sodium (50 mg/kg body wt) and the skin covering the right and left hindlimbs was removed. A midline incision was made, and both the inferior vena cava and the abdominal aorta were exposed. The abdominal aorta was cannulated and immediately perfusate was delivered via the abdominal aorta at a rate of 0.32 ml·min−1·g−1 (4, 17, 45, 51) to the hindlimb musculature. The inferior vena cava below the level of the renal vasculature was then cannulated. The first 50 ml of perfusate passing through the hindlimb was discarded. The inferior vena cava cannula was then connected to the perfusion system, and recirculation of the perfusate began. After perfusion for an additional 5 min, [3H]phenylalanine was introduced into the perfusate to give a final concentration of 2 μCi/ml, and perfusion continued for 30 min. After perfusion, gastrocnemius were excised, weighed, and frozen between aluminum blocks precooled to the temperature of liquid nitrogen. All tissues were stored at −85°C until analysis. A perfusate sample was withdrawn and centrifuged to remove red blood cells. Plasma samples were stored at −20° until they were analyzed to determine phenylalanine-specific radioactivity.
Perfusate (250 ml/hindlimb) consisted of a modified Krebs-Henseleit bicarbonate buffer containing 30% (vol/vol) washed bovine erythrocytes, 4.5% (wt/vol) BSA (BSA fraction V), 11 mM glucose, 1 mM phenylalanine, and all other amino acids at normal rat plasma concentrations as previously described (4, 17, 45, 51). The concentration of human recombinant IGF-I in the perfusate was 10 nM (17, 45, 51). The medium was maintained at 37°C and gassed with humidified 95% O2-5% CO2.
The rate of protein synthesis in vivo was measured by incorporation of radioactive phenylalanine into protein. The frozen tissue samples were powdered under liquid nitrogen with a mortar and pestle, and a portion was used to estimate the amount of [3H]phenylalanine incorporated into total mixed protein as previously described (4, 17, 45, 51). The specific radioactivity of phenylalanine in the plasma was measured by high-performance liquid chromatography analysis of supernatants from trichloroacetic acid extracts of the plasma (9). Rates of protein synthesis were calculated by dividing the amount of [l-3H]phenylalanine incorporated in protein per hour by the specific radioactivity of the perfusate phenylalanine as the precursor pool. Bylund-Fellenuis and coworkers (4) have provided evidence that at perfusate concentrations >0.8 mM, the specific radioactivity of tRNA-bound phenylalanine is equivalent to the extracellular and intracellular pools of free phenylalanine. Therefore, the specific radioactivity of perfusate phenylalanine provides an accurate estimate of the specific radioactivity of phenylalanine tRNA. Protein content in tissue homogenates was measured using the Biuret method with crystalline BSA used as a standard.
Frozen powdered gastrocnemius was homogenized in 7 volumes of buffer A (in mM: 20 HEPES, pH 7.4, 100 KCl, 0.2 EDTA, 2 EGTA, 1 DTT, 50 NaF, 50 β-glycerolphosphate, 0.1 PMSF, 1 benzamidine, and 0.5 sodium vanadate, plus 1 μM microcystin LR) using a Polytron PT10 homogenizer. The homogenate was either used directly or centrifuged at 10,000 g for 10 min at 4°C, and the pellet was discarded. The supernatant was used to evaluate the regulation of the eIF4E system. eIF4E, as well as 4E-BP1·eIF4E and eIF4G·eIF4E complexes, were immunoprecipitated from aliquots of 10,000 g supernatants using an anti-eIF4E monoclonal antibody. Another aliquot of the 10,000 g supernatant was mixed with an equal volume of 2× Laemmli SDS sample buffer (65°C) and was then subjected to protein immunoblot analysis of S6K1, mTOR, eIF4G, and 4E-BP1 phosphorylation. Another aliquot was used to measure the protein concentration by the Biuret method with crystalline BSA as a standard.
Quantification of 4E-BP1·eIF4E and eIF4G·eIF4E complexes.
Association of eIF4E with 4E-BP1 or eIF4G was determined in gastrocnemius after immunoprecipitation of eIF4E using immunoblot techniques as previously described in our laboratory (48, 49, 51, 52, 54, 56). The antibody-antigen complex was collected by incubation for 1 h with BioMag goat anti-mouse IgG beads (PerSeptive BioSystems, Framingham, MA). Before use, the beads were washed in 1% nonfat dry milk in buffer B [in mM: 50 Tris·HCl (pH 7.4) 150 NaCl, 5 EDTA, plus 0.1% β-mercaptoethanol and 0.5% Triton X-100]. Beads were captured using a magnetic sample rack and washed twice with buffer B and once with buffer B containing 500 mM rather than 150 mM NaCl. Resuspending in SDS-sample buffer and boiling for 5 min eluted protein bound to the beads. Beads were captured using a magnetic sample rack, and the supernatants were subjected to electrophoresis either on a 7.5% polyacrylamide gel for quantization of eIF4G or on a 15% polyacrylamide gel for quantization of 4E-BP1 and eIF4E. Proteins were then electrophoretically transferred to a PVDF membrane (Biotrace, PALL, Pensacola, FL). PVDF membranes were incubated with a mouse anti-human eIF4E antibody (kindly provided by Dr. Leonard S. Jefferson, Penn State University College of Medicine, Hershey, PA), a rabbit anti-rat 4E-BP1 antibody, or a rabbit anti-eIF4G antibody (Bethyl Laboratories, Montgomery, TX) overnight at 4°C. Blots were then developed using an enhanced chemiluminesence (ECL) Western blot analysis kit as per the manufacturer's instructions (Amersham Pharmacia Biotech, Piscataway, NJ). Films were scanned using a Microtek ScanMaker III scanner equipped with a transparent media adaptor connected to a Macintosh computer. Images were obtained using the ScanWizard Plugin (Microtek) for Adobe Photoshop and quantitated using NIH Image 1.63 software. Abundance of eIF4G and 4E-BP1 was normalized to the amount of eIF4E in the immunoprecipitate.
Determination of phosphorylation state of eIF4G.
To measure the relative extent of phosphorylation of eIF4G, homogenate proteins were separated by 7.5% SDS-PAGE. After electrophoresis, the proteins were transferred to PVDF membranes (Biotrace, PALL, Pensacola, FL). The membranes were incubated overnight at 4°C with antibodies specific for phosphorylated eIF4G (Ser1108) (Cell Signaling Technology, Beverly, MA). The blots were then developed using an ECL Western blot analysis kit as per the manufacturer's instructions (Amersham Pharmacia Biotech). Films were scanned and analyzed as described in Quantification of 4E-BPI·eIF4E and eIF4G·eIF4E. After development of the immunoblot, membranes were treated with a solution containing 62.5 mM Tris·HCl (pH 6.7), 100 mM β-mercaptoethanol, and 2% (wt/vol) SDS to remove antibodies as per the manufacturer's instructions. This procedure effectively removed all signals resulting from incubation with the phospho-eIF4G antibody (data not shown). Membranes were blocked with nonfat dry milk and then were immunoblotted with the antibody that recognizes eIF4G independently of its phosphorylation state (Bethyl Laboratories). Blots were developed using ECL (Amersham Pharmacia Biotech), and the autoradiographs were scanned and analyzed as described above. The phosphorylated eIF4G signal densities were normalized to the respective total eIF4G signal to reflect the relative ratio of phosphorylated eIF4G to total eIF4G.
Determination of phosphorylation state of 4E-BP1.
The phosphorylated forms of 4E-BP1 were measured in homogenates of gastrocnemius after boiling of an aliquot (200 μl) of muscle tissue homogenates for 5 min. The boiled homogenate was centrifuged in a microcentrifuge at room temperature and supernatant was decanted. An equal volume of 2× Laemmli SDS buffer (65°C) was then added to the supernatant. The various phosphorylated forms of 4E-BP1 (designated α, β, and γ) were separated by SDS-PAGE and quantitated by protein immunoblot analysis as described previously (48, 49, 51, 52, 54, 56).
Determination of phosphorylation state of S6K1.
Avruch and coworkers (62) have provided evidence that after phosphorylation, S6K1 activity in vivo is most closely related to the phosphorylation of residue Thr389. To examine the phosphorylation of S6K1, homogenates of skeletal muscle were mixed with 2× Laemmli SDS sample buffer and subjected to electrophoresis on 12.5% SDS-PAGE. Proteins were then electrophoretically transferred onto PVDF membranes and blocked with Tris-buffered saline supplemented with 0.1% Tween and containing 5% (wt/vol) nonfat dry milk. Initially, the PVDF membranes were probed with antibody that only recognizes S6K1 phosphorylated on amino acid Thr389 (Cell Signaling Technology), the phosphorylation site required for activation of the kinase (62). Blots were developed using an ECL Western blot analysis kit according to the manufacturer's (Amersham Biosciences) instructions. We used this property as an indicator of the effect of IGF-I on the activation of the kinase. After quantification of the relative intensity of the signal for phosphorylation at Thr389, the site-specific anti-phosphopeptide antibodies were removed from PVDF membranes as described above in Determination of phosphorylation state of eIF4G. Blots were then probed with an antibody (Santa Cruz Biotechnology, Santa Cruz, CA) that recognizes total S6K1 (i.e., both phosphorylated and unphosphorylated forms). Results are presented as the ratio of the densitometric analysis of blot for phosphorylated S6K1(Thr389) divided by total S6K1 performed on the same gel.
Determination of phosphorylation of mTOR.
Another portion of the homogenate was electrophoresed and transferred as described above in Determination of phosphorylation state of eIF4G. The PVDF membranes were then incubated with an antibody that recognized the phosphorylated form of mTOR(Ser2448) or mTOR(Ser2481) (Cell Signaling Technology). Blots were developed and analyzed as described above in Determination of phosphorylation state of eIF4G. After development of the immunoblot, the PVDF membranes were treated to remove the phosphospecific antibodies as described above in Determination of phosphorylation state of eIF4G. The PVDF membranes were blocked with Tris-buffered saline supplemented with 0.1% Tween containing 5% (wt/vol) nonfat dry milk and then were immunoblotted with the antibody that recognizes mTOR independently of its phosphorylation state (Bethyl Laboratories). Blots were developed using ECL, and the autoradiographs were scanned and analyzed as described in Determination of phosphorylation state of eIF4G. The phosphorylated mTOR signal densities were normalized to the respective total mTOR signal to reflect the relative ratio of phosphorylated mTOR to total mTOR.
Results are expressed as means ± SE for 5–12 animals in each group. Statistical evaluation of data was performed using Student's t-test to test for differences among the groups. Differences among means were considered significant when P < 0.05. Least squares linear regression analysis was performed using INSTAT Software from GraphPad Software (San Diego, CA).
IGF-I stimulates protein synthesis during sepsis.
Perfused hindlimb preparations allow for examination of the effects of IGF-I on skeletal muscle independent of systemic perturbations in plasma components either 1) resulting from septic insult (e.g., lower plasma IGF-I concentrations and higher plasma IGFBP-1 concentrations) or 2) secondary to in vivo administration of IGF-I itself (e.g., alterations in plasma amino acid and glucose concentrations). Hence, the perfused hindlimb preparation was studied to investigate more directly the effects of IGF-I on control of protein synthesis in gastrocnemius from septic animals. Inclusion of IGF-I (10 nM) in the perfusate stimulated protein synthesis more than twofold in gastrocnemius of septic rats (Fig. 1), consistent with our previous finding, showing the inhibition of protein synthesis in fast-twitch muscles from septic rats was reversed by perfusion or incubation with buffer containing IGF-I (17, 45).
IGF-I-induced formation of active eIF4G·eIF4E complex is associated with an increased phosphorylation of eIF4G and dissociation of eIF4E from 4E-BP1.
Abundance of the eIF4G·eIF4E complex was determined by immunoprecipitating eIF4E from gastrocnemius using a monoclonal antibody followed by Western immunoblot analysis for eIF4E and eIF4G. IGF-I administration brought about a pronounced (twofold) increase in the abundance of eIF4G associated with eIF4E in skeletal muscle of septic rats (Fig. 2). Increased assembly of eIF4G·eIF4E complex was not the result of an increased abundance of eIF4E, because there were no significant differences in the eIF4E content between the two groups.
Interaction between eIF4E and eIF4G can be regulated, at least in part, by several mechanisms including availability of eIF4E and/or phosphorylation of eIF4G. Increased phosphorylation of eIF4G on Ser1108 is associated with enhanced formation of active eIF4G·eIF4E complex (37). We examined the phosphorylation state of eIF4G in gastrocnemius extracts from septic animals perfused with buffer supplemented with IGF-I. eIF4G phosphorylation on Ser1108 was analyzed by Western blot and corrected for the total amount of eIF4G in homogenates of skeletal muscle from septic rats. The extent of phosphorylation of eIF4G was increased twofold after perfusion with buffer supplemented with IGF-I (Fig. 3). Augmented phosphorylation of eIF4G after perfusion with buffer supplemented with IGF-I is not the result of alterations in skeletal muscle content of eIF4G, because the total cellular content of eIF4G is not different between the two experimental groups (data not shown).
Increasing the availability of eIF4E for formation of the eIF4G·eIF4E complex can occur through diminished binding of eIF4E to the translational repressor 4E-BP1. Binding of eIF4E to 4E-BP1 prevents the formation of an active eIF4E·eIF4G complex, presumably because 4E-BP1 blocks the binding site for eIF4G (13, 33). A shift in the distribution of eIF4E should lead to a lowering of the abundance of the inactive 4E-BP1·eIF4E complex after perfusion with buffer supplemented with IGF-I. The association of eIF4E with 4E-BP1 was determined in gastrocnemius by immunoprecipitating eIF4E with a monoclonal antibody followed by immunoblot analysis for eIF4E and 4E-BP1. Results were normalized to the total eIF4E in the immunoprecipitate. Perfusion of gastrocnemius with buffer supplemented with IGF-I decreased the formation of the inactive 4E-BP1·eIF4E complex (Fig. 4, top).
IGF-I increases phosphorylation of 4E-BP1.
The interaction of 4E-BP1 with eIF4E is regulated, in part, through phosphorylation of 4E-BP1 (11, 26, 35, 36). Dissociation of the eIF4E·4E-BP1 complex occurs when 4E-BP1 becomes hyperphosphorylated. Therefore, we examined the ability of IGF-I to induce phosphorylation of 4E-BP1 to determine whether the decreased assembly of 4E-BP1·eIF4E resulted from an increased phosphorylation of 4E-BP1. When phosphorylated, 4E-BP1 resolves into distinct electrophoretic forms (α, β, and γ) with the γ-form representing the most highly phosphorylated form. The results reveal that ∼13% of the total 4E-BP1 was present in the γ-form in gastrocnemius of septic rats (Fig. 4, bottom). The extent of phosphorylation of 4E-BP1 was significantly elevated after perfusion with buffer supplemented with IGF-I as indicated by retardation in the mobility of 4E-BP1 on SDS-PAGE. The percentage of 4E-BP1 present in the γ-form increased to 26% in muscles of septic rats perfused with buffer supplemented with IGF-I.
IGF-I increases phosphorylation of S6K1.
S6K1 is a threonine/serine kinase that phosphorylates ribosomal protein S6 and has been implicated in the stimulation of translation of mRNAs containing a 5′-TOPs sequence. Changes in phosphorylation of S6K1 often change in parallel with the phosphorylation of 4E-BP1 under a variety of conditions. S6K1 is activated by multisite phosphorylation that results in isoforms exhibiting retarded electrophoretic mobility when subjected to SDS-PAGE (10, 20). Analysis of the multisite phosphorylations of S6K1 indicates that its activity is associated with phosphorylation of Thr389 (62, 63). Therefore, we examined the effect of IGF-I on the phosphorylation of Thr389 in gastrocnemius of skeletal muscle from septic rats (Fig. 5). The percentage of S6K1 kinase phosphorylated at Thr389 increased approximately twofold in gastrocnemius perfused with buffer supplemented with IGF-I. The magnitude of the stimulation in phosphorylation of S6K1 in septic rats in the presence of IGF-I was similar to that observed in gastrocnemius from control animals (−IGF-I 340 ± 40 vs. +IGF-I 670 ± 95 arbitrary units of phospho-S6K1/total S6K1; P < 0.05).
IGF-I increases phosphorylation of mTOR.
mTOR is a serine/threonine kinase that phosphorylates both S6K1 and 4E-BP1. mTOR itself is phosphorylated on multiple sites. Phosphorylation of mTOR on residues Ser2448 and Ser2481 is a surrogate measure of mTOR activity (2, 31). As such, altered mTOR phosphorylation appears to represent an important nexus for the observed changes in translation initiation. In control animals, perfusion with buffer containing IGF-I enhanced the phosphorylation of mTOR at both Ser2448 (3.6-fold; −IGF-I 1.2 ± 0.1 vs. +IGF-I 4.2 ± 0.9 arbitrary units of phospho-mTOR/total mTOR; P < 0.05) and Ser2481 (4-fold; −IGF-I 3 ± 1 vs. +IGF-I 13 ± 3 arbitrary units phospho-mTOR/total mTOR; P < 0.01). There is no information of the effect of IGF-I on mTOR phosphorylation in skeletal muscle of septic rats. We examined the phosphorylation of mTOR using phosphospecific antibodies after perfusion of gastrocnemius with buffer supplemented with IGF-I. Phosphorylation of both Ser2448 and Ser2481 is significantly increased twofold after perfusion with buffer supplemented with IGF-I (Fig. 6).
Results of the present study provide additional evidence that IGF-I stimulates protein synthesis in gastrocnemius of chronic, hypermetabolic septic rats. Perfusion of hindlimb muscles with buffer supplemented with 10 nM IGF-I accelerates protein synthesis twofold, consistent with our previous reports (17). Because these studies were performed in an isolated perfused hindlimb preparation, the effects of IGF-I to enhance rates of protein synthesis represent direct effects of IGF-I on skeletal muscle and are not related to systemic perturbations secondary to infusion of IGF-I or changes in IGF binding proteins as a result of the septic insult (25). Furthermore, binding studies in gastrocnemius provide evidence that cross-reactivity of IGF-I with the insulin receptor is unlikely, given the concentration of IGF-I used in the present study (7). Likewise, twice daily injections of IGF-I/IGFBP-3 binary complex partially restores protein content in gastrocnemius through a restoration of rates of protein synthesis (43). These observations are in contrast with those in liver (8, 43), where elevations in plasma IGF-I did not modulate hepatic protein synthesis, yet improved cumulative nitrogen balance compared with untreated septic rats. Thus protein synthesis in gastrocnemius seems more responsive to IGF-I than other organs of the body during sepsis.
IGF-I accelerates protein synthesis by reversing the inhibition of translation initiation characteristic of the septic response in muscles composed of fast-twitch fibers (17, 43). Analysis of ribosomal subunits indicated IGF-I improved translational efficiency by relieving the restraint on peptide-chain initiation (17, 43). These changes are similar to those observed in skeletal muscle from nonseptic rats treated with IGF-I (52).
eIF4F is considered an essential eIF in promoting mRNAs for translation. In addition to present findings, previous reports have indicated that sepsis (51, 54) or sepsislike conditions (22) reduce the association of eIF4G with eIF4E, thereby contributing to mechanisms responsible for the inhibition of protein synthesis in skeletal muscle. eIF4E combines with eIF4G and eIF4A to form the heterotrimeric complex eIF4F. eIF4G is a large polypeptide that serves as a scaffold for eIF4E, eIF4A, mRNA, and the ribosome. eIF4G is intimately involved with regulating the translation apparatus during the initiation phase of mRNA translation. eIF4G, in addition to other translation functions, recruits the 40S ribosomal subunit to the 5′ end of mRNA, coordinates the circularization of mRNA through eIF4E and Poly(A)+-binding protein interactions (61), and assists in Mnk1 and eIF4E association (34, 60). Phosphorylation of eIF4G (28, 37) is associated with assembly of the eIF4F complex. Availability of eIF4E to form the eIF4F complex appears to be regulated by a family of small, acid- and heat-stable translation repressor proteins referred to as eIF4E binding proteins. 4E-BPs of which 4E-BP1 is the most prominent in skeletal muscle compete with eIF4G for binding with eIF4E. Binding of 4E-BP1 and eIF4G with eIF4E is mutually exclusive because both proteins have overlapping binding sites on eIF4E (13). Phosphorylation of 4E-BP1 dictates the binding of 4E-BP1 to eIF4E. Hence, assembly of active eIF4G·eIF4E complex may be augmented by an increased phosphorylation of eIF4G and/or 4E-BP1.
Growth-promoting hormones, such as IGF-I, acutely stimulate protein synthesis in skeletal muscle from control animals by controlling the activity and/or activation of proteins involved in regulating mRNA translation initiation. In skeletal muscle from control animals, IGF-I did not alter either the amount of eIF4E associated with the eIF4E binding protein 4E-BP1 or the phosphorylation state of 4E-BP1. In contrast, the amount of eIF4E associated with eIF4G was increased threefold by IGF-I, suggesting that IGF-I regulates protein synthesis in skeletal muscle of control animals by enhancing formation of the active eIF4E·eIF4G complex (52). However, the potential sites in the pathway of protein synthesis stimulated by IGF-I during sepsis remain unresolved.
This is the first report to document an increased assembly of active eIF4G·eIF4E complex in skeletal muscle of chronic sepsis rats after treatment with IGF-I. The increased assembly of eIF4G·eIF4E complex was associated with a stimulation of protein synthesis in skeletal muscle of chronic sepsis rats after treatment with IGF-I compared with animals perfused in the absence of IGF-I in the medium. Least squares linear regression analysis revealed a positive linear relationship between rates of protein synthesis and assembly of active eIF4G·eIF4E complex in gastrocnemius from septic rats perfused with and without IGF-I in the medium and controls (r2 = 0.952; F = 39.302, P < 0.05). This observation provides one possible mechanism to account for the increased rates of protein synthesis observed after administration of IGF-I to septic rats. During translation initiation, mRNA binds either directly to eIF4E already associated with 40S ribosomal subunits or to free eIF4E with subsequent binding of the mRNA·eIF4E·eIF4G complex to the ribosome. With either scenario, the increased amount of eIF4E associated with eIF4G after perfusion of gastrocnemius from septic rats with IGF-I would increase association of mRNA with the ribosome (36, 39, 42). We have previously proposed that IGF-I stimulates protein synthesis by increasing the formation of eIF4G·eIF4E complex in isolated perfused rat skeletal muscle from control rats (52). The results of the present set of investigations using gastrocnemius from septic rats perfused with medium containing IGF-I are consistent with that hypothesis. Thus IGF-I may accelerate protein synthesis in part by improving the ability of eIF4E to bind mRNA to the 43S preinitiation complex through an enhanced assembly of the active eIF4G·eIF4E complex in skeletal muscle of septic rats. In contrast, a previous report (51) failed to demonstrate a stimulation in protein synthesis when the assembly of eIF4G·eIF4E complex was enhanced after perfusion of gastrocnemius with buffer containing insulin from septic animals. In that study, the abundance of eIF4G associated with eIF4E was the same in septic animals treated with insulin as nontreated controls, yet rates of protein synthesis remained depressed in the septic rats. Reasons for this disparity remain unresolved.
Reasons for the enhanced formation active eIF4G·eIF4E complex in gastrocnemius of septic rats in response to IGF-I are unknown. In the present set of investigations, we examined the potential role of the two mechanisms proposed to lead to an increased assembly of an active eIF4G·eIF4E complex. Factors affecting the assembly of active eIF4G·eIF4E in gastrocnemius of septic rats include enhanced availability of eIF4E (18, 19) and enhanced phosphorylation of eIF4G (28, 37).
eIF4E availability is controlled through a family of 4E-BPs (26, 30, 44) that function as translational repressors by competing with eIF4G for a common binding site on eIF4E (13, 27). When eIF4E binds with 4E-BP1, eIF4G binding with eIF4E is blocked, thereby rendering eIF4E unavailable for binding with eIF4G.
Ability of 4E-BPs to bind to eIF4E is controlled through phosphorylation. The nonphosphorylated isoforms of 4E-BPs bind to eIF4E with high affinity and prevent it from binding to eIF4G to form the translationally active eIF4F complex (13, 27). Conversely, the phosphorylation of 4E-BPs reduces the binding affinity for eIF4E and thereby relieves translational repression because eIF4E is able to bind to eIF4G. IGF-I stimulates the formation of eIF4F complex in part through decrease in 4EBP1 associated with eIF4E in C2C12 myotubes in culture (40). In the present study, IGF-I increased the phosphorylation of 4E-BP1 with a corresponding decrease in the abundance of the 4E-BP1·eIF4E complex. Hence, a reciprocal change between the eIF4E·4E-BP1 complex and eIF4G·eIF4E complex was observed after perfusion with buffer containing IGF-I in the present set of investigations.
In contrast, no significant increase in 4E-BP1 phosphorylation was observed in gastrocnemius from control animals perfused in the presence of 10 nM IGF-I (52) and administered IGF-I/IGFBP-3 complex in vivo (24). Reasons for these differences are not readily apparent. In addition, chronic elevation of plasma IGF-I concentrations was not associated with an enhanced phosphorylation of 4E-BP1 in gastrocnemius of septic rats (43). Several potential reasons may help explain this apparent discrepancy compared with our findings. First, muscle perfused in situ is no longer exposed to elevated circulating cytokines and catabolic hormones observed in vivo in response to the septic insult. Inflammatory cytokines and hormones may modulate the effects of IGF-I on 4E-BP1 phosphorylation in vivo. In this regard, infusion of TNF for 24 h is associated with a profound reduction in the 4E-BP1 phosphorylation and increase in the amount of 4E-BP1 bound to eIF4E (23). Second, pathways responsible for phosphorylation of 4E-BP1 may be modulated when the gastrocnemius from septic rats is exposed to chronic stimulation by daily administration of IGF-I/IGFBP-3 complex. Third, there may be modulating effects of sepsis-induced elevations in plasma IGFBP-1 concentrations on 4E-BP1 phosphorylation to respond to daily injections of IGF-I/IGFBP-3 complex.
However, changes in abundance of inactive 4E-BP1·eIF4E complex do not always correlate with corresponding alterations in formation of active eIF4G·eIF4E complex. A low degree of correlation between inactive 4E-BP1·eIF4E complex and active eIF4G·eIF4E complex is observed in skeletal muscle from ovine fetus (41). We have previously reported there is either no significant change (43, 54) or a significant increase (51) in the abundance of the inactive 4E-BP1·eIF4E complex in gastrocnemius during chronic (5-day) sepsis. In addition, no changes in the abundance of 4E-BP1·eIF4E complex is observed in gastrocnemius 24 h after administration of endotoxin despite reduced amounts of eIF4G associated with eIF4E (22). In contrast, a significant decrease in the 4E-BP1 associated with eIF4E is observed after in vivo infusion of the proinflammatory cytokine TNF-α (23). Therefore, additional mechanisms may also be involved in the formation of active eIF4G·eIF4E complex.
In particular, we examined the potential role of phosphorylation of eIF4G in the assembly of an active eIF4G·eIF4E complex. At the present, no information exists concerning the ability of IGF-I to enhance the phosphorylation of eIF4G in gastrocnemius during chronic sepsis. The present findings show that acute IGF-I administration augmented the phosphorylation of eIF4G in gastrocnemius approximately twofold. Hence, IGF-I leads to an enhanced phosphorylation of eIF4G in muscles from septic rats. However, increased phosphorylation of eIF4G correlates with conditions known to stimulate protein synthesis (37). In contrast, phosphorylation of eIF4G was not enhanced in gastrocnemius from septic rats perfused with buffer containing insulin compared with no insulin present in the perfusate (−insulin 8.5 ± 1.2 arbitrary units vs. +insulin 7.1 ± 0.6 arbitrary units; P > 0.8). The mechanism by which IGF-I-induced eIF4G phosphorylation stimulates assembly of an active eIF4G·eIF4E complex remains unknown.
Other signaling pathways also appear to modulate translation initiation after perfusion with buffer containing IGF-I. S6K1 phosphorylates ribosomal protein S6 and accelerates the translation of specific mRNAs containing 5′-TOPs sequences, including translation components that make up the protein synthetic apparatus. According to the prevailing model of activation for S6K1, sites in the autoinhibitory domain (Ser411, Ser418, Thr421, and Ser424) are phosphorylated by an upstream kinase whose identity remains unresolved (12). Phosphorylation of these residues disrupts the interaction between the COOH-terminal and NH2-terminal domains, thereby permitting S6K1 to unfold and exposing additional sites in the linker and kinase domains. Subsequently, the Thr389 residue in the linker domain is phosphorylated, and this step has been demonstrated to be necessary for full and functional activation of S6K1 (63). In the present study, IGF-I administration stimulated the phosphorylation of Thr389 of S6K1.
Proline-directed serine/threonine protein kinase mTOR is reported to be a common intermediate involved in mRNA translation control produced by growth factors (for review see Ref. 14). As such, altered mTOR phosphorylation appears to represent an important nexus for the observed changes in translation initiation. mTOR is believed to be the upstream kinase responsible for phosphorylating 4E-BP1 and S6K1 (3, 35). Phosphorylation of mTOR on residues Ser2448 and Ser2481 has been used to monitor the activity of mTOR (2, 31). Therefore, we examined the phosphorylation state of Ser2448 and Ser2481 after IGF-I administration. IGF-I caused a significant increase in phosphorylation of mTOR at both Ser2448 and Ser2481.
Results of the present set of experiments are consistent with our previous report documenting that rates of protein synthesis are correlated with formation of active eIF4G·eIF4E complex in gastrocnemius (51). Although such a correlation does not prove cause and effect, the relationship between protein synthesis and amount of eIF4G associated with eIF4E is consistent with the proposed role of an active eIF4G·eIF4E complex in the overall regulation of protein synthesis. IGF-I may increase the active eIF4G·eIF4E complex through increased availability of eIF4E and phosphorylation of eIF4G. Thus these findings provide further evidence that assembly of an active eIF4G·eIF4E complex may be important in controlling rates of protein synthesis in gastrocnemius of septic rats.
National Institute of General Medical Sciences Grant GM-39277 (to T. C. Vary) supported these investigations.
These investigations could not be performed without the expert technical assistance provided by Heather Hubler and Gina Deiter.
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.
- Copyright © 2006 the American Physiological Society