Signal Transduction Pathways That Regulate Eukaryotic Protein Synthesis*

The last several years have witnessed an explosion in the published literature on two topics, the pathways that transduce extracellular signals to their intracellular targets and modification of the core translational apparatus in response to these signals. Most of these pathways result in cell growth and cell division. Synthesis of the entire complement of proteins is necessary to double the cell size, but synthesis of the so-called “growth-regulated” proteins (1) is needed for cell division. This article summarizes recent advances in our understanding of how a single mitogenic stimulus can simultaneously lead to an increase in both global and growth-regulated protein synthesis.

The last several years have witnessed an explosion in the published literature on two topics, the pathways that transduce extracellular signals to their intracellular targets and modification of the core translational apparatus in response to these signals. Most of these pathways result in cell growth and cell division. Synthesis of the entire complement of proteins is necessary to double the cell size, but synthesis of the so-called "growth-regulated" proteins (1) is needed for cell division. This article summarizes recent advances in our understanding of how a single mitogenic stimulus can simultaneously lead to an increase in both global and growth-regulated protein synthesis.

Mechanism of Protein Synthesis
The three stages of protein synthesis are catalyzed by initiation, elongation, and release factors (Ref. 2; a guide to current and previous nomenclature can be found in Ref. 3). A ternary complex of eIF2⅐GTP⅐Met-tRNA i 1 binds to the 40 S ribosomal subunit to form the 43 S initiation complex (Fig. 1). The eIF4 factors plus poly(A)-binding protein recognize the 5Ј-terminal cap or 3Ј-terminal poly(A) tract of mRNA, unwind mRNA secondary structure, and transfer it to the 43 S initiation complex, resulting in the 48 S initiation complex. Scanning for the first initiation codon in good sequence context requires eIF4A and the presence of eIF1 and eIF1A (4). Then eIF5 stimulates GTP hydrolysis by eIF2, after which the initiation factors are replaced by the 60 S subunit to form the 80 S initiation complex. The released eIF2⅐GDP is recycled to eIF2⅐GTP by the GEF eIF2B. The first elongator aminoacyl-tRNA is brought to the A-site by eEF1, followed by a cycle of GTP hydrolysis and exchange analogous to that of eIF2. Translocation is catalyzed by eEF2, again with a GTP hydrolysis cycle.

Signaling Intermediates Involved in Protein Synthesis
RTKs-The binding of growth factors to the extracellular domain of RTKs causes a conformational change that induces oligomerization and activation of the intracellular protein Tyr kinase domain (Ref. 5; Fig. 2). Substrates for the kinase can be either the RTK itself or a separate RKS. The SH2 domains of several different signaling molecules dock to the resulting Tyr(P)s in a sequence-specific manner, thereby activating separate downstream signaling cascades.
RKS-These are docking proteins for downstream effectors of RTKs (7). The best studied RKS are the insulin receptor substrates, which include IRS-1, IRS-2, IRS-3, Gab-1, and p62 DOK . Members of the IRS family bind to insulin receptor via an NH 2 -terminal PH domain and a Tyr(P)-binding domain. The COOH-terminal portions of the proteins contain numerous Tyr phosphorylation sites. IRS-1 alone provides docking sites for PI3-K, SH-PTP2, Grb-2, Fyn, Nck, and Crk.
SH-PTP2-This phosphatase contains two SH2 domains, and enzyme activity is maximally activated when both are occupied by Tyr(P)-containing peptides (8). SH-PTP2 is activated by docking to EGF receptor, platelet-derived growth factor receptor, c-kit, insulin receptor, IRS-1, IRS-2, and IRS-3 and may serve to attenuate the Tyr(P) signal in these molecules (7).
Ras-This G-protein is bound to the plasma membrane by COOH-terminal prenylation and myristoylation (9). GEF activity is provided by SOS, which associates constitutively with the SH2and SH3-containing protein Grb-2. The Grb-2⅐SOS complex is recruited to the plasma membrane by binding to specific Tyr(P)s in IRS-1, IRS-2, Shc, or SH-PTP2 (7). Another GEF, Ras-GEF, is stimulated by Ca 2ϩ /calmodulin (CaM) downstream of GPCR (10). The hydrolysis of GTP by Ras is stimulated by GTPase-activating proteins such as p120 GAP and NF1 (9).
MAPKs-Ras⅐GTP activates the Ser/Thr kinase Raf-1 by recruiting it to the plasma membrane. Raf-1, in turn, phosphorylates and activates MEK1 and MEK2. The MEKs are dual specificity kinases, phosphorylating both Thr and Tyr residues in ERK1 and ERK2 (p42 and p44 MAPKs).
PI3-K-This kinase is composed of a catalytic subunit and a * This minireview will be reprinted in the 1999 Minireview Compendium, which will be available in December, 1999. This work was supported by Grant GM20818 from the National Institute of General Medical Sciences.
PKC-There are at least 10 isoforms of PKC (␣-) that differ in responsiveness to phospholipids and Ca 2ϩ (21). Classical PKCs (␣, ␤, and ␥) are activated and eventually down-regulated by phorbol esters, which are structural analogs of the physiological signal DAG, but atypical isoforms ( and ) are not. Insulin activates both classical and atypical isoforms (22,23). PKC is activated downstream of PI3-K (24) via direct phosphorylation at Thr-410 by PDK1 (25).
Ca 2ϩ -Changes in cytosolic Ca 2ϩ levels can occur by at least two mechanisms (26). First, GPCR operate Ca 2ϩ channels that allow influx from the extracellular space. Second, Ins(1,4,5)P 3 , released in response to both RTKs and GPCR, binds to receptors in the endoplasmic reticulum and releases Ca 2ϩ into the cytosol.
GSK-3-This kinase regulates numerous cellular processes besides phosphorylation of glycogen synthase, including insulin-stimulated protein synthesis (27). Phosphorylation of GSK-3 at Ser-9 inactivates the enzyme and is correlated with the activation of protein synthesis. GSK-3 is phosphorylated directly by PKB in vivo (28,29). mTOR (FRAP, RAFT-1)-mTOR is a 290-kDa protein kinase that is activated by PKB (30) and phosphorylates Ser/Thr-Pro motifs (31). As the name implies, its kinase activity is inhibited by the immunosuppressant rapamycin (31). The ability of insulin to cause phosphorylation and activation of mTOR is attenuated by cAMP (32).
Ribosomal S6 Kinases-The major activity responsible for phosphorylation of ribosomal protein S6 is p70 S6K (33), which is activated by hierarchical phosphorylation of seven Ser/Thr sites. Phosphorylation of COOH-terminal sites occurs first, making Thr-389 available for phosphorylation by a rapamycin-sensitive pathway (34) and culminating in phosphorylation of Thr-229 by constitutively active PDK1 (35). Multipotential S6 kinase is activated and becomes membrane-associated in response to insulin (36). p90 S6K is activated in response to growth factors by forming complexes with ERK1 and -2 and undergoing phosphorylation (37). Additional S6 kinases continue to be discovered (38,39).
PHAS (4E-BP)-PHAS-I was originally observed as a phosphorylated protein that increased upon insulin treatment of cells but was later found to bind eIF4E specifically (40). Two additional isoforms have since been found, PHAS-II and PHAS-III (41). Phosphorylation of PHAS occurs at six or more sites, two to five of which appear to result from direct phosphorylation by mTOR (34,42).
Mnk1 and -2-These kinases, homologous to p90 S6K , were identified as binding partners of ERK1 and -2 (43) and in a screen for ERK1 substrates (44). Mnk is activated by phosphorylation at Thr-197 and Thr-202 both by a mitogen-activated pathway via ERK1 or -2 and also by a stress-activated pathway via p38 (45). Interestingly, Mnk1 also binds to the COOH terminus of eIF4G (45,46). eEF2K (CaMK III)-This kinase is activated by Ca 2ϩ /CaM and contains a putative CaM-binding domain COOH-terminal to the catalytic domain (47). eEF2 kinase is also activated by cAMP, independently of Ca 2ϩ (48).

Components of the Core Translational Machinery
That Are Targets of Signaling Pathways Two types of signal-induced modifications of translation factors have been described to date: changing the intrinsic activity or binding properties of the factor by phosphorylation and sequestration of the factor in an inactive complex, the formation or dissociation of which may be controlled by phosphorylation.
eIF2-Most situations that lead to phosphorylation of eIF2␣ represent cellular stress, but eIF2␣ phosphorylation also changes as a result of normal signaling pathways, e.g. those regulated by amino acids (49) and interleukin-3 (50). Phosphorylation of eIF2␣ on Ser-51 by several different kinases causes formation of a stable complex with eIF2B ( Fig. 1), thereby reducing the concentration of active eIF2B (51).
eIF2B-This GEF is phosphorylated on the ⑀ subunit and inactivated by GSK-3␣ and -␤ (52). Phosphorylation occurs at Ser-540, although evidence suggests that a priming phosphorylation at Ser-544 by some other kinase is needed (53).
eIF4E-The mRNA cap-binding protein is phosphorylated at Ser-209 in vivo (54, 55) by Mnk1 and -2 (43,45), resulting in an increase of its affinity for caps (56). eIF4E variants unable to bind eIF4G are poorly phosphorylated (46). The availability of eIF4E is also regulated by formation of an inactive complex with PHAS, and phosphorylation of PHAS causes dissociation of the complex (57,58). The relationship between these two eIF4E regulatory mechanisms is unclear. In vitro, PHAS inhibits eIF4E phosphorylation by Mnk1 (59), but in vivo, eIF4E is extensively phosphorylated in cells overexpressing PHAS, suggesting that eIF4E phosphorylation is independent of PHAS (45).
eIF4G-This linking protein, as part of the eIF4F complex, is phosphorylated at unknown sites by multipotential S6 kinase (60). The phosphorylated complex is more stimulatory for in vitro protein synthesis and binding of mRNA to the 43 S initiation complex.
S6 -This ribosomal protein is phosphorylated at five Ser residues, located at 235, 236, 240, 244, and 247 (33). However, the effect of S6 phosphorylation on protein synthesis is unclear. There is a correlation between activation of p70 S6K and translation of mRNAs containing a 5Ј-TOP tract (61), and treatment of 80 S ribosomes with multipotential S6 kinase increases the rate of elongation 2-fold (36). A recent study implicates S6 phosphorylation in the interaction between 40 S subunits and mRNA (62).
eEF1-The first of the elongation factors is composed of ␣, ␤, ␥, and ␦ subunits. Various subunits are phosphorylated in vitro by several kinases, including casein kinase II, multipotential S6 kinase, and PKC (36). In the latter two cases, these phosphorylations result in a stimulation of elongation and GDP/GTP exchange, respectively.
eEF2-This elongation factor is inhibited by phosphorylation via eEF2 kinase on Thr-56 and Thr-58, the phosphorylation of the latter site requiring prior phosphorylation of the former (26,63).
Amino Acids-Addition of amino acids to starved mammalian cells causes many of the same modifications as insulin, but some of the pathways appear to be different. The activity of eIF2B is increased, partly through dephosphorylation of eIF2␣ and partly through phosphorylation of eIF2B⑀, but GSK-3 activity is not altered (75). Like insulin, amino acids cause phosphorylation of p70 S6K and PHAS through a mTOR-dependent pathway (76 -79). Unlike insulin, however, the pathway does not involve PKB (78,77) but does involve tRNA aminoacylation (80). The insulin-or IGF-1mediated phosphorylation of p70 S6K and PHAS-I requires amino acids (76,79). In the presence of insulin, amino acids activate eIF2B but not eIF4E (75).
Ca 2ϩ -Changes in the level of intracellular Ca 2ϩ and its distribution among compartments affect protein synthesis in at least two ways (26). Depletion of Ca 2ϩ from the sarcoplasmic/endoplasmic reticulum by hormones like vasopressin inhibits initiation by phosphorylation of eIF2␣ via PKR (87). Elevation of cytosolic Ca 2ϩ by agents such as bradykinin inhibits elongation by phosphorylation of eEF2 via eEF2K (26). Similarly, stimulation of the NMDA receptor causes eEF2 phosphorylation (88).

Modification of Initiation Factors Can Affect the Spectrum of mRNAs Translated
Signaling pathways do not stimulate translation of all mRNAs equally. Messenger RNAs differ widely in translational efficiency. Factors contributing to low efficiency of translation include a highly structured 5Ј-UTR, the presence of upstream AUGs, and poor sequence context for the initiating AUG (89), all of which are found in the 5Ј-UTRs of mRNAs for scarce proteins (90). mRNAs with these properties encode a disproportionate share of proteins involved in cell growth and cell cycle progression (89). These mRNAs are poorly translated in quiescent cells but preferentially recruited to ribosomes after a mitogenic signal (91)(92)(93). Overexpression of eIF4E in cultured cells preferentially stimulates translation of a number of mRNAs with high 5Ј-UTR secondary structure that are involved in cell growth and division (89). Expression of cell cycle-dependent proteins like c-Myc (64) and cyclin D (94) requires mTOR, and inhibition of mTOR with rapamycin prolongs the G 1 phase in both T-cells (95) and yeast (96). These results suggest that pathways which activate the unwinding machinery, i.e. by phosphorylation of eIF4E, eIF4G, or PHAS, disproportionately stimulate translation of growth-regulated mRNAs with high 5Ј-UTR secondary structure.
5Ј-TOP mRNAs are also differentially regulated in response to extracellular signals (97). These encode many translational components, including ribosomal proteins, eEF1␣, eEF2, and poly(A)binding protein. 5Ј-TOP mRNAs are recruited to polysomes in a growth-dependent fashion that is selectively inhibited by rapamycin (72,98). This finding alone does not distinguish between signaling through PHAS and signaling through p70 S6K , because the pathway bifurcates downstream of mTOR (99) (see Fig. 2). However, ectopic expression of a dominant interfering p70 S6K blocks both activation of p70 S6K and 5Ј-TOP mRNA translation, indicating a direct role for p70 S6K (61).
The insulin-stimulated pathways to general protein synthesis and to growth-regulated protein synthesis in 32D cells can be dissected with rapamycin (64). Insulin-stimulated total protein synthesis is inhibited only ϳ10%, and actin synthesis is not affected at all, but insulin-stimulated c-Myc synthesis is completely inhibited. Also, expression of constitutively active PKC in the absence of IRS-1, which bypasses the mTOR pathway, permits insulin-stimulated general protein synthesis but not c-Myc synthesis (23). This suggests that the insulin signal bifurcates at some point after PI3-K, one pathway stimulating general protein synthesis through eIF2B and one stimulating growth-regulated protein synthesis, accounting for ϳ10% of the total, through eIF4E and S6. Similarly, T-cell activation causes 13% of the pre-existing mRNAs to be recruited to ribosomes (100), and rapamycin causes a 15% decrease in protein synthesis in activated T-cells (95).

Conclusions and Future Directions
Protein synthesis is one of the most complicated biochemical processes undertaken by the cell, requiring roughly 150 different polypeptides and 70 different RNAs. Yet only seven polypeptides (eIF2␣, eIF2B⑀, eIF4E, eIF4G, S6, eEF1, and eEF2) have been identified as targets for regulatory pathways to date. Early observations that multiple initiation and elongation factors were phosphorylated in response to a single extracellular signal (101,102) may have suggested unnecessary redundancy. This now seems more comprehensible when it is realized that modification of some factors affects the overall rate of translation whereas modification of others affects the spectrum of mRNAs translated. Understanding the pathways for regulation of protein synthesis holds promise for novel approaches for cancer intervention (e.g. Ref. 103), especially if pathways leading to growth-dependent protein synthesis can be selectively inhibited.