The regulation of initiation of protein synthesis by phosphorylation of ...

Communication

THEJOURNAL OF BIOLOGICAL CHEMISTRY Vol. 259 No. 11 Issue of June 10 pp. 6708-6111 1984 0 1984 by ThiAmerican Society of Bioiopcal Chamis&, Inc. Printed in U.S.A.

The Regulationof Initiation of Protein Synthesis by Phosphorylation of eIF-Z(a) and the Role of Reversing Factor in the Recyclingof eIF”2*

the concentration of eIF-2 (20): Earlier studieshad suggested that heme-regulated eIF-2(a) kinase acts toimpair the recycling of eIF-2 inthe mechanism of initiation (21). eIF-2 is thought to be released as a binary complex with GDP (eIF-2.GDP) in the reaction joining the 48 S initiation complex and the 60 S ribosomal subunit (22, 23). At physiological MgZ+ co-icentrations, the recycling of (Received for publication, January 16,1984) eIF-2 requires that GDPin the binary complex be exchanged with GTP, which has a much lower affinity than GDP for Robert L. Matts$ and Irving M. London eIF-2 (24,25). Clemens et al. (26) observed that theexchange From the Haruard-MlT Diukwn of Health Sciences and of GDP bound to eIF-2 for GTP was inhibited if eIF-2(a) was Technology and Department of Bwbgy, Massachusetts phosphorylated. Subsequently, RF was found to catalyze this Institute of Technology, Cambridge,Massachusetts 02139 exchange, thereby stimulating eIF-2. Met-tRNAf. GTP ternary complex formation in the presence of Met-tRNA, (17, The capacity of whole reticulocyte lysates to catalyze 18, 27, 28). If eIF-2(a) is phosphorylated, however, the RFthe dissociation of exogenously added eIF-2*[’H]GDP was determined as a measure of theirreversing factor catalyzed exchange of GTP for GDP bound to eIF-2 is inhib(RF) activity in the recycling of eIF-2 for the mainte- ited (17, 18,27, 28). We proposed (14), and subsequently showed in in vitro nanceorrestoration of protein synthesis. We have examined the relationship RF of activity to the protein studies, that the phosphorylated binary complex eIF-2(aP). synthetic activity of the lysate under normal conditions GDP interactswith RF toform a RF.eIF-Z(aP) complex that and on inhibition of protein synthesis by heme defi- is not readily dissociable (17). As a result, the RFis unavailciency, double-stranded RNA, or oxidized glutathione. able to catalyze the dissociation of unphosphorylated binary A direct correlation was found between a lysate’sca- complex. Since the concentration of RF in the reticulocyte pacity tosynthesize protein andits ability to stimulate lysate is much lower than the concentration of eIF-2, the the dissociation of eIF-2.GDP. Thesefindings further phosphorylation of only a relatively modest portion of the support the proposed mechanism by which the phoseIF-2 could suffice to inhibit the initiation of protein synthesis phorylation of only 30-40% oftheeIF-B(a)inthe (14, 17). Similar proposals have been made by Ochoa (4) and lysate renders thelimiting amount of RFpresent non- Safer (29). functional,impairing the recycling of eIF-2 and In this report, we present studies that are designed to test thereby inhibiting theinitiation of protein synthesis. this hypothesis. We have determined the capacity of whole reticulocyte lysates to catalyze the dissociation of exogenously added eIF-2. [3H]GDP as a measure of their RFactivity and we have examined the relationship of the RF activity to the The inhibition of initiation of protein synthesis in reticu- protein synthetic activity of the lysate under normal condilocytes and theirlysates which is observed in heme deficiency tions and on inhibition of protein synthesis by heme defihas been shown to result from the phosphorylation of the a ciency, double-stranded RNA, or oxidized glutathione. We subunit of eIF-2’ (for reviews, see Refs. 1-4). This phospho- have found a direct correlation between a lysate’s capacity to rylation is catalyzed by a heme-regulated CAMP-independent synthesize protein and its ability to effect the dissociation of protein kinase which is specific for eIF-2(a) (5-11). The fact eIF-2. GDP, i.e. its RF activity. These findings provide further that inhibition of initiation occurs when only 30-40% of the support for the mechanism by which the phosphorylation of eIF-B(a) is phosphorylated (12, 13) prompted the suggestion eIF-S(a) renders RF rate limiting, the recycling of eIF-2 is that the phosphorylation of eIF-2(a) renders another factor impaired, and theinitiation of protein synthesis is inhibited. limiting inthe cycle of initiation (13,14).The reversing factor’ seemed to be a likely candidate to serve as a rate-limiting EXPERIMENTAL PROCEDURES factor. It is a multipolypeptide which forms a 1:l complex Assay for RF Activity in the Reticulocyte Lysate-RF activities in with eIF-2 (15-19), it acts catalytically to reverse the inhibi- rabbit reticulocyte lysates were measured by their ability to stimulate tion of protein synthesis inheme-deficient or heme-regulated the dissociation of added eIF-2.13H]GDP binary complex. ReticuloeIF-2(a) kinase-inhibited lysates(17,19,20), and theconcen- cyte lysates were incubated under conditions of protein synthesis tration of RF in the reticulocyte lysate is much lower than essentially as described by Pelham and Jackson (30). Standard pro* This work wassupported by United States Public Health Service Grant AM-16272. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $Supported by a fellowship from the National Foundation for Cancer Research. The abbreviations used are: eIF-2, eukaryotic initiation factor 2; eIF-B(a),a subunit (38,000daltons) of eIF-2; Ac, acetyl; RF, reversing factor; eIF-2. GDP,binary complex containing eIF-2; RF.eIF-2(aP), RF.eIF-2 phosphorylated in the eIF-B(a) subunit. 2Also referred to as eukaryotic initiation factor 2B, eIF-2B (19, 29), and GDP exchange factor (GEF) (4,18).

tein synthesis mixtures (100 pl) contained 65.4 pl of rabbit reticulocyte lysate (1:1lysate prepared as described in Ref. 31), 80 mM KCl, 0.5 mM MgCI2, 4 pg of creatine kinase, 8 mM creatine phosphate, 20 amino acids at 40 p~ each, 0.4 mM dithiothreitol, 0.5 mM Mg“+-GTP, ~ as indicated. Lysate dilution buffer contained and 20 y c hemin-C1 40 mM Tris-HC1 (pH 7.6), 100 mMKC1, 50 mM KF, 2 mM Mg(OAc)z, 10% glycerol, and 40 p~ unlabeled GDP. KF was present to prevent dephosphorylation of eIF-Z(a) and release of RF. GDP was added to ensure the presence of excess unlabeled nucleotide for exchange with [WIGDP from binary complex being dissociated. eIF-2. [3H]GDP was preformed (10 min, 30 “C) in a reaction mixture containing 20 mM Tris-HC1 (pH 7.6), 100 mM KC1, 1 mM dithiothreitol, 100 pg/ml

6708

R. L. Matts andI. M. London, unpublished observations.

6709

Role of RF inthe Recycling of eIF-2 of creatine kinase as carrier, 2 p~ [3H]GDP (6500 cpm/pmol), and eIF-2 (-80% pure). Mg(OAc)z wasthen added to a concentration of 1 mM and the preformed eIF-2. [3H]GDP complexes were stored on ice prior to use. Protein synthesis mixtures were incubated a t 30 "C and diluted with an equal volume of dilution buffer (30 "C) at the times indicated. A 10-pl volume, containing 1-2 pmol of preformed eIF-2. [3H]GDPfor each 2 0 0 4 volume of diluted protein synthesis mixture, was then added immediately, and theincubation was continued a t 30 "C. Undissociated eIF-2. ['HIGDP remaining in each 200 p1 of diluted protein synthesis mixtures was measured at the times indicated by the retention of the complex on Millipore filters (HAWP 02500). Assays werestopped by the addition of 3 ml of ice-cold wash buffer containing 20 mM Tris-HC1 (pH 7.6), 100 mM KCl, and 1 mM Mg(OAc)*.Remaining eIF-2. ['HIGDP was collected by filtration on Millipore filters. Each tube was then rinsed three times with 3 ml of wash buffer, followed by two 3-ml washes of each filter. Filters were heat-dried and radioactivity was determined in 5 ml of Econofluor (New England Nuclear). The total picomoles of eIF-2.13H]GDP added was measured a t various timesduring the course of each experiment by immediately pouring a diluted assay mixture containing heme-deficient lysate onto a Millipore filter upon addition of the eIF-2. [3H]GDPcomplex. Total picomoles of eIF-2. 13H]GDPadded was always determined in heme-deficient lysates because some rapid dissociation of the complex was found to occur in heme-supplemented lysates between the addition of the complex and theinitial zero time filtration. Picomoles of eIF-2. [3H]GDPdissociated was determined by calculating the difference between the total eIF-2.[3H]GDPadded and thatremaining in an assay mixture after incubation for the time stated in each figure. eIF-2 (-80% pure), free of RF, was provided by Dr. Daniel H. Levin, Massachusetts Institute of Technology. The RFused in these experiments was a partially purified preparation, substantially free of eIF-2, derived from a reticulocyte ribosomal salt wash. The 0.2-0.4 M KC1 eluate from the phosphocellulose step of the eIF-2 preparation (32) was dialyzed against standard buffer (20 mM Tris. HCl, pH 7.5, 1 mM dithiothreitol, 0.2 mM EDTA 10% glycerol) containing 0.1 M KCl. This preparation was applied to a Sepharose GB-heparin (33) column (1 ml) previously equilibrated with the above buffer. The column was washed with standard buffer containing 0.3 M KC1. RF was then eluted with buffer containing0.5 M KC1. RF was precipitated by addition of (NH4)$304to a concentration of 80%. The RF was resuspended in standardbuffer containing 0.1 M KC1 and theremaining (NH4),S04was removed by dialysis. Materials-GTP, GDP, GSSG, and creatine kinase were purchased from Sigma. [8-3H]GDP (9 Ci/mmoI) and ["Clleucine(338mCi/ mmol) were purchased from New England Nuclear. Reovirus doublestranded RNA was provided by Dr. Ray Petryshyn, Massachusetts Institute of Technology. RESULTS AND DISCUSSION

RF activity inwhole lysates was measured by their capacity to stimulate the dissociation of exogenously added eIF-2. [3H] GDP complexes under various conditions of protein synthesis. In heme-deficient lysates, when protein synthesis is inhibited by over 90%, the rate of eIF-2. [3H]GDP dissociation is reduced to 5% of that found inheme-supplemented lysates (Fig. 1). Addition of hemin to heme-deficient lysates after the shutoff of protein synthesis restores the rate of protein synthesis and the rate of eIF-2-[3H]GDP dissociation to those found in heme-supplemented lysates(Fig. 1).To examine the relationship between RF activity and protein synthesis, we have used the ability of lysates to dissociate eIF-2. [3H]GDP in 2min as a measure of RF function. Under these conditions, the rateof dissociation of eIF-2. GDPwas found to be linear: with decreasing concentrations of lysate present in theassay, there was a corresponding decrease in the amountof eIF-2. [3H]GDP dissociated, an indication that the amount of eIF2. [3H]GDP dissociation activity measured by this method reflects the concentration of active RF present in the incubation mixture (Fig. 2). The relationship between protein synthesis and eIF-2.[3H] GDP dissociation activity was then examined to determine the correlation between these two parameters in lysates as

TIME (mtnutesl

FIG. 1. Kinetics of eLF-2-['H]GDP dissociation inhemesupplemented, heme-deficient, andheme-restoredreticulocyte lysates. Protein synthesis reaction mixtures (700 pl, 30 'C) were diluted with an equal volume of lysate dilution buffer (30 "C), followed by the immediate addition of20 pl, containing 9 pmol of preformed eIF-2. 13H]GDP complex.2 0 0 4 aliquots were taken at the times indicated in the figure and the amount of 13H]GDPremaining bound to eIF-2 was determined by the retention of the complex on Millipore filters as described under "Experimental Procedures." A, standard protein synthesis mixture preincubated for 15 min with 20 ptM hemin-Cl added at 0 min; 0, standard protein synthesis mixture preincubated for 15 min without added hemin-C1; 0,standard protein synthesis mixture preincubated 30 min with 20 p~ hemin-C1 added at 10 min.

i Lysateconcentration

(%I

FIG. 2. The relationship between lysate concentration and eIF-2-[aH]GDP dissociationactivity. Protein synthesis mixture (100 pl) containing 20 p~ hemin-C1 and 0-65.4 p1 of reticulocyte lysate were preincubated at 30 "C for 30 s. Reticulocyte lysate was replaced in the reaction mixtures with buffer containing 20 mM TrisHCl (pH 7.6),100 mM KC1, 1 mM Mg(OAc)z, and 100pg/ml of creatine kinase as carrier. Reaction mixtures were diluted with 100 pl of lysate dilution buffer (30 "C), followed bythe immediate addition of 20 p1 containing 1.7 pmol of preformed eIF-2. 13H]GDPcomplex. Incubations were continued at 30 "C for 2 min and the amount of eIF-2. 13H]GDPcomplex remaining was determined by retention of the complex of Millipore filters as described under "Experimental Procedures." Reticulocyte lysate in the final dissociation assay represented between 0-32% of the total volume of the reaction mixture. Recovery of eIF-2. ['HIGDP from each reaction mixture was found to be unaffected by differences in protein concentration.

inhibition and restoration of proteinsynthesis are taking place. Incubation of heme-deficient lysates for various lengths of time results in a progressive loss of the lysates' ability to catalyze the dissociation of eIF-2. [3H]GDP (Fig. 3). At 10 min, the loss of the ability of heme-deficient lysates to stimulate eIF-2. [3H]GDP dissociation corresponds to thetime at which protein synthesis becomes maximally inhibited. At 7 min, when the rateof protein synthesisis beginning to decline

6710

I

:A !2

Role of RF in the Recycling of eIF-2 Protein Synthesis 40

A

IB

Protein Synthesis

Lysate RF Activity

x

0

c .-

al

10V

-

I IO30

Y

TIME (minutes) FIG. 3. Restoration of lysate eIF-2*[’H]GDP dissociation activity upon reversal of protein synthesis inhibition by the addition of hemin or RF. A, standard lysate-protein synthesis assays (50 pl) containing [“C]leucine (148 mCi/mmol) were carried out for 30 min at 30 “C. Incorporation of [14C]leucineinto protein was measured as described in Ref. 38. A, with 20 p~ hemin-C1added at 0 min; 0, without hemin-C1; 0, with 20 p~ hemin-C1 added at 7 min; 0,without hemin-Cl but with 1pmol of RF added at 10 min. E, RF activity was determined in standard protein synthesis mixture similarly incubated at 30 “C in the absence of [“Clleucine. At the times indicated, 100-pl aliquots were diluted with an equal volume of lysate dilution buffer. 20 p1 containing 1.2 pmol of preformed eIF-2. 13H]GDPwas immediately added, and the[3H]GDP remaining bound to eIF-2 after 2 min was determined by retention on Millipore filters as described under “Experimental Procedures.” A, with 20 p~ heminCI added a t 0 min; 0, without hemin-C1; 0, with 20 p~ hemin-C1 added at 7 min; 0, without hemin-C1 but with 2 pmol of RF added per 100 p1 of protein synthesis mix a t 10 min.

rapidly, heme-deficient lysates show a substantial decrease in eIF-2.GDP dissociation activity. Addition of hemin to a heme-deficient lysate at 7 min results in a restorationof eIF2. GDP dissociation activity in the lysate which is well correlated with the restoration of proteinsynthesis (Fig. 3). Addition of an RF preparation to a heme-deficient lysate at 10 min predictably restores protein synthesisand eIF-2.[3H] GDP dissociation activity. The dissociation of eIF-2. [3H] GDP in the presence of exogenously added RF also indicates that the eIF-2. [3H]GDP complexes remain available to interact with RF added to heme-deficient lysates. Inhibition of proteinsynthesisin lysates by doublestranded RNA or oxidized glutathione is also known to be accompanied by the phosphorylation of eIF-2 (10, 11). On addition of double-stranded RNA or oxidized glutathione, there is a rapid loss of RF activity in the reticulocyte lysate which corresponds to thetime of shutoff of protein synthesis (Fig. 4). The loss of RF activity in thereticulocyte lysate, measured as eIF-2. [3H]GDP dissociation activity, correlates with the known activation of eIF-2(a) kinases (5-11), the phosphorylation of eIF-2(a), and theinhibition of protein synthesis (31, 34-37). At the time when protein synthesisbecomes inhibited, the lack of significant eIF-2. GDPdissociation activity demonstrates that this function is specific for RF and that RF has become nonfunctional or sequestered. The findings that RF becomes nonfunctional in inhibited lysates when only a limited portion of the eIF-2(a)is known to be phosphorylated (12, 13) demonstrates that RF has become the limiting component in the initiation cycle when protein synthesis is inhibited in lysatesdue to theactivation of eIF-B(a) kinases. These observations are in agreement with the previously proposed hypothesis (4, 17, 29) summarized below, which was based on

I

I

20

TIME (MINUTES) FIG. 4. Effect of double-stranded RNA and oxidized glutathione on protein synthesis and lysate RF activity. A, protein synthesis assay was carried out as described in Fig. 3A. A, with 20 p M hemin-C1 added at 0 min; 0, with 20 p~ hemin-CI and doublestranded RNA (20 ng/ml) added at 0 min; 0, with 20 p~ hemin-C1 and 500 p~ GSSG added a t 0 min. B, RF activity was determined in lysates incubated as above as described in Fig. 3B.

in vitro data using purified components. i) Thecritical role of RF in the lysate is to catalyze the dissociation of eIF-2. GDP to permit the formation of the ternary complex. ii) eIF-2. GDP is the primary site of eIF-P(a) kinase action. iii) RF interacts with phosphorylated binary complex to form an RF. eIF-2(aP) complex that effectively sequesters RF so that it cannot function in the recycling of eIF-2. eIF-2.GDP

+ RF -+

RF.eIF-2

+ GDP

+ ATP eIF-B(a) kinase eIF-B(aP).GDP + ADP eIF-2(aP).GDP + RF + RF.eIF-2(aP) + GDP(iii)

eIF-2.GDP

9

(i) (ii)

This hypothesis further predicts that restoration of protein synthesis in inhibited lysates requires the presence of phosphatase(s) which can dephosphorylate theRF.eIF-Z(aP) complex to release functional RF, and which can dephosphorylate any excess eIF-B(aP) .GDP to eliminate the lysate’s ability to sequester RF. The control of such phosphatase activities is also likely to be important in the regulation of protein synthesis by phosphorylation of eIF-B(a). Acknowledgments-We thank Dr. N. Shaun Thomas and Dr. D. H. Levin for their generous help. REFERENCES 1. Rosen, 0. M., and Krebs, E. G., eds (1981) Cold Spring Harbor Conf. Cell Proliferation 8,931-998 2. Grunberg-Manago, M., and Safer, B., eds (1982) Deu. Biochem. 24,297-358 3. Goldwasser,E. (ed) (1983)Regulation ofHemoglobinBiosynthesis, pp. 165-267, Elsevier Biomedical, New York 4. Ochoa, S. (1983) Arch. Biochem. Biophys. 223,325-349 5. Levin, D. H., Ranu, R. S., Ernst, V., and London, I. M. (1976) Proc. Natl. Acad. Sci. U. S. A. 73, 3112-3116 6. Kramer, G., Cimadevilla, M., and Hardesty, B. (1976) Proc. Natl. Acad. Sci. U. S. A. 73,3078-3082 7. Ranu, R. S., and London, I. M. (1976) Proc. Natl. Acad. Sci. U. S. A. 73,4349-4353 8. Farrell, P. J., Balkow, K., Hunt, T., Jackson, R. J., and Trachsel, H. (1977) Cell 11, 187-200 9. Gross, M., and Mendelewski, J. (1977) Biochem. Biophys. Res. Commun. 74,559-569 10. Levin, D. H., and London, I. M. (1978) Proc. Natl. Acad.Sci. U. S. A. 76,1121-1125

Role of RF in theRecycling of eIF-2 11. Ernst, V., Levin, D. H. and London, I. M. (1978)Proc. Natl. Acad.

12. 13. 14.

15. 16. 17. 18.

19. 20.

21.

22.

Sci. U. S. A. 75,4110-4114 Farrell, P. J., Hunt, T., andJackson, R.J. (1978) Eur. J. Biochem. 89,517-521 Leroux, A., and London, I. M. (1982) Proc.Natl.Acad.Sci. U. S. A. 79, 2147-2151 London, I. M., Fagard, R., Leroux, A., Levin, D. H., Matts, R., and Petryshyn, R. (1983) in Regulntion of Hemoglobin Biosynthesis (Goldwasser, E., ed) pp. 165-183, 209, Elsevier Biomedical, New York Amesz, H., Goumans, H., Haubrich-Morre, T., Voorma, H. O., and Benne, R. (1979) Eur. J. Biochem. 98,513-520 Siekierka, J., Mitsui, K., and Ochoa, S. (1981) Proc. Natl. Acad. Sci. U. S. A. 78, 220-223 Matts, R. L., Levin, D. H., and London, I. M. (1983) Proc. Natl. Acad. Sci. U. S. A. 80, 2559-2563 Panniers, R., and Henshaw, E. C. (1983) J. Bioi. Chem. 258, 7928-7934 Konieczny, A., and Safer, B. (1983) J. Biol. Chem. 2 5 8 , 34023408 Safer, B., Jagus, R., Konieczny, A., and Crouch, D. (1982) in Developments in Biochemistry (Grunberg-Manago, M., and Safer, B., eds) Vol. 24, pp. 311-325, Elsevier Biomedical, New York Cherbas, L., and London, I. M. (1976) Proc.Natl.Acad.Sci. U. S. A. 73,3506-3510 Trachsel, H., and Staehelin, T. (1977) Proc.Natl.Acad.Sci. U. S. A. 75,204-208

6711

23. Peterson, P. T., Safer, B., and Merrick, W.C.. (1979) J. Biol. Chem. 254, 7730-7735 24. Walton, G.M., and Gill, G. N. (1975) Biochim. Biophys.Acta 390,231-245 25. Walton. G.M.. and Gill.G. N. (1976) . , Biochim. Bioohvs.Acta 418,'195-203 26. Clemens, M.J., Pain, V. M., Wong, S., and Henshaw, E. C. (1982) Nature (Lord.)206.93-95 27. Siekierka, J., Mauser, L., and Ochoa, S. (1982) Proc. Natl. Acad. Sci. U. S. A. 79, 2537-2540 28. Pain, V.M., and Clemens, M. J. (1983) Biochemistry 22, 726733 29. Safer, B. (1983) Cell 33, 7-8 30. Pelham. H. R.B.. and Jackson. R. J. (1976) . . Eur. J. Biochem. 67. 247-256 31. Hunt, T., Vanderhoff, G., and London, I. M. (1972) J. Mol. Biol. 66,471-481 32. Ranu, R. S., and London, I. M. (1979) Methods Enzymol. 60, 459-484 33.Van der Mast, C., Thomas, A., Goumans, H.,Amesz,H., and Voorma, H. 0.(1977) Eur. J. Biochem. 75, 455-464 34. Zucker, W. V., and Schulman, H. M. (1968)Proc. Natl. Acad. Sci. U. S. A. 59,582-589 35. Rabinovitz, M., Freedman, M. L., Fisher, J. M., and'Maxwell, C. R. (1969) Cold Spring Harbor Symp. Quant. Biol. 34,567-578 36. Ehrenfeld, E., and Hunt, T. (1971) Proc. Natl. Acad. Sci.U. S. A. 68,1075-1078 37. Kosower, N. S., Vanderhoff, G.A., and Kosower, E. M. (1972) Biochim. Biophys. Acta 272, 623-627 38. Ernst, V., Levin, D. H., and London, I. M. (1978) J. Biol. Chem. 253,7163-7172

."