Phosphorylation of Ribosomal Protein S6 by CAMP-dependent Protein Kinase and Mitogen-stimulated S6 Kinase Differentially Alters Translation of Globin mRNA*

The effects of phosphorylation of ribosomal protein S6 by two different protein kinases, the CAMP-de- pendent protein kinase and the mitogen-stimulated S6 kinase, or translation of globin mRNA in a reconsti- tuted system and on binding of globin mRNA to 40 S ribosomal subunits were examined. The CAMP-de- pendent protein kinase incorporated 1.5 mol of phos-phate/mol of 40 S ribosomal subunits. Phosphorylation of S6 by the CAMP-dependent protein kinase had no effect on binding of 3' terminus-labeled globin mRNA to 40 S ribosomal subunits. ['HILeucine incorporation with 40 S ribosomal subunits phosphorylated by the CAMP-dependent protein kinase was identical to that observed with nonphosphorylated 40 S ribosomal sub- units, although on occasion, a slight inhibition (<lo%) was observed; there was no effect on the rate of syn- thesis of either the a or /3 chains of globin. Phosphorylation with the mitogen-stimulated S6 kinase (2.5 mol/ mol) did not alter binding of globin mRNA to 40 S ribosomal subunits; however, translation of globin mRNA in the reconstituted protein-synthesizing system was stimulated up to 4-fold over that observed with nonphosphorylated subunits. Synthesis of both the a and @ chains of globin was enhanced by phospho- rylation as shown by electrophoretic analysis. Since the sites phosphorylated by the mitogen-stimulated S6 kinase are


Phosphorylation of Ribosomal Protein S6 by CAMP-dependent Protein
Kinase and Mitogen-stimulated S6 Kinase Differentially Alters Translation of Globin mRNA* (Received for publication, September 9, 1985) Emilia Palen* and Jolinda A. Traugh From the Department of Biochemistry, university of California, Riverside, California 92521 The effects of phosphorylation of ribosomal protein S6 by two different protein kinases, the CAMP-dependent protein kinase and the mitogen-stimulated S6 kinase, or translation of globin mRNA in a reconstituted system and on binding of globin mRNA to 40 S ribosomal subunits were examined. The CAMP-dependent protein kinase incorporated 1.5 mol of phosphate/mol of 40 S ribosomal subunits. Phosphorylation of S6 by the CAMP-dependent protein kinase had no effect on binding of 3 ' terminus-labeled globin mRNA to 40 S ribosomal subunits. ['HILeucine incorporation with 40 S ribosomal subunits phosphorylated by the CAMP-dependent protein kinase was identical to that observed with nonphosphorylated 40 S ribosomal subunits, although on occasion, a slight inhibition (<lo%) was observed; there was no effect on the rate of synthesis of either the a or / 3 chains of globin. Phosphorylation with the mitogen-stimulated S6 kinase (2.5 mol/ mol) did not alter binding of globin mRNA to 40 S ribosomal subunits; however, translation of globin mRNA in the reconstituted protein-synthesizing system was stimulated up to 4-fold over that observed with nonphosphorylated subunits. Synthesis of both the a and @ chains of globin was enhanced by phosphorylation as shown by electrophoretic analysis. Since the sites phosphorylated by the mitogen-stimulated S6 kinase are identical to those observed in vivo in response to insulin and growth-promoting compounds, the data support the hypothesis that enhanced synthesis of specific proteins may be due to phosphorylation of S6 and that differential phosphorylation of S6 can alter translation of natural mRNA.
The components comprising the eucaryotic translational machinery contain a large number of proteins which can be phosphorylated by a variety of protein kinases (1, 2). Ribosomal protein S6, the major phosphoprotein in the 40 S ribosomal subunit, is phosphorylated at multiple sites. Evidence has been provided that hormones and growth factors promote the phosphorylation of ribosomal protein S6 in uiuo.
These include hormones which enhance cAMP levels, insulin, growth factors, tranforming viruses, and compounds which promote cell division (1-7).
In response to elevated levels of CAMP, the CAMP-dependent protein kinase has been shown to incorporate up to two *This research was supported by United States Public Health Service Grant GM21424. 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. phosphates into S6 in vivo to form the diphosphorylated derivative (8). The phosphate is incorporated into two major sites on S6. The same sites are phosphorylated in vitro by the CAMP-dependent protein kinase (8,9). These sites have been identified by 11) to be in the sequence Arg-Arg-Leu-Ser(P)-Ser(P)-Leu-Arg, where Ser(P) is phosphoserine. Insulin and growth-promoting compounds stimulate incorporation of 4-5 phosphates into S6 (4, 6, 12-15). One of the seryl groups phosphorylated in response to these compounds is the same as one of the residues phosphorylated by the CAMP-dependent protein kinase (16). The remainder are distinct and appear to be independent of the type of growth-promoting compound utilized (6,7,14,15,17,18). Several laboratories have isolated protein kinases which may be responsible for phosphorylation of these sites in uiuo.
These S6 kinases include protease-activated kinase I1 from reticulocytes (19) and the proenzyme and activated form of protease-activated kinase I1 from liver (20) and 3T3-Ll cells (14), protein kinase C from rat brain (21-23), H4 kinase from lymphosarcoma cells (24), and S6 kinase from 3T3-Ll cells (5) and from Xenopus oocytes (25, 26). Previously, we have shown that the proenzyme form of protease-activated kinase I1 and the enzyme activated in response to insulin phosphorylate the same sites on S6 (14). These sites are identical to those phosphorylated in cultured cells in response to insulin, epidermal growth factor, and tumor-promoting phorbol esters (14,15,17).
Multiple phosphorylation of S6 appears to be associated with enhanced protein synthesis. Synthesis of specific proteins has been observed within 30 min after stimulation of 3T3 cells with serum (27); this stimulation is not due to enhanced production of mRNA. Evidence has also been presented which suggests that ribosomes containing highly phosphorylated S6 are preferentially incorporated into polysomes (28,29). However, stimulation of phosphorylation of S6 by the CAMP-dependent protein kinase with &butyryl cAMP has little or no effect on protein synthesis in reticulocytes (30).
In recent studies, Burkhard and Traugh (31) have examined the effects of phosphorylation on binding of poly(A,U,G) to 40 S ribosomal subunits and on translation in a reconstituted protein-synthesizing system. With 40 S ribosomal subunits phosphorylated by protease-activated kinase 11, binding and translation of synthetic mRNA was enhanced. The opposite results, inhibition, were obtained when S6 was phosphorylated with the CAMP-dependent protein kinase. In this paper, we have extended investigation of the effects of differentially phosphorylated 40 S ribosomal subunits on binding of natural globin message to 40 S ribosomal subunits and on translation of the mRNA in a reconstituted translation system.

Phosphorylation of 40 S Ribosomul Subunits by the CAMPdependent Protein Kinase-Purified 40 S ribosomal subunits
were maximally phosphorylated with the purified catalytic subunit of the CAMP-dependent protein kinase; up to 1.5 mol of phosphate were added per mol of 40 S ribosomal subunits.
The degree of phosphorylation was quantified by counting the radioactive phosphate in ribosomal protein S6 from onedimensional polyacrylamide gels, and this value was confirmed by two-dimensional polyacrylamide gel electrophoresis. Since endogenous phosphate in S6 was determined to be approximately 0.5 mol/mol of S6 (9), total phosphate in S6 was 2.0 mol/mol. The diphosphorylated form of S6 was the primary product, with a small amount of monophosphorylated S6 observed. Tryptic phosphopeptide maps of S6 contained two phosphopeptides which were identified as A and B. The ratio between these two forms was dependent on the amount of phosphate incorporated.
Effects of Phosphorylation of 40 S Ribosomal Subunits by the CAMP-dependent Protein Kinase on Translation-The effects of phosphorylation on translation were monitored in a reconstituted protein-synthesizing system. As shown in Fig.  1, the system was dependent on the addition of globin mRNA. Protein synthesis was inhibited by the cap analog, m7GTP, indicating initiation was correct. The degree of stimulation observed upon addition of mRNA was dependent on the degree of contamination of 60 S subunits by 40 S subunits containing mRNA. In all experiments, a stimulation of at least 2-fold was observed upon addition of mRNA.
The reconstituted system was used to examine the effects of phosphorylation by the CAMP-dependent protein kinase on protein synthesis with ribosomal subunits containing 0, 1.0, and 1.5 mol of added phosphate (Fig. 2, left panel). Reactions were carried out in the presence and absence of 1 pg of globin mRNA/assay. This concentration of mRNA was limiting; saturation was observed at 2 pg/assay. Phosphorylation of S6 by the CAMP-dependent protein kinase had no effect on translation of globin mRNA. Incorporation of [3H] ' Portions of this paper (including "Experimental Procedures" are presented in miniprint at the end of this paper. The abbreviations used are: PAK 11, protease-activated kinase 11; GMP-PNP, guanyl-5"yl imidodiphosphate. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are available from the Journal of Biological Chemistry, 9650 Rockville Pike, Bethesda, MD 20814. Request Document No. 85M-3032, cite the authors, and include a check or money order for $2.00 per set of photocopies. Full size photocopies are also included in the microfilm edition of the Journal that is available from Waverly Press. leucine was virtually identical with all three forms of 40 S ribosomal subunits. On occasion, a small inhibition of protein synthesis was observed following phosphorylation with the CAMP-dependent protein kinase (Fig. 2, middle panel). In these experiments, 40 S ribosomal subunits were isolated from the remainder of the components of the phosphorylation reaction prior to addition to the protein-synthesizing system. Similar results were obtained when the entire assay mixture was added to the translation reaction mixture (Fig. 2, right  panel).
It is important to note that these experiments were carried out with a number of different preparations of 40 S ribosomal subunits, pH 5 fraction, initiation factors, mRNA, and protein kinase. Each time a component in the assay was changed, conditions were optimized for protein synthesis and a saturation curve for mRNA was carried out; it was important that suboptimal concentrations of mRNA were used in the experiments.
The effects of phosphorylation on the synthesis of the CY and / 3 chains of globin were monitored using 40 S ribosomal subunits containing 0 and 1.5 mol of phosphate (Fig. 3). In the absence of added mRNA, twice as much CY chain as p chain was produced. Upon addition of 1 wg of globin mRNA, synthesis of /3 globin was greatly stimulated whereas the synthesis of CY globin was stimulated less than 2-fold. This coincides with the results of Kabat and Chappell(44) showing that production of CY and /3 globin i s dependent on the amount of mRNA present, with translation of / 3 chain predominant at higher levels of message. As observed with total globin synthesis, the effects of phosphorylation of S6 by the CAMPdependent protein kinase had little effect on the synthesis of either CY or fi globin.
Binding of Globin mRNA to 40 S Ribosomal Subunits Phosphorylated by the CAMP-dependent Protein Kinase-The effects of phosphorylation of S6 were investigated by monitoring the initiation factor-dependent attachment of purified globin mRNA, labeled in the 3' teminus with cytidine 3', 5'-bi~[~'P]phosphate, to 40 S ribosomal subunits in the presence and absence of m7GTP. 40 S ribosomal subunits phosphorylated with the CAMP-dependent protein kinase were examined and compared with control subunits incubated in parallel in the absence of protein kinase. The amount of mRNA associated with the 40 S ribosomal subunits was monitored by sucrose density gradient centrifugation and calculated as the percent of total mRNA in the gradient. As shown in Table I, phosphorylation of S6 had little effect on binding of mRNA to 40 S ribosomal subunits either in the presence or absence of m7GTP. This coincides with the translation data, where phosphorylation of S6 by the CAMP-dependent protein kinase had little effect on translation.
Phosphorylation of 40 S Ribosomal Subunits by Mitogen-Stimulated S6 Kinase-Ribosomal subunits were phosphorylated with the endogenously activated S6 kinase from liver under conditions where 2.5 mol of phosphate were added. This form of the mitogen-stimulated S6 kinase was used to eliminate any potential problems that would result from the addition of trypsin/trypsin inhibitor to the protein-synthesizing system. Under these conditions, S6 was divided equally between the diphosphorylated and triphosphorylated derivatives when analyzed by two-dimensional gel electrophoresis. Phosphopeptide maps of phosphorylated S6 gave a pattern identical to that previously published for protease-activated kinase I1 from reticulocytes (16)   tion of globin mRNA in the reconstituted system with phosphorylated and nonphosphorylated 40 S ribosomal subunits is shown in Fig. 4. A small stimulation was observed with 40 S ribosomal subunits containing 2.4 mol of phosphate in the absence of added mRNA (left and middlepanels). When globin mRNA (1 pg/assay) was added to the reconstituted system, greater than a 2-fold increase in [3H]leucine incorporation was observed with nonphosphorylated S6 and a 4.5-fold stimulation with phosphorylated 40 S ribosomal subunits. At 15 min, globin synthesis with phosphorylated 40 S ribosomal subunits was 3.4-fold greater than that observed with nonphosphorylated subunits, with a 4.0-fold stimulation at 30 min. The effects were consistent when the synthesis of a and , f 3 globin was analyzed (Fig. 5). Phosphorylation of 40 S ribosomal subunits increased the synthesis of ,f3 globin by 3.9fold over that observed with nonphosphorylated subunits after 15 min of incubation, while the synthesis of a globin was stimulated 1.8-fold. The degree of stimulation of protein synthesis was dependent on the amount of phosphate added by the S6 kinase. With 1.5 mol of phosphate added, stimulation of translation was only %fold (Fig. 4, right panel). With less than one phosphate, no stimulation was observed (data not shown).

Effects of Phosphorylation of 40 S Ribosomal Subunits by Mitogen-stimulated S6
Kinase on Binding of rnRNA-The effects of phosphorylation of S6 by the mitogen-stimulated kinase on binding of 3'-labeled mRNA to 40 S ribosomal subunits was examined and compared with that of nonphosphorylated subunits. With subunits containing 2.5 mol of phosphate, the same amount of mRNA was bound to phosphorylated as to nonphosphorylated subunits (Table 11). In the presence of m7GTP, 25% more mRNA was associated with the phosphorylated subunits over that observed with nonphosphorylated. These differences are small in comparison with the stimulation observed with phosphorylated S6 in translation.

DISCUSSION
Ribosomal protein S6 is phosphorylated at two major sites in response to compounds elevating CAMP levels (8-10) and  a t four to five sites in response to insulin and growth-promoting compounds (4,6,(12)(13)(14)(15). Utilizing the cyclic AMPdependent protein kinase, we have produced S6 containing two maximally phosphorylated seryl residues (9, 13). With the mitogen-stimulated S6 kinase isolated as the proenzyme (protease-activated kinase 11) from reticulocytes or liver or as the endogenously activated form from liver, approximately 2.5 phosphoryl groups are incorporated into S6. At least five phosphopeptides are produced following extensive tryptic digestion of S6 (16). These phosphopeptides are the same as those obtained from cultured cells stimulated with insulin, tumor-promoting phorbol esters, or epidermal growth factor (14, 15, 17).
In recent years, only a few attempts have been made to probe the effects of phosphorylation of 40 S ribosomal subunits on protein synthesis. Leader et ~l .
(45) used ribosomes, endogenously phosphorylated in response to different stimuli, and highly purified initiation and elongation factors to examine binding of initiator tRNAs and synthesis of polyphenylalanine and globin. They found no significant differences. However, they did not monitor the amount of phosphate in S6 following isolation of the subunits, and the lack of an effect may be due to dephosphorylation during purification. With the purified protein kinases, it was possible to examine the effects of phosphorylation on binding and translation of mRNA. Burkhard and Traugh (31) found an inhibition of binding and translation of poly(A,U,G) with 40 S ribosomal subunits phosphorylated by the CAMP-dependent protein kinase. When 40 S ribosomal subunits were phosphorylated with S6 kinase, stimulation of both binding and translation was observed.
We have continued these analyses by examining binding of 3' terminus-labeled globin mRNA to 40 S ribosomal subunits phosphorylated either by the CAMP-dependent protein kinase or by the endogenously activated form of S6 kinase. The effects of phosphorylation have also been examined in a reconstituted protein-synthesizing system with the addition of a crude initiation factor fraction to make the system closely resemble endogenous conditions. This system is inhibited by the addition of the cap analogue, m7GTP. Phosphorylation of S6 by the CAMP-dependent protein kinase has little or no effect on binding of globin mRNA and little effect or a slight inhibition on translation when monitoring total globin production or synthesis of the (Y and 6 chains. Phosphorylation of S6 by the mitogen-stimulated S6 kinase has little effect on binding of mRNA to 40 S ribosomal subunits but significantly affects globin synthesis. Up to 4-fold stimulation is observed, and this is reflected by increased production of both the CY and 6 chains. These data are consistent with the results obtained previously with poly(A,U,G) (31).
Thus, the in vitro system mimmicks that observed in vivo where growth-promoting compounds enhance phosphorylation and stimulate protein synthesis (46). Floyd and Traugh (30) have shown that S6 is phosphorylated in reticulocytes in response to dibutyryl CAMP, but no alteration of globin synthesis is detected, reflecting the observations obtained with the reconstituted system. The effects of phosphorylation appear to be specific for specific mRNAs since phosphorylation of S6 by either protein kinase does not alter collagen synthesis (47) in a reconstituted protein-synthesizing system. In addition, Thomas et al. (27) have shown that the synthesis of a few specific proteins is enhanced immediately following serum stimulation of Swiss 3T3 cells.
Several groups have shown that S6 can be cross-linked to initiation factors, mRNA, and ribosomal proteins associated with the site of initiation on the ribosome (48-50). Thus, localized conformational changes in the region of mRNA binding or in the entire subunit could result from differential phosphorylation of S6. Kisilevsky et al. (51) have data indicating that phosphorylation of S6 following intoxication with ethionine leads to a conformational change of the whole ribosome as monitored by changes in reductive methylation. The proteins in the small subunit involved in the conformational change (S3, S4, S7, and S23/24) are topographically close to S6, and some have been shown to be involved in initiation by cross-linking reactions (48-50).
Specific classes of mRNA such as those stimulated in response to growth-promoting compounds or inhibited by cAMP could have a distinctive leader sequence which preferentially binds to the preferred conformation, whether it be due to the charged phosphoryl groups or due to a resulting structural change. Previously, we proposed that translation of different classes of mRNA are specifically enhanced under conditions of growth promotion where S6 is highly phosphorylated (20). Translation of another class of mRNA coding for housekeeping proteins such as collagen would be unaffected by the state of phosphorylation of S6, while a third class of proteins could be stimulated under conditions where levels of cAMP are elevated. We are currently examining other natural mRNAs to determine the effects of phosphorylation on their synthesis.