Fractionation and Partial Characterization of the Protein Synthesis System of Wheat Germ

Wheat germ SlOO supernatant was resolved into 10 components, all of which are required for tobacco mosaic virus RNA-directed incorporation of amino acid into protein. Two of the components, C2b and C2a, are, respectively, elongation factors 1 and 2, having molec- ular weights of 51,000 and 72,000. A third factor (C2e) binds Met-tRNA in a GTP-requiring reaction and is absolutely required for the formation of both 40 S ribosome .Met-tRNAp and 80 S ribosome *Met-tRNAyt complexes. This factor is considered to be eucaryotic initiation factor 2. Formation of 80 S ribosome Met- tRNAy complexes is augmented 2 to %fold by the simultaneous addition of an mRNA, factor C1, and frac- tion D2(a + b). Factor C1 also reverses the inhibition of translation by an increased concentration of monova- lent cations and promotes the translation of a "competitively inhibited'' mRNA. These observations suggest that factor C1 and either factor D2a or D2b (or both) function in mRNA attachment reactions. Factor C1 has a molecular weight of 115,000 and appears to be made up of two or three subunits. Factor D2b has been purified to essential homogeneity and has a molecular weight of 55,000. Its interaction with mRNA and its molecular weight suggest that this factor is the wheat germ equivalent of reticulocyte eucaryotic initiation factor 4A. Eucaryotic protein synthesis is a complex process involving the interaction of a substantial number of initiation and elongation factors (1, 2). As part of a detailed study of the initiation process, we have fractionated wheat germ lysate into a number of components based on their function in the primary reaction of protein synthesis, namely amino acid polymerization. This report describes the resolution of the system

* This work was supported by Grants GM-15122, CA-06927, and RR-05539 from the National Institutes of Health, by Grant 8100166 from the United States Department of Agriculture, and by an appropriation from the Commonwealth of Pennsylvania. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact, Portions of this paper (including "Materials and Methods" and Figs. [1][2][3][4][5][6][7][8] are presented in miniprint as prepared by the authors. Miniprint is easily read with the aid of a standard magnifying glass.

RESULTS
Resolution of the Soluble Factors-The initial step in the fractionation of the soluble factors resolves the SlOO supernatant into two fractions, C and D, each of which can be used as an assay reagent allowing the purification of the factors in the other fraction ( Fig. 1). In practice, fraction C was first resolved into the components indicated, i.e. C1 and C2a-e, and then, using a partially resolved system of the C components (Cl, C2(b + c), C2e), the D components were fractionated. The requirement of the system for the TMV-RNA2directed incorporation of amino acids into protein is presented in Table I. Where appropriate, the different components have been designated according to the standardized system (12) to allow a comparison to the fractionated reticulocyte system. The data of Experiment 1 demonstrate the requirement for components C1 and C2b (EF-1) and a strong stimulation by components C2c and C2e (eIF-2). The data of Experiment 2 demonstrate a strong requirement for components C2a (EF-2), D2a (eIF-4B), D2b (eIF-4A), D l (eIF-3), and D2(c + d).
Functional Characterization of Factors C26 (EF-1) and C2c"Fraction C2(b + c) (fraction C2 purified through the heparin-Sepharose step) has sufficient EF-1 activity such that 2 pg provide a saturating level of activity in the poly(U)catalyzed polymerization of phenylalanine. When chromatographed on a column of AcA 44, its activity in the protein synthesis assay was resolved into two components (Fig. 2). and 2 pg of C2e as indicated. In experiment 2, the complete system contained the basic assay system with KmMs ribosomes (100 pg of RNA), 6 pg of C1, 6 pg of C2(b + c), and 2 pg of C2e. Other additions were 1 pg of C2a, 4 pg of D2a, 1.5 pg of D2b, 2 pg of Dl, and 2 pg of D2(c + d) as indicated.

Component omitted
Incorporation pmol The heavier component eluted at a position corresponding to a molecular weight of 51,000 (Fig. 3), and its activity was coincident with the EF-1 activity. Analysis by SDS-acrylamide gel electrophoresis showed a single band with molecular weight of 56,000 (Fig. 4, lane a). The lighter component in the AcA fractionation (factor C2c) eluted at a position corresponding to a molecular weight of 24,000 (Fig. 3). It had no EF-1 activity, and SDS-acrylamide gel electrophoresis showed several bands ranging in molecular weight from 15 to 20,000 (Fig.  4, lane f ) . Functional Characterization of Factor C2e (eIF-2)"The binding of Met-tRNA" to ribosomes as determined by sucrose gradient analysis was completely dependent on factor C2e (see legend to Fig. 7). When the reaction was assayed by retention of radioactivity on nitrocellulose filters in the presence of 5 mM Mg2', a similar strong requirement for this factor was obtained (Table 11). The fiiter assay for ribosome-dependent binding of Met-tRNAye' could also be used to monitor two other components, factor D2(c + d) and fraction Dl.
Factor D l was required to almost the same extent as factor C2e, while factor D2(c + d) stimulated the ribosome-dependent reaction 2.5-fold. At a low Mg" concentration (0.2 mM), factor C2e bound Met-tRNA," in a GTP-dependent reaction

Requirements for the binding of Met-tRNA? to ribosomes as assayed by retention of radioactivity on nitrocellulose filters
The conditions were those of the standard assay for the formation of ribosome. Met-tRNA? complexes as described under "Materials and Methods." Where indicated, 4 pg of factor C2d were added in place of C2e.  in the absence of ribosomes and any other factor (Table 111). Based on these observations, we have designated C2e as eIF-2, the factor that binds Met-tRNA" in a ternary complex with GTP (1,2). The data of Tables I1 and I11 also show that factor C2e can be replaced in the Met-tRNA binding reaction by fraction C2d (see Fig. 1). Apparently, both of these fractions contain the eIF-2 activity. Of the remaining components, only fraction D2(c + d) demonstrated significant Met-tRNAF binding activity (Table 111). The reaction with this component, however, was independent of G T P and, under the 5 mM Mg2+ condition of the ribosome binding assay (Table I1 and Fig. 7), this fraction could not replace C2e or C2d in functioning as an eIF-2-like component. On SDS-gel electrophoresis, factor C2e showed a considerable number of bands, with the predominant ones having molecular weights of 36,000 and 50,000 (data not shown). Functional Characterization of Factor C2a (EF-2)"The requirement for this component was realized when the resolved system was assayed with ribosomes washed with 0.6 M KC1 (KmMa ribosomes). The purification steps on DE23 and hydroxylapatite were initially developed to provide EF-2 from an otherwise discarded fraction. Having observed that this fraction restored the reaction with the KmMs ribosomes, it was chromatographed on a column of AcA 44. The results presented in Fig. 5 show that the TMV-RNA-catalyzed amino acid polymerizing activity co-elutes with the EF-2 (poly(U)dependent Phe-tRNA polymerizing) activity. Based on its elution from the AcA 44 column, the molecular weight of this component is 70,000 (Fig. 3). On an SDS-acrylamide gel, a major band with molecular weight of 72,000 was obtained together with two minor bands with molecular weights of 63,000 and 93,000 (Fig. 4, lane c).
Functional Characterization of Factors D2a (eIF-4B) a n d D2b (eIF-4A)"The resolution of these components on a n AcA 44 column is shown in Fig. 6. The two peaks of protein synthesis activity correspond to molecular weights of 80,000 and 50,000, respectively (Fig. 3). The fraction from which these components were obtained, D2(a + b), is required for the mRNA-dependent augmentation of the formation of 80 S ribosome-Met-tRNA? complexes (Fig. 7), suggesting that one or both of the subfractions are involved in mRNA attachment reactions. This characteristic would place these factors in the eIF-4 series (12). The specific designations of these components as eIF-4B and eIF-4A are based on the molecular weight analogy to the reticulocyte factors. On SDS-acrylamide gels, factor D2b (eIF-4A) is homogeneous (Fig. 4, lane b), migrating with a molecular weight of 55,000. Factor D2a (eIF-4B) showed a substantial number of bands with no obviously major component (data not shown).
Functional Characterization of Factor Cl-Factor C1 has no clear counterpart among the components of the reticulocyte system. In the Met-tRNA? binding reaction, this factor is required for the mRNA-catalyzed augmentation of the formation of 80 Sribosome.Met-tRNATe' complexes ( Fig. 7), suggesting a function in a reaction necessary for mRNA attachment to ribosomes. In an earlier study (7), we showed that amino acid incorporation is inhibited by increased concentration of K' and that this inhibition varied with the mRNA being translated. In addition, we demonstrated that the mRNAs whose translation was inhibited to a greater extent by the high salt concentration were weaker in competitive translation (see below) and were more susceptible to inhibition by pm7G (7). Sensitivity to competitive translation and to pm7G inhibition is a likely consequence of a situation in which an mRNA attachment reaction is rate-limiting to translation. By analogy, the high K' effect might also be a consequence of the same reaction, becoming rate-limiting to translation. Supplementing the reaction with a component functioning in the mRNA attachment reaction might then be expected to reverse the high salt inhibition and to augment the translation of an mRNA whose translation was being competitively inhibited. The data of Tables IV and V test this idea for the high K' inhibition. In an assay with limiting TMV-RNA (Table IV), 62.5 m~ KC1 was optimum, and the addition of the different factors, either alone or in combinations of two (Cl + D l , C2b + D l , C2e + Dl), had no effect (data shown only for factor Cl). When the monovalent K' ' T Factors of Wheat Germ concentration was increased, either with KC1 or KOAc, incorporation was inhibited. Addition of factor C1 now reversed the inhibition and in fact established an optimum at an increased salt concentration. Supplementing the reaction with the other components of the translation system did not reverse the inhibition except for factor D l (Table V). Further purification of this factor on DEAE-cellulose (Dl-DE) abolished its stimulatory activity. The reversal by factor C1 was also obtained at a saturating level of TMV-RNA, again more strikingly when the salt was raised to an inhibitory level (Table   IV) .
Typical results for an mRNA competition assay are presented in Table VI. The experimental conditions are designed so that the presence of AlMV-RNA 4 inhibits globin mRNA translation by 80-90%. Addition of factor C1 reverses this inhibition and in sufficient amount can completely restore globin translation to the noncompetitive level (data not shown). Tests of the other fractionated components indicated some activity in fractions D l and C2(b + c). Resolved factor C2b and factor D l purified on DEAE, however, lacked any stimulatory activity.
Chromatography of fraction C1 on a column of Sephadex G-150 is shown in Fig. 8. Either 1 mM Mg'+ or 2 M urea had to be included in the chromatography to prevent the activity from spreading throughout the elution. The major peak of activity (tubes 21-23) had a specific activity twice that of the starting C1 in the assay for I4C-aminoacid incorporation, and its molecular weight calculated from gel filtration on AcA 44 was 115,000 (Fig. 3). SDS-acrylamide gel electrophoresis    Factors of Wheat Germ n showed three prominent bands of molecular weights ranging between 45,000 and 50,000 (Fig. 4, lanes d and e ) . DISCUSSION This report together with the accompanying report (3) describe the resolution of 10 separate protein factors, all of which are required for amino acid polymerization. Two of these, C2b and C2a, catalyze the poly(U)-dependent polymerization of phenylalanine and are consequently designated as elongation factors EF-1 and EF-2. None of the other components can replace these two factors in the poly(U)-catalyzed reaction. Thus, the remaining factors appear to be functioning in the initiation of the protein chain.
With regard to specific functions, the strong requirement for factor C2e in all reactions in which Met-tRNA? binding is involved (Tables I1 and I11 and legend to Fig. 7 ) makes it likely that this component is eIF-2, the factor that forms a ternary complex with Met-tRNAp and GTP (1,2,14). The stimulation of the formation of 80-S ribosome. M e t t R N A P t complexes by the combination of C1, D2(a + b), and mRNA and the dependence of this stimulation on all three of these components (Fig. 7 ) suggest that at least two of the factors are functioning in the mRNA attachment reaction. The unique ability of factor C1 to reverse the inhibition of translation by high salt (Tables IV and V) and to promote the translation of a "competitively inhibited" mRNA (Table VI) provides strong support for a function of this factor at the level of mRNA interaction. In addition, in an assay that measures directly the binding of radioactive mRNA to 40 S ribosomes, we have found a strong requirement for factors C1 and D2b.3 The functions of these two factors can therefore be relegated to reactions in which mRNA attaches to the ribosome. When chromatographed on Sephadex G-150, factor C1 has two peaks of activity (Fig. 8). We are currently trying to determine the difference between these fractions.
The primary impetus for the resolution of the different protein synthesis components is our intent to analyze the process of mRNA binding. Several attempts to delineate the details of this process with partially resolved components have indicated the necessity to fist obtain total resolution and purification of the different components. Two points illustrate these considerations. The fiist is the finding that Met-tRNA? binding to the 40 S ribosome is not required for mRNA binding. Proving this requires a system strongly dependent upon eIF-2, a situation only now attained in the ~ S. Seal, A. Schmidt, and A. Marcus, manuscript in preparation.

Elongation and
Initiation Factors of Wheat Germ 865 wheat germ system (see Table I Throughout this work, a loss of activity that occurred during an attempted purification often indicated the existence of a previously unrecognized component. As a consequence, established purification procedures had to be modified or abandoned. Nevertheless, the achievement of extensive reconstitution (Table I) is an indication of the usefulness of this approach. An obvious additional advantage is that we can be reasonably certain that the components used in studying partial reactions are indeed functional in the protein synthesis process.