Eur. J. Biochem. 59, 137-143 (1975)

Specificity of ATP-Dependent and GTP-Dependent Protein Kinases with Respect to Ribosomal Proteins of Escherichia coli

Olaf-Georg ISSINGER, Michael C. KIEFER, and Robert R. TRAUT

Department of Biological Chemistry, School of Medicine, University of California, Davis

(Received April 15, 1975)

Two protein kinases differing in substrate specificity were used to phosphorylate the 3 0 3 and the 50-S ribosomal subunits of Escherichia coli. The catalytic subunit from the rabbit skeletal muscle protein kinase phosphorylates proteins S1, S4, S9, S13 and S18 of the 30-S subunit and proteins L2, L4, L5, L16, L17, L18 and L23 of the 50-S subunit with [p3’P]ATP as phosphoryl donor. A second protein kinase isolated from rabbit reticulocytes, formerly shown to phosphorylate preferentially acidic proteins and to use GTP as well as ATP, strongly phosphorylated protein S6, an acidic protein of the small ribosomal subunit, and to a lesser extent proteins L7 and L12 ofthe large subunit. Evidence is presented showing different phosphorylation patterns when either whole subunits or the extracted proteins were used as substrate for the protein kinase. Kinetic studies showed proteins S1 and S4 to become most rapidly phosphorylated. Although most proteins incorporated less than stoichio- metric amounts of phosphate, it is shown that with a high excess of ATP L2 bound 1 mol phosphate/ mol protein.

The catalytic subunit of the cyclic-AMP-dependent protein kinase isolated from rabbit skeletal muscle [I] has been shown to phosphorylate selectively certain proteins in both the intact 30-S and 5 0 4 ribosomal subunits of Escherichia coli [2]. The substrate specifi- city of the reaction was established by analysis of the phosphorylated ribosomal proteins by electrophoresis in polyacrylamide disc gels containing either sodium dodecylsulfate, or urea. Unambiguous identification of all the individual ribosomal proteins which were phosphorylated was not possible using these analytical techniques, neither of which resolves all the ribosomal proteins present in either of the two subunits.

It has been shown that phosphorylated ribosomal proteins do not occur in vivo in E. coli under a variety of growth conditions [J. Gordon, 1971 ; J. A. Traugh and R. R. Traut, unpublished results; P. Sypherd, M. Nomura, and H. G. Wittmann, personal com- munications]. However, it was recently reported that certain ribosomal proteins of E. coli do become phosphorylated during infection of E. coli with bac- teriophage T7 [3]. Despite the lack of evidence for any general physiological role of ribosomal phospho- proteins in uninfected E. coli cells, it was considered worthwhile to continue the earlier studies. Protein kinases of different specificities may serve as valuable tools for the study of the protein topography of E. coli ribosomes ; they provide a means of selectively labeling ribosomal proteins in order to facilitate other experi-

ments, such as those involving cross-linking with bifunctional reagents. Finally the ribosomal proteins provide a spectrum of primary structures with which to explore the determinants of the specificity of pro- tein kinases.

It has recently been shown that a newly charac- terized protein kinase from rabbit reticulocytes, which phosphorylates acidic and not basic proteins and can use GTP in addition to ATP as a phosphoryl donor, phosphorylates selectively protein(s) L12 and/or L7 from the 50-S subunit of E. coli [4]. This protein has been shown to have functional and immunchemical properties in common with certain eucaryotic ribo- somal proteins [5] and is involved in many ribosomal functions related to GTP hydrolysis [6]. The action of this enzyme on the 30-S subunit has been investi- gated in the work reported here.

The experiments reported here extend the previous results in two major respects. First, a more powerful analytical technique, electrophoresis in polyacrylamide gels in two dimensions [7,8] has been employed to identify unambiguously the phosphorylated ribosomal proteins. Second, attempts have been made to find conditions for the incorporation of stoichiometric amounts of phosphate into the ribosomal proteins. In the earlier experiments cited above, only fractional amounts of any given ribosomal protein species were phosphorylated. This observation was particularly puzzling, since there was no evidence for the presence

138 Phosphorylation of Ribosomal Proteins of E. c d i

of endogenous phosphate in the ribosomal proteins from E. coli. Possible explanations were that [ Y - ~ ~ P I - ATP, the phosphoryl donor in the in vitro reaction, became limiting, that phosphatase activity was present in the reaction mixtures, or that the putative ribosomal protein substrates existed in heterogeneous confor- mational states either in the intact ribosome or even after extraction in urea. The conditions in the experi- ments reported here were varied in an attempt to decide among these possibilities and to achieve stoichiometric phosphorylation of the ribosomal pro- tein substrates.

Evidence is presented here for the stoichiometric phosphorylation in the intact 50-S ribosomal sub- units of protein L2 as well as of proteins L4, L5, L16, L17, L18, L73 and L7/L12 in lower amounts; and in the intact 30-S ribosomal subunits of proteins S1, S4, S9, S13 and S18 in less than stoichiometric amounts. Further studies with the partially purified protein kinase, having preferential substrate specificity for acidic proteins, showed that protein S6 from the 30-S subunit becomes rapidly phosphorylated.

MATERIALS AND METHODS

R fbosomes

The preparation of ribosomes and ribosomal sub- units was carried out as described previously [2]. The method for the extraction of ribosomal proteins was that of Hardy et al. [9]. The extracted proteins were lyophilized and suspended in the buffer for the first electrophoresis.

Protein Kinases

Rabbit skeletal muscle kinase was a gift from the laboratory of E. G. Krebs. Only the catalytic subunit was employed. Rabbit reticulocyte kinases were pre- pared as described by Traugh and Traut [lo]. The fractions previously characterized by their behavior during chromatography on DEAE-cellulose as peaks I1 and 111 were used in the present experiments without further purification.

Preparation of [ 3 2 P ] A T P and ( 3 2 P ] G T P

[y-32P]ATP was prepared by a modification of the method of Glynn and Chappel [ l l ] . [y-32P]GTP was prepared by a modification of the procedure for ATP following a suggestion by Dr Earle Davie.

Phosphorylat ion

The reaction mixture was as described earlier [4]. When extracted ribosomal proteins were used instead of ribosomal subunits, the reaction mixture was

similar to that described, except that it contained 400 mM NH,Cl in order to solubilize the extracted ribosomal proteins. The higher salt concentration did not markedly affect the kinetics of phosphoryla- tion. About 1 pg of highly purified catalytic sub- unit of the protein kinase from rabbit muscle [l] was added to each reaction mixture. The enzyme was diluted in a 0.05% bovine serum albumin solution in 25 mM 2-(N-morpholine)-ethane sulphonic acid buffer, pH 6.5 [P. Bechtel, personal communication]. About 1 SO pg of the less purified DEAE-cellulose peak I1 and 111 fractions were used in each reaction mixture.

Polyacrylamide Gel Electrophoresis

a) Polyacrylamide gel electrophoresis in two dimen- sions was carried according to Howard and Traut [8] as modified from the method originally described by Kaltschmidt and Wittmann [7] . Samples were divided into two portions of 30 p1 and applied to two different gel tubes of identical composition of pH 8.7 with an inner diameter of 3 mm. The basic proteins were separated in tubes 75 mm long with a 10-mm spacer gel, the acidic proteins in tubes 50 mm long and also with a 10-mm spacer gel. The cathode was at the bottom for the former and at the top for the latter. Gels were run at constant current with 5 mA per tube at 25 'C for 3 -4 h. Following electrophoresis in the first dimension, the gels were used directly for electro- phoresis in the second dimension. The dialysis step originally described by Kaltschmidt and Wittmann [6] was omitted. After the electrophoresis of the proteins in the second dimension at pH 4.5 the gels were stained with Coomassie brilliant blue R250 [12], photo- graphed and dried under vacuum at 90°C according to the method described by Maize1 [ 131. The gels could also be used for autoradiography prior to drying. For this purpose they were placed on a glass plate, wrapped in a plastic bag and incubated with X-ray film.

b) Polyacrylamide - dodecylsulhte gel electro- phoresis in polyacrylamide slab gels 2 mm thick was performed as described by Studier [14]. The acryl- amide concentration of the separating gel was 15 x. The best resolution of 30-S and 5 0 3 ribosomal proteins was obtained when samples were applied in volumes not exceeding 20 pl and containing 30- 50 pg of protein.

RESULTS

Proteins Phosphorylated by Rabbit Skeletal Muscle Protein Kinase

30-S Proteins in the Intact Subunit. When intact 30-S subunits were incubated with the protein kinase from rabbit skeletal muscle and [Y-~'P]ATP as phos-

0.-G. Issinger. M. C. Kiefer. and R. R. Traut 139

phoryl donor, proteins S1, S4, S9, S13 and S18 were phosphorylated (Fig. 1 B).

Extracted 30-S Proteins. When proteins extracted from the 30-S subunits were incubated with the skeletal muscle protein kinase, the number of phosphorylated proteins was reduced to two major components: pro- teins S9 and S18 (Fig. 1 C). The phosphorylation of S4 was greatly reduced and neither S1 nor S13 was detectable. However, protein S21, not phosphoryl- ated in the intact subunit, was phosphorylated in solution.

50-S Proteins in the Intact Subunit. When 50-S ribosomal subunits were incubated with protein kinase from rabbit skeletal muscle and ATP (see above), 7 major radioactive spots could be observed (Fig. 2 B). Six of these could be identified unambiguously: L2, L4, L5, L16, L17 and L23. The seventh spot below L16 does not correspond to any stained spot corre- sponding to known 50-S ribosomal proteins. This protein may represent a minor contaminant present in the 50-S preparation, which is nevertheless a good substrate for the protein kinase. However, it is also possible that this protein is identical with S18 from the 30-S subunit, as it is located at exactly the same place in the two-dimensional gel as S18.

E-xtracted 50-S Proteins. The total extracted 50-S ribosomal proteins showed the same specificity for protein kinase as those in the intact particle, with the exceptions that L5 was less strongly phosphorylated in the extracted proteins than in the intact particle, and that L18 and L28, not phosphorylated in the intact particle, are phosphorylated after extraction (Fig. 2C). The strong phosphorylation of the unidenti- fied protein also occurs in the extracted proteins; moreover its relative intensity of phosphorylation is greater in the latter experiment.

Eiyect of Excess A TP on the E.utent of Phosphorylution of 50-S Proteins

With the amounts of 32P-labeled nucleotide tri- phosphate used in the preceding experiments, about 1.0 pmol of phosphorus was incorporated into the total ribosomal proteins which were phosphorylated ; i .e., the seven phosphorylated proteins of the 50-S subunit together contained 1.0 pmol and no single protein was phosphorylated stoichiometrically. This estimate is based on the amount of [Y-~~PIATP added to the incubation mixture compared to the amount of radioactivity extracted from the two-dimensional slab gels of the stained ribosomal proteins. The esti- mate does not consider the radioactivity which remains at the origin of the two-dimensional gel. The ratio of ribosomes to ATP in the incubation mixtures de- scribed above was 1.4 to 1 (about 1400 pmol of ribo- somes and 975 pmol of [y-32P]ATP). The ratio be- tween the total of seven proteins of the 504 subunit

which are susceptible to phosphorylation (seven pro- teins) and [‘p3*P]ATP is therefore about 10 : 1, as- suming one copy of protein per particle and one phosphorylatable site per protein.

With these conditions under which ATP is limiting, the maximum molar yield of any specific phosphoryl- ated product would be on the average less than 10 2). The limiting concentration of ATP may be further reduced by hydrolysis. The yield of phosphoprotein may be further lowered due to the action of phos- phatases. The fact that only a fraction of any given protein species was phosphorylated is consistent with the fact that there is not a perfect coincidence between the radioactive spots revealed by the autoradiograms with the stained protein patterns. The addition of negatively charged residue would lower the mobility of ribosomal proteins in both dimensions of the gel, mainly in the first dimension.

An experiment was performed in order to determine whether more extensive phosphorylation of 50-S ribosomal proteins was possible by increasing the ratio of ATP to ribosomes. After the standard in- cubation of 504 subunits with [Y-~*P]ATP, 200 nmol of nonradioactive ATP was added and the incubation was continued for 15 min. The reaction was then stopped by the addition of 20 p11 M MgClz and 300 p1 glacial acetic acid. The results of this experiment are shown in Table 1. In the preceding experiments with [.u’-~~P]ATP present in concentrations of slightly less than 1 nmol/l.4 nmol ribosomal subunits, about 60- 80% of the radioactivity was located to the left of the stained spot. As can be seen in Table 1 , the chase experiment with an excess of nonradioactive ATP led to the almost complete coincidence of the radioactive and stained spots representing protein L2. By contrast the yields of the phosphorylated forms of proteins L4, L5 and L17 were not increased significantly.

Kinetics of Phosphorylution with Skeletal Muscle Protein Kinuse

Extracted 30-S Proteins. The extracted total pro- teins from 30-S subunits were incubated in the pres- ence of 400mM NH,CI with rabbit skeletal muscle protein kinase and [p3’P]ATP under the incubation conditions previously described. The reaction was initiated with the addition of [y-32P]ATP.

Aliquots of 10 p1 were taken at various times (see Fig. 3) from a 100-p1 sample, pipetted into 300 pl of dodecylsulfate gel sample buffer to stop the kinase reaction and incubated at 80°C for 30 min. The samples were then dialyzed against propionic acid and lyophilized. Samples containing 30- 50 pg of protein were then suspended in 20p1 of the sample buffer for electrophoresis in dodecylsulfate. Aliquots of 10 p1 were applied to a polyacrylamide - dodecyl- sulfate slab gel. The autoradiogram in Fig. 3A shows

140 Phosphorylation of Ribosomal Proteins of E . tali

Fig. 1. ( A ) A stained two-dimensional gel pattern of’ the 30-S sub- unit. ( B ) Autoradiogram of phosphorylated 3 0 3 subunits with the protein kinase from rabbit skeletal muscle and [y-32P]ATP as phosphorous donor. ( C ) Autoradiogram of phosphorylated extructed 3 0 3 ribosomal proteins with ;he protein kinase from rabbit skeletal muscle and [y-”PJATP as phosphorous donor. (A) Separation o f the proteins was as described under Methods. Identification of the proteins was according to the nomenclature of Wittmann et al. [14]

Fig. 2. ( A ) A two-dimensional gel pattern of the SO-S subunit. ( B ) Autoradiogram showing !he proteins which become phosphoryl- ated when 50-S subunits were subjected to phosphorylation. (C) Autoradiogram showing the proteins which become phosphorylated when extracted 50-S ribosomal proteins were used as suhslrates for phosphorylation. In both cases [y-”PIATP was the phosphoryl donor and a protein kinase from rabbit skeletal muscle was used

0.-G. Issinger. M. C. Kicfer, and R. R. Traut 141

Table 1. Efyect of ~ S C Y S S A T P on the stoichiometry ofphosphorylation as indicated by the distribution of 23P in the proteins coincident with the stainrd spot or left brside it

spot Radioactivite ATP only (3.2 nrnol) + cold ATP (200 nrnol)

coincident with left of stained spot stained spot

coincident with left of stained spot stained spot

L2 L4 L5 L16 L17

counts/rnin (7" recovered)

SO0 (23) 1651 (77) 1056 (88) 150 (12) 1102 (38) 1776 (62) 380 (33) 774 (67) 170 (25) 515 (75) 180 (17) 883 (83) 185 (18) 830 (82) 150 (82) 789 (84) 651 (26) 231 (74) 559 (74) 200 (26)

Fig. 3. ( A ) 7ii~ie-course ( I / the e.vtructed ribosomal proteins oj the 30-S subunil. ( B ) A time course of the e.rtructed SOLS riho.tomal proteins

about seven phosphorylated proteins. This number is greater by two than five 30-S proteins identified by two-dimensional gel electrophoresis. The phosphoryl- ated proteins range in molecular weight from 11 000 to 70000. S1 is the 3 0 4 ribosomal protein with the highest molecular weight that becomes phosphoryl- ated. The next radioactive band which corresponds to a known ribosomal protein is S4. Its kinetics of phosphorylation is unlike S1. Protein S9, which was shown to be the most highly phosphorylated protein in the experiments described previously, was also the major phosphorylated protein in these experi- ments. However in contrast to S4 and S1, which be- came phosphorylated within 0.5 min, protein S9 was not significantly phosphorylated until 2 min of in- cubation. Phosphorylation was detectable after 0.5 min and increased until 12 min, after which no further increase took place. Proteins S13 and S18 have the same molecular weight and migrated in the same band in polyacrylamide - dodecylsulfate gels. Interpretation based on the analyses of the two-dimensional gels suggests that the major phosphorylated component is S18. In contrast to the other proteins, S13/SlS

became phosphorylated relatively slowly. The appear- ance of radioactivity was visible only after 2 min of incubation.

E.xtracted 5 0 4 Proteins. Fig.3B shows a time- course experiment of extracted 5 0 4 proteins. All the proteins appeared gradually after an incubated period of about 0.5 min. The extent of phosphorylation of all the proteins increased until 12min and then remained constant.

Protein Substrate Specificity of the Protein Kinase Specific for Acidic Proteins and GTP

By comparison with the rabbit skeletal muscle kinase, for with GTP is not a substrate [l], the DEAE- cellulose peak 111 enzyme from rabbit reticulocytes utilizes GTP as well as ATP as a phosphoryl donor. The ribosomal protein substrate specificity of the latter enzyme was investigated by the procedures de- scribed for the skeletal muscle enzyme. The DEAE- cellulose peak 111 enzyme, prepared as previously described, is a mixture of two protein kinase activities, one with relatively high specificity for casein (or acidic

142 Phosphorylation of Ribosomal Proteins of E . coli

k-ig.4. An autorodiogram oj' the 304 .subunits nlwn phosphorjlari~d uith DEAE-cellulose 111 ,fraction and [y-32P]GTP as phosphorous donor. The spot below the notation coincides with the stained spot S6

proteins) as opposed to histone (basic proteins) as phosphoryl acceptors [4].

a ) 30-S Subunit. Whereas the skeletal muscle pro- tein kinase with ATP as phosphoryl donor led to the phosphorylation of protein S1, S4, S9, S13 and S18, a similar incubation with the DEAE-cellulose peak 111 enzyme with GTP as phosphoryl donor led to the phosphorylation only of protein S6 as the major product (Fig. 4).

b ) 50-S Subunit. With the enzyme fraction DEAE- cellulose peak I11 and ATP as phosphoryl donor, proteins L2, L4, L5, L16, L17, L23 and L7/L12 were phosphorylated [4]. These proteins are identical to those found for the rabbit skeletal muscle protein kinase, except for proteins L7 and L12. When GTP was substituted for ATP as phosphoryl donor, only proteins L7 and/or L12 were phosphorylated. Since the crude DEAE-cellulose peak 111 enzyme has been fractionated by chromatography on phosphocellulose into two components [lo], one specific for histone and ATP, the other for casein and ATP or GTP, it is concluded that the casein-specific protein kinase present in the fraction phosphorylates selectively proteins L7 and L12. A preliminary report of this observation has been reported previously [4]. How- ever, it should be noted that the phosphorylation of protein S6 by the casein-specific protein kinase using ATP or GTP as a phosphoryl donor was several times higher than the observed phosphorylation of proteins L7/L12.

DISCUSSION

The present report extends previous results from this laboratory on the phosphorylation of proteins from the ribosomal subunits of E.coli by protein

kinases from eucaryotic sources. First, two-dimen- sional polyacrylamide gel electrophoresis has been employed; second, the substrate specificities of two protein kinases, that from rabbit skeletal muscle previ- ously studied, and a new kinase from rabbit reticulo- cytes which preferentially utilizes GTP over ATP, have been compared. Experiments were performed both with intact ribosomal subunits and on the proteins extracted from them.

The previous report gave evidence for the phospho- rylation of the following 30-S ribosomal proteins, both in the intact subunit and in the extracted proteins with rabbit skeletal muscle protein kinase [4] : S4, S9, S18 and S19. The present work, employing analytical techniques of greater resolving power, extends and slightly alters the earlier conclusions. Proteins S1, S4, S9, S13 and S18 were phosphorylated in the intact subunit. No evidence for the phosphorylation of S19 was found. Only two of these proteins, S9 and S18 were phosphorylated when total 30-S proteins previ- ously extracted from the particles were tested. In addition, protein S21, not found to be phosphorylated in the intact particle, was phosphorylated in the extracted proteins. The failure to find phosphorylated S1, S4 and S13 cannot be attributed to inactivation of the enzyme due to the higher salt concentration employed, since S9 and S18 appear more strongly phosphorylated than in the intact particles. In addition protein S21 is phosphorylated only in the extracted proteins. The failure to confirm the phosphorylation of S19 reported previously is most likely explained by the use of two-dimensional gel electrophoresis in the present study. Differences in the specificity of protein kinase between intact subunits and extracted proteins are likely due to the fact that the structural determinants of kinase specificity reside not only in primary structure, but also upon conformational states imposed upon or maintained by the integration of proteins in the intact ribosome structure. It is clear that the action of the protein kinase cannot be ex- plained by the existence of exposed and protected proteins in the ribosomal subunit.

The previous report gave evidence for the phos- phorylation of a group of 50-S proteins identified at that time in terms of a nomenclature based only upon their behavior in one-dimensional gel electrophoresis (Traut et al. [15,16]). The relationship of this nomen- clature with that of Kaltschmidt and Wittmann, based upon two-dimensional gel electrophoresis, has never been established. Only three of the 50-S proteins previously found, L2, L3, L5 were readily correlated with the new nomenclature. The present study shows proteins L2, L4, L5, L16. L17 and L23 to become phosphorylated when intact 504 subunits are in- cubated with ATP and protein kinase from rabbit skeletal muscle. Of these only L17 is not also phos- phorylated when extracted proteins as opposed to

0.X. Issingcr, M. C . Kiefer, and R. R. Traut 143

intact particlcs are employed. In addition proteins L18 and L28 become phosphorylated only in the extracted 50-S proteins.

In the second dimension of the two-dimensional electrophoresis procedure, proteins are separated primarily according to molecular weight, owing to a relatively high acrylamide concentration, although charge plays a role in the separation obtained. These considerations lead to the prediction that all phos- phorylated protein spots should be displaced to the left and very slight above the unmodified proteins. This has been found to be the case in the present study, and is most graphically apparent when indivi- dual proteins are fractionally phosphorylated such that radioactive spots do not coincide with the stained spots representing the major fraction of the protein species. This phenomenon introduces certain difficulties in unambiguously identifying the proteins that are phosphorylated in cases in which one of the proteins belongs to the cluster of L14, L15, L16, L17 and L18. The possibility that there may be proteins with amounts of serine and threonine residues, consti- tuting active sites for phosphorylation greater than 1 /polypeptide chain further complicates the identifi- cation of such proteins. Fortunately, identification of the proteins listed here has been unambiguous, both because most of the spots are well resolved, and secondly because in some cases stoichiometric con- version of the protein into the phosphorylated form has been possible.

If it is true that the phosphorylation of ribosomal proteins plays no significant physiological role in E. coli, it is worthwhile to explain the interest in the continued study of phosphorylation in the hetero- logous system employed here. First, the protein kinases may provide tools for investigating protein topography and possible conformational changes in ribosomes. The results reported here show that no proteins which are substrates for protein kinases when they have been extracted from the particle are pro- tected when they are integrated in the intact particles. On the other hand the converse is not true: that certain proteins are substrates only in the intact par- ticle suggests that conformational determinants of kinase specificity are maintained by the organized ribosome structure. The accessibility of the phos- phorylated proteins has been investigated in 70-S ribosomes as opposed to the free 50-S and 30-S sub- units. We have found (results not show) that 50-S

proteins L4, L5, L17 and L23 and 3 0 4 proteins Sl , S4, S13 and S18 are protected against phosphorylation in the 70-S ribosome.

Second, although the structural determinants of the specificity of protein kinases have not yet been extensively investigated, it is likely, given the evidence for a high degree of specificity in kinase action, that comparative studies on the phosphorylated sites in a group of proteins, such as that represented by pro- caryotic and eucaryotic ribosomal proteins, will prove invaluable in further studies on the specificity of protein kinases.

We would like to thank Dr P. Bechtel for his generous gift of the catalytic subunit of rabbit skeletal muscle protein kinase as well as for his many helpful suggestions, James Kenny for the prepa- ration of the ribosomal subunits, and Dr E. G. Krebs for critically reading the manuscript. This work was supported by grants from the American Heart Association (69461) and the Damon Runyon Memorial Fund (DR6-1140).

REFERENCES 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11. 12.

13. 14. 15.

16.

Beavo, J. A,, Bechtel, P. J. & Krebs, E. G. (1974) Methods

Traugh, J. A. & Traut, R. R. (1972) Biochemistry, I f , 2503

Rahmsdorf, H. J., Herrlich, P., Pai, S. H., Schweiger, M. & Wittmann, H. G. (1973) Mol. Gen. Genet. 127, 259-271.

Issinger, 0.-G. & Traut, R. R. (1974) Biochem. Biophyr. Rrs. Commun. 59,829-836.

Sherton. C. & Wool, 1. G. (1974) J . Biol. Chem. 248, 2258- 2267.

Kischa, K., Moeller, W. & Stoffler, G. (1971) Nai. New Bid. 233,62 - 63.

Kaltschmidt, E. & Wittmann, H. G. (1970) Anal. Biochem. 36,

Howard, G. & Traut, R. R. (1974) Methods Enzjwol . 30.

Hardy, S. J . S.. Kurland, C. C., Voynow, P. & Mora, G. (1969)

Traugh. J. A. & Traut. R. R. (1974) J. Biol. Chem. 249, 1207

Glynn, J. H. & Chappell. J. B. (1964) Biochem. J . 90, 147 - 149. Casjens, S., Hohn, T. & Kaiser, A. D. (1970) Virology, 42.

Maizel. J. V. (1971) Methods Virol. 5. 179-196. Studier, F. W. (1973) J. Mol. Biol. 79, 237-248. Traut, R. R., Delius, H., Ahmad-Zadeh, C., Bickle, T. A,,

Pearson. P. & Tissieres, A. (1969) Cold Spring Harbor Sjmp. Quant. Biol. 34, 25 - 38.

Wittmann, H. G., Stoffler, C . , Hindennach, I., Kurland, C. G.. Randall-Hazelbauer. L., Birge, E. A,, Nomura, M., Kalt- Schmidt, E., Hizushima, S. & Traut, R. R. (1971) Mol. Gen. Genet. I l l , 327-332.

Enzymol. 36,299 - 307.

- 2508.

401 -412.

526 - 531.

Biochemistry, 8, 2897 - 2905.

- 1212.

496- 507.

0.-G. Issinger. M. C. Kiefer, and R. R. Traut, Department of Biological Chemistry. University of California, School of Medicine, Davis, California, U.S.A. 95616