Reprint from Recent Progress in Microbiology, Symposia held at VII Intern. Congr. for Microbiology, 1958 PROTEIN AND NUCLEIC ACID SYNTHESIS IN SUBCELLULAR FRACTIONS OF BACTERIAL CELLS! S. SPIEGELMAN Department of Bacteriology, University of Illinois, Urbana, Illinois 1, INTRODUCTION It is now widely accepted that significant advances in our understanding of protein synthesis requires the development of in vitro systems permitting a direct experimental analysis. As a consequence, recent years have witnessed an increasingly intensive search for methods designed to isolate cell com- ponents capable of carrying out reactions relevant to the fabrication of the biologically specific macromolecules. The almost incredible optimism which characterized the initiation of such projects has been justified with breath- taking speed. In illustration, one need merely mention the pioneering experiments of Gale (1955), the more recent accumulation of data on amino acid activation (Hoagland, 1955; DeMoss and Novelli, 1955; Novelli, 1958; Lipmann, 1958) and their transfer via the soluble RNA fraction? (Hoagland et al., 1957; Ogata and Nokara, 1957; Schweet et al., 1958; Berg and Offengand, 1958; Lipman, 1958). To these must be added the astonishing advances in our understanding of nucleic acid polymer synthesis we owe to the efforts of Ochoa and his group on RNA (Grunberg-Manago et al., 1956; Ochoa and Heppel, 1956) and of the Kornberg (1956) school on DNA. The last two investigations, in particular, pose an interesting question of methodology for all investigators in these and contiguous fields. In both instances, the synthesis of the relevant macromolecules were obtained with highly purified enzyme preparations, an achievement one would have thought on a priori grounds would be reached in the distant future. However, this having been accomplished, does it make any sense to continue the study of cruder prepa- rations? In answering this question, one must immediately grant the obvious 1 The original investigations reported here were made possible by grants from the US. Public Health Service, Office of Naval Research and the National Science Founda- 1D The following abbreviations are used: RNA, ribonucleic acid; DNA, deoxyribonucleic acid, RNAase and DNAase for the corresponding nucleolytic enzymes; PRN for poly- ribonucleotide composed of ribonucleotide subunits with no implications of its relation to normal “RNA”; ATP, GTP, CTP, UTP, for adenosine, guanosine, cytidine and uridine triphosphates respectively. The corresponding diphosphates will be denoted by ADP, GDP, etc. Tris, tris (hydroxymethyl) amino methane. 6-—588815 Microbiology Symposia 82 Symposium II - 8. SPIEGELMAN advantages which stem from the use of a system purified to a level of simpli- city which permits unequivocal interpretation of data derived from its use. There is, however, the danger that in efforts at purification the system may be simplified to the point where it will no longer possess properties of im- portance for the solution of other central biological issues. Of particular pertinence here is the problem of the interrelation amongst the biosynthetic mechanisms of RNA, DNA and protein. It is true that in these days of unlimited optimism, one can justifiably hope that such interrelations will be revealed by reconstructive additions of the purified systems as we know them today. This does presuppose that they contain all the necessary com- ponents, an assumption for which no guarantee can at present be offered. In its absence it would seem desirable to continue some investigations with preparations which have not been carried to the state of purity which charac- terizes the goal of the enzyme chemist. Jt was our feeling that the latter approach was being so vigorously and effectively prosecuted in other labora- tories that ours could add little to these efforts. It seemed likely that a useful function could still be served by continuing the study of systems which were sufficiently crude to retain the inherent potentialities of exhibiting interrela- tions amongst the three macromolecules of major interest. To these more philosophical justifications must be added the more practical one that no purified enzyme system has thus far been isolated which can synthesize protein molecules. Before embarking on a brief description of our most recent efforts it may be of interest to summarize briefly the evolution of the in vitro system which at present engages our major attention. Our primary interest from the outset was the development of a cell-free system capable of carrying out whatever sequence of reactions might be necessary for the ultimate production of a recognizable protein molecule. Our search for such a preparation began with the bacterial protoplast system described by Weibull (1953). The osmotic fragility of the protoplast promised to provide a uniquely suitable departure point for the derivation of subcellular fractions by procedures far gentler than any required to disrupt intact bacterial cells. All the initial exploratory experiments were performed with protoplasts of B. megaterium. Subsequently, methods of achieving similar preparations with E. coli were devised (Lederberg, 1956; Repaske, 1957; Zinder and Arndt, 1956) and our attention was turned to this organism in view of the wealth of genetically marked strains available in this form. All the experiments to be detailed in the present report were performed with material obtained from protoplasts of E. coli. The first successes in observing synthetic activity were obtained with total lysates of B. megaterium. These preparations could fabricate enzymati- cally active protein and both types of nucleic acid. Extensive investigations Protein and nucleic acid synthesis in subcellular fractions 83 were made to determine the optimal conditions and supplementation re- quired by the lysates to exhibit maximal activities. These results have been previously summarized (Spiegelman, 1956). The information gained was extremely useful for later studies since many of the properties and require- ments exhibited by the lysates of B. megaterium were also possessed by the similar preparations prepared from E. coli. Once reproducibility was achieved, it became clear that the crude lysates were too complex and heterogeneous to permit the performance of ex- periments readily interpretable in terms of either known cellular compo- nents or defined enzymatic reactions. As prepared, lysates were mixtures of soluble enzyme systems, ribonucleoprotein particles and membrane fragments of various sizes. Further progress demanded the evolvement of procedures which would separate these different components and so permit an identification of which ones were necessary and responsible for the synthetic functions being studied. After a large number of exploratory ex- periments which tested a variety of conditions, a reproducible procedure was ultimately developed which permitted the separation of defined fractions possessing synthetic activities of interest. It is the purpose of the present paper to describe the preparation and properties of these fractions and sum- marize the information obtained in the course of these studies with particular emphasis on those aspects which are relevant to the problem of the relation between nucleic acid and protein synthesis. 2. THE FRACTIONATION OF OSMOTIC LYSATES As a necessary prelude to a discussion of the chemical and synthetic properties of the fractions studied we begin with a brief description of the methods employed in their isolation. In what follows attention is confined to E. coli protoplasts prepared by the Lederberg penicillin procedure (1956). After harvesting by centrifugation, the protoplasts are washed with 10% sucrose buffered with 0.05 M Tris at pH 7.4 and supplemented with 1 x 10-8 M MsgCl,. They are then resuspended in 9 % glycerol containing Tris and MgCl, at the levels just noted, frozen and stored at — 25°C. The density of protoplasts in the freezing mixture is 40 times that which obtains at the time of harvesting. If kept in the frozen state they retain their activity without detectable loss for at least 3 months. When needed, tubes containing the desired amount of protoplasts are removed, thawed and the protoplasts recovered by centrifugation. To attain uniformity, differences which may exist from one batch of protoplasts to another are randomized by choosing tubes from different dates to provide the material for a given experiment. The use of this procedure has yielded results with satisfactory reproducibility. 84 Symposium II : S. SPrEGELMAN PROTOPLASTS lysed and fet stand at O°C. for 10-15 min. and spun at (50006 for 15min. SUPERNATE [I5G15S] PELLET Ground in Spun at 100,0006 pre-chilled mortar for 120 min. with oluming end spun ot 50006 for 5 min. SUPERNATE PELLET SUPERNATE PELLET [ioosi205] [ioos 120 P] Spun at 50006 for 5 min. 7 PELLET SUPERNATE Spun (50006 for 15 min. 7 PELLET SUPERNATE [Frogments] [isc 1sP]} Fic. 1. Flow sheet of centrifugal fractionation of total lysates. Exposure of protoplasts to a medium lacking an osmotic stabilizer (e.g., sucrose) leads to extensive lysis. The resulting lysates provide the starting material for the preparation of the various fractions of interest. The proce- dure followed in the centrifugal fractionation is depicted as a flow diagram in Fig. 1. Lysis is accomplished by exposing the pellet from 5 ml of proto- plasts to 5 ml of 0.05 M Tris buffer (pH 7.4) containing 1 x 10-* M MgCl, and 5 x 10-3 M MnCl,. This is followed by a centrifugal separation at 15,000 g into a low-speed pellet and a supernate fraction (15G15S). The Mg concentration of the supernatant is raised to | x 10-? M and it is then further separated by high-speed centrifugation into a 100G120S fraction, containing the bulk of the soluble proteins, and a pellet fraction (100G120P). The low-speed pellet corresponds essentially to the fraction designated as the ‘‘shockate” in the earlier investigations (Spiegelman, 1956). It con- tains membrane fragments of various sizes. Grinding this pellet with a small amount of alumina (ca. 1/2 wet weight of the pellet) converts it into a prepa- TABLE 1. Composition of centrifugal fractions. Fractions obtained as in Fig. 1. Nucleic acid analyzed on hot acid hydrolysate by di- chromatic readings on the UV, the orcinol reaction (Dische and Schwartz, 1937) and Burton’s (1956) modification of the Dische reaction. Over 80% of the DNA is lost as acid-precipitable polymer during lysis. 10% of the nucleic acid in the 15G15P fraction is DNA. Protein was analyzed by the Lowry et al. (1951) modification of the Folin reaction. % of total Fraction NA Protein 15G15P 6 8 100G120P 80 36 100G120S 14 56 Protein and nucleic acid synthesis in subce lular fractions 85 ration from which relatively pure and uniform membrane material can be readily obtained in fair yield. The grinding step serves two functions. It ensures a virtually complete disruption of any comparatively intact material and in addition fragments the membranes to sizes which do not pellet at low speeds. After the grinding, the material is resuspended in lysing medium pe Fic. 2. Ultracentrifuge pattern of 15G15S fraction. Centrifugation at 59,740 rpm in a Spino Analytical Centrifuge. Under conditions described in text only one major particu- late component is observed. and subjected to two successive clearing spins at 5000 g which remove the alumina and any residue of large fragments. The remaining supernatant is centrifuged at 15,000 g for 15 minutes to yield the membrane I5G15P fraction. The latter is routinely subjected to at least one wash with lysing medium prior to use or analysis. Table 1 summarizes the distribution of protein and ribonucleic acid in the three fractions. It will be noted that the bulk of the RNA is found in the high-speed pellet although some appears in both the high-speed supernate and membrane fractions. In the case of protein the picture is somewhat reversed in that the bulk of the protein is found in the high-speed supernate. It may be useful to append here a few details concerning the reproduci- bility of the chemical and morphological characteristics of these fractions. It must be emphasized that the distribution of components summarized in table 1 is characteristic of the fractions only if they are isolated from the material and by the procedures described. Introducing any of the methods usually employed in disrupting intact cells (e.g., sonication or grinding with 2-3 times the wet weight of alumina) has in our hands led to prepara- tion possessing markedly different chemical and biosynthetic properties. Further, unless the Mg level is maintained at a high level (0.01 M) during the isolation the RNA content of the 100G120S can rise to between 25 and 35 per cent of the total. Such modifications in the distribution of the chemical constituents are accompanied by changes in the structure of the 100G120P fraction. 86 Symposium II - S. SPrEGELMAN Fic. 3. Electron photomicrograph of membrane fraction. Electron microphotographs of the 15G15S fractions prepared in the presence of 0.01 M MgCl, show spherical bodies of predominantly one size. The uniformity of these preparations is illustrated by the ultracentrifuge pattern shown in Fig. 2. Aside from the soluble proteins there is a single dominant peak of material corresponding to an 70S component. The 70S particles can be reversibly broken down into 30S and 50S components by dialysis against 0.1 M phosphate buffer at pH 7.4. Dialysis of the resulting mixture against 0.01 M MgCl, buffered with 0.05 M Tris at pH 7.4 results in what appears to be a complete reassembly as evidenced by the disappea- rance of the 30S and 50S peaks and the reappearance of the 70S component (Spiegelman, 1958). Comparative studies carried out in our laboratory with protoplasts as the starting material suggests that much of the confusion existent in the literature on the particulate composition of E. coli cells may be ascribed to the variety of ionic mixtures and methods of cell rupture used by different authors in the preparation of extracts for centrifugal analysis. Fig. 3 is an electron microphotograph of the 15G15P fraction and illustra- tes the heterogeneous mixture of fragment sizes of which it is comprised. Protein and nucleic acid synthesis in subcellular fractions 87 TaBLe 2. Molar ratios of RNA in fractions. Fractions obtained as in Fig. 1. Analysis of bases by electrophoretic separation of nucleo- tides (Davidson and Smellie, 1952). Fraction Cytidylate Adenylate Guanylate Uridylate 100G120P 21 21 38 20 100G120S 20 20 30 30 15G15P 18 23 34 25 Total 21 21 37 21 None of them possess the electron density of a protoplast. When prepared under the standard conditions described, the chemical composition of this fraction is uniform and reproducible. It may be noted that, if fresh, rather than glycerol-frozen protoplasts are used, consistently higher RNA to pro- tein ratios are obtained (Ben-Porat, 1958). The method employed in their separation and constant composition gave reason to hope that the three fractions described were distinct entities which had some relevance to the structure of the cells from which they were derived. Certain distinguishing features served to strengthen this supposition. Thus, the particle fraction of both E. coli (Spiegelman, 1958) and B. megaterium (Aronson and Spiegelman, 1958a, 19585) are rich in basic proteins characterized by solubility in dilute mineral acid. Further, compared to the other fractions, the particle proteins are extremely poor in sulphur containing amino acids, corresponding to less than 5 % of that characteristic of the bulk of the proteins of the cell. According to the analysis of Roberts, Britten and Boton (1958) no cysteine is detectable in particle proteins. Table 2 shows that the base composition of the RNA also serves to dis- tinguish the three fractions (Seaman and Spiegelman, 1958). Both the particle and membrane fractions are comparatively rich in guanylate. However, the uridylate ratio is higher in the 15G15P fraction. The soluble 100G120S fraction possesses the lowest molar ratio of guanylate and the highest rela- tive uridylate content. It is the only component containing RNA in which the sum of the purines is equal to that of the pyrimidines. These differen- tiating features of base composition disappear when the fractions are iso- lated by procedures involving sonication or low Mg levels in the environment. 3. THE BEHAVIOUR OF THE FRACTIONS IN SITU Before considering the capacity of the three fractions described to synthe- size protein and nucleic acid in vitro it is of some interest to examine their behaviour in the intact cell. 88 Symposium II - S. SPIEGELMAN 800 600 KE °o a a z 400- /- ~ = / y 7 4/ ——/5GI5P 200 if ---- /006 1208 if —-— /006120P 0 I 1 t ; J 0 | 2 3 MINUTES Fic. 4. Incorporation7of C*-leucine in intact protoplasts. Fractionation according! to diagram of Fig. 1; membranes (15G15P), soluble pro- teins (100G120S), ribonucleoprotein particles (100G120P). Fig. 4 shows the results of a representative short-term labeling experiment with C-leucine. Intact protoplasts were exposed to labeled amino acid for the periods indicated, synthesis being stopped by chilling and the use of NaN3. The samples were then centrifuged and the protoplasts fractionated as described in Fig. 1. It is evident that the 15G15P fraction is the one most rapidly labeled, followed by the soluble proteins and finally by the parti- culate components. An analogous experiment (Ben-Porat, 1958) designed to obtain the same sort of information with respect to RNA metabolism is summarized in or T I i" 2 ISG ISP OQ 5P Lc | x a qt qe “ z a 7 & ? s 1006 12057 ~N 7 = yp Fic. 5. Incorporation of C14- oO 2f- b __4 1006 120P | uridine in intact protoplasts. _ Fractionation according to / = diagram of Fig. 1 to yield ly a —| membranes (15G15P), solu- / —_ ble proteins (100G120S), a ribonucleoprotein particles Tag ! | (100H120P). Numbers rep- Oo 3 10 20 resent thousands of counts MINUTES per minute. Protein and nucleic acid synthesis in subcellular fractions 89 C2 Uridine A 10061208 Fic. 6. Pulse experiment with C'-uridine in intact protoplasts. Fractionation according to diagram of Fig. 1 to yield membranes (15G 15P), soluble proteins (100G 120S) and ribonucleoprotein particles (100G120P). At one 2 minute C-uridine diluted ! with Curidine. Numbers Lg \ represent hundreds of counts 12 5 15 25 per minute. MINUTES i Em 4 1006 120P ~ oa oe 0 56 15 P CPM/MG RNA X 102 Fig. 5. Here C™-uridine was used and a similar fractionation carried out. Again, we see that it is the membrane fraction which is the most rapid in macromolecular synthesis. The results obtained with C14-uridine are in qualitative agreement with the experiments of Volkin and Astrachan (1956a, 6) on the distribution of P® in phage-infected bacteria. This assumes that we can equate our 15G15P with their P, fraction. Other experiments (Ben-Porat, 1958) have analyzed the flow of labeled material in pulse experiments and one of these is summarized in Fig. 6. Protoplasts were exposed to C14-uridine for one minute, at the end of which a sample was removed for fractionation and specific activity determination. To the remainder an amount of C!*-uridine was added sufficient to achieve a 100-fold dilution of the label. Samples were then removed at the time periods indicated to determine the subsequent flow of the material labeled in the first minute. It will be noted that in the period of labeling the 15G15P fraction achieved the highest specific activity. Following dilution of the label, the activity of the membranes falls while the other two fractions are rising. Such results are consistent with the interpretation that the membrane fraction is concerned with the initial act of fabricating RNA macromole- cules which ultimately are transferred to the soluble and particulate fraction. 4. AMINO ACID INCORPORATION INTO FRACTIONS ISOLATED FROM OSMOTIC LYSATES From the experiments described in the preceding paragraphs, as well as others not detailed, one is led to conclude that the principal site of protein and RNA synthesis is physically associated with the protoplast membranes. 90 Symposium II - S. SPIEGELMAN On the basis of such evidence, one would be led to predict that attempts to isolate subcellular fractions possessing the ability to fabricate complete protein molecules are most likely to succeed if attention is focused on the membrane fraction. A comparative study was made of the abilities of these three fractions to carry out the various steps which are presumed to constitute recognizable stages in protein synthesis. Particular attention was paid to the features mentioned in the introductory paragraphs. These included the ATP depen- dent carboxyl activation mediated by the amino acid activating enzymes, the presence of the soluble RNA acceptor and, finally, the formation of peptide bonds. To distinguish between the last two, advantage was taken of the fact that amino acids fixed to the soluble RNA fraction are stable to cold acid, but are liberated by exposure to hot acid (cf. Berg and Offengand, 1958). Finally, an examination was always made of the response of the incorporation of any given amino acid to supplementation with a complete mixture of the others. It was felt that a positive response represented a good diagnostic indication that the reaction being studied was more likely to represent the formation of a complete or nearly complete protein molecule. (a) The high-speed supernate (100G120S). The high-speed supernatant fraction contains the major portion of the amino acid activating enzymes as measured by the pyrophosphate exchange reaction (DeMoss and Novelli, 1955). As may be seen in Fig. 7 it can carry out a reaction resulting in the fixation of amino acids into a linkage which is stable to cold, but not to hot acid. It is further noteworthy that the amino acids are incorporated into a linkage which is highly labile under the condi- tions of the incubation. Direct evidence has been obtained that the amino acids fixed in the 100G120S fraction are extremely sensitive to RNAase by the following sort of experiment. The reaction was run for 5 minutes and then stopped by the addition of 2.5 volumes of cold alcohol and 10-? M MgCl,. After precipitation was complete, the precipitate was recovered and washed several times. The pellet was then redissolved in Tris buffer at 7.5 and one aliquot was exposed to 100 ug RNAase per ml for 5 minutes. Treated and control preparations were then precipitated with cold 10% TCA, the pellets washed with cold acid and then dissolved and counted. The RNAase-treated material contained no detectable counts as compared to a 75 % recovery of the counts in the untreated control. Maximal amino acid fixation in this system requires the presence of ATP and 5’-ribotides. The reaction is not inhibited by even elevated levels (500 ug/ml) of chloramphenicol. The fixation of a given amino acid is not aug- mented by the presence of a mixture of others. Finally, it may be noted that no combination of supplements in the form of various intermediates Protein and nucleic acid synthesis in subcellular fractions 91 600 A cl4-aa, 1006 1208 Loony | . fy ° r | ew f00;- | 1 a ! \ a ! \ E r th ~N i \ = t \ a ! \ © 200}-; \ ! ! \ CA Ss. Fic. 7. Incorporation of C1*-leucine r# @~---Le------- 8. -< into isolated soluble protein fraction / HAS (100G120S). Counts stable to cold 0 + “s t acid (CAS); counts stable to hot 0 30 60 90 120 acid (HAS). MINUTES served to confer on this fraction the ability to insert amino acids into linkages stable to hot acid. Insofar as amino acid incorporation is concerned, the 100G120S fraction possesses features which previous authors (Hoagland ef al., 1957; Lipman, 1958; Berg and Offengand, 1958) have reported as characterizing the “soluble RNA” acceptor system. (b) The high-speed pellet fraction (100G120P). The particulate fraction, prepared as described, contains amino acid activating enzymes. It can incorporate labeled amino acids into linkages stable to both hot and cold acids. Comparison of the relevant curves in Figures 6 and 7 reveals that fixation in the linkage stable only to cold acid (CAS) precedes in time the appearance of label in hot acid stable (HAS) form. Thus, in 15 minutes, incorporation of CAS counts is complete. At that time only half the counts are retained after hot acid extraction. During the next 15 minutes all the counts precipitable with cold acid are converted into a form which is stable to hot acid. Optimal incorporation requires the presence of ATP and 5’-ribotides. The incorporation of any given amino acid is uninfluenced by supplementa- tion with a complete mixture of the others. Neither is it affected by the presence of chloramphenicol. Again, the activity observed and its properties do not encourage the belief that one is here studying a system capable of carrying all the stages of protein synthesis. (c) The low-speed supernatant (15G15S) This fraction is the one which yields the 100G120P and 100G120S frac- tions by high-speed centrifugation. It is, therefore, a combination of the 92 Symposium II + S. SPIEGELMAN 500 4 CAS 0 1006 120P 400 — 4 1006120$ © I5SGISS 300 |- . 100 — a SPECIFIC ACTIVITY CPM/MG PROTEIN o Fic. 8. Incorporation of C4- 0 | | | | | leucine into cold acid stable 0 10 20 30 40 50 60 (CAS) linkage in isolated TIME IN MINUTES fractions, two fractions we have discussed in detail in the two preceding sections. Its amino acid incorporating abilities are, not surprisingly, a composite of the high-speed pellet and supernate fractions. The CAS fixing abilities of the three fractions all derived from the same preparation are given in Fig. 8. The corresponding kinetics of incorporation into hot acid stable linkages are shown in Fig. 9. Quantitatively the 15GI5S fraction is somewhat more active than one would expect from adding the observed activities of its components functioning in isolation. Nevertheless, like them, the incorpora- tion of a particular amino acid shows no response to supplementation with a complete mixture of the others. Neither does it indicate by any other property that it is capable of manufacturing complete protein molecules. Considering only the data derived from examination of hot acid stable counts, the extent of the activity observed is discouragingly small in all of the fractions described thus far. In no case did the “synthesis” correspond > 9 1006 120P z ae 4 1006 120S 2b E 2 s00L ° 15G15S ta o o uw = 200b _ uz ° ao _ 100 + “ ° Za ' I r Fic. 9. Incorporation of C4- 0 leucine into hot acid stable 0 10 «820 30 40 50 69 (HAS) linkage in isolated TIME IN MINUTES fractions. Protein and nucleic acid synthesis in subcellular fractions 93 4000 ISG ISP cl4 AA 3000 |- E ° a a a L— z 2000 ~ = a oO 1000 -— 9 * CAS Fic. 10. Incorporation of C1*-leucine ° ° HAS into isolated membrane fraction. Open circles represent counts stable 0 ° L | to hot acid, and closed circles counts 0 60 120 180 stable to cold acid. MINUTES to more than 0.1% of the protein input and in many instances it was con- siderably less. If any of these fractions contain active protein synthesizing mechanisms, our procedures have failed to reveal them. In view of the success attained with the next fraction to be described, one is tempted to conclude, at least tentatively, that the comparative inactivity of the other fractions reflects a deficiency in the preparations rather than in the methods employed in their isolation and the supplementation used to test them. (d) The membrane fragment preparation (ISGI5P) The membrane fragments were the only fraction found to possess signifi- cant capacity to synthesize protein as measured by either incorporation or induced enzyme synthesis. This is a finding we previously suggested might have been expected from the experiments which compared the activi- ties of the three fractions in the intact protoplast. Like all the other fractions examined, preparations of membrane frag- ments contain enzymes capable of carrying out a pyrophosphate exchange reaction with ATP in the presence of amino acids. These activities are re- tained even after several washings. It must, however, be noted that the bulk of these activities are found in the 100G120S fraction. The amino acid incorporating activity of the membrane fraction exceeds by a factor of 100 that observed with the other fractions when examined under comparable conditions. Fig. 10 shows the kinetics of incorporation with a membrane preparation and compares insertion into linkages stable to hot and cold acid. Unlike the high-speed pellet and supernatant fraction there is here no observable incorporation of CAS counts which are not also stable to hot acid. It will be noted further that the kinetics of the incorporation is very diffe- 94 Symposium II - S. SpmcELMAN rent from that seen with the other fractions. There is a slight lag at the onset followed by a linear rate of synthesis which continues for more than 3 hours in the vast majority of preparations examined. The specific activity attained at the 3-hour point corresponds to 15% synthesis based on the protein input. A variety of combinations of various nucleic acid intermediates were tested for their effect on this incorporation in an attempt to achieve an optimal mixture. The experiment described in Fig. 10 was carried out under optimal conditions of supplementation detailed in table 3 which compares the effects of removing various components on incorporation of C!-leucine. The omission of either ATP or the 5’-nucleotides results in drastic loss of the incorporation observed over a 2.5-hour period. Mn appears to be a manda- tory requirement, an observation made in our earlier experiments with crude lysates of B, megaterium (Spiegelman, 1956) and E. coli. It will be further noted that the presence of other amino acids is necessary if any significant incorporation of leucine is to be observed. The same situation was obtained with other labeled amino acids tested. This dependence of incorporating activity on the presence of a virtually complete mixture of amino acids serves to distinguish the incorporation observed in this fraction from the others described. Another unique property noted in table 3 is its sensitivity to chloramphenicol. None of the other fractions are inhibited by this agent. One other finding, relating to the lag apparent in Fig. 8, may be noted. An attempt to identify the cause of this delay led to the discovery that it could be eliminated by including the other three riboside triphosphates. The corresponding diphosphates had no detectable effect on the extent or kinetics of amino acid incorporation. TABLE 3. C1*-Amino acid incorporation into membrane fraction 15GI5P. Complete incubation mixture contains per ml: 20 4M KCl; 50 LM Tris; 50 uM maleate; pH 6.5; 5 uM Mn; 1 uM Mg; four 5’-ribotides, 0.1 4M each; four 5’-deoxyribotides, 0.1 uM each. Amino acids mixture in ratio corresponding to E. coli protein, 2 mg. ATP 2 uM between 100-200 wg of enzyme preparation. Counting done on automatic micromil gas-flow counter, Nuclear-Chicago. Incorporation Medium muM AA/mg prot. % of complete Complete 600 100 — ATP 120 20 — 5’-ribotides 220 37 — 5’-deoxytides 200 33 — (5’-ribotides + 5’-deoxytides) 100 17 — Other amino acids 20 3 + Chloramphenicol (200 g/ml) 40 7 —Mn (5x 10-3 M) 0 0 Protein and nucleic acid synthesis in subcellular fractions 95 5. POLYRIBONUCLEOTIDE SYNTHESIS The supplemental responses of amino acid incorporation yielded per- sistent evidence for the active involvement of nucleic acid metabolism. Particularly interesting was the finding of consistent stimulation which occurred on the addition of the 5’-deoxytides. These observations are reminiscent of the preliminary reports of Beljansky (1954) and Lester (1953) who observed stimulation of incorporation in total lysates on the addition of DNAase. Such results suggested the intriguing possibility that we might here ultimately have the opportunity of studying the interaction between the synthetic mechanisms of the two types of nucleic acid. However remote, such a possibility seemed worth exploring, for, despite the startling advances recorded in recent years, the direct experimental analysis of the biosynthetic interrelations between DNA and RNA has thus far remained an unexplored aspect of the central problem. Indications of active polyribonucleotide metabolism was searched for employing both net increments of acid-precipitable nucleic acid polymer and the incorporation of P®?-labeled 5’-ribonucleotides. Evidence was quickly obtained indicating that both the 15GI5S and 15G15P fractions possessed extensive synthetic activities worthy of further analysis. Many features of these systems are still under investigation and it is difficult at the present time to specify with certainty which are relevant and which are biochemical artefacts reflecting the comparative crudity of the preparations being studied. We summarize here those data which illustrate the possible potentialities of these systems as tools for the further investigation of certain problems of obvious interest. Because of our previous experience, much of our initial efforts were concentrated on the 15G15P fraction. Consequently, we here detail only the experiments performed with it. (a) Stabilization of polyribonucleotide synthesized Preliminary experiments with the membrane fraction made it quickly evident that polyribonucleotide synthesis could be readily studied in terms of net increases of acid-precipitable polymer. There was, however, at the beginning a disturbing amount of inconsistency and poor reproducibility. Kinetic analysis suggested that this was due to a degradation competing with the synthetic reaction leading under certain circumstances to loss of the product. A variety of conditions and agents were tested in an effort to overcome this difficulty. The attempt was either to inhibit the nucleolytic enzymes present or introduce a substance which would combine with the polymer and thus protect it against destruction. Ultimately, it was found that the use of the polyamine, spermine, at a concentration of 5 x 10-3 M, 96 Symposium II - S. SPIEGELMAN TABLE 4, Effect of supplements on PRN synthesis in ISGIS5P. Complete system contains per ml: 5 uM Mn; 1 uM Mg; 20 uM KCl; 50 uM Tris; 50 uM of maleate at pH 6.5; 0.2 uM of each 5’-ribotide and deoxytide; 2 uM ATP; 0.4 uM of each of the other riboside triphosphates; 5 4M spermine; ca. 200 wg protein of enzyme preparation containing 20 ug RNA. Analysis for PRN as in Table 1. A PRN Incubation in wg/mg prot. % Change % of complete Complete 340 + 380 100 -— ATP 0 0 0 —Riboside triphosphates (TOP) 70 + 80 21 — TOP; +riboside diphosphates 120 +130 34 —(ribotides and deoxytides) 10 + 11 3 —Mn 0 0 0 provided the protecting effect which was sought. Fig. 11 shows the difference of synthetic activity observed in the presence and absence of this agent at two levels of substrate input. In the absence of spermine, synthesis of polyri- bonucleotide occurs for 90 minutes followed by loss of the product. If excess substrate is provided, the presence of spermine insures a linear syn- thesis for at least three hours. It should be noted that at the three-hour point the synthesis corresponds to over a 10-fold increase of the input of polyribo- nucleotide. (b) Nutritional and ionic requirements for polyribonucleotide synthesis. With this problem resolved, it became possible to inquire into the condi- tions and supplements required for optimal synthesis of polyribonucleotide. The pH optimum for the synthesis was found to be between 6.0 and 6.5 with a plateau in this range. In view of the effects observed with the ribose triphosphates on amino acid incorporation, they were compared with cor- responding diphosphates in the system synthesizing polyribonucleotide. The results observed were clear and consistent. The ribose-disphosphates were definitely inferior and their use was invariably attended by a lag period of 1 hour before any significant polyribonucleotide synthesis was observed. In contrast, as may be seen from Fig. 10, synthesis in the presence of tri- phosphates begins immediately. Another property which was in agreement with the information gathered during the study of amino acid incorpocation was the finding that the polyribonucleotide formation had a mandatory requirement for Mn. The optimal level of this ion was found to be 1 x 10°? M. These and other characteristics of the system are summarized briefly in table 4. It will be noted that ATP is required and must be present at levels ex- ceeding 0.5 wM per ml. While excellent synthesis can be observed with Protein and nucleic acid synthesis in subcellular fractions 97 r fo; Oxg/ml ping 300F / = / £ Z + Spermine 4 t 5x105M 200+ 7 = / - y L AO Smg