Vol. 124, No. 3 Printed in U.S.A.
JOURNAL OF BACTERIOLOGY, Dec. 1975, p. 1122-1127 Copyright © 1975 American Society for Microbiology
Ribosomal Distribution in a Polyamine Auxotroph of Escherichia coli I.
D. ALGRANATI,* G. ECHANDI, SARA H. GOLDEMBERG, SUSANNA CUNNINGHAM-RUNDLES,' AND W. K. MAAS
Instituto de Investigaciones Bioquimicas "Fundacion Campomar" and Facultad de Ciencias Exactas y Naturales, Obligado 2490, Buenos Aires (28), Argentina,* and Department of Microbiology, School of Medicine, New York University, New York, New York 10016 Received for publication 29 May 1975
The distribution of ribosomal particles has been studied in a polyamine-deficient mutant of Escherichia coli by sucrose gradient centrifugation analysis. Lysates from starved cells contained less 70S monomers and 30S subunits but more 50S particles than those prepared from bacteria supplemented with putrescine. The addition of the polyamine to putrescine-depleted cells induced a rapid change of the ribosomal profile. A similar effect could be obtained in vitro by equilibrium dialysis against a polyamine-containing solution. The ribosomal pattern obtained from starved bacteria was specific for polyamine deficiency. We conclude that the changes in ribosomal profiles upon restoration of putrescine levels in previously starved cells denote a shift of the equilibrium between 30S-50S couples and ribosomal subunits.
The distribution of the ribosomal pool between polysomes, 70S couples, and subunits is related to the physiological state of bacteria, as well as to the intracellular ionic levels and the presence of initiation (24), dissociation (5, 28) and, perhaps, also association factors (11). The association factor, obtained from extracts of Bacillus stearothermophilus, has been fractionated into two components, one of which behaves like a polyamine (11-13). This fact and the recent isolation of Escherichia coli mutants blocked in the biosynthesis of putrescine (17, 18, 21) (which is a precursor of spermidine) prompted us to investigate the ribosomal distribution in extracts prepared from a mutant strain grown in the absence and presence of exogenous polyamines. Our aim was to elucidate whether putrescine starvation would modify the equilibrium between ribosomal subunits and 70S couples, which in normal cells is strongly shifted towards the latter. The present paper deals with changes in ribosomal profiles after the addition of putrescine to a culture of an E. coli polyamine auxotroph in an attempt to elucidate the physiological role of polyamines and their possible relationship with the association factor of ribosomal subunits. Although multiple functions have been suggested for polyamines in relation to ribonucleic I Present address: Division of Immunology, Memorial Sloan-Kettering Cancer Center, New York, N.Y. 10021.
acid (RNA), deoxyribonucleic acid, and protein synthesis (4, 8, 27), membrane stability (4, 8), aggregation of ribosomal subunits (9, 23, 25), and osmotic adaptation (22), the biological role of these widespread organic cations is still not well understood. MATERIALS AND METHODS Bacterial strain and growth media. E. coli MA 261 Thr-, Leu-, Ser-, Thi- was used in all of the experiments described in the present work. This strain is, in addition, a double mutant deficient in two enzymes: agmatine ureohydrolase (AUH-) and constitutive ornithine decarboxylase (OD,-). Therefore, E. coli MA 261 requires an exogenous polyamine supply for growth (9a). The media used were those described by Srinivasan et al. (26) and designated as MMO and MMA. They contained salts supplemented with 0.5% glucose, thiamine, biotin, leucine, threonine, methionine, serine, glycine, and either ornithine or arginine. The putrescine (100 jg/ml)-supplemented media are designated MMOP and MMAP, respectively. Chemicals. Putrescine dihydrochloride and sucrose, density gradient grade (ribonuclease free), were obtained from Schwarz/Mann, Orangeburg, N. Y.; spermidine trihydrochloride was from New England Nuclear Corp., Boston, Mass. All other reagents were of the highest purity commercially available. Starvation for putrescine. For putrescine starvation the method reported by Young and Srinivasan (31) was slightly modified as follows. A culture inoculated from a slant was grown in MMO medium; after refrigeration for several hours, it was diluted 50-fold in MMA medium and allowed to grow again.
1122
VOL. 124, 1975
RIBOSOMAL PROFILES AND POLYAMINES IN E. COLI
The new culture was used in the ratio 1:10 to inoculate a MMO medium. After growing to a density of 2 x 108 to 3 x 108 cells per ml, it was diluted 1:3 with fresh MMO medium. At this point, the intracellular level of putrescine was depleted in such a way that the bacteria required exogenous polyamines for normal growth. All cultures utilized in this method were grown overnight at 37 C with aeration. The starved cells were cultivated in the absence or presence of putrescine and then used for the different experiments. Bacterial growth was measured spectrophotometrically at 550 nm. Preparation of bacterial lysates and analysis of ribosomal distribution. Cells were collected from exponentially growing cultures after slow cooling to allow completion of protein synthesis and accumulation of run-off ribosomes. Lysis of bacteria was carried out essentially as described previously (2). The supernatant fluids obtained after centrifugation of lysates at 7,000 x g for 10 min were layered on top of 15 to 40% (wt/vol) linear sucrose density gradients containing 10 mM tris(hydroxymethyl)aminomethanehydrochloride buffer (pH 7.8), 5 mM magnesium acetate, and 50 mM KCl and were centrifuged for 90 min at 45,000 rpm in a Spinco SW65 rotor. Where indicated, 60 mM NaCl was used instead of KCl in the sucrose solutions. Gradients were analyzed by monitoring at 254 nm with an ISCO ultraviolet analyzer. RNA concentrations were determined by the orcinol method (7). Dissociation of ribosomes. To study the state of association of ribosomes under different ionic conditions, we incubated them for 15 min at 37 C in the presence of different Mg2+ concentrations. They were then analyzed by sucrose gradient centrifugation and the percentages of dissociation were calculated as described previously (5). Binding of polyamines to ribosomes. The binding of ["4Clspermidine to ribosomes was determined by equilibrium dialysis. Samples (0.4 ml) containing 4 to 5 absorbancy units at 260 nm (A2**) of ribosomes in a solution of 10 mM tris(hydroxymethyl)aminomethane-hydrochloride buffer (pH 7.8), 5 mM magnesium acetate, 50 mM KCl, and 2 mM 2-mercaptoethanol were dialyzed at 4 C for 24 h against 30 ml of the same buffer made with 1 mM ["4C]spermidine (0.5 1sCi). In all of these experiments Visking dialysis tubing was carefully prewashed (19). For the assay of radioactivity, triplicate 50-ul samples were taken from solutions inside and outside of the dialysis bag and counted in a liquid scintillation spectrometer after addition of 0.2 ml of water and 3 ml of Bray mixture. The amount of polyamine bound was calculated from the difference between the radioactivity of solutions inside and outside of the dialysis bag. One A2* unit of 70S ribosomes was taken as 23 pmol (16).
1123
the intracellular level of this polyamine decreased more than 100-fold, whereas spermidine was reduced to about one-half of its normal concentration (17, 21, 26, 31). Under these conditions the mutants showed a requirement for exogenous polyamine to attain normal growth rate. The double mutant E. coli MA 261 behaved in a similar way. Addition of putrescine to MMO medium containing previously starved bacteria enhanced the growth rate considerably after a lag period of about 60 min. The generation time in several experiments varied from 180 to 300 min in the absence of putrescine and from 60 to 90 min in the presence of the polyamine (MMOP medium). Effect of putrescine addition on the ribosomal distribution. It is well known that polyamines induce the association of ribosomal subunits in vitro (9, 23, 25) and decrease the optimal Mg2+ concentration required for protein synthesis in cell-free systems (3, 30). These facts and the finding that the association factor contains polyamines (13, 14) led us to investigate whether or not the addition of putrescine to a culture previously starved could modify the ratio between 70S couples and subunits. For this purpose, putrescine-depleted bacteria were grown in MMO medium. After several hours putrescine was added to a portion of the culture in a separate flask and incubated along with the original culture for 30 or 60 min. The bacterial lysates were analyzed by sucrose gradient centrifugation. The results (Fig. 1A) show that in the absence of putrescine (MMO medium) there were fewer 70S ribosomes and the amount of 30S seemed to be small in comparison with the 50S peak. Perhaps the biosynthesis or assembly of the small subunit was somehow damaged during polyamine starvation. The addition of putrescine produced a marked change of ribosomal pattern (Fig. 1B). The lower total ribosomal content observed in the profiles obtained from starved cells is compensated for by a higher absorbancy in the supernatant fluid (not shown in figures). To have a quantitative comparison, the areas under the 70S and 50S peaks were measured. The ratio 70S/50S was 1.54 in Fig. 1A (MMO medium) and 2.89 in Fig. 1B (MMOP medium). Similar results were obtained when the bacteria were collected after fast cooling; in this case, polyribosomes were also seen and were more RESULTS abundant in the culture with putrescine. The different ribosomal profiles obtained Putrescine requirement of starved bacteria. Previous investigations carried out from putrescine-starved or -supplemented bacwith several conditional polyamine auxotrophs teria could indeed reveal a shift of the equilibhave shown that upon starvation of putrescine rium between 30S-50S couples and ribosomal
1124
J. BACTERIOL.
ALGRANATI ET AL.
Ec
AN
Iq
4
3
2
1
ml
4
3
21
from top
FIG. 1. Effect of polyamine addition on the ribosomal profile. Bacteria were harvested and lysed as described in the text. Cell extracts (about 0.6 A... unit) were analyzed after sucrose gradient centrifugation. (A) Ribosomal pattern corresponding to a lysate obtained from cells cultivated in MMO medium. (B) Profile corresponding to bacteria grown in MMO medium to which putrescine was added 30 min before harvesting of the culture.
subunits occurring in vivo. If this were the case, we should rule out several other possibilities as sources of artifacts, namely: (i) a different efficiency of extraction of monomers and subunits obtained with our method for lysis applied to both kinds of bacteria; (ii) a different content
taining Na+ ions instead of K+ ions (1, 6). When such experiments were performed, only the run-off ribosomes dissociated (Fig. 2). Comparison of these results with the patterns obtained with sucrose gradients containing K+ (Fig. 1) allowed us to calculate that the percentages of run-off ribosomes in the fractions containing monomers were 82 and 87% in the lysates prepared from starved and supplemented cells, respectively. We have purified ribosomal subunits from extracts of bacteria grown in the absence or in the presence of putrescine. The sucrose gradient analysis of both kinds of 50S particles showed that they are at least 95% pure even after a heating for 5 min at 55 C in the presence of 2 mM dithiothreitol. This treatment is able to dissociate 30S dimers, as has been shown in our laboratory (M. Gracia-Patrone, unpublished data). On the other hand, the analysis of ribosomal RNA obtained by phenol extraction of 50S subunits prepared from polyaminestarved and unstarved bacteria demonstrated that similar and small amounts of 16S RNA appeared in both cases along with the 23S RNA component. The presence of 16S RNA could be attributed to a partial breakdown of 23S RNA by endonuclease activity (29; 0. P. Van Diggelen, Ph.D. thesis, Univ. of Leyden, Leyden, A
70S
50s
30s B 705
50S 305
0.4
of run-off and complexed ribosomes in the 70S monomers, depending upon whether lysates were prepared from starved or unstarved bacte0.3 ria; and (iii) a different proportion of 30S dimers present as contaminants of the 50S fractions. All of these possibilities were consid02 ered and studied in detail. Quantitative determinations of RNA showed that our preparations of cell lysates gave similar 0.1 yields (80 to 90%) of total RNA obtained either from starved or unstarved bacteria. Since the ribosomal RNA is a high proportion of total RNA, we can conclude that the differences in 0 321f 4 4 321f the ribosomal patterns shown in Fig. 1 could not n/from top be attributed to selective extraction of subunits FIG. 2. Effect of polyamine addition on the riboor 70S monomers from both kinds of cells. It has been demonstrated that run-off ribo- somal profiles obtained by centrifugation in sucrose gradients containing Na+ ions. Cell lysates were somes, which are released from polysomes at the and analyzed as described in the text and end of translation, and complexed monomers, prepared the Fig. 1. (A) Ribosomal pattern correlegend which are still bound to messenger RNA, be- sponding to tobacteria grown in MMO medium. (B) have in different ways under different condi- Profile corresponding to cells cultivated in MMO tions. These two kinds of monomers can be medium to which putrescine was added 30 min before distinguished by using sucrose gradients con- harvesting.
VOL. 124, 1975
1125
RIBOSOMAL PROFILES AND POLYAMINES IN E. COLI
The Netherlands, 1973) rather than to a contamination of 50S subunits with 30S dimers or modified 70S particles. However, even assuming the possibility of contamination, the 50S particles obtained from starved and unstarved cells should contain about the same amount of 30S dimers and/or modified 70S monomers, and these contaminations would not alter the general pattern of ribosomal profiles. Specificity of the polyamine effect. We have investigated whether the effect on the ribosomal profile of putrescine depletion was specific for polyamine deficiency and whether it could also be obtained after starvation for a carbon or a nitrogen source. For this purpose, bacteria grown in MMOP medium were washed several times with glucose-free medium and then suspended either in complete medium or in glucose-free MMOP medium. After further incubation for 30 min at 37 C, the corresponding lysates were prepared and analyzed as usual. The results (Fig. 3B and C) show that the profile of run-off ribosomes obtained from cultures in the presence of putrescine did not change after starvation for glucose. A parallel culture in MMO medium gave a pattern characteristic of polyamine starvation (Fig. 3A). The same results were obtained when the nitrogen source was omitted. Dissociation of ribosomes. The sensitivity to dissociation of ribosomes obtained from bacteria grown either in MMO or in MMOP medium was studied by decreasing the Mg2+ levels. The results (Fig. 4) demonstrate that, although the ribosomal particles prepared from cultures without putrescine were more highly dissociated than those of bacteria cultivated in MMOP at 8 70S 505 30S
A. 70s 50S 30s
4
4
C. 70s
different Mg2+ concentrations, the dissociation curves were parallel, indicating that both types of 70S particles were equally sensitive to low Mg2+ levels. Binding of spermidine to ribosomes. The binding of[14C]spermidine to ribosomal particles was measured by equilibrium dialysis and was of the same order (920 4 35 molecules per ribosome) for particles obtained from cultures with and without putrescine. After dialysis for 24 h against a buffer containing 1 mM spermidine, there was a partial association of subunits, so that the patterns obtained from bacteria grown in the absence and presence of polyamine were more similar than those obtained before the equilibrium dialysis. Furthermore, both kinds of ribosomes became more stable to dissociation induced by a decrease of Mg2+ concentration (Fig. 5). 70 60
2 50 E- 40 ° 30 Q 20
10
1
0
2
3
6
5
4
7
9
8
10
tg+g concentration (mM) FIG. 4. Stability of ribosomes at different Mg2" concentrations. Extracts obtained from bacteria grown in MMO (0) or in MMOP (0) medium were adjusted to the indicated levels of Mg2+. Samples were then analyzed, and the resulting percentages of dissociation were calculated as outlined in the text.
50S 30s
4
0 rt0
c It in
N Iq
).3
l
A
0 0
O2
,,,
,,, -4 3 2
f
4 ml
3 from
2 1
4
3 2
f 0
top
FIG. 3. Specificity of the effect of polyamine ribosomal distribution. Lysates from bacteria under different conditions were analyzed as indicated in the text. Bacteria were grown in the following media: (A) MMO; (B) MMOP; (C) MMOP, but subsequently cells were washed and reincubated for 30 min in glucose-free MMOP medium. on
1
2
3
4
5
Mg't"' concentrotion
6
7
(mm)
FIG. 5. Stability of ribosomes after dialysis against spermidine. Samples were submitted to equilibrium dialysis against I mM spermidine and then adjusted to the indicated Mg2+ concentrations. Analysis and calculations were done as described in the text. Symbols as in Fig. 4.
1126
J. BACTERIOL.
ALGRANATI ET AL.
DISCUSSION The isolation of mutants unable to synthesize putrescine permitted us to investigate whether or not the ribosomal profile is related to the intracellular polyamine content and eventually to the association factor. Our results indicate that, in some cases, as early as 30 min after the addition of putrescine to previously depleted cells the distribution of ribosomes appeared markedly altered. Before the addition of polyamine, the amounts of 70S ribosomes and 30S subunits were relatively low in comparison with the high 50S peak. Moreover, preliminary results have demonstrated that the broad 30S peak which appeared in the ribosomal profiles of starved-cell lysates indicates the existence of a mixture of abnormal and normal 30S particles. The restoration of putrescine changed the pattern rapidly, provoking a marked increase of 70S couples and a normalization of the relative amounts of both subunits (Fig. 1). The modification of the ribosomal pattern occurred even before an increase in growth rate was detectable. The ribosomal profiles obtained from polyamine-starved and -supplemented bacteria were not due to a different efficiency of extraction of 70S ribosomes and subunits, to a different content of run-off particles in the 70S monomers (Fig. 2), or to a different amount of 30S dimers in the fractions of 50S peaks. A similar modification of the ribosomal profile could be obtained in vitro by dialysis against spermidine. This fact indicates that polyamines at physiological levels have a direct effect on the equilibrium between 70S couples and subunits. Furthermore, the well-known effect of ribosomal stabilization by polyamines (15) was confirmed by measuring the sensitivity towards decreasing Mg2+ concentrations (Fig. 4 and 5). The profile of ribosomes after growth in MMO medium seems to be characteristic of polyamine starvation, since the pattern brought about by MMOP medium did not change after starvation for a glucose or a nitrogen source (Fig. 3). We have recently reported some studies on polypeptide synthesis induced by polyuridylic acid in cell-free systems obtained from starved and unstarved bacteria (10). Under these conditions, it is possible to detect at least one effect of polyamines on protein synthesis that is independent of transcription. Many biological functions have been suggested for polyamines in living cells. Our results indicate that the translation process and the
assembly of its machinery, including the distribution pattern of ribosomes, are good candidates for targets of a primary effect of these organic cations. ACKNOWLEDGMENTS We are grateful to L. F. Leloir and all members of the Instituto de Investigaciones Bioquimicas for helpful discussions. This work was supported by grants from the University of Buenios Aires, the Ministerio de Bienestar Social, the Programa Regional de Desarrollo Cientifico y Tecnol6gico de la O.E.A., and the Consejo Nacional de Investigaciones Cientificas y Tecnicas (Argentina). I. D. A. and S. H. G. are Career Investigators of the latter institution, and G. E. is a Fellow of the Organization of American States.
LITERATURE CITED 1. Algranati, I. D. 1970. Occurrence of termination ribosomes as free 70S particles. FEBS Lett. 10:153-155. 2. Algranati, I. D., N. S. Gonzalez, and E. Bade. 1969. Physiological role of 70S ribosomes in bacteria. Proc. Natl. Acad. Sci. U.S.A. 62:574-580. 3. Algranati, I. D., and P. Lengyel. 1966. Polynucleotide dependent incorporation of amino acids in a cell-free system from thermophilic bacteria. J. Biol. Chem. 241:1778-1783. 4. Bachrach, U. 1970. Metabolism and function of spermine and related polyamines. Annu. Rev. Microbiol. 24:109-134. 5. Bade, E., N. S. Gonzalez, and I. D. Algranati. 1969. Dissociation of 70S ribosomes: some properties of the dissociation factor from Bacillus stearothermophilus and Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 64:654-660. 6. Beller, R. J., and B. D. Davis. 1971. Selective dissociation of free ribosomes of Escherichia coli by sodium ions. J.
Mol. Biol. 55:477-485. 7. Campbell, P. N., and J. R. Sargent. 1967. Miscellaneous
8. 9. 9a.
10.
11.
12.
13.
14.
techniques, p. 299-311. In P. N. Campbell and J. R. Sargent (ed.), Techniques in protein biosynthesis, vol. 1. Academic Press Inc., London. Cohen, S. S. 1971. Introduction to the polyamines. Prentice-Hall, Inc., Englewood Cliffs, N.J. Cohen, S. S., and J. Lichtenstein. 1960. Polyamines and ribosome structure. J. Biol. Chem. 235:2112-2116. Cunningham-Rundles, S., and W. K. Maas. 1975. Isolation, characterization, and mapping of Escherichia coli mutants blocked in the synthesis of ornithine decarboxylase. J. Bacteriol. 124:791-799. Echandi, G., and I. D. Algranati. 1975. Protein synthesis and ribosomal distribution in a polyamine auxotroph of Escherichia coli: studies in cell-free systems. Biochem. Biophys. Res. Commun. 62:313-319. Garcia-Patrone, M., N. S. Gonzalez, and I. D. Algranati. 1971. Association factor of ribosomal subunits from Bacillus stearothermophilus. Proc. Natl. Acad. Sci. U.S.A. 68:2822-2825. Garcia-Patrone, M., N. S. Gonzalez, and I. D. Algranati. 1972. Association of ribosomal subunits: mechanism of the reaction induced by association factor. FEBS Lett. 24:126-130. Garcia-Patrone, M., N. S. Gonzalez, and I. D. Algranati. 1973. Association of ribosomal subunits. III. Properties of the 30S-50S couples produced in uitro by the association factor from Bacillus stearothermophilus. Biochim. Biophys. Acta 299:452-459. Garcia-Patrone, M., N. S. Gonzalez, and I. D. Algranati. 1975. Association of ribosomal subunits. IV. Poly-
VOL 124, 1975
15. 16. 17.
18. 19.
20.
21. 22. 23.
RIBOSOMAL PROFILES AND POLYAMINES IN E. COLI
amines as active components of the association factor from Bacillus stearothermophilus. Biochim. Biophys. Acta 395:373-380. Hardy, S. J. S., and G. Turnick. 1971. Stabilization of 70S ribosomes by spermidine. Nature (London) New Biol. 299:17-19. Hill, W. E., G. P. Rossetti, and K. E. Van Holde. 1969. Physical studies of ribosomes from Escherichia coli. J. Mol. Biol. 44:263-277. Hirshfield, I. N., H. J. Rosenfeld, Z. Leifer, and W. K. Maas. 1970. Isolation and characterization of a mutant of Escherichia coli blocked in the synthesis of putrescine. J. Bacteriol. 101:725-730. Maas, W. K., Z. Leifer, and J. Poindexter. 1970. Studies with mutants blocked in the synthesis of polyamines. Ann. N.Y. Acad. Sci. 171:957-967. Modolell, J. 1971. The S-30 system from Escherichia coli, p. 1-65. In J. A. Last and A. Y. Laskin (ed.), Protein biosynthesis in bacterial systems. Marcel Dekker Inc., New York. Morris, D. R. 1973. RNA and protein synthesis in a polyamine-requiring mutant of Escherichia coli, p. 111-112. In D. H. Russell (ed.), Polyamines in normal and neoplasic growth. Raven Press, New York. Morris, D. R., and C. M. Jorstad. 1970. Isolation of conditionally putrescine-deficient mutants of Escherichia coli. J. Bacteriol. 101:731-737. Munro, G. F., and C. A. Bell. 1973. Polyamine requirements of a conditional polyamine auxotroph of Esecherichia coli. J. Bacteriol. 115:469-475. Norton, J. W., V. A. Erdmann, and E. J. Herbst. 1968. Polyamine-inorganic cation interaction with ribo-
24.
25.
26.
27. 28.
29. 30.
31.
1127
somes of Escherichia coli. Biochim. Biophys. Acta 155:293-295. Revel, M. 1972. Polypeptide chain initiation: the role of ribosomal protein factors and ribosomal subunits, p. 87-131. In L. Bosch (ed.), The mechanism of protein synthesis and its regulation. North-Holland Publishing Co., Amsterdam. Silman, N., M. Artman, and H. Engelberg. 1965. Effect of magnesium and spermine on the aggregation of bacterial and mammalian ribosomes. Biochim. Biophys. Acta 103:231-240. Srinivasan, P. R., D. V. Young, and W. Maas. 1973. Stable ribonucleic acid synthesis in stringent (rel') and relaxed (relI) polyamine auxotrophs of Escherichia coli K-12. J. Bacteriol. 116:648-655. Stevens, L. 1970. The biochemical role of naturally occurring polyamines in nucleic acid synthesis. Biol. Rev. 45:1-27. Subramanian, A. R., E. Z. Ron, and B. D. Davis. 1968. A factor required for ribosome dissociation in Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 61:761-767. Szer, W. 1969. Enzymatic degradation of ribosomal RNA in isolated purified ribosomes. Biochem. Biophys. Res. Commun. 35:653-658. Takeda, Y. 1969. Polyamines and protein synthesis. II. The shift in optimal concentration of Mg++ by polyamines in the MS2 phage RNA-directed polypeptide synthesis. Biochim. Biophys. Acta 179:232-234. Young, D. V., and P. R. Srinivasan. 1972. Regulation of macromolecular synthesis by putrescine in a conditional Escherichia coli putrescine auxotroph. J. Bacteriol. 112:30-39.
JOURNAL OF BACTERIOLOGY, Dec. 1975, p. 1122-1127 Copyright © 1975 American Society for Microbiology
Ribosomal Distribution in a Polyamine Auxotroph of Escherichia coli I.
D. ALGRANATI,* G. ECHANDI, SARA H. GOLDEMBERG, SUSANNA CUNNINGHAM-RUNDLES,' AND W. K. MAAS
Instituto de Investigaciones Bioquimicas "Fundacion Campomar" and Facultad de Ciencias Exactas y Naturales, Obligado 2490, Buenos Aires (28), Argentina,* and Department of Microbiology, School of Medicine, New York University, New York, New York 10016 Received for publication 29 May 1975
The distribution of ribosomal particles has been studied in a polyamine-deficient mutant of Escherichia coli by sucrose gradient centrifugation analysis. Lysates from starved cells contained less 70S monomers and 30S subunits but more 50S particles than those prepared from bacteria supplemented with putrescine. The addition of the polyamine to putrescine-depleted cells induced a rapid change of the ribosomal profile. A similar effect could be obtained in vitro by equilibrium dialysis against a polyamine-containing solution. The ribosomal pattern obtained from starved bacteria was specific for polyamine deficiency. We conclude that the changes in ribosomal profiles upon restoration of putrescine levels in previously starved cells denote a shift of the equilibrium between 30S-50S couples and ribosomal subunits.
The distribution of the ribosomal pool between polysomes, 70S couples, and subunits is related to the physiological state of bacteria, as well as to the intracellular ionic levels and the presence of initiation (24), dissociation (5, 28) and, perhaps, also association factors (11). The association factor, obtained from extracts of Bacillus stearothermophilus, has been fractionated into two components, one of which behaves like a polyamine (11-13). This fact and the recent isolation of Escherichia coli mutants blocked in the biosynthesis of putrescine (17, 18, 21) (which is a precursor of spermidine) prompted us to investigate the ribosomal distribution in extracts prepared from a mutant strain grown in the absence and presence of exogenous polyamines. Our aim was to elucidate whether putrescine starvation would modify the equilibrium between ribosomal subunits and 70S couples, which in normal cells is strongly shifted towards the latter. The present paper deals with changes in ribosomal profiles after the addition of putrescine to a culture of an E. coli polyamine auxotroph in an attempt to elucidate the physiological role of polyamines and their possible relationship with the association factor of ribosomal subunits. Although multiple functions have been suggested for polyamines in relation to ribonucleic I Present address: Division of Immunology, Memorial Sloan-Kettering Cancer Center, New York, N.Y. 10021.
acid (RNA), deoxyribonucleic acid, and protein synthesis (4, 8, 27), membrane stability (4, 8), aggregation of ribosomal subunits (9, 23, 25), and osmotic adaptation (22), the biological role of these widespread organic cations is still not well understood. MATERIALS AND METHODS Bacterial strain and growth media. E. coli MA 261 Thr-, Leu-, Ser-, Thi- was used in all of the experiments described in the present work. This strain is, in addition, a double mutant deficient in two enzymes: agmatine ureohydrolase (AUH-) and constitutive ornithine decarboxylase (OD,-). Therefore, E. coli MA 261 requires an exogenous polyamine supply for growth (9a). The media used were those described by Srinivasan et al. (26) and designated as MMO and MMA. They contained salts supplemented with 0.5% glucose, thiamine, biotin, leucine, threonine, methionine, serine, glycine, and either ornithine or arginine. The putrescine (100 jg/ml)-supplemented media are designated MMOP and MMAP, respectively. Chemicals. Putrescine dihydrochloride and sucrose, density gradient grade (ribonuclease free), were obtained from Schwarz/Mann, Orangeburg, N. Y.; spermidine trihydrochloride was from New England Nuclear Corp., Boston, Mass. All other reagents were of the highest purity commercially available. Starvation for putrescine. For putrescine starvation the method reported by Young and Srinivasan (31) was slightly modified as follows. A culture inoculated from a slant was grown in MMO medium; after refrigeration for several hours, it was diluted 50-fold in MMA medium and allowed to grow again.
1122
VOL. 124, 1975
RIBOSOMAL PROFILES AND POLYAMINES IN E. COLI
The new culture was used in the ratio 1:10 to inoculate a MMO medium. After growing to a density of 2 x 108 to 3 x 108 cells per ml, it was diluted 1:3 with fresh MMO medium. At this point, the intracellular level of putrescine was depleted in such a way that the bacteria required exogenous polyamines for normal growth. All cultures utilized in this method were grown overnight at 37 C with aeration. The starved cells were cultivated in the absence or presence of putrescine and then used for the different experiments. Bacterial growth was measured spectrophotometrically at 550 nm. Preparation of bacterial lysates and analysis of ribosomal distribution. Cells were collected from exponentially growing cultures after slow cooling to allow completion of protein synthesis and accumulation of run-off ribosomes. Lysis of bacteria was carried out essentially as described previously (2). The supernatant fluids obtained after centrifugation of lysates at 7,000 x g for 10 min were layered on top of 15 to 40% (wt/vol) linear sucrose density gradients containing 10 mM tris(hydroxymethyl)aminomethanehydrochloride buffer (pH 7.8), 5 mM magnesium acetate, and 50 mM KCl and were centrifuged for 90 min at 45,000 rpm in a Spinco SW65 rotor. Where indicated, 60 mM NaCl was used instead of KCl in the sucrose solutions. Gradients were analyzed by monitoring at 254 nm with an ISCO ultraviolet analyzer. RNA concentrations were determined by the orcinol method (7). Dissociation of ribosomes. To study the state of association of ribosomes under different ionic conditions, we incubated them for 15 min at 37 C in the presence of different Mg2+ concentrations. They were then analyzed by sucrose gradient centrifugation and the percentages of dissociation were calculated as described previously (5). Binding of polyamines to ribosomes. The binding of ["4Clspermidine to ribosomes was determined by equilibrium dialysis. Samples (0.4 ml) containing 4 to 5 absorbancy units at 260 nm (A2**) of ribosomes in a solution of 10 mM tris(hydroxymethyl)aminomethane-hydrochloride buffer (pH 7.8), 5 mM magnesium acetate, 50 mM KCl, and 2 mM 2-mercaptoethanol were dialyzed at 4 C for 24 h against 30 ml of the same buffer made with 1 mM ["4C]spermidine (0.5 1sCi). In all of these experiments Visking dialysis tubing was carefully prewashed (19). For the assay of radioactivity, triplicate 50-ul samples were taken from solutions inside and outside of the dialysis bag and counted in a liquid scintillation spectrometer after addition of 0.2 ml of water and 3 ml of Bray mixture. The amount of polyamine bound was calculated from the difference between the radioactivity of solutions inside and outside of the dialysis bag. One A2* unit of 70S ribosomes was taken as 23 pmol (16).
1123
the intracellular level of this polyamine decreased more than 100-fold, whereas spermidine was reduced to about one-half of its normal concentration (17, 21, 26, 31). Under these conditions the mutants showed a requirement for exogenous polyamine to attain normal growth rate. The double mutant E. coli MA 261 behaved in a similar way. Addition of putrescine to MMO medium containing previously starved bacteria enhanced the growth rate considerably after a lag period of about 60 min. The generation time in several experiments varied from 180 to 300 min in the absence of putrescine and from 60 to 90 min in the presence of the polyamine (MMOP medium). Effect of putrescine addition on the ribosomal distribution. It is well known that polyamines induce the association of ribosomal subunits in vitro (9, 23, 25) and decrease the optimal Mg2+ concentration required for protein synthesis in cell-free systems (3, 30). These facts and the finding that the association factor contains polyamines (13, 14) led us to investigate whether or not the addition of putrescine to a culture previously starved could modify the ratio between 70S couples and subunits. For this purpose, putrescine-depleted bacteria were grown in MMO medium. After several hours putrescine was added to a portion of the culture in a separate flask and incubated along with the original culture for 30 or 60 min. The bacterial lysates were analyzed by sucrose gradient centrifugation. The results (Fig. 1A) show that in the absence of putrescine (MMO medium) there were fewer 70S ribosomes and the amount of 30S seemed to be small in comparison with the 50S peak. Perhaps the biosynthesis or assembly of the small subunit was somehow damaged during polyamine starvation. The addition of putrescine produced a marked change of ribosomal pattern (Fig. 1B). The lower total ribosomal content observed in the profiles obtained from starved cells is compensated for by a higher absorbancy in the supernatant fluid (not shown in figures). To have a quantitative comparison, the areas under the 70S and 50S peaks were measured. The ratio 70S/50S was 1.54 in Fig. 1A (MMO medium) and 2.89 in Fig. 1B (MMOP medium). Similar results were obtained when the bacteria were collected after fast cooling; in this case, polyribosomes were also seen and were more RESULTS abundant in the culture with putrescine. The different ribosomal profiles obtained Putrescine requirement of starved bacteria. Previous investigations carried out from putrescine-starved or -supplemented bacwith several conditional polyamine auxotrophs teria could indeed reveal a shift of the equilibhave shown that upon starvation of putrescine rium between 30S-50S couples and ribosomal
1124
J. BACTERIOL.
ALGRANATI ET AL.
Ec
AN
Iq
4
3
2
1
ml
4
3
21
from top
FIG. 1. Effect of polyamine addition on the ribosomal profile. Bacteria were harvested and lysed as described in the text. Cell extracts (about 0.6 A... unit) were analyzed after sucrose gradient centrifugation. (A) Ribosomal pattern corresponding to a lysate obtained from cells cultivated in MMO medium. (B) Profile corresponding to bacteria grown in MMO medium to which putrescine was added 30 min before harvesting of the culture.
subunits occurring in vivo. If this were the case, we should rule out several other possibilities as sources of artifacts, namely: (i) a different efficiency of extraction of monomers and subunits obtained with our method for lysis applied to both kinds of bacteria; (ii) a different content
taining Na+ ions instead of K+ ions (1, 6). When such experiments were performed, only the run-off ribosomes dissociated (Fig. 2). Comparison of these results with the patterns obtained with sucrose gradients containing K+ (Fig. 1) allowed us to calculate that the percentages of run-off ribosomes in the fractions containing monomers were 82 and 87% in the lysates prepared from starved and supplemented cells, respectively. We have purified ribosomal subunits from extracts of bacteria grown in the absence or in the presence of putrescine. The sucrose gradient analysis of both kinds of 50S particles showed that they are at least 95% pure even after a heating for 5 min at 55 C in the presence of 2 mM dithiothreitol. This treatment is able to dissociate 30S dimers, as has been shown in our laboratory (M. Gracia-Patrone, unpublished data). On the other hand, the analysis of ribosomal RNA obtained by phenol extraction of 50S subunits prepared from polyaminestarved and unstarved bacteria demonstrated that similar and small amounts of 16S RNA appeared in both cases along with the 23S RNA component. The presence of 16S RNA could be attributed to a partial breakdown of 23S RNA by endonuclease activity (29; 0. P. Van Diggelen, Ph.D. thesis, Univ. of Leyden, Leyden, A
70S
50s
30s B 705
50S 305
0.4
of run-off and complexed ribosomes in the 70S monomers, depending upon whether lysates were prepared from starved or unstarved bacte0.3 ria; and (iii) a different proportion of 30S dimers present as contaminants of the 50S fractions. All of these possibilities were consid02 ered and studied in detail. Quantitative determinations of RNA showed that our preparations of cell lysates gave similar 0.1 yields (80 to 90%) of total RNA obtained either from starved or unstarved bacteria. Since the ribosomal RNA is a high proportion of total RNA, we can conclude that the differences in 0 321f 4 4 321f the ribosomal patterns shown in Fig. 1 could not n/from top be attributed to selective extraction of subunits FIG. 2. Effect of polyamine addition on the riboor 70S monomers from both kinds of cells. It has been demonstrated that run-off ribo- somal profiles obtained by centrifugation in sucrose gradients containing Na+ ions. Cell lysates were somes, which are released from polysomes at the and analyzed as described in the text and end of translation, and complexed monomers, prepared the Fig. 1. (A) Ribosomal pattern correlegend which are still bound to messenger RNA, be- sponding to tobacteria grown in MMO medium. (B) have in different ways under different condi- Profile corresponding to cells cultivated in MMO tions. These two kinds of monomers can be medium to which putrescine was added 30 min before distinguished by using sucrose gradients con- harvesting.
VOL. 124, 1975
1125
RIBOSOMAL PROFILES AND POLYAMINES IN E. COLI
The Netherlands, 1973) rather than to a contamination of 50S subunits with 30S dimers or modified 70S particles. However, even assuming the possibility of contamination, the 50S particles obtained from starved and unstarved cells should contain about the same amount of 30S dimers and/or modified 70S monomers, and these contaminations would not alter the general pattern of ribosomal profiles. Specificity of the polyamine effect. We have investigated whether the effect on the ribosomal profile of putrescine depletion was specific for polyamine deficiency and whether it could also be obtained after starvation for a carbon or a nitrogen source. For this purpose, bacteria grown in MMOP medium were washed several times with glucose-free medium and then suspended either in complete medium or in glucose-free MMOP medium. After further incubation for 30 min at 37 C, the corresponding lysates were prepared and analyzed as usual. The results (Fig. 3B and C) show that the profile of run-off ribosomes obtained from cultures in the presence of putrescine did not change after starvation for glucose. A parallel culture in MMO medium gave a pattern characteristic of polyamine starvation (Fig. 3A). The same results were obtained when the nitrogen source was omitted. Dissociation of ribosomes. The sensitivity to dissociation of ribosomes obtained from bacteria grown either in MMO or in MMOP medium was studied by decreasing the Mg2+ levels. The results (Fig. 4) demonstrate that, although the ribosomal particles prepared from cultures without putrescine were more highly dissociated than those of bacteria cultivated in MMOP at 8 70S 505 30S
A. 70s 50S 30s
4
4
C. 70s
different Mg2+ concentrations, the dissociation curves were parallel, indicating that both types of 70S particles were equally sensitive to low Mg2+ levels. Binding of spermidine to ribosomes. The binding of[14C]spermidine to ribosomal particles was measured by equilibrium dialysis and was of the same order (920 4 35 molecules per ribosome) for particles obtained from cultures with and without putrescine. After dialysis for 24 h against a buffer containing 1 mM spermidine, there was a partial association of subunits, so that the patterns obtained from bacteria grown in the absence and presence of polyamine were more similar than those obtained before the equilibrium dialysis. Furthermore, both kinds of ribosomes became more stable to dissociation induced by a decrease of Mg2+ concentration (Fig. 5). 70 60
2 50 E- 40 ° 30 Q 20
10
1
0
2
3
6
5
4
7
9
8
10
tg+g concentration (mM) FIG. 4. Stability of ribosomes at different Mg2" concentrations. Extracts obtained from bacteria grown in MMO (0) or in MMOP (0) medium were adjusted to the indicated levels of Mg2+. Samples were then analyzed, and the resulting percentages of dissociation were calculated as outlined in the text.
50S 30s
4
0 rt0
c It in
N Iq
).3
l
A
0 0
O2
,,,
,,, -4 3 2
f
4 ml
3 from
2 1
4
3 2
f 0
top
FIG. 3. Specificity of the effect of polyamine ribosomal distribution. Lysates from bacteria under different conditions were analyzed as indicated in the text. Bacteria were grown in the following media: (A) MMO; (B) MMOP; (C) MMOP, but subsequently cells were washed and reincubated for 30 min in glucose-free MMOP medium. on
1
2
3
4
5
Mg't"' concentrotion
6
7
(mm)
FIG. 5. Stability of ribosomes after dialysis against spermidine. Samples were submitted to equilibrium dialysis against I mM spermidine and then adjusted to the indicated Mg2+ concentrations. Analysis and calculations were done as described in the text. Symbols as in Fig. 4.
1126
J. BACTERIOL.
ALGRANATI ET AL.
DISCUSSION The isolation of mutants unable to synthesize putrescine permitted us to investigate whether or not the ribosomal profile is related to the intracellular polyamine content and eventually to the association factor. Our results indicate that, in some cases, as early as 30 min after the addition of putrescine to previously depleted cells the distribution of ribosomes appeared markedly altered. Before the addition of polyamine, the amounts of 70S ribosomes and 30S subunits were relatively low in comparison with the high 50S peak. Moreover, preliminary results have demonstrated that the broad 30S peak which appeared in the ribosomal profiles of starved-cell lysates indicates the existence of a mixture of abnormal and normal 30S particles. The restoration of putrescine changed the pattern rapidly, provoking a marked increase of 70S couples and a normalization of the relative amounts of both subunits (Fig. 1). The modification of the ribosomal pattern occurred even before an increase in growth rate was detectable. The ribosomal profiles obtained from polyamine-starved and -supplemented bacteria were not due to a different efficiency of extraction of 70S ribosomes and subunits, to a different content of run-off particles in the 70S monomers (Fig. 2), or to a different amount of 30S dimers in the fractions of 50S peaks. A similar modification of the ribosomal profile could be obtained in vitro by dialysis against spermidine. This fact indicates that polyamines at physiological levels have a direct effect on the equilibrium between 70S couples and subunits. Furthermore, the well-known effect of ribosomal stabilization by polyamines (15) was confirmed by measuring the sensitivity towards decreasing Mg2+ concentrations (Fig. 4 and 5). The profile of ribosomes after growth in MMO medium seems to be characteristic of polyamine starvation, since the pattern brought about by MMOP medium did not change after starvation for a glucose or a nitrogen source (Fig. 3). We have recently reported some studies on polypeptide synthesis induced by polyuridylic acid in cell-free systems obtained from starved and unstarved bacteria (10). Under these conditions, it is possible to detect at least one effect of polyamines on protein synthesis that is independent of transcription. Many biological functions have been suggested for polyamines in living cells. Our results indicate that the translation process and the
assembly of its machinery, including the distribution pattern of ribosomes, are good candidates for targets of a primary effect of these organic cations. ACKNOWLEDGMENTS We are grateful to L. F. Leloir and all members of the Instituto de Investigaciones Bioquimicas for helpful discussions. This work was supported by grants from the University of Buenios Aires, the Ministerio de Bienestar Social, the Programa Regional de Desarrollo Cientifico y Tecnol6gico de la O.E.A., and the Consejo Nacional de Investigaciones Cientificas y Tecnicas (Argentina). I. D. A. and S. H. G. are Career Investigators of the latter institution, and G. E. is a Fellow of the Organization of American States.
LITERATURE CITED 1. Algranati, I. D. 1970. Occurrence of termination ribosomes as free 70S particles. FEBS Lett. 10:153-155. 2. Algranati, I. D., N. S. Gonzalez, and E. Bade. 1969. Physiological role of 70S ribosomes in bacteria. Proc. Natl. Acad. Sci. U.S.A. 62:574-580. 3. Algranati, I. D., and P. Lengyel. 1966. Polynucleotide dependent incorporation of amino acids in a cell-free system from thermophilic bacteria. J. Biol. Chem. 241:1778-1783. 4. Bachrach, U. 1970. Metabolism and function of spermine and related polyamines. Annu. Rev. Microbiol. 24:109-134. 5. Bade, E., N. S. Gonzalez, and I. D. Algranati. 1969. Dissociation of 70S ribosomes: some properties of the dissociation factor from Bacillus stearothermophilus and Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 64:654-660. 6. Beller, R. J., and B. D. Davis. 1971. Selective dissociation of free ribosomes of Escherichia coli by sodium ions. J.
Mol. Biol. 55:477-485. 7. Campbell, P. N., and J. R. Sargent. 1967. Miscellaneous
8. 9. 9a.
10.
11.
12.
13.
14.
techniques, p. 299-311. In P. N. Campbell and J. R. Sargent (ed.), Techniques in protein biosynthesis, vol. 1. Academic Press Inc., London. Cohen, S. S. 1971. Introduction to the polyamines. Prentice-Hall, Inc., Englewood Cliffs, N.J. Cohen, S. S., and J. Lichtenstein. 1960. Polyamines and ribosome structure. J. Biol. Chem. 235:2112-2116. Cunningham-Rundles, S., and W. K. Maas. 1975. Isolation, characterization, and mapping of Escherichia coli mutants blocked in the synthesis of ornithine decarboxylase. J. Bacteriol. 124:791-799. Echandi, G., and I. D. Algranati. 1975. Protein synthesis and ribosomal distribution in a polyamine auxotroph of Escherichia coli: studies in cell-free systems. Biochem. Biophys. Res. Commun. 62:313-319. Garcia-Patrone, M., N. S. Gonzalez, and I. D. Algranati. 1971. Association factor of ribosomal subunits from Bacillus stearothermophilus. Proc. Natl. Acad. Sci. U.S.A. 68:2822-2825. Garcia-Patrone, M., N. S. Gonzalez, and I. D. Algranati. 1972. Association of ribosomal subunits: mechanism of the reaction induced by association factor. FEBS Lett. 24:126-130. Garcia-Patrone, M., N. S. Gonzalez, and I. D. Algranati. 1973. Association of ribosomal subunits. III. Properties of the 30S-50S couples produced in uitro by the association factor from Bacillus stearothermophilus. Biochim. Biophys. Acta 299:452-459. Garcia-Patrone, M., N. S. Gonzalez, and I. D. Algranati. 1975. Association of ribosomal subunits. IV. Poly-
VOL 124, 1975
15. 16. 17.
18. 19.
20.
21. 22. 23.
RIBOSOMAL PROFILES AND POLYAMINES IN E. COLI
amines as active components of the association factor from Bacillus stearothermophilus. Biochim. Biophys. Acta 395:373-380. Hardy, S. J. S., and G. Turnick. 1971. Stabilization of 70S ribosomes by spermidine. Nature (London) New Biol. 299:17-19. Hill, W. E., G. P. Rossetti, and K. E. Van Holde. 1969. Physical studies of ribosomes from Escherichia coli. J. Mol. Biol. 44:263-277. Hirshfield, I. N., H. J. Rosenfeld, Z. Leifer, and W. K. Maas. 1970. Isolation and characterization of a mutant of Escherichia coli blocked in the synthesis of putrescine. J. Bacteriol. 101:725-730. Maas, W. K., Z. Leifer, and J. Poindexter. 1970. Studies with mutants blocked in the synthesis of polyamines. Ann. N.Y. Acad. Sci. 171:957-967. Modolell, J. 1971. The S-30 system from Escherichia coli, p. 1-65. In J. A. Last and A. Y. Laskin (ed.), Protein biosynthesis in bacterial systems. Marcel Dekker Inc., New York. Morris, D. R. 1973. RNA and protein synthesis in a polyamine-requiring mutant of Escherichia coli, p. 111-112. In D. H. Russell (ed.), Polyamines in normal and neoplasic growth. Raven Press, New York. Morris, D. R., and C. M. Jorstad. 1970. Isolation of conditionally putrescine-deficient mutants of Escherichia coli. J. Bacteriol. 101:731-737. Munro, G. F., and C. A. Bell. 1973. Polyamine requirements of a conditional polyamine auxotroph of Esecherichia coli. J. Bacteriol. 115:469-475. Norton, J. W., V. A. Erdmann, and E. J. Herbst. 1968. Polyamine-inorganic cation interaction with ribo-
24.
25.
26.
27. 28.
29. 30.
31.
1127
somes of Escherichia coli. Biochim. Biophys. Acta 155:293-295. Revel, M. 1972. Polypeptide chain initiation: the role of ribosomal protein factors and ribosomal subunits, p. 87-131. In L. Bosch (ed.), The mechanism of protein synthesis and its regulation. North-Holland Publishing Co., Amsterdam. Silman, N., M. Artman, and H. Engelberg. 1965. Effect of magnesium and spermine on the aggregation of bacterial and mammalian ribosomes. Biochim. Biophys. Acta 103:231-240. Srinivasan, P. R., D. V. Young, and W. Maas. 1973. Stable ribonucleic acid synthesis in stringent (rel') and relaxed (relI) polyamine auxotrophs of Escherichia coli K-12. J. Bacteriol. 116:648-655. Stevens, L. 1970. The biochemical role of naturally occurring polyamines in nucleic acid synthesis. Biol. Rev. 45:1-27. Subramanian, A. R., E. Z. Ron, and B. D. Davis. 1968. A factor required for ribosome dissociation in Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 61:761-767. Szer, W. 1969. Enzymatic degradation of ribosomal RNA in isolated purified ribosomes. Biochem. Biophys. Res. Commun. 35:653-658. Takeda, Y. 1969. Polyamines and protein synthesis. II. The shift in optimal concentration of Mg++ by polyamines in the MS2 phage RNA-directed polypeptide synthesis. Biochim. Biophys. Acta 179:232-234. Young, D. V., and P. R. Srinivasan. 1972. Regulation of macromolecular synthesis by putrescine in a conditional Escherichia coli putrescine auxotroph. J. Bacteriol. 112:30-39.
