Dependence of the Putrescine Content of Escherichia coli on the Osmotic Strength of the Medium*

Abstract Putrescine is the polyamine present in highest concentration in Escherichia coli; however, few specific functions for it are known. We found that cells growing in nutrient broth supplemented with high concentrations of NaCl, KCl, MgCl2, or sucrose had greatly reduced levels of cellular putrescine. On the basis of osmotic strength, all four solutes produced similar decreases in putrescine content. In contrast, glycerol had little effect on the amount of cellular putrescine. Cellular spermidine content was not affected by any of the additives. When cells were grown in high NaCl, pools for most amino acids increased; a few remained the same or decreased slightly. A sudden increase in the osmolarity of the medium led to a rapid excretion of cellular putrescine while there was no decrease in spermidine or free amino acids. This loss of putrescine could be blocked by sodium azide or sodium arsenate. The mechanism of putrescine excretion has two components; one is dependent on high concentrations of potassium ion in the medium, and the other is not. Cells grown in the presence of 11 mm potassium and then put into high osmolarity medium lacking potassium failed to excrete [14C]putrescine rapidly. Replacing the potassium greatly stimulated putrescine excretion within 1 min. The rate of [14C]putrescine excretion was half-maximal at a potassium concentration of 0.69 mm. E. coli which were grown in medium containing 0.1 mm potassium and resuspended in 11 mm potassium initially could excrete [14C]putrescine only slowly but regained the ability to excrete putrescine maximally after 15 min. Sodium, magnesium, ammonium, and rubidium ions and pH from 6.4 to 7.3 did not affect the rate of [14C]putrescine excretion. A sudden increase in the osmolarity of the medium produces a rapid increase in the cellular potassium content. However, this rapid uptake of potassium was not dependent on excretion of large quantities of cellular putrescine. When cells growing in high salt medium were transferred to low salt medium, the putrescine content approached the low salt value only after 140 min. However, the ability of cells to take up [14C]putrescine from the medium did increase 6-fold within 2 min after resuspension in low salt medium. [14C]Spermidine was not metabolized to a measurable extent over 70 min in either low salt or high salt nutrient broth cultures. Because the cellular spermidine contents were similar in the two cultures and because [14C]spermidine was not catabolized in either culture, the rates of spermidine synthesis must also be similar, even though the precursor (putrescine) pool sizes are quite different.

Fyona the Department of Biophysics and Genetics, University o,f Colorado Medical Center, Demer, Coloyarlo 80220 SUMMARY Putrescine is the polyamine present in highest concentration in Escherichia coli; however, few specific functions for it are known.
We found that cells growing in nutrient broth supplemented with high concentrations of NaCl, KCl, MgCl*, or sucrose had greatly reduced levels of cellular putrescine.
On the basis of osmotic strength, all four solutes produced similar decreases in putrescine content.
In contrast, glycerol had little effect on the amount of cellular putrescine. Cellular spermidine content was not affected by any of the additives.
When cells were grown in high NaCl, pools for most amino acids increased; a few remained the same or decreased slightly.
A sudden increase in the osmolarity of the medium led to a rapid excretion of cellular putrescine while there was no decrease in spermidine or free amino acids. This loss of putrescine could be blocked by sodium azide or sodium arsenate.
The mechanism of putrescine excretion has two components; one is dependent on high concentrations of potassium ion in the medium, and the other is not. Cells grown in the presence of 11  cells to take up [i4C]putrescine from the medium did increase 6-fold within 2 min after resuspension in low salt medium.
[i4C]Spermidine was not metabolized to a measurable extent over 70 min in either low salt or high salt nutrient broth cultures.
Because the cellular spermidine contents were similar in the two cultures and because [14C]spermidine was not catabolized in either culture, the rates of spermidine synthesis must also be similar, even though the precursor (putrescine) pool sizes are quite different.
Putrescine (1,4-diaminobutane) and spermidine (l-aminopropyl-1 , 4-diaminobutane) are the polyamines found in highest concentrations in Escherichia coli. The putrescine content is usually several-fold greater than that of spermidine (1) ; in vivo both polyamines turn over slowly if at all (2,3). In E. co& both polyamines have been implicated in various aspects of in vitro RNA and protein synthesis (4). Spermidine is more effective than putrescine in all such systems in which both polyamines have been examined.
Few specific functions for putrescine are known. It is a precursor of spermidine (2,5). Hirshfield et al. (6) described a mutant of E. coli in which putrescine synthesis could be blocked by growth on arginine-containing medium.
The slow growth of these cells was greatly stimulated by adding putrescine, and much less by spermidine.
Morris and Jorstad (7) found that reducing the intracellular putrescine of another mutant to 1% of normal decreased the doubling rate by only 10%. Inouye and Pardee (8) found that if an arginine auxotroph of E. coli was starved for arginine for 1 hour, addition of arginine caused the cells to divide synchronously.
Putrescine was specific in abolishing this synchrony.
The authors suggested that an increase in the molar ratio of putrescine to spermidine might be a critical factor for cell division; however, this ratio was at least as much affected by a decrease in the relative spermidine content as by an increase in the relative putrescine content. All other proposed specific functions for putrescine require relatively low concentrations of putrescine and do not explain the high levels found in autotrophic E. coli.
We postulated that high concentrations of putrescine might play some role in maintaining the cellular ionic or osmotic balance. Evidence presented here demonstrates that mono-and divalent Issue of February 2.5, 1972 G. P. Afunro, K. Hercules, J. Morgan, and W. Xauediel 1273 ions and several sugars, present in the medium, can reduce the active spermidine and putrescine ranged from 88 to 95%. The cellular putrescine content without affecting the spermidine method of Lowry et al. (16) was used to determine total cellular content.
Glycerol, which freely penetrates the cell (9, lo), does protein in the trichloroacetic acid precipitates, using bovine not reduce the cellular putrescine content. plasma albumin (Armour Pharmaceutical Company) as standard. A sudden increase in the osmolarity of the medium leads to a prompt excretion of cellular putrescine, but not spermidine.
Because the cellular K+ content rises under the same circumstances (II), we suspected that K+ uptake and putrescine excretion might be coupled.
Evidence presented here demonstrates that the rate of putrescine excretion is dependent on the I(+ concentration in the medium. A preliminary report of this work has appeared (12). XATERIALS AND Media with different concentrations of K+, NH4+, or 1\Ig+2 were made by replacing some or all of the KCl, NH&l, or &IgS04 lvith NaCl or Na2S04, so that constant Cl-and Sod+ concentrations were maintained. Medium at pH 8.0 was made by adjusting the phosphate buffer stock solution to pH 8.0 with NaOH.
The final phosphate concentration in medium at pH 8.0 was the same as in the standard 11 rnM K+ medium.
Viable cell concentrations were determined by diluting culture aliquots in phosphate buffer (3.16 g of Na2HP04, 3.0 g of KH2POI, and 4.0 g of NaCl per liter, pH 7.2) and plating on tryptone agar.

Analysis of Amino Acid Pools-Cultures
for analysis of amino acid pools were rapidly chilled on ice and centrifuged at 3000 X g and 4" for 3 min. Pellets were taken up in ice-cold 0.25 N HC104 to a total volume of Ti ml (17), allowed to stand in an ice bucket for 10 min, and centrifuged at 2000 x g and 4" for 10 min. Pellets were saved for protein estimation (16). Supernatants were adjusted to pH 7.1 with 1 N KOH and the KC104 precipitates collected by centrifugation at 2000 x g and 4' for 10 min. Supernatants were then adjusted to pH 2.0 with concentrated HCl, and 1.0 to 2.0 ml aliquots were analyzed for amino acid composition with a Beckman model MS instrument.
Radioactive Chemicals and of Polyamines-illiquots of 50 ml R-ere removed from the bacterial culture and chilled on ice. Optical density (at 600 nm; Beckman DU spectrophotometer), pH, and viable cell concentration were measured. Culture aliquots were then centrifuged at 4" and 2000 x g for 10 min (International PR-1 centrifuge), resuspended in 7 ml of cold 0.15 1% NaCI, and centrifuged at about 2000 x g (Sorvall desk centrifuge) for 10 min in a cold room (4") or incubator (37").
After decanting supernatants, cell pellets were frozen.  (Instrumentation  Laboratories) were added to each vial, and potassium values were determined with a flame photometer (Instrumentation Laboratories, model 143). Potassium contents were corrected for blank filters which contained an average of 0.088 Fmole of potassium.

RESULTS
The method of Dubin and Rosenthal (13) was used to quantitate polyamines.
Polyamines were extracted from bacteria into 0.3 N trichloroacetic acid, hydrolyzed in 6 N HCI to convert acetylated polyamines to the parent compounds, and separated by descending paper chromatography on Whatman No. 1 paper in butanol-l-acetic acid-pyridine-water (4 : 1: 1: 2). (Sonication of the trichloroacetic acid extracts (Branson sonifier, model 575, setting 4, two 30-set bursts on ice) did not increase the amount of polyamine recovered.) Polyamine bands were stained, extracted, and estimated by the method of Raina and Cohen (14), as modified by Raina et al. (15). Putrescine and spermidine standards (Calbiochem) were chromatographed in every experiment. As an aid to identification of bands, 0.1 PCi of [14C]putrescine or [14C]spermidine was added to some trichloroacetic acid extracts, and the areas of radioactivity were located with a radiochromatogram scanner (Packard, model 7200). This method also established that very little if any of the radioactive polyamines were degraded during the analysis.
Recovery of radio-Polyamine Content of Bacteria Grown at Different Osmotic Strengths-Cells were initially grown in low salt nutrient broth which contains low concentrations of several cations including Na+, 10 mM; Kf, 2.5 m&r; Ca+2, 0.63 mM; Mg+2, 0.31 InM (determined on the single bottle of Difco nutrient broth used for all experiments).
Growth at increased concentrations of NaCl, KCl, MgC12, or sucrose, up to 0.6 osmole per liter, reduced the putrescine content of E. coli B by more than 5-fold ( Fig. 1). Growth at high concentrations of glycerol did not bring about a comparable reduction in putrescine content.
The spermidine content ( Fig. 1) varied only slightly with growth in media of different osmotic strengths.
E. coli K12(h) showed similar decreases in putrescine content with increasing concentrations of NaCl and MgC12; again, spermidine content was unaffected (data not shown).
The following experiments show that the low putrescine content observed with cells grown in high osmolarity media is not due to artificial loss of cell constituents.
(a) Centrifugation at  decreased about 20% (Fig. 2). (n-Xylose, D-arabinose, and maltose were similar to sucrose in their ability to bring about a loss of 14C from the cells (data not shown).) In a similar esperiment the 14C released from cells after a shift to high salt nutrient broth (0.6 M NaCl) was identified as [14C]putrescine by extraction into butanol-1 (13) and chromatography.
To confirm these results, cellular putrescine and spermidine contents were measured chemically after a shift to high salt nutrient broth (0.6 M NaCU. Spermidine values remained at the low salt level for 2 hours after adding NaCl, whereas the putrescine content fell A-fold in 9 min and remained depressed for at least 2 hours (Fig. 3, circles). As a control for the experiment in Fig. 2, cells were shifted to 0.3 M NaCl, 0.6 RI sucrose, or 0.6 M glycerol on ice. After 15 min intracellular 14C was measured as for the other samples. Little or no 14C was lost from the cells during this time. The fact that cold could inhibit the rapid putrescine loss suggested that the process might be energy dependent.
Therefore, the effect of a sudden increase in salt concentration was repeated in the presence of 6 m&I sodium azide or 10 mnr any increase in cell tit'er without killing cells in either high or low salt nutrient broth.) Sodium azide by itself caused a small loss of cellular putrescine (Fig. 3), but comparison of the curves for NaCl alone and sodium azide plus NaCl indicates that azide inhibited the loss of putrescine, even at the low concentration used. Sodium arsenate (neutralized to pH 6.8 before use) also blocked the putrescine loss (data not shown).
None of the experiments with inhibitors affected the spermidine level. The fact that two inhibitors could block putrescine loss implies that the process may directly or indirectly require energy.
It was possible that the excretion of [l%]putrescine following a shift to high salt reflects the normal rate of putrescine excretion for cultures grown in high salt and not a response to the change in salt concentration per se. However, when cells were grown and previously labeled with ['*C]putrescine in high salt nutrient broth (0.3 M NaCl), they excreted 14C in high salt at a very slow rate similar to cells which were grown, labeled, and incubated in low salt medium.
A similar experiment using II mu K+ medium gave identical results. as usual. At -5 min another culture grown in low salt nutrient broth was made 6 mM in sodium azide (A-A). A third culture received sodium azide at -5 min and NaCl at 0 min (A-A).
Cells maintained in low osmotic strength medium lost putrescine very slowly (about 4% in 3 min).
In contrast to loss of putrescine, amino acid pools were not depleted by transfer of cells to high osmolarity medium (Table I, column B) .
As shown in Fig. 4B, reducing the external K+ concentration significantly reduced the rate of excretion of cellular [14C]putrestine; cells lost only 29% of the 14C after 3 min in no K+, 0.4.~ NaCl medicm.
To establish that this rather slow excretion rate was not affected by the particular solute used to increase the osmolarity of the medium, the experiment shown in Fig. 4B was repeated with additions of 0.35 M NaCl or 0.7 M sucrose. The same rates of [14C]putrescine excretion were observed, indicating that the high concentration of Na+ present in the experiment shown in Fig. 4B could not partially substitute for K+.
The reduction in the rate of [%]putrescine excretion is reversible. When cells were suspended in 0.4 M NaCl medium containing no K+ and made 11 mM in KC1 2 min later, they promptly excreted 61% of the cellular 14C in the next 3 min (Fig. 4C)  At the times shown l-ml aliquots were filtered onto Schleicher and The pellets were then resuspended in various media (Table II).
The cells in suspension A (low osmolarity control) lost W at the slow rate characteristic of cells in media of low osmotic strength.
Cultures B, C, and D, which all contained 11  Thus, the absence of MgS2 and NH4+ or the absence of Naf did not significantly alter the excretion patterns. In similar experiment,s 30 maI Mg+2 or 11 ml\% Rb+ could not aubstitute for I(+ (data not shown).
The effect of pH on the rate of putrescine excretion xas measured over the range of pH 6.4 to 7.3. E. coli B was grown in 11 mM K+ medium (pH 6.9), labeled with [14C]putrescine, washed in growth medium, and split into four portions before the final centrifugation.
Pellets were resuspended in 10 ml of (a) 11 mM K+ medium (pH 6.9), (b) 11 rnM K+, 0.7 M sucrose medium (pH 6.9), (c) 11 m&I K+ medium (pH 8.0), and (d) 11  for example, in 0.01 mM K+ medium, the optical density increased with a doubling time of 44 min to optical density 0.132 and then decreased over the following 30 min to a doubling time of 220 min. To avoid these changes in growth rate, cells were grown in 0.1 mM K+, a concentration which is low and yet maintains cells in rapid growth to an optical density greater than 0.700.
Excretion Addition of 11 mM K+ after 7 min initially produced a slow rate of putrescine excretion.
This rate of excretion appeared to increase at later times. Thus, cultures grown in 0.1 mv K+ medium are not immediately able to excrete putrescine rapidly in 11 mM K+, 0.4 M NaCl medium.
To quantitate the time required for regeneration of the Kfdependent excretion mechanism, a culture of E. coli B was grown in 0.1 mM K+ medium, labeled with [14C]putrescine, washed twice, and then resuspended in 11 mM K+ medium.
Periodically, aliquots were made 0.4 M in NaCl and incubated for 3 min to measure the rate of [14C]putrescine excretion.
The ability of cells to excrete [14C]putrescine began to increase immediately after resuspension in 11 InM K+ medium; the fraction of 1% excreted in 3 min doubled 3 min after resuspension and tripled by 10 min after resuspension.
The maximal rate of excretion (70% of i4C excreted in 3 min) was reached after 15 min.
To determine the time required for cells to lose the ability to excrete [i4C]putrescine rapidly, cells were grown in 11 mM K+ medium, labeled with [14C]putrescine, and then incubated in 0.1 mM K+ medium.
At various times aliquots were made 0.4 M in NaCl and 11 mM in K+ and incubated for 3 min. The cells retained the ability to excrete [i4C]putrescine rapidly for more than 50 min of growth in low K+ medium (data not shown).
Excretion of [i*C]Putrescine by E. coli B Grown in Low Concentrations of Ammonium, Magnesium, and Sodium Ions-In contrast to low Kf medium, E. coli B grown in low concentrations of NH4+, Mg+", or Na+ produced the usual rapid excretion of [14C]putrescine in 11 mM K+, 0.7 M sucrose medium lacking the specific cation under study.
The lowest concentration of NH*+ which would support logarithmic growth past optical density 0.400 was 7 mM. Cells in 0.050 mM Mg+2 could grow logarithmically to optical density 0.600. Concentrations of Na+ from 0.1 mM to 6.0 X 1OW mM allowed the optical density to increase logarithmically to 0.175; the growth rate then gradually slowed over several generations. Therefore, low Naf cultures were grown only to optical density 0.150 before labeling.
Polyamine Content after Decreasing Salt Concentration oj Nutrient Broth and M9 Medium-E. coli B growing in high salt nutrient broth (0.6 M NaCl) was centrifuged at 37" and resuspended in low salt nutrient broth.
Cells began to grow logarithmically within 2 min. The putrescine content rose from a high salt value of 0.69 pg ((total putrescine.2 HCl)/(milliliters of culture x optical density of culture)) to 1.33 pg of putrescine at 9 min after resuspension in low salt nutrient broth. This was followed by a gradual increase to 3.0 pg of putrescine at 80 min after resuspension, two generations later. This putrescine value was only half of the final low salt value. There was little change in the spermidine content.
The putrescine content of E. coli 13 in low salt (regular 319 medium ((3.9 pig of putrescine. 2 HCl)/(milliliters of culture x optical density of culture)) was only about half the value for low salt, rmt,rient broth; however, high conoent,rations of ;2-:rCl resulted in reduction of the putrescine content.
The putrescine content increased slowly ((from 0.8 to 1.0 .ug of putrescine.2 HCl)/(milliliters of culture x optical density of culture)) and approached the value for cells grown in low salt M9 medium (3.2 pg of putrescine) 2 hours after resuspension (two generations of growth in low salt medium).
The changes in polyamine levels after transfer from high to low salt medium were similar in both nutrient broth and the chemically defined medium, M9. The putrescine content returned to the low salt level very slowly in both media; therefore, high putrescine content is not a prerequisite for resumption of rapid growth.
If putrescine has specific cellular functions other than an adjustment to the osmolarity of the medium, these functions must require very low levels of putrescine.
Uptake of [14C]Putrescine after Decreasing Salt Concenfration-The ability of E. coli B to take up [i*C]putrescine from the medium was studied before and after transferring cells from high salt (0.6 M NaCl) to low salt nutrient broth (Fig. 5). Cells were centrifuged out of the original high salt medium and resuspended in fresh high salt medium at the start of the experiment to remove any putrescine which the cells might have excreted during growth.
After resuspension the cells were able to take up very little radioactive putrescine.
After a second centrifugation and resuspension in low salt nutrient broth, the cells took up g-fold more [i*C]putrescine after only 2 min and 15.fold more after 20 min.
Occasional low values in the plateau region (40 t,o 120 min, Fig. 5) are probably due to lysis of cells during filtration. The plateau itself might have been caused by dilution of isotope with unlabeled putrescine excreted from the cells. To eliminate this possibility cells were again centrifuged and resuspended in low salt medium at the time shown by the arrow (Fig. 5). No change in uptake occurred; therefore, isotope dilution was not significant.
Putrescine uptake was not affected by centrifugntion and resuspension in any part of the experiment.
Potassium Uptake in Bacteria with EIigh or Low Pufrescine Content-Because a sudden increase in the osmolarity of the medium causes both a rapid uptake of K+ (11) and rapid excretion of putrescine, K+ uptake might be influenced by the level of putrescine in the cell. It is possible to obtain cells growing logarithmically in 11 mM K+ medium which contain small amounts of putrescine.
If a culture in 11 maI K+ medium is made 0.4 M in NaCl, the cells rapidly excrete put,rescine and begin to grow logarithmically after a lag period of about 40 min. If these growing cells are centrifuged at 37" and resuspended in 11 mM K+ medium (no NaCl supplement), they continue log:trithmic growth, but the putrescine content remains at most 25 y0 of normal for more than 40 min.
Increases in cellular K+ of a control culture and a culture At the times shown (-40 to -20 min), aliquots were removed for optical density determination and incubation with /W]putrescine as described below.
At -15 min cells were centrifuged at 37" and then resuspended at 0 min in low salt nutrient broth.
At +59 min the culture was again centrifuged at 37" and at +70 min was resuspended in warm, low salt nutrient broth. Uptake was measured by incubating 1 ml of culture with 1 ml of high or low salt nutrient broth containing 2 pCi of [Wlputrescine for 3 min at 37".
The incubation mixture was filtered onto Schleicher and Schuell filters and washed with 10 ml of cold 0.15 M N&l; 30 set after adding cells to filters, the filters were placed in 5 ml of counting solution.
Washing the filters with cold 0.6 M NaCl did not alter the radioactivity recovered.
If the filters were allowec! to dry on the suction apparatus, they lost radioactivity.  (2,3), so that rapid fiuctuations in the putrescine pool should have only long term effects on the spermidine level. Maintenance of a high spermidine content in the presence of a reduced putrescine (precursor) pool might require an alteration in the 1 This concentration of sucrose was used to make the results comparable to those of Epstein and Schultz (11). High external Kf concentrations were used to assure that putrescine excretion would be rate limiting in the proposed coupling of K+ uptake and putrescine excretion. enzyme system which forms spermidine (2,5) or an alteration in the catabolism of spermidine.
To measure the catabolism of spermidine by E. coli B, a 30. ml logarithmically growing culture in low salt nutrient broth and a 30-ml logarithmically growing culture in high salt nutrient broth (0.6 M NaCI) were incubated at 37" for 20 min with 1 PCi of [%]spermidine (41.7 PCi per mg) and centrifuged at 2160 X g and 37" for 10 min. Then, the pellets from low salt and high salt nutrient broth were washed twice in 20 ml of low salt nutrient broth or high salt nutrient broth, respectively. They were next resuspended and incubated at 37" in 35 ml of low salt or high salt nutrient broth.
At various times 5-ml samples were removed, and polyamines were isolated and separated by chromatography.
Radioactivity in the paper strips was estimated with a radiochromatogram scanner.
In both cultures no detectable change in [14C]spermidine occurred over 70 min. This indicates that spermidine turnover was not increased in cells in in high salt nutrient broth.
Because the spermidine content was the same for both low and high salt cultures and spermidine was not metabolized in either type of culture, the rate of spermidine synthesis must parallel the generation time of cells in both media.
The enzyme system which forms spermidine is capable of maintaining a constant spermidine level even though the pre- This could come about by close regulation of enzyme activity or by rate-limiting amounts of enzyme being present in both cultures.
Also, the K, foi putrescine of the spermidine biosynthetic enzyme could be so low that the observed changes in putrescine concentration might not alter enzymatic activity. DISCUSSION We have shown that the putrescine content of E. coli varies inversely with the osmotic strength of the growth medium (Fig.  1). Medium containing glycerol was an exception.
We have also shown that the sudden addition of any of a variety of charged or uncharged solutes (but not glycerol) produces a rapid loss of intracellular putrescine (Figs. 2 and 3). It is unlikely that the rapid loss of intracellular putrescine which occurs following a sudden increase in the osmotic strength of the medium is due to passive leakage through a nonspecifically damaged membrane.
First, treatment of cells with toluene 01 butanol-1, which are known to damage cell membranes, releases both putrescine and spermidine from the cell (19). If high osmotic strength damaged the membrane similarly, it should bring about a release of spermidine as well as putrescine.
Second, if the membrane were damaged, the cell might lose its amino acid pools. However, addition of NaCl to cultures of E. coli and other gram-negative organisms does not cause loss of amino acids from the cellular pool (Ref. 17 and our Table I). Third, if the loss of putrescine were passive, this loss should not be blocked by metabolic inhibitors. However, metabolic inhibitors and cold did block loss of putrescine (Figs. 2 and 3).
Even with an intact membrane, increasing the osmotic strength of the medium could produce a rapid loss of cell water which might carry polyamines out of the cell. This decrease in cell volume can be observed as an increase in turbidity of the culture (20). In our experiments (Fig. 3) an increase in turbidity did occur without loss of putrescine.
The ratio of culture optical density to milligrams of protein increased 47y0 after the culture was made 0.6 M in NaCl; the ratio increased 697, in the presence of NaCl and sodium azide. Thus, azide-treated cells lost water without losing putrescine.
Only two agents, other than high pH (13) and chilling (3), have been reported to reduce the cellular putrescine content of E. coli without decreasing spermidine. These nonphysiologic agents, levorphanol (21) and streptomycin (14), were effective in E. coli 15 TAT:, which normally excretes much more putrestine into the medium than do strains B and K12 (22). The loss of substantial amounts of putrescine in the presence of levorphanol, streptomycin, or chilling took more than an hour, whereas our cells exposed to high osmotic strengths lost putrestine within 3 min (Figs. 2, 3, and 4).
Only a few studies have been published on the relationship of polyamine levels and salt concentrations in the medium. Hur-witz2 reported that the spermidine content of E'. coli grown in 0.01 mM Mg+Z was 20 times greater than that of cells grown in 1 mM Mg+2.
Hurwitz and Rosano (23) found that over the range of 0.001 to 10 mM Mg+2, the amount of spermidine bound to ribosomes varied inversely with the amount of ribosomebound Mg+2, and ribosome-bound putrescine was unchanged.
2 According to Cohen and ltaina (22) C. Hurwitz reported these studies in a paper given at the Los Angeles meeting of the American Society for Microbiology, May 1966.
Smith and Richards (24) reported another example of salt coilcentrations affecting polyamines; I(+-deficient barley and cabbage leaves and red clover plants contained more putrescine than the normal plants.
The experiments in our paper help to define the mechanism by which E. coli excretes putrescine after an increase in osmolarity.
B change in turgor pressure is a likely trigger for putrescine excretion.
Bolton et al. (25) found that increasing the osmotic strength of the medium produced a temporary decrease in turgor pressure of I<. coli and cessation of nucleic acid and protein synthesis.
Later, macromolecular syntheses resumed as the turgor pressure increased.
Of the various compounds used in our experiments, sucrose and the salts are known to penetrate the cell very slowly (20) and would be expected to reduce the turgor pressure of the cell. All of these agents produce rapid putrescine excretion in the presence of adequate K+. Glycerol, which does not cause putrescine excretion, penetrat,es the cell freely (9, 10) and would not be expected to lower the turgor pressure.
Thus, a change in turgor pressure appears to be a likely signal for the excretion of putrescine.
Epstein and Schultz (11) proposed that the cell achieves an increase in turgor pressure through an osmoregulatory pump which takes up K+ in exchange for Hf, the II+ being provided by an increase in metabolic acids. We have demonstrated that, in addition to uptake of K+, the cell also excretes putrescine after an increase in the osmolarity of the medium.
Fulthermore, the rapid excretion of putrescine is K+ dependent (Fig. 4). Lack of K+ in the medium caused a severe decrease in the rate of putrescine excretion.
Resupplementing K+ restored the high rate of putrescine excretion within a few minutes. Na+, NH4+, Rb+, and Mg2 did not substitute for K+ (Table II).
In fact, putrescine excretion may be coupled with the uptake of K+. The K, for potassium uptake after a sudden increase in the medium osmolarity is approximately 1 rnM (26); the Z<, for potassium stimulation of putrescine excretion is 0.69 nlM. Also, the rates of putrescine excretion (Fig. 4) and K+ uptake (Ref. 26 and Fig. 6) are similar.
Both K+ uptake (11) and putrescine excretion (Fig. 3) appear to require metabolic energy. However, it seems that K+ uptake is not dependent upon putrescine excretion. Fig. 6 shows that cells with reduced content of putrescine took up almost normal amounts of K+ following an increase in medium osmolarity.
Therefore, if cation excretion is necessary for K+ uptake and if putrescine facilitates this uptake, substitutes must exist when the intracellular putrescine concentration is low. Tempest et al. (17) have demonstrated that the concentrations of free amino acids, and particularly glutamate, increased when E. coli was transferred to high osmolarity medium.
Our observations confirm these data (Table I). Since glutamate is a precursor of putrescine, a reduction in putrescine synthesis could account for part of the increase in glutamate.
Pools of ornithine and arginine, direct precursors of putrescine, also increase under these conditions (Table I).
The data on putrescine excretion, K+ uptake (II), and glutamate increase can be integrated in the following hypothesis; shifts from low osmolarity media to high osmolarity media force gram-negative bacteria to suddenly increase their interior osmolarity in order to maintain a positive turgor pressure. The mechanisms for increasing the interior osmolarity include K+ uptake (11) and synthesis of amino acids, primarily glutamate (Ref. 17 and our Table I). After an increase in medium osmolarity, putrescine+2 escretion could be used by the cell to help balance the increase in internal positive charge from K+ uptake.
I'utrescine+ excretion would play an even more important role in reducing the large increases in internal ionic strength resulting from K+ uptake, with a minimum expense of osmotically active solute. l-'utrescine+2 excretion would reduce the net increase in internal positive charge by 28% and internal ionic strength by 56a/ while reducing the net increase in internal osmolarity by only 14y0 (calculated from Fig. 1 and Ref. 11 for an increase of 0.8 osmole per liter).
For an increase of 0.4 osmole per liter, the loss of internal ionic strength by putrescine+2 excretion would almost exactly balance the increase in ionic strength from K+ uptake.
Synthesis of negatively charged molecules, such as glutamate, would serve to maintain charge balance and would increase internal osmolarity at the same time. By coordinated K+ uptake, putrescine+2 excretion, and glutamate synthesis, the cell could achieve substantial increases in internal osmolarity while minimizing changes in internal positive charge and ionic strength.