Infection of Actinomycin-Permeable Mutants of

Transcription

1 JOURNAL OF VIROLOGY, Jan. 97, p Copyright 97 American Society for Microbiology Vol. 7, No. Printed in U.S.A. Infection of Actinomycin-Permeable Mutants of Escherichia coli with Urea-Disrupted Bacteriophage SHOKO IIDA AND MUTSUO SEKIGUCHI Departmenit of Biology, Faculty of Science, Kyushu University, Fukuoka, Japan Received for publication 2 October 97 Intact cells of actinomycin-permeable mutants of Escherichia coli could be infected with urea-disrupted phage T4 (designated as T47r). The parental strains and the revertants, which are impermeable to actinomycin, were not susceptible to T4ir unless they had been treated with agents which altered their permeability. The permeable mutants developed competence for 7r infection during the growth cycle; cells in the early stationary phase produced - to -fold more plaques on plating with T47r than did eponentially growing cells. Colistin (polymyin E) was effective in converting noncompetent cells of either permeable or nonpermeable strains to the competent state. Treatment with lysozyme resulted in a considerable increase in susceptibility to T4ir of permeable mutants but not of nonpermeable cells. It appears that development of competence for ir infection is mainly due to alterations in the permeability barriers of the cell. Normal cells of Escherichia coli cannot be infected with viral nucleic acid or subviral particles. However, when cells are converted to spheroplasts by treatment with ethylenediaminetetraacetate (EDTA) and lysozyme, they become susceptible to the infecting agents (3-5, 3, 4). Thus, changes in the permeability barrier of E. coli cells are required for infection with nucleic acid or subviral particles. In a previous paper (2), we showed that some strains of actinomycin-sensitive mutants of E. coli have anomalous permeability properties. The mutants are permeable to actinomycin and are readily lysed by lysozyine. It is known that normal E. coli cells are resistant to these agents unless they are treated with EDTA or chloroform to alter surface layers of the cell wall. It was suggested that the mutants are defective in some part of the cell wall structure, possibly in the outer lipoprotein or lipopolysaccharide layers, which serve as permeability barriers. There is a possibility that intact cells of actinomycin-permeable mutants can be infected with viral nucleic acid or subviral particles. In a preliminary eperiment, we showed that the mutant can be infected with OX74 deoyribonucleic acid (DNA), although the efficiency of infection is very low (2). In the present eperiments, we have studied conditions for infection of the mutants with urea-disrupted phage, a preparation of subviral particles (3). The results indicate that intact cells of the mutants, but not of wild-type E. coli, can be infected with ureatreated T4 phage. The efficiency of infection depended on the growth phase of cells and was enhanced by treatment with lysozyme (EC ) or colistin. MATERIALS AND METHODS Bacteria and phage. The isolation and properties of actinomycin-sensitive mutants of E. coli have been described previously (2). AS9 and AS27, actinomycin-permeable mutants, were isolated from strain B; ASI and AS2, which are sensitive but not permeable to the antibiotic, were isolated from strain AB57 of E. coli K-2. AS9R and AS27R are actinomycin-resistant revertants from AS9 and AS27, respectively. Bacteriophage T4B, obtained from S. Benzer, was used throughout the eperiments. Growth of organisms. The bacteria were grown in broth. The medium contained g of polypeptone, 5 g of meat etract, and 3 g of NaCl in liter of distilled water. The ph of the medium was adjusted to 7. with NaOH. Unless otherwise noted, an overnight culture of cells was diluted 2-fold with fresh broth and shaken at 37 C. Eponentially growing cells (logarithmicphase cells) were harvested after 3 hr of incubation; stationary phase cells, after hr. The density of the bacterial culture was measured in a Klett-Summerson photoelectric colorimeter with filter 66. Preparation of urea-disrupted phage. Urea-disrupted T4 phage, designated as T47r, was prepared by a modification of the method of Fraser et al. (3, 6). Solid urea was added to a suspension of T4B (ca. "Y plaque-forming units/ml) in broth to give a final urea concentration of 8 M. The miture was incubated at 37 C for 6 min and then dialyzed overnight in the cold against 3 liters of. M NaCl-.2 2 Downloaded from on November 5, 28 by guest

2 22 IIDA AND SEKIGUCHI J. VIROL. M sodium citrate (ph 7.). Such a preparation showed a residual titer against E. coli B of less than 3 plaque-forming units/ml. When phage was suspended in a salt solution [e.g.,. M tris(hydroymethyl)- aminomethane buffer, ph 8., or. M sodium citrate, ph 7.-2% serum albumin] instead of broth during incubation with urea, a considerably lowered activity of 7r was obtained. Determination of activity of 7. The ordinary phage techniques of Adams () were applied. A.-mi amount of an appropriately diluted suspension of T47 and.3 ml of bacterial culture were mied and immediately plated on 3 ml of soft agar. The plate was incubated at 37 C overnight, and the number of plaques formed was scored. Chemicals. Crystalline egg white lysozyme was purchased from Sigma Chemical Co. Colistin sulfate (polymyin E) is the product of the Kayaku Antibiotic Research Co., Ltd., Tokyo, Japan, and was kindly provided by Atsuo Nakata of Osaka University. RESULTS Competence of actinomycin-permeable mutants for 7r infection. There is a considerable difference in the susceptibility of various E. coli strains to the urea-disrupted phage T47r. As shown in Table, AS9 and AS27, permeable to actinomycin and sensitive to lysozyme (2), produced many plaques when they were plated with a preparation of T47r. The original actinomycin-resistant strains, B and AB 57, and the other type of actinomycinsensitive mutants, ASI and AS2, which are not permeable to actinomycin or sensitive to lysozyme (2), produced only a few plaques under the same conditions. In spite of the great difference in susceptibility to T47r, all of these strains plated intact T4 phage with amost equal efficiency. Of much interest is the finding that cells of the actinomycin-permeable strains in the early stationary phase are most competent for 7r infection. In the eperiment shown in Fig., an overnight culture of E. coli AS9 was diluted 2-fold with fresh broth and incubated at 37 C with shaking. At the intervals indicated,.3-ml portions of the culture were withdrawn and plated with. ml of a T4ir preparation. Only a few infective centers were formed when cells harvested at the beginning of the incubation were used for plating. The number of infective centers increased rapidly with time of incubation and reached the maimum level after about hr of incubation, at which time most of cells enter the stationary phase. The efficiency of plating with T47r of cells in the early stationary phase was 5 times that of cells in the early logarithmic phase. This presents a striking contrast to the finding that eponentially growing cells are more susceptible to intact T4 phage than are stationary-phase cells (Fig. ). TABLE. Comparison of susceptibility to T47r of various E. coli strains No. of Strain Sensitivity to Permeabil- plaques actinomycin ity to formed on actinomycin plating with T47r' B Resistant _ 4 AS9 Sensitive + 23 AS27 Sensitive + 8 AB57 Resistant < ASI Sensitive < AS2 Sensitive < a A.3-ml amount of overnight culture of cells was plated with. ml of T4r (original phage titer: 5.5 X 7/ml). c l 5 8 /4 2 g 6 _ 3 a. 5O E z 2, 5 5 Time of incubation (hr.) FIG.. Infection with T47r of E. coli ASJ9 cells at different stages of growth. An overnight culture of E. coli ASI9 was diluted 2-fold with fresh broth and incubated at 37 C with shaking. At the intervals indicated,.3 ml of the culture was withdrawn and plated with T47r. (@) T47r; () normal T4; (X) turbidity. Table 2 summarizes the results of additional eperiments in which three E. coli strains, harvested at two different phases of the growth cycle, were compared with respect to their abilities to produce plaques on plating with T47r. At all of the concentrations of cells eamined, AS9 and AS27 cells harvested during early stationary phase produced more plaques than did eponentially growing cells of the same strains. However, only a few plaques were produced by E. coli B, even when high concentrations of the stationary- Downloaded from on November 5, 28 by guest

3 VOL. 7, 97 PHAGE INFECTION OF E. COLI MUTANTS 23 phase cells were used for plating. From these results, it is evident that the susceptibility to T47r depends on an inherent permeability property of the strain, but is greatly affected by the growth phase of cells. There is a possibility that the competent culture contains some factor(s) which makes cells susceptible to T47r. To eamine this possibility, an eperiment was performed in which logarithmicand stationary-phase cultures of AS9 were centrifuged, and the cells and the supernatant fluids were echanged. The results indicated, however, that no such factor is present in the culture medium. Effects of colistin on 7r infection. It has been shown that treatment of E. coli with colistin causes alterations in the permeability of the cell; it induces a release from the cells of ultravioletabsorbing materials and renders the cells sensitive to actinomycin (7, 8, ). Recently, Nakata, Hara, and Kawamata (9) reported that colistintreated E. coli can be infected with T47r with high efficiency. Thus, we tested whether colistin is effective on AS9 cells, which are originally sensitive to T47r, in increasing the efficiency of infection with T47r. E. coli B and AS9 were grown in broth and harvested during logarithmic and early stationary phases of growth. Cells were suspended in broth containing various levels of colistin, incubated at 37 C for min, and then plated with an appropriate concentration of T47r. As shown in Table 3, treatment of cells with colistin resulted in TABLE 2. Infection with T47r of three strainis of E. coli in logarithmic and stationiary phases of growth Strain Growth phasea Concn of cellsb B Logarithmic 4c 4 4 Stationary 8 23 AS9 Logarithmic Stationary AS27 Logarithmic Stationary Bacterial cultures were withdrawn after 3 hr (logarithmic phase) or hr (stationary phase) of incubation, and the concentration of cells was adjusted to three levels. ' The turbidity of the culture was measured in a Klett-Summerson colorimeter with filter 66. A density of 25 corresponds to 9 viable cells/ml for a logarithmic culture of strain B. r Number of plaques formed on plating with T4,r. TABLE 3. Effect of colistin on intfectioni of E. coli with T47ra Strain Growth phase Concn of colistin (jug/ml) 5 B Logarithmic Stationary AS9 Logarithmic 7 4,2 Stationary 2,93 4,7 - a Cells (5 Klett units) were suspended in broth containing different levels of colistin and incubated at 37 C for min. A.3-ml sample of the cell suspension was plated with. ml of T47r. Results are epressed as the number of infective centers. a considerable increase in number of plaques, but the degree of increase differed considerably depending on the type of cells as well as the growth phase of the cultures. AS9 cells in logarithmic phase or B cells in either phase of growth, all of which have relatively low susceptibility to T47r, became more susceptible to T47r after treatment of the cells with colistin. On the other hand, AS9 cells harvested during the stationary phase of growth, which are originally highly susceptible to T47r, were little affected by colistin. Thus colistin is effective in converting the noncompetent cells to a competent state but has little effect on the cells which are already competent to T47r. Since the antibiotic is known to induce alterations in surface layers of the cell wall, it appears that AS9 cells undergo changes in permeability barriers, similar to those induced by colistin, during their growth cycle. Effects of lysozyme on 7r infection. The actinomycin-permeable mutants are sensitive to lysozyme, which attacks glycopeptide in a rigid layer of the cell wall (2). In the following eperiments, we investigated how lysozyme affects the susceptibility of cells to T47r. Table 4 summarizes the results of an eperiment in which the effects of lysozyme on the colonyand plaque-forming abilities of E. coli B and AS9 were compared. Judging from inactivation of the colony-forming ability, AS9 cells harvested during the logarithmic phase of growth and those in the stationary phase are equally sensitive to lysozyme. It was found that treatment of AS9 cells with lysozyme resulted in a considerable increase in the efficiency of infection with T47r; however, the same treatment impaired the formation of plaques by intact T4 phage in the same strain. The enhancement occurred both in eponentially growing cells and in stationary-phase Downloaded from on November 5, 28 by guest

4 24 IIDA AND SEKIGUCHI J. VIROL. TABLE 4. Effect of lysozyme oni colony-forming ability of cells atid plaque-formintg abilities of cells intfected with niormal T4 or T4r Strain of E. coli Concn of Iysozyme (/g,/ml) Colony-forming Pof cells infectedaith Pla(lue-forming ability of cells abiliy of ells" intact T4b infected wvith T4wb INo. of No. of colonies/ Relative infective plate value centers/ plate Relative No. of infective value centers plate Relative value AS9 (stationary phase) AS9 (logarithmic phase) B (logarithmic phase) Oi ca. 2, ca.,8 25 ca., ,2 a Diluted suspensions of bacteria were plated on nutrient plates which contain various amounts of lysozyme. I Lysozyme solution was added to soft agar kept at 5 C, a temperature at which no lysozyme activity was inactivated. Samples of.3 ml of bacterial cultures, adjusted to the density of 2, and.-ml amounts of appropriate concentrations of T4 or T47r were mied and plated with 3 ml of the soft agar. cells of AS9. Figure 2 indicates that AS9 cells, harvested during the two periods of growth, ehibit almost the same high efficiency of infection with T47r in the presence of lysozyme at concentrations of more than 3 Agg/ml. No such dramatic effects of lysozyme were observed with E. coli B; the susceptibility to T47r, as well as the colony- and the plaque-forming abilities of B, were not significantly affected by the treatment. There is a correlation between permeability to actinomycin and susceptibility to T47r in the presence of lysozyme. Actinomycinresistant revertants, AS9R and AS27R, did not respond to T47r even when they were eposed to a high concentration of the enzyme (Table 5). The relation between concentration of T47r and number of infective centers was studied by use of stationary-phase cells of AS9. The results are shown in Fig. 3. The efficiency of plating in the presence of lysozyme (3,ug/ml) was about 25- fold more than that in the absence of the enzyme. Within a certain range of T47r concentration, the number of infective centers formed was proportional to the amount of T47r suspension used, indicating that a single particle of 7r can produce a plaque and that this system can be used for quantitative determination of T47r. DISCUSSION The present eperiments demonstrate that intact cells of some strains of actinomycin-sensitive mutants can be infected with T47r, whereas ordinary E. coli cells cannot. This result had been anticipated, since it is known that the mutants have increased permeability and may be defective in outer layers of the cell wall, which prevent access of large or charged molecules to inner parts of the cell (2). However, it was beyond our epectation that a competent state of the cell is required for infection of the mutants with T47r, namely, that the most effective infection with T4wr occurs when cells are in the early stationary phase of the growth cycle. Development of competence of cells for transformation or transfection at certain growth periods has been demonstrated in E. coli and other bacterial systems (2, 5). One of the most plausible eplanations for the appearance of competence is that during the growth cycle there are some alterations in permeability barriers of the cells that facilitate more effective penetration of the infecting agent into the cell. Thus, the investigations were focused upon the effects of colistin and of lysozyme on the susceptibility of cells to T47r. It has been demonstrated that colistin induces alterations in Downloaded from on November 5, 28 by guest

5 VOL. 7, 97 PHAGE INFECTION OF E. COLI MUTANTS 25 TABLE 5. Inifection of various strains of E. coli with T47r in the absentce or presentce of lysozymea Infective centers a E U, v) B E z v Lysozyme ( pg/mi) FIG. 2. Effect of lysozyme oni 7r infection of different phases of cells. Samples of.3 ml of cell suspension (5 Klett units) of E. coli AS9 and.-mi amounts of T4'r were plated with 3 ml of sofi agar containing various concentrations of lysozyme. () Stationaryphase cells; () logarithmic-phase cells. the outer surface layers (7, 8, ) and that lysozyme attacks glycopeptide in a rigid layer of the cell wall (). Colistin is effective in converting the noncompetent cells of both B and AS9 to a competent state, suggesting that the development of competence in actinomycin-permeable mutants during the growth cycle is due to alterations in the surface layers of the cell wall. Lysozyme, on the other hand, increases the susceptibility to T47r of AS9 cells, in either phase of growth, but not of B. From these results, it appears that not only the outer layers but also the glycopeptide layer in the cell wall are obstacles to penetration of the infecting agent into the cell, and thus removal or alteration of these layers is necessary for the effective infection of cells with subviral particles. The reason for the enhancement of competence of the logarithmic-phase cells of permeable mutants by lysozyme is not clear at present, but it Strain Stationary-phase cells Logarithmic-phase cells No With No With lysozyme lysozyme lysozyme lysozyme B AS9, 4, 9 24, AS9R AS , 5 86, AS27R a The procedures were identical to those described in Fig. 2, and the concentration of lysozyme was 3 Ag/ml Relative concentration of L FIG. 3. Correlation betweenl concenltratiot of T4ir antd number of infective centers fbrmed. The procedures were identical to those described in Fig. 2. () In the presenice of lysozyme (3,ug/ml); () in the absence of lysozyme. may be eplained by the hypothesis that digestion of glycopeptide by lysozyme causes release of some components of the outer layers, which interfere with infection with T47r. There is a linear relationship between concentration of T47r and number of plaques formed on the actinomycin-permeable mutants, indicating that the procedure can be used for quantitative assay of T47r. In the presence of appropriate concentrations of lysozyme, a preparation of T47r -ii E N tn to U' tn -t > c Downloaded from on November 5, 28 by guest

6 26 IIDA AND SEKIGUCHI J. VIROL. produces plaques with an efficiency as high as - per original T4 titer, which is comparable to the value obtained by using the conventional spheroplast system. Moreover, the procedure has advantages in that there are no complications of preparing and handling fragile spheroplasts, and the ordinary phage techniques can be used. It would be of great interest to apply the procedure to the assay of infectious nucleic acid. Preliminary eperiments with 4X74 DNA and MS2 ribonucleic acid indicated, however, that more etensive studies are necessary to establish conditions for the effective infection of permeable mutants with free nucleic acid. ACKNOWLEDGMENTS We epress our gratitude to Yasuyuki Takagi, in whose lalboratory a part of this work was performlled. We also thank Atsuo Nakata for stimulating discussions. LITERATURE CITED. Adams, M. H Bacteriophages, p Interscience Publishers, Inc., New York. 2. Anagnostopoulos, C., and J. Spizizen. 96. Requirements for transformation in Bacillus subtilis. J. Bacteriol. 8: Fraser, D., H. R. Mahler, A. L. Shug, and C. A. Thomias, Jr The infection of sub-cellular Escherichia coli, straiin B, with a DNA preparation fromi T2 bacteriophage. Proc. Nat. Acad. Sci. U.S.A. 43: Guthrie, G. D., and R. L. Sinslheimiier. 96. Ilfectioni of protoplasts of Escherichia coli by subviral particles of bacteriophage 4X74. J. Mol. Biol. 2: Hofschneider, P. H. 96. Uber ein infektbises Deoyribonucleinsaure-Agens aus dem phagen lx74. Z. Naturforsch. 5b: Mahler, H. R., and D. Fraser Studies in partially resolved bacteriophage-host systems. IV. Some properties of the protoplast-infecting agent derived from T2 bacteriophage. Virology 8: Nakajima, K., and J. Kawamata Effect of colistin on actinomycin sensitivity of Escherichia coli. Biken J. 8: Nakajima, K., and J. Kawamata Studies on the mechanism of action of colistin. II. Alteration of permeability of Escherichia coli by colistin. Biken J. 8: Nakata, A., K. Hara, and J. Kawamata. 97. Infection of polymyin E-treated Escherichia coli with urea-disrupted bacteriophage T4, p In Progress in antimicrobial and anticancer chemotherapy (Proc. 6th Int. Congr. Chemother.), University of Tokyo Press, Tokyo.. Newton, B. A The properties and mode of action of the polymyins. Bacteriol. Rev. 2: Repaske, R Lysis of gram-negative organisms and the role of Versene. Biochim. Biophys. Acta 3: Sekiguchi, M., and S. lida Mutants of Escherichia coli permeable to actinomycin. Proc. Nat. Acad. Sci. U.S.A. 58: Sekiguchi, M., A. Taketo, and Y. Takagi. 96. An infective deoyribonucleic acid from bacteriophage OX Biochim. Biophys. Acta 45: Spizizen. J Infection of protoplasts by disrupted T2 virus. Proc. Nat. Acad. Sci. U.S.A. 43: Taketo, A., and S. Kuno Sensitivity of Escherichia coli to viral nucleic acid. It. Development of competence in Escherichia coli CT during growth in a hypertonic medium. J. Biochem. 65: Downloaded from on November 5, 28 by guest