Structure, and Assembly of the Cell Wall

Transcription

1 JOURNAL OF BACTERIOLOGY, August, American Society for Microbiology Role of Multivalent Cations in the Organization, Structure, and Assembly of the Cell Wall of Pseudomonas aeruginosa MARY A. ASBELL AND R. G. EAGON Department of Bacteriology, University of Georgia, Athens, Georgia Received for publication 29 March 1966 Vol. 92, No. 2 Printed in U.S.A. ABSTRACT ASBELL, MARY A. (University of Georgia, Athens), AND R. G. EAGON. Role of multivalent cations in the organization, structure, and assembly of the cell wall of Pseudomonas aeruginosa. J. BacterioL 92: Incubation of Pseudomonas aeruginosa with ethylenediaminetetraacetate induced the formation of osmotically fragile rods termed osmoplasts. These could be restored to osmotically stable forms by multivalent cations. Only those cells restored by divalent cations normally found in the cell wall were capable of multiplication. The respiration of restored cells, however, was unimpaired, irrespective of whether they were capable of multiplication. Moreover, the permeability characteristics of osmoplasts and restored cells were unimpaired. When multivalent cations were chelated from the cell wall and replaced by sodium, a weakened cell wall and an osmotically fragile cell resulted. This was apparently caused by the absence of cross-linkages in the cell wall via multivalent cations. Tris(hydroxymethyl)aminomethane buffer compounded the lethal effects of ethylenediaminetetraacetate. The lipopolysaccharide component was inferred to be the site of attack by ethylenediaminetetraacetate. A mechanism for the synthesis of the lipopolysaccharide sacculus was proposed whereby negatively charged subunits are "trapped" by forming ionic and coordinate bonds intermediated by multivalent cations. Recently, Eagon and Carson (7) and Gray and Wilkinson (10) reported that Pseudomonas aeruginosa dies rapidly in the presence of ethylenediaminetetraacetate (EDTA). Evidence was presented by both groups that the structural integrity of the cell wall is impaired The former authors concluded, moreover, that the binding of divalent cations is essential for the integrity of the cell wall of this microorganism Furthermore, Eagon, Simmons, and Carson (8) reported that Ca++, Mg++, and Zn++ are components of the cell wall of P. aeruginosa. On the other hand, evidence was presented by Carson and Eagon (6) that the mucopeptide component is not solely responsible for the structural integrity of the cell wall of P. aeruginosa. The investigations described in this paper, therefore, were undertaken to provide further information on the role of multivalent cations in the organization and structure of the cell wall of P. aeruginosa. A role for multivalent cations in the nonenzymatic assembly of subunits of lipopolysaccharide or lipoprotein components of the cell 380 wall, or both, was inferred as well from results of our investigations. A preliminary report of this work has been published (2). MATERIALS AND METHODS Cultivation of organism. P. aeruginosa strain OSU 64 was cultivated at 37 C on a rotary shaker in a basal salts-glucose-yeast extract medium as previously described (6). Cells were harvested after 14 to 16 hr of cultivation and were washed with distilled water before use. Experimental procedures. Cells of P. aeruginiosa were incubated with EDTA, tris(hydroxymethyl)- aminomethane (Tris) buffer, and lysozyme according to techniques previously reported (7). Modifications of these techniques are described as appropriate in the protocols of the figures and tables. RESULTS Descriptions of terms used. Microscopic examination of cells of P. aeruginosa incubated with EDTA or with EDTA plus lysozyme in a hypertonic sucrose solution revealed rod-shaped bacilli indistinguishable from normal cells; these were,

2 VOL. 92, 1966 CELL WALL OF P. AERUGINOSA 381 however, osmotically fragile. Voss (21) has previously observed this phenomenon. Thus, we have proposed the term, osmoplast, to describe these osmotically fragile rods (2). Restored cells are defined as the osmotically stable cells resulting from the incubation of osmoplasts with multivalent cations. Restoration of EDTA-induced osmoplasts by multivalent cations. The experimental results depicted in Table 1 clearly indicate that EDTA-induced osmoplasts were restored to an osmotically stable state by the addition of a variety of multivalent cations, as evidenced by the absence of lysis when the restored cells were removed from hypertonic sucrose and resuspended in water. Whereas 10 ;umoles of cations per ml was used in several early experiments to ensure an excess of cations, it was evident that I ;umole of cations per ml was equally effective. Lower quantities than this were not used. The data indicate that all multivalent cations employed in these experiments restored osmoplasts to an osmotically stable state irrespective of whether they were di-, tri-, or tetravalent cations. Similarly, neither the variations in the ionic radii of the multivalent cations nor whether they were hard or soft Lewis acids, as defined by Pearson (16), appeared to affect the process. Finally, cations that are not normally found in the cell wall of P. aeruginosa were also able to restore osmoplasts to the osmotically stable state. For example, Mn i is known not to be a component of the cell wall of P. aeruginosa (8). It is unlikely, however, that many of the other cations used occur naturally in the cell wall. The data in Table 1 also indicate that monovalent cations were not able to restore osmoplasts even when used in high concentrations. Finally, lysozyme alone was not effective in inducing osmoplast formation, confirming previous reports (6, 7). When this reagent was used in combination with EDTA, however, irreversible osmoplasts were formed which could not be restored with Ca++ nor, presumably, with other multivalent cations either. This indicates that the damage done to the cell wall by the combination of EDTA and lysozyme was too extensive for repair by multivalent cations. It is considered unlikely, moreover, that the sensitive linkages cleaved by lysozyme could be repaired by multivalent cations. Effect of cations and of composition of medium on the viability of restored osmoplasts of P. aeruginosa. The results shown in Table 2 indicate that cells restored from osmoplasts prepared in a reaction mixture buffered with Tris buffer were not able to multiply extensively, as evidenced by their failure to form colonies. The highest percentage of survivors most consistently noted was achieved when a mixture of Ca++, Mg++, and Zn++ in the same ratio as found in the cell wall (8) was used TABLE 1. Restoration of ED TA-induced osmoplasts of Pseudomonas aeruginosa by multivalent cationsr System Optical density (660 my) Complete EDTA.. i EDTA + 3O0jg of lysozyme ug of lysozyme ,umoles of Na ,umoles of Li pAmoles ofk psmoles of Ba pmole of Ca ,umoles of Ca rmole ofcd ,umole of Co ,umole of Cu ,umoles of Fe b + 1 samole of HG pmoles of Mg ,umole of Mni ,umoles of Mn amole of Ni ,moles of Sr ,Amoles of UO lmoles of VO ;moles of Zn Amoles of Bi b + 10 umoles of Fe IpAmole of Las pamoles of La ,umole of Pt!! * ,umole of CaF; 1.3 pamoles of MgF+; 0.6 pmole of Zn basal salts lmoles of Ca4; 30,g of lysozyme a Protocol: Complete system contained 1,umole of EDTA (ph 8), 33 pmoles of Tris buffer (ph 8), and 2.9 X 10' cells per milliliter 0.55 M sucrose. After incubation of the complete system for 10 min at room temperature, cations were added to give a final concentration per milliliter as indicated above. The resulting reaction mixtures were then incubated for an additional 10 min. Restored cells or osmoplasts were collected by centrifugation at 3,500 X g for 15 min and resuspended in water. Lysozyme was added to complete system where indicated at time zero to give a final concentration of 30,g/ml. Basal salts were the same multivalent cations in the same final concentration as used in media for propagation of P. aeruginosa (i.e., Ca++, 1 pamole; Fe++, 0.07 pmole; Mg+, 0.5 Amole; Mn++, 0.01,umole; and ZnH, 0.05 ;pmole). b High readings were due to colored solutions or to insoluble materials.

3 382 ASBELL AND EAGON J. BAC-1 RIOL. TABLE 2. Effect of cations and of composition of medium on viability of restored osmoplasts of Pseudomonas aeruginosaa No. of bacteria/ml Per cent ~~~survivors in Nutrient agar Basal salts + 1% glucose Nutrient agar + 2.5% NaCl Penassay medium medium Complete 1 X X x X X X X No EDTA 3.3 X X X X X X X Mumole of Ca X X X X X X X I gemole of Ca++; 1.3,umoles of Mg++; 0.6,umoles of Zn 9.3 X X X X X X X ,mole of Mn X X X X X X X jamole of La' 3.6 X X X 105 5X X X a Protocol: Experimental conditions were the same as described for Table 1. The reaction mixtures were diluted into sterile water blanks and numbers of bacteria were estimated by the plate count method. The cultures were incubated at 35 C for 48 to 72 hr. Results from two experiments are recorded, representing high and low values. to restore osmoplasts. For example, results from several additional experiments with a variety of media indicated an average of 35% more survivors from osmoplasts restored with Ca++, Mg++, and Zn++ than from osmoplasts restored with Ca++ alone. Similarly, this effect has been previously noted (2). The percentage of survivors of osmoplasts restored with either Mn++ or La++ was not greater than those survivors observed in the unrestored system. It is noteworthy that Mn++ has not been detected in the cell wall of P. aeruginosa (8), whereas La+++ is assumed not to be a component of the cell wall. The results shown in Table 2 also indicate that the greatest viability occurred in culture media containing salts. The percentage of survivors from the complete, unrestored system and from osmoplasts restored with Ca++ or with a mixture of Mg++, and Zn++, for example, was increased approximately 10-fold when cultivated in media containing salts as compared with survivors cultivated in nutrient agar. Normal cells (i.e., no EDTA) were not affected by salts. On the other hand, composition of the medium other than salts was unimportant, since approximately equal numbers of colonies were observed in glucosebasal salts medium and in Penassay agar. The latter is an enriched medium containing 0.75% NaCl and phosphate salts. It was prepared by solidifying Penassay Broth (Difco) with 1.5% agar. Thus, it is suggestive that the salts gave osmotic protection. Evidence for the lethal effect of the EDTA- Tris buffer combination. When osmoplasts were formed and restored in a system containing phosphate buffer, there was a several-fold increase in percentage of survivors in comparison with survivors from systems containing Tris buffer (Table 3). Thus, Tris buffer contributed to the lethal effects of EDTA. Similarly, the lethal effect of the EDTA-Tris combination for cells and cysts of Azotobacter vinelandii has been reported recently (9). On the other hand, we noted no toxicity of Tris buffer in the absence of EDTA. Evidence for the competition of Na+ with multivalent cations. The data in Table 4 show that material absorbing at 260 m,a was released into the hypertonic environment when osmoplasts were formed and restored. Similarly, additional material absorbing at 260 m, was released into water in which restored cells had been resuspended. From these experiments, it could not be discerned whether this released material represented a small amount of lysis of all osmoplasts and restored cells or whether a few cells had been lysed extensively. We feel that the latter is the most

4 VOL. 92, 1966 probable explanation, for when sucrose was used as the hypertonic agent, electron microscopic observations of restored cells indicated that approximately 10 to 15% of these cells had been lysed, TABLE 3. Increased viability from cation restoration by the use of phosphate buffer for the preparation of osmoplasts of Pseudomonas aeruginosaa No. of System bacteria/ml (Penassay survivors Per cent medium) Complete 3.5 X X X No EDTA 3.5 X X X I klmole of Ca++; 1.3,umole of Mg++; X /Amole of Zn X X jamole of Ca X a Protocol: Experimental conditions were the same as described for Tables 1 and 2 with the exception that 33 jumoles of phosphate buffer (ph 8) were used per ml in place of Tris buffer for the preparation of osmoplasts, and the time of incubation was 30 min. Results of three experiments are indicated with the exception of Ca++ restorat ion. TABLE 4. CELL WALL OF P. AERUGINOSA as evidenced by numbers of fragments of cell walls and ghost cells compared to intact restored cells. Similarly, Gray and Wilkinson (10) reported that the permeability characteristics of the cell membrane of P. aeruginosa were unimpaired by EDTA. The 0.5 M NaCl environment appeared to give better osmotic protection than the 0.55 M sucrose environment, as evidenced by the weaker absorption at 260 m,u demonstrated by the NaCl supernatant fluids (Table 4). On the other hand, osmoplasts formed in the presence of 0.5 M NaCl could not be restored with multivalent cations. Evidence for lysis is shown by the decrease in optical density at 660 m,u when these osmoplasts were resuspended in water and by the amount of material absorbing at 260 m,u which was released into the water environment. Evidence for the unimpaired respiration of restored cells. That restored cells are capable of respiration is shown by data in Fig. 1. Osmoplasts restored either by 1 j,mole of Ca++ per ml or by the mixture of divalent cations used in the basal salts solution took up oxygen at a lower rate and to a lesser extent than did normal cells. Osmoplasts restored with 10,umoles of Ca++ per ml aggregated into large spherules (approximately 1 to 2 mm in diameter) when added to phosphate buffer in the Warburg vessel, possibly due to the formation of a calcium-phosphate complex at the cellular surface. Thus, the slow rate of oxygen uptake may have been due to slow diffusion of glucose throughout these spherules. Nevertheless, these data show that the respiration of restored cells was not extensively impaired, especially when Failure of cation restoration of osmoplasts of Pseudomonas aeruginosa formed in hypertonic NaCL solution and detection of substances absorbin-r at 260 mu in all svstemsa Sucrose (0.55 i) as hypertonic solution NaCl (0.5 s) as hypertonic solution System Supernatant fluid Supernatant fluid Cell Cell suspension S suspension NaCl Water Sucrose Water NC ae Complete > > 2 No EDTA Amole of Ca++; 1.3,umoles of Mg++; 0.6 jamoles of Zn >2 + 1 Amole of MoO O >2 + 1 iumole of La a Protocol: Experimental conditions were the same as described for Table 1 except that the reactions were carried out in 0.5 M NaCl for one series of experiments. Sucrose and NaCl supernatant fluids were the supernatant fluids resulting from removal of osmoplasts and restored cells by centrifugation at 3,500 X g for 15 min. These sedimented osmoplasts and restored cells were then resuspended in an equal volume of water, incubated for 15 min at room temperature, and again sedimented by centrifugation. The resulting supernatant fluids are indicated as the water supernatant above. Results are expressed as optical density at 660 in/a for the cell suspensions and at 260 mnu for the supernatant fluids. 383

5 384 ASBELL AND EAGON J. BACTERIOL. it is considered that 10 to 15% of the cells may have been lysed during these experimental manipulations. Evidence that restored cells can form an induced permease system and that cell membrane permeability is not impaired by EDTA. The data in Fig. 2 compare the rate of formation of an induced permease to citrate and of citrate dissimilation by resting normal cells and by restored cells. It is apparent that permease induction required approximately the same time interval in each case. Neither formation of an induced permease to citrate nor citrate dissimilation by osmoplasts suspended in either the 0.55 M sucrose or in the 0.5 M NaCl systems could be demonstrated even after the addition of cations. This suggests that the hypertonic environment may have been inimical to citrate dissimilation. The low rate of oxygen uptake by the lysed system and by systems restored with Zn++ and La-l w, on the other hand, appeared to be related to the number of surviving, viable cells and not to release of enzymes to the medium. 480 NORMAL 400 BASAL SALTS I'Ole Co++ w y 320 CY240- O umole Co TIME, HOURS FIG. 1. Comparison of rate and extent of glucose dissimilation by resting normal cells and by restored cells of Pseudomonas aeruginosa. Experimental procedures for the preparation of restored cells were the same as described for Table 1. Normal cells are defined as those cells that were handled in the same manner as restored cells but were not exposed to EDTA or to cations. Oxygen uptake was measured by the Warburg respirometer at 30 C. Each Warburg vessel contained 4,Amoles ofglucose, 800 pnoles ofphosphate buffer (ph 7), 2 ml of cell suspension (4.1 X 109 cells per milliliter, and 0.2 ml of 40% KOH in the center well, to give a total volume of 3.2 ml. Endogenous activity was substracted. 150 ~~~~~6,u MOLE Zn@++ 1,0MOLE ZnM 0 LYSED 100 -IpMOLE La'++ 50 ITAlMOLECo" 6,p MOLE Zn++ ~~OSMOPLASTSE 0~~~~~~~~ TIME, HOURS FIG. 2. Comparison of rate of formation of an induced permease to citrate and of citrate dissimilation by resting normal cells and by restored cells of Pseudomonas aeruginiosa. Experimental conditions were the same as described for Fig. 1, except that osmoplasts were prepared in phosphate buffer, ph 8 (33,umoles/ml) and that 4 umoles of citrate were used as substrate. The lysed system indicated above was composed of lysed osmoplasts that had been resuspended in an equal volume of water. Endogenous activity was subtracted. The data presented represent average values from three separate experiments. We also observed that endogenous respiration of osmoplasts suspended in these hypertonic solutions was unimpaired, however, as compared with intact cells suspended in the same hypertonic environment. These findings compare favorably, therefore, with those reported by Smith (20), who examined the effect of external sucrose concentration on the respiration of spheroplasts of Bacillus subtilis. She concluded that electron transport was not inhibited by the stretching of the cytoplasmic membrane after changes in osmotic pressure within the spheroplast. These results, moreover, provide additional evidence that the permeability characteristics of the cell membrane were unimpaired by EDTA. DIscussIoN We have shown that incubation of P. aeruginosa with EDTA inducesthe formation of osmoplasts (i.e., osmotically fragile rods). These can be restored to osmotically stable forms by multivalent cations. Only those cells that are restored with divalent cations normally found in the cell wall are capable of multiplication. This suggests that the inability of alien cations to restore osmoplasts may be the result either of their toxicity or of the conformational changes which render the cell wall incapable of further metabolic interactions and of replication.

6 VOL. 92, 1966 CELL WALL OF P. AERUGINOSA 385 Osmoplasts so prepared are capable of endogenous respiration but incapable of forming an induced permease to citrate. This is a clear indication that EDTA does not alter the permeability characteristics of P. aeruginosa. This is in agreement with previous observations by Gray and Wilkinson (10) but in opposition to those by Leive (13) for Escherichtia coli. We have also presented evidence, however, that specific divalent cations are necessary for permease induction. Thus, osmoplasts restored with Ca++, Mg++, or Mn++, or a combination of Ca++, Mg++, and Zn+, form an induced permease to citrate. Those restored with La or Zn i, however, show no greater permease induction than the unrestored, lysed system. Osmoplasts restored with Mn++, which is not a component of the cell wall, are capable of permease induction. Whether Mn++ is a component of the cell membrane cannot be discerned from these experiments. On the other hand, Mn++ may be able to substitute for Ca++ and Mg++ in the cell membrane. Finally, osmoplasts restored with Zn++ do not form the induced permease even though this cation is a component of the cell wall. These results cannot be fully interpreted at this time. The fact that osmoplasts failed to form an induced permease may have resulted from effects of the change in osmotic pressure within the osmoplast, resulting in stretching of the cytoplasmic membrane. On the other hand, Ca++ and Mg++ are considered to be present in cell membranes of all cells in nature (19). Similarly, Brown (4) presented evidence for the presence of Mg++ in the cytoplasmic membrane of Sarcina lutea. These divalent cations are thought to provide for a stiffening mechanism for lipoprotein membranes by forming salt bridges between neighboring -COO- groups (3) and to increase surface potentials by a contractile effect due to intra- and intermolecular interactions with phospholipids (19). Thus, it is possible that chelation of divalent cations from the cytoplasmic membrane of P. aeruginosa results in a conformationally altered structure which is no longer capable of forming induced permeases. Many of the cations found to be deleterious to multiplication of restored cells are present in the normal environment. Thus, a high degree of selectivity by the cells for specific cations is indicated. The respiration ofrestored cells was found to be unimpaired, irrespective of whether they were capable of multiplication. Thus, the factors causing inhibition of permease induction and of multiplication appeared not to affect glucose dissimilation and electron transport. Finally, we have confirmed the observation by Goldschmidt and Wyss (9) of the lethal effect of the EDTA-Tris combination. Examination of restored cells with a light microscope revealed bacilli indistinguishable from normal cells. Preliminary examinations with the electron microscope, on the other hand, revealed alterations of the external morphology of restored cells when compared with normal cells. Cells restored with cations which have been shown to be normal components of the cell wall (i.e., Ca++, Mg++, Zn++) appear more similar to normal cells than those cells restored with "alien" cations (e.g., Mn++ or La± 1 1). A study of the morphology of restored cells is currently under investigation. Osmoplasts formed in a hypertonic solution of NaCl could not be restored with multivalent cations. This observation was interpreted to indicate that Na+ reacts with negative charges exposed when EDTA chelates multivalent cations from the cell wall. Thus, these negative charges are "screened" by Na+ and, therefore, are no longer free to react with added multivalent cations. Since the monovalent Na+ cannot form cross-linkages, a weakened cell wall results. Our observations may also explain why E. coli will form spheroplasts when incubated in a growth medium containing LiCl (17). If the Li+ were able to compete with multivalent cations for negative charges on the replicating cell wall, a weakened structure would result due to the absence of cross-linkages via multivalent cations within the cell wall. Sensitivity to EDTA is not restricted to P. aeruginosa. Gray and Wilkinson (10, 11) reported that, in addition to P. aeruginosa, Alcaligenes faecalis is highly sensitive to EDTA and E. coli is moderately sensitive. Lipopolysaccharide is solubilized by this reagent (11). These authors also indicated that EDTA sensitizes a wide variety of bacterial species to chloroxylenol preparations or potentiates its effects. Leive (14) reported that EDTA liberates lipopolysaccharide from E. coli. Unpublished observations in this laboratory indicate that lipopolysaccharide per se from P. aeruginosa is altered by EDTA. Burton and Carter (5) detected Ca++ and Mg++ in the lipid A component of lipopolysaccharide of E. coli. Divalent cations have also been implicated in the structural organization of lipoprotein-type membranes. Abram and Gibbons (1), Brown (3), and Kushner and Onishi (12) concluded that the lipoprotein cell walls of the halophilic Halobacterium may be held together in part by divalent cations. Razin, Morowitz, and Terry (18) reported that the lipoprotein cell membranes of the pleuropneumonia-like organism Mycoplasma can

7 386 ASBELL AND EAGON J. BACTERIOL. be dissolved into subunits with sodium lauryl sulfate. These subunits can be reaggregated to form membrane-like structures in the presence of multivalent cations. Similarly, Onishi and Kushner (15) found that salts added to dissociated envelopes of H. cutirubrum lead to a nonspecific aggregation. Thus, there is sufficient evidence to postulate a unitary role for the association of divalent cations with the lipopolysaccharide and lipoprotein components of cell walls of gram-negative bacteria. The results of our experiments suggest that a component of the cell wall of P. aeruginosa is attacked by EDTA. There is a strong inference from our studies and from those of Gray and Wilkinson (11) that the lipopolysaccharide component is liberated. Thus, the lipopolysaccharide may be composed of subunits cross-linked via divalent cations. The lipopolysaccharide layer may also be cross-linked to other components of the cell wall via multivalent cations. Similarly, hydrophobic attraction to other cell wall components is likely as well. It is probable that divalent cations are associated with phospholipids as proposed for other systems (3, 12, 19). In this respect, a preliminary study (unpublished data) of the phospholipids of the cell wall of P. aeruginosa revealed phosphatidylethanolamine as the major component. There is also evidence that diphosphatidylglycerol is a component of the cell wall. Finally, it may be speculated that the formation of a complete bacterial cell wall sacculus containing lipopolysaccharide and lipoprotein may be formed in vivo via physicochemical properties by which negatively charged subunits are "trapped" by forming ionic and coordinate bonds intermediated by multivalent cations. ACKNOWLEDGMENTS This investigation was supported by Public Health Service pre-doctoral fellowship 1-FM-GM-30, from the National Institute of General Medical Services awarded to the senior author and by Public Health Service research grant AI from the National Institute of Allergy and Infectious Diseases awarded to the junior author. LITERATURE CITED 1. ABRAM, D., AND N. E. GIBBONS The effect of chlorides of monovalent cations, urea, detergents and heat on morphology and turbidity of suspensions of red halophilic bacteria. Can. J. Microbiol. 7: ASBELL, M. A., AND R. G. EAGON The role of multivalent cations in the organization and structure of bacterial cell walls. Biochem. Biophys. Res. Commun. 22: BROWN, A. D Aspects of bacterial response to the ionic environment. Bacteriol. Rev. 28: BROWN, A. D Evidence for a magnesiumdependent dissociation of bacterial cytoplasmic membrane particles. Biochim. Biophys. Acta 94: BURTON, A. J., AND H. E. CARTER Purification and characterization of the lipid A component of the lipopolysaccharides from Escherichia coli. Biochemistry 3: CARSON, K. J., AND R. G. EAGON Further evidence for the role of the non-peptidoglycan components in cell wall rigidity. Can. J. Microbiol. 12: EAGON, R. G., AND K. J. CARSON Lysis of cell walls and intact cells of Pseudomonas aeruginosa by ethylenediamine tetraacetic acid and by lysozyme. Can. J. Microbiol. 11: EAGON, R. G., G. P. SIMMoNS, AND K. J. CARSON Evidence for the presence of ash and divalent metals in the cell wall of Pseudomonas aeruginosa. Can. J. Microbiol. 11: GOLDSCHMIDT, M. C., AND 0. WYss Chelation effects on Azotobacter cells and cysts. J. Bacteriol. 91: GRAY, G. W., AND S. G. WILKINSON The action of ethylenediaminetetra-acetic acid on Pseudomonas aeruginosa. J. Appl. Bacteriol. 28: GRAY, G. W., AND S. G. WILKINSON The effect of ethylenediaminetetra-acetic acid on the cell walls of some gram-negative bacteria. J. Gen. Microbiol. 39: KUSHNER, D. J., AND H. ONISHI Contribution of protein and lipid components to the salt response of envelopes of an extremely halophilic bacterium. J. Bacteriol. 91: LEIvE, L A nonspecific increase in permeability in Escherichia coli produced by EDTA. Proc. Natl. Acad. Sci. U.S. 53: LEIVE, L Release of lipopolysaccharide by EDTA treatment of E. coli. Biochem. Biophys. Res. Commun. 21: ONISHI, H., AND D. J. KUSHNER Mechanism of dissolution of envelopes of the extreme halophile Halobacterium. J. Bacteriol. 91: PEARSON, R. G Acids and bases. Science 151: PITZURRA, M., AND W. SZYBALSKI Formation and multiplication of spheroplasts of Escherichia coli in the presence of lithium chloride. J. Bacteriol. 77: RAZIN, S., H. J. MOROWITZ, AND T. M. TERRY Membrane subunits of Mycoplasma laidlawii and their assembly to membrane-like structures. Proc. Natl. Acad. Sci. U.S. 54: SHAH, D. 0., AND J. H. SCHULMAN Binding of metal ions to monolayers of lecithins, plas-

8 VOL. 92, 1966 CELL WALL OF P. AERUGINOSA 387 malogen, cardiolipin, and diacetyl phosphate. J. Lipid Res. 6: SMITH, L Structure of the bacterial respiratory chain system. Respiration of Bacillus subtilis spheroplasts as a function of osmotic pressure of the medium. Biochim. Biophys. Acta 62: Voss, J. G Lysozyme lysis of gram-negative bacteria without production of spheroplasts. J. Gen. Microbiol. 35: