Lipopolysaccharides of Pseudomonas aeruginosa - Antimicrobial ...

ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Apr. 1993, p. 715-721

Vol. 37, No. 4

0066-4804/93/040715-07$02.00/0 Copyright ©) 1993, American Society for Microbiology

Interaction of Gentamicin with the A Band and B Band Lipopolysaccharides of Pseudomonas aeruginosa and Its Possible Lethal Effect JAGATH L. KADURUGAMUWA,* JOSEPH S. LAM, AND TERRY J. BEVERIDGE Center for Canadian Bacterial Diseases Network, Department of Microbiology, University of Guelph, Guelph, Ontanio NIG 2W1, Canada Received 4 November 1992/Accepted 25 January 1993

The lipopolysaccharide (LPS) of Pseudomonas aeruginosa PA01 possesses two distinct types of 0 polysaccharide, A and B band LPSs, but the majority of clinical isolates from cystic fibrosis patients who are infected with the organism possess only the A band as the major LPS antigen. The initial step in a series of events during the uptake of aminoglycoside antibiotics such as gentamicin is the ionic binding of the molecule to the cell surface. In an attempt to elucidate the role of A and B band LPSs of P. aeruginosa in this passive ionic binding of gentamicin to the outer membrane and its possible lethal effects, strains PA01 (A' B+) and LPS isogenic derivatives (A' B-, A- B+, A- B-) were treated with the antibiotic. Ionic binding of gentamicin appeared to be subtly different in PAO1 and its LPS derivatives; a lethal dose of drug was bound to all strains, although the degree of binding varied with each strain. The outer membrane affinity for gentamicin was higher in strains possessing the B band than in strains with A band LPS, and these B band strains were more prone to antibiotic-induced killing. Strains with both A and B band LPSs bound the most gentamicin of all strains, and this binding caused an almost 50%o loss in viability. Ionic binding of aminoglycoside antibiotics to the outer membrane of cell surfaces must not only weaken the cell surface (R. E. W. Hancock, Annu. Rev. Microbiol. 38:237-264, 1984; N. L. Martin and T. J. Beveridge, Antimicrob. Agents Chemother. 29:1079-1087, 1986; S. G. Walker and T. J. Beveridge, Can. J. Microbiol. 34:12-18, 1988) but it must also be more important in cell death than was originally thought.

Uptake of aminoglycoside antibiotics such as gentamicin into Pseudomonas aerugino2a is a multifactorial process that involves ionic interaction with the exterior of the cell and then two energy-dependent phases that involve an energized cytoplasmic membrane and electron transport (8, 47). It is generally accepted that the first, crucial step in the self-promoted uptake of gentamicin is a rapid electrostatic binding to the outer membrane which is a passive, nonenergy-dependent event initiated by an interaction between cationic moieties of the antibiotic and negatively charged residues in the lipopolysaccharide (LPS) of the bacterial membrane (15, 21, 32, 38, 47, 50). The polycationic antibiotic competitively displaces the essential divalent cations Mg2+ and Ca2+ which cross-bridge adjacent LPS molecules (14, 35, 39). This binding can so rearrange the LPS packing order that it results in outer membrane blebbing, the formation of transient holes in the cell wall, and the disruption of the wall's innate permeability to antibiotics (32, 45). Disruption of outer membrane permeability by this mechanism has been implicated as providing a self-promoted pathway for the aminoglycoside antibiotic's own uptake (14, 21, 31, 32, 39, 50). The ionic binding of gentamicin is related to drug concentration, and the postantibiotic effect has been shown to be bactericidal (31). The relationship of ionic binding to the drug concentration-dependent bactericidal action dictates the clinical selection of the most effective and least toxic drug regimens for treatment (21). Most P. aeruginosa strains possess two chemically and immunologically distinct types of LPS 0 polysaccharides termed the A and B bands. The high-molecular-weight B

*

band is highly anionic and determines the serotype specificity of a particular strain, whereas the A band contains shorter chains of predominantly uncharged sugars composed primarily a polymer of a1--2-, al1-+3-, ot1--3-linked D-rhamnose and low levels of 2-keto-3-deoxyoctulosonic acid (KDO). Unlike B band LPS, it lacks phosphate but contains stoichiometric amounts of sulfate (2, 29, 41). P. aenrginosa is known to undergo phenotypic alterations during in vivo growth (5, 28), so that the B band LPS becomes entirely replaced with A band LPS and is therefore the major LPS antigen among most clinical isolates from cystic fibrosis patients with chronic Pseudomonas pulmonary infections (17, 19, 28, 46). Prolonged antibiotic therapy has been implicated as one of the reasons for this conversion (2, 13, 37). Many of these clinical isolates show altered susceptibilities to aminoglycosides, and the mechanism(s) which confers this resistance is unclear (19). Laboratory studies have shown that permeability resistance to aminoglycosides is partly due to alterations in the LPS structure which decrease antibiotic passage across the outer membrane (1, 8, 9, 43, 49) or is due to defects in the cytoplasmic transport system that are related to inefficient energy generation (3, 7, 36). Mutagenesis with ethyl methanesulfonate and an LPSspecific phage has generated P. aeruginosa strains possessing A' B-, A- B+, and A- B- LPS attributes (4, 29). The contribution of either A or B band LPS in the ionic binding of aminoglycosides to the outer membrane of P. aeruginosa is not known. Accordingly, we investigated the influence of A and B band LPSs in ionic binding of gentamicin after disrupting the energy-dependent phases of membrane transport so that passive ion binding could be separated from active transport. This experimental technique also allowed

Corresponding author. 715

716

ANTimICROB. AGENTS CHEMOTHER.

KADURUGAMUWA ET AL.

TABLE 1. Bacterial strains, their sources, and their LPS characteristics

P.saeraginosa strain PAO1 AK 1401 dps 89 rd 7513

propertyA

and B band LPS prpIrcharacteristic

Wild type; 05 (B band), A band LPS Rough mutant that has core plus one 0-antigen, A band LPS Revertant of rd 7513, B band LPS A band LPS-deficient mutant of AK 1401 has core plus one 0 antigen

to determine how the ionic binding contributed to the antibiotic's bactericidal action. (Parts of the present study were presented at the 92nd General Meeting of the American Society for Microbiology [22].)

us

MATERUILS AND METHODS Bacterial strains and growth conditions. The nonmucoid strains of P. aeruginosa with various A band or B band LPS compositions used in the present study are described in Table 1. All bacterial strains were grown to the mid-exponential growth phase at 37°C with vigorous shaking, as described previously by Bryan and Van Den Elzen (10). Antibiotics and chemicals. The [3H]gentamicin used in the present studies was obtained from Amersham Canada Ltd., Oakville, Ontario, Canada, and had a specific activity of 510 mCi/mmol. Gentamicin sulfate, lysozyme (EC 3.2.1.17), and 2,4-dinitrophenol (DNP) were from Sigma Chemical Co. (St. Louis, Mo.), and sodium azide and potassium cyanide (KCN) were from the Fisher Scientific Co. (Nepean, Ontario, Canada). Antibiotic susceptibility. The susceptibilities of P. aeruginosa strains to gentamicin were determined by a dilution method in Mueller-Hinton broth (33). The MICs for A' B+, A' B-, A- B+, and A- B- strains were 2, 4, 4, and 0.5 pLg/ml, respectively. Measurement of ionic binding of gentamicin. The ionic binding of gentamicin to bacterial cell surfaces was determined with [3H]gentamicin by the method of Bryan and Van Den Elzen (10), with slight modifications. Exponentialgrowth-phase cultures were adjusted to 108 CFU/ml (0.15 optical density unit at 460 nm) and were incubated for 10 min in medium containing KCN (1 mM) and DNP (2 mM). [3H]gentamicin was mixed with unlabeled gentamicin as needed to obtain various final concentrations (5, 10, and 20 ,ug/ml) of gentamicin. After 2 min, 2.0-ml samples were removed and filtered through a 0.45-,um-pore-size membrane filter pretreated with a 3.0-ml solution of appropriate concentrations (5, 10, and 20 ,ug/ml) of unlabeled gentamicin. The quantity of radiolabeled gentamicin in the filtrate was measured, and the amount of gentamicin passively bound to each strain was determined. Effect of ionic binding on cell viability. The effect of ionic binding on cell viability was determined as described by Jackson et al. (21), with slight modifications. Exponentialgrowth-phase cultures were diluted in nutrient broth to produce a bacterial suspension of 108 CFU/ml, and the suspension was placed in an ice bath for 30 min. After reaching equilibrium with the bath temperature (4°C), gentamicin prepared in cold medium from stock solutions of various concentrations was added to obtain various final concentrations (5, 10, and 20 ,ug/ml) in test cultures. Identical bacterial cultures without antibiotic served as controls. After exposing the culture to drug for 2 min in the cold, the numbers of viable bacteria in the postexposure test and

A+ B+ A+ BA- B+ A- B-

Source or reference

R. E. W. Hancock (16) A. M. Kropinski (4) J. Lightfoot and J. S. Lam (29a) J. Lightfoot and J. S. Lam (29)

control cultures were determined. Briefly, an equal volume was withdrawn from the culture and unbound drug was eliminated by absorption to 20 mg of phosphocellulose per ml in the first two of six tubes in nutrient broth at 4°C, and then the cells were plated onto antibiotic-free agar. Survival values were determined by calculating the average viability for each strain and comparing them with the viabilities for control cultures. Permeabilization of whole cells to lysozyme. Lysozymemediated lysis of P. aeruginosa strains by gentamicin was carried out by the method described by Angus et al. (1). Preparation of LPS. LPS was prepared by the whole-cell lysis method described by Hitchcock and Brown (18). SDS-PAGE analysis and silver staining. The various LPSs were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) as described by Rivera et al. (41) by using 13% acrylamide gels. Silver staining of the separated LPS bands was done by the method of Tsai and Frasch (48). Western immunoblotting. LPS samples were transferred from SDS-polyacrylamide gels to nitrocellulose and reacted with monoclonal antibody (MAb) MF 15-4 (specific for B band serotype 05) or MAb NlF 10 (specific for A band polysaccharide) as described by Lightfoot and Lam (29). RESULTS SDS-PAGE banding pattern and immunoreactivities of LPSs from wild-type and mutant strains. LPSs from P. aeruginosa strains were separated by SDS-PAGE and silver stained to characterize the carbohydrate moieties and banding patterns (Fig. la). To identify the A and B band LPS antigens, electrophoretic blots of LPSs from the SDS-polyacrylamide gel described in Fig. la were reacted with either A band- or B band-specific MAbs (Fig. lb and c). The ladder-like banding pattern of LPS from strain PA01 (A' B+) was similar to the patterns reported previously (29, 41) and showed irregularities in the spacings and intensities of the bands. Immunoblotting of the PAO1 (A' B+) LPSs with both A band- and B band-specific MAbs showed coexpression of the A band and the B band (Fig. lb and c). The B band was not well resolved and reacted as more of a smear, with less distinction between separate bands, particularly in the higher-molecular-weight region containing the longer lengths of 0 repeating units. A band units were observed in a different position in the gel, with the spacing and intensity of the ladder pattern being much more regular than those seen with B band LPS, confirming the results of earlier studies (29, 41). In contrast, the silver-stained LPS profile from strain AK 1401 (A' B-) also indicated ladder-like bands, but these were not as extensive as those of PA01 (A' B+), indicating that the bands were of lower-molecularweight LPS than the B band (Fig. la). Furthermore, these AK 1401 bands reacted only with the A band-specific MAb; the placement of most bands was similar to that of the A band LPS from PAO1 (29, 41). The silver-stained LPS profile

VOL. 37, 1993

INTERACTION OF GENTAMICIN WITH LPS OF P. AERUGINOSA

XC

717

25

20

15

*0

0) co15

*0

0

5

10

15

20

25

Concentrabon of gentaricin 1lg/ml FIG. 2. Ionic binding of gentamicin to P. aeruginosa PA01

FIG. 1. (a) Silver-stained SDS-polyacrylamide gel of LPS from P. aeruginosa PA01 (A' B+) (lane 1), AK 1401 (A' B-) (lane 2), dps 89 (A- B+) (lane 3), and rd 7513 (A- B-) (lane 4). Western immunoblots of this LPS profile reacted with B band-specific MAb (b) and with A band-specific MAb (c) are shown.

of strain dps 89 (A- B+) was similar to that of PAO1 (A' B'). However, the doublet bands seen in the middle position of the gel for PAO1 (A' B+) were absent in the gel for dps 89 (A- B+). Immunoblots of LPS profiles of dps 89 (A- B+) indicated that the high-molecular-weight material was B band LPS and confirmed that the A band was not present. Silver-stained LPS from strain rd 7513 (A- B-) lacked the ladder-like bands, and this LPS was not reactive with either A band- or B band-specific MAb, confirming the absence of A and B band LPSs. There were no significant differences seen in the outer membrane protein profiles of PAO1 and its LPS derivatives, although a trace amount of an additional protein (-70 kDa) in rd 7513 was seen (data not shown). The bacterial strains and their LPS characteristics are summarized in Table 1. Ionic binding of gentamicin and its antibacterial effect on P. aeruginosa PAO1 and LPS derivatives. The ionic binding of gentamicin to the cell surface of P. aenrginosa strains was investigated by using [3H]gentamicin. Figure 2 demonstrates that ionic binding of gentamicin was related to the drug concentration and that binding was nonsaturable at the highest drug concentration tested. This suggests that there is some cooperativity of binding as gentamicin complexes to the surface. Other investigators have reported similar observations in P. aeruginosa (21, 30). The ionic binding of gentamicin appeared to be different between PAO1 (A' B+) and its LPS derivatives. PAO1 (A' B+) bound the most

(A'

B+) (-) and its LPS derivatives AK 1401 (A' B-) (-), dps 89 (AB') (A), and rd 7513 (A- B-) ( ). Exponential-growth-phase cultures (10' CFU/ml) were exposed to KCN (1 mM) and DNP (2 mM) to inactivate energy-dependent drug transport and were then exposed to various concentrations of ['H]gentamicin for 2 min. Samples (2.0 ml) were removed and filtered through a 0.45 Wtm-poresize membrane which was treated with appropriate concentrations of unlabeled gentamicin. The amount of gentamicin bound to each strain was determined by counting the radioactivity in the filtrate. Data are means -+ standard deviations from three independent experiments.

gentamicin at all concentrations tested; this was followed by dps 89 (A- B+), AK1401 (A' B-), and rd 7513 (A- B-). The degree of binding varied slightly among the strains, and strains that possessed the B band LPS appeared to bind the most gentamicin. Strain PA01 (A' B+) and derivative strains A' B-, AB+, and A- B- precooled to 4°C to inactivate energy-v dependent drug transport were exposed to gentamicin to assess the antibacterial effect of passive ionic binding of gentamicin to the cell envelope. Precooled cells not treated with gentamicin served as controls. Removal of unbound drug and reenergization of the cells by incubation at 37°C in medium showed the postexposure effect to be bactericidal (Fig. 3). The number of viable organisms decreased with an increase in the concentration of gentamicin. The maximum bactericidal action of passively bound gentamicin was seen on strain A+ B+ at all concentrations tested and resulted in a 50% reduction in the population at 20 5g of gentamicin per ml; this was followed by A- B+ (30%), A+ B- (22%), and A- B- (11%). These results demonstrate that cells that possess both A and B band LPSs are the most prone to antibiotic-induced killing, while the A- B- strain is least affected. Strains A+ B- and A- B+ were intermediate in

susceptibility. Ionic binding of gentamicin and lysis of whole cells by isozyme. The addition of increasing concentrations of gentamicin to whole cells in the presence of lysozyme (and azide to inhibit active uptake of gentamicin) caused increased lysis in all strains. Cells not treated with gentamicin were not lysed by the lysozyme. This indicatesthfat ee binding of gentamicin to cells causes permeabilization of the outer

718

ANTiMICROB. AGENTS CHEMOTHER.

KADURUGAMUWA ET AL. 125

1001 E 75 0

3 50

7

(D 0

25

0 0

5

10

15

20

25

Concentration of gentamicin ,ug/mI

FIG. 3. Antibacterial effect of gentamicin caused by ionic binding on P. aeruginosa PAO1 (A' B+) (-), AK 1401 (A' B-) (@), dps 89 (A- B+) (A), and rd 7513 (A- B2 (*). After cooling to 4°C, exponential-growth-phase cultures (10 CFU/ml) were exposed to various concentrations of gentamicin for 2 min. The decrease in viable bacteria was measured, after removal of unbound drug by phosphocellulose, by plating in quadruplicate. Means ± standard deviations of three independent determinations are shown.

membrane to relatively large molecules such as lysozyme (i.e., increased lysis is due to increased murein degradation, which appeared to be proportional to increased access to lysozyme). Lysozyme-mediated lysis in strain PAO1 (A' B+) appeared to be higher than that in mutant strains. A concentration of 20 jig of gentamicin per ml resulted in almost 68% lysis in A' B+ compared with 57% for A- B+, 50% for A' B-, and 30% for A- B-; the mutant strains had lower affinities for gentamicin binding (Fig. 4). The Ionic binding, viability, and lysozyme-mediated lysis studies revealed the affinity of gentamicin binding to PAO1 and its LPS derivatives to be as follows: A' B+ > > > A- B+ > > A+ B- > A- B-. DISCUSSION The results of the present study demonstrate that the strains that possess both A and B band LPSs bound more gentamicin than the LPS-defective strains. Ionic binding of 1 to 20 ,ug of gentamicin per ml to the cell surface of the wild-type strain caused almost 5 to 50% loss in viability, indicating that the passive, couloumbic interaction of antibiotic with the outer membrane can be a crucial step in the killing action of gentamicin. A lethal dose of drug was bound to all strains, and the degree of binding varied with the strain. Strains with A band LPS bound less gentamicin and were less prone to antibiotic-induced killing than strains with B band LPS. This implies that the accessibility of ionic binding sites differs between the two types of LPS molecules. Previous studies have demonstrated that the ionic binding of the aminoglycoside is an interaction between antibiotic and the outer membrane (11, 20, 21). In gram-negative bacteria, the ionic binding sites include charged LPS substituents, polar heads of phospholipids, and outer membrane proteins (47). The ability of cationic antibiotics to penetrate

0

5

10

15

20

25

Concentralon of gentamidn lAg/mI FIG. 4. Ionic binding of gentamicin and lysozyme-mediated lysis of P. aeruginosa PAO1 (A' B+) (U) and its LPS derivatives AK 1401 (A' B-) (0), dps 89 (A- B+) (A), and rd 7513 (A- B-) (*).

Gentamicin was added to cell suspensions in the presence of

lysozyme and NaN3. Lysis was followed as the decrease in optical density at 600 nm. Means t standard deviations of three independent determinations are shown.

the outer membrane is related to their ability to bind LPS near the lipid A head group, destroying the LPS-LPS crossbridging and destabilizing the outer membrane bilayer structure (14, 35). Earlier studies in our laboratory have demonstrated that ultrastructural and biochemical alterations occur in the cell wall of P. aeruginosa during the self-promoted uptake of gentamicin. During this process, the outer membrane lost significant amounts of protein, LPS, Ca2+, and Mg2+ (32, 50). The chemical structure of A and B band LPSs from P. aeruginosa have been elucidated (2, 24, 41). The B band LPS of P. aeruginosa is believed to contain approximately 12 phosphate groups and two to three KDO residues per molecule, and it is also often high in amino derivatives of uronic acid or in fucose residues (25, 51). As a consequence, the LPS carries a net negative charge at physiological pHs, resulting in a strong negative charge on the surface of the cell (14). However, the chemical composition of the A band has been found to contain low levels of heptose, KDO, and amino sugars and is composed mainly of uncharged D-rhamnose. A band LPS replaces its phosphate with sulfate (2, 29, 41, 42, 44). Stoichiometric differences in the constituents of the outer core or lipid A of A and B band LPSs have also been reported (41, 42). In total, these differences ensure that the A band is much less negatively charged than the B band. Indeed, results from our laboratory have demonstrated that strains with A band LPS do not stain with electron-dense uranyl ions (UO22+) during the preparation of specimens for electron microscopy (27). On the other hand, strains that possess the B band LPS stain darkly because the electronegative sites of this LPS readily complex UO22+ from solution. The ionic binding of gentamicin appeared to be nonsaturable at the highest concentration of gentamicin. This observation is in agreement with previous published work (21, 30). It is believed that as the cation binds to LPS, the molecule alters its conformation to the extent that other

VOL. 37, 1993


sites of interaction subsequently occur (39). Walker and Beveridge (50) proposed that during binding of large cations such as gentamicin to sites previously occupied by Ca2" and Mg2+, the outer membrane components are so rearranged that the normal packing order is destabilized and more potential aminoglycoside-binding sites are exposed. Although the B band LPS clearly bound more gentamicin than the A band LPS, the latter was also capable of some binding. Since the side chain polysaccharide has little electronegativity, this gentamicin binding must rely on reactive sites in the core lipid A region of the LPS molecule. The lower affinity of gentamicin to A band LPS is probably due to a shielding of cation-binding sites by the uncharged 0 antigen chains or because of a decrease in negative charge because of the loss or esterification of phosphate moieties deep within the LPS molecule. Presumably, these A band characteristics also result in decreased levels of metal cations such as Ca2+ and Mg2+, stabilizing intermolecular associations. Accordingly, gentamicin as a cation neither can bind effectively to A band LPS nor can it compete for Ca or Mg salt bridges. The presence of esterified phosphate in the A band has not been chemically verified. We are investigating the levels of Ca and Mg in parent and LPS-defective strains and the loss of these metals from intact cells during ionic binding of gentamicin. Previous studies have demonstrated that bacteria with less negatively charged LPSs have a decreased affinity for polycations and increased resistance to a variety of cations (1, 6, 9, 12, 30, 38, 43). The absence or reduction in 0-specific polymer and, especially, truncation of the LPS core are also known to increase resistance to a variety of cations, including gentamicin (9). Bryan et al. (9) proposed that a reduction in repeating side chain sugars could reduce the overall length of the LPS molecule extending from the surface and could make a bacterium less hydrophilic. This could have two effects; first, the decreased exposure of the LPS on the cell surface would limit the number of sites for gentamicin interaction, and second, the general decrease in hydrophilicity would discourage access to highly polar molecules such as gentamicin. Loss of A or B band LPS could also have similar consequences. More recently, Kastowsky et al. (23) reported that, depending on the length of the 0 repeats in the LPS, the core- portion of LPS could exist in a well-defined cylindrical shape whose long axis runs parallel to the outer membrane surface. On the other hand, the 0 antigen (including the Ra portion) runs perpendicular to the cell surface. The lower affinity of gentamicin toward LPS-defective strains in relation to that toward the wild-type strain could be partly due to an architectural alteration in the bacterial outer membrane. For example, the absence of 0 side chains could alter the number and spatial arrangement of the negative charges, LPS fluidity, the presence or absence of hydrophobic sites, the specificities of binding sites for certain cations, the number, position, and composition of LPS phosphates, and outer membrane proteins such as OpH (H1) (16, 34, 38). Alterations in the LPS molecule are also known to induce changes in the fatty acid composition of the acyl chains. Kropinski et al. (26) showed that an antibiotic-supersusceptible strain had a markedly lower level of C12 and C16 fatty acids compared with the number in the parent strain. The fatty acid compositions of the mutants used in the present study were not investigated. A high proportion of P. aeruginosa isolates from cystic fibrosis patients lack the repeating ladder pattern of LPS, and the majority of clinical isolates from these patients possess the A band as the major LPS antigen (17, 37, 40).

719

Furthermore, these isolates show increased resistance to aminoglycoside antibiotics (19, 46). Our results demonstrate that the highest affinity for gentamicin was to strains possessing both A and B band LPSs. This indicates that the permeability type resistance among clinical strains of P. aeruginosa could partly be due to lower affinities of aminoglycosides to their A band LPSs. Traditionally, gentamicin has been considered to be an antibiotic which inhibits protein synthesis by binding to the 30S ribosomal subunit. However, our results demonstrate that surface binding alone could reduce substantially the number of viable cells. We believe that the binding of gentamicin to the outer membrane contributes to cell death by weakening the cell envelope. To test this hypothesis further, we are attempting to modify the gentamicin molecule chemically so that it can be directly visualized by electron microscopy on the binding sites of intact bacterial cells before, during, and after cell death. ACKNOWLEDGMENTS This work was supported by a grant (to T.J.B.) from the Canadian Bacterial Diseases Network, which is funded as a National Center of Excellence. We are grateful to Jeff Lightfoot for the P. aeruginosa strains.

REFERENCES 1. Angus, B. L., J. A. M. Fyfe, and R. E. W. Hancock. 1987. Mapping and characterization of two mutations to antibiotic

susceptibility in Pseudomonas aeruginosa. J. Gen. Microbiol. 133:2905-2914. 2. Arsenault, T. L., D. W. Hughes, D. B. Maclean, W. A. Szarek, A. M. B. Kropinski, and J. S. Lam. 1991. Structure studies on the polysaccharide portion of "A-band" lipopolysaccharide from a mutant (AK1401) of Pseudomonas aeruginosa strain PA01. Can. J. Chem. 69:1273-1280. 3. Bayer, A. S., D. C. Norman, and K. S. Kim. 1987. Characterization of impermeability variants of Pseudomonas aeruginosa isolated during unsuccessful therapy of experimental endocarditis. Antimicrob. Agents Chemother. 31:70-75. 4. Berry, D., and A. M. Kropinski. 1986. Effect of lipopolysaccharide mutations and temperature on plasmid transformation efficiency in Pseudomonas aeruginosa. Can. J. Microbiol. 32:436438. 5. Brown, M. R. W., H. Anwar, and P. A. Lambert. 1984. Evidence that mucoid Pseudomonas aeruginosa in the cystic fibrosis lung grows under iron-restricted conditions. FEMS Microbiol. Lett. 21:113-117. 6. Brown, M. R. W., and J. Melling. 1969. Role of divalent cations in the action of polymyxin B and EDTA on Pseudomonas aeruginosa. J. Gen. Microbiol. 59:263-274. 7. Bryan, L. E. 1988. General mechanisms of resistance to antibiotics. J. Antimicrob. Chemother. 22(Suppl. A):1-15. 8. Bryan, L. E., and S. Kawan. 1983. Roles of ribosomal binding, membrane potential, and electron transport in bacterial uptake of streptomycin and gentamicin. Antimicrob. Agents Chemother. 23:835-845. 9. Bryan, L. E., K. O'Hara, and S. Wong. 1984. Lipopolysaccharide changes in impermeability-type aminoglycoside resistance in Pseudomonas aeruginosa. Antimicrob. Agents Chemother. 26:250-255. 10. Bryan, L. E., and H. M. Van Den Elzen. 1976. Streptomycin accumulation in susceptible and resistant strains of Escherichia coli and Pseudomonas aeruginosa. Antimicrob. Agents Chemother. 9:928-938. 11. Bryan, L. E., and H. M. Van Den Elzen. 1977. Effects of membrane-energy mutations and cations on streptomycin and gentamicin accumulation by bacteria: a model for entry of streptomycin and gentamicin in susceptible and resistance bac-

720

KADURUGAMUWA ET AL.

teria. Antimicrob. Agents Chemother. 12:163-177. 12. Gilleland, H. E., Jr., and L. B. Farley. 1982. Adaptive resistance to polymyxin in Pseudomonas aeruginosa due to an outer membrane impermeability mechanism. Can. J. Microbiol. 28: 830-840. 13. Govan, J. R. W., and J. A. M. Fyfe. 1978. Mucoid Pseudomonas aeruginosa and cystic fibrosis: resistance of the mucoid form to carbenicillin, flucloxacillin, and tobramycin and the isolation of mucoid variants in vitro. J. Antimicrob. Chemother. 4:233240. 14. Hancock, R. E. W. 1984. Alterations in outer membrane permeability. Annu. Rev. Microbiol. 38:237-264. 15. Hancock, R. E. W., and A. Bell. 1988. Antibiotic uptake into gram-negative bacteria. Eur. J. Clin. Microbiol. Infect. Dis. 7:713-720. 16. Hancock, R. E. W., and A. M. Crey. 1979. Outer membrane of Pseudomonas aeruginosa: heat- and 2-mercaptoethanol-modifiable proteins. J. Bacteriol. 140:902-910. 17. Hancock, R. E. W., L. M. Mutharia, L. Chan, R. P. Darveau, D. P. Speert, and G. B. Pier. 1983. Pseudomonas aeruginosa isolates from patients with cystic fibrosis: a class of serumsensitive, nontypable strains deficient in lipopolysaccharide 0 side chains. Infect. Immun. 42:170-177. 18. Hitchcock, P., and T. M. Brown. 1983. Microheterogeneity among Salmonella lipopolysaccharide chemotypes in silverstained polysaccharide gels. J. Bacteriol. 154:269-277. 19. H0iby, N., and K. Rosendal. 1980. Epidemiology of Pseudomonas aenrginosa infection in patients treated at a cystic fibrosis centre. Acta Pathol. Microbiol. Scand. Sect. B 88:125131. 20. Hurwitz, C., C. L. Rosano, and J. V. Landau. 1962. Kinetics of loss of viability of Escherichia coli exposed to streptomycin. J. Bacteriol. 83:1210-1216. 21. Jackson, G. E., V. T. Lolans, and G. L. Daikos. 1990. The inductive role of ionic binding in the bactericidal and postexposure effects of aminoglycoside antibiotics with implications for dosing. J. Infect. Dis. 162:408-413. 22. Kadurugamuwa, J., T. Mok, J. Lam, G. Southham, and T. J. Beveridge. 1992. Interaction of A-band and B-band lipopolysaccharide of Pseudomonas aeruginosa with gentamicin and its possible lethal effects. Program Abstr. 92nd Gen Meet. Am. Soc. Microbiol. abstr. A3, p. 1. American Society for Microbiology, Washington, D.C. 23. Kastowsky, M., T. Gutberlet, and H. Bradaczek. 1992. Molecular modeling of the three-dimensional structure and confirmational flexibility of bacterial lipopolysaccharide. J. Bacteriol. 174:4798-4806. 24. Knirel, Y. A., E. V. Vinogradov, N. A. Kocharova, N. A. Paramonov, N. K. Kochetkov, B. A. Dmitriev, E. S. Stanislavsky, and B. Lanyi. 1988. The structure of 0-specific polysaccharides and serological classification of Pseudomonas aeruginosa. Acta Microbiol. Hung. 35:3-24. 25. Kropinski, A. M., L. C. Chan, and F. H. Milazzo. 1979. The extraction and analysis of lipopolysaccharide from Pseudomonas aeruginosa strain PAO1, and three rough mutants. Can. J. Microbiol. 25:390-398. 26. Kropinski, A. M., J. Kuzio, B. L. Angus, and R. E. W. Hancock. 1982. Chemical and chromatographic analysis of lipopolysaccharide from an antibiotic-supersusceptible mutant of Pseudomonas aeruginosa. Antimicrob. Agents Chemother. 21:310319. 27. Lam, J. S., L. L. Graham, J. Lightfoot, T. Dasgupta, and T. J. Beveridge. 1992. Ultrastructural examination of the lipopolysaccharide of strains and their isogenic rough mutants by freeze substitution. J. Bacteriol. 174:7159-7167. 28. Lam, M. Y. C., E. J. McGroarty, A. M. Kropinski, L. A. MacDonald, S. S. Pedersen, N. H0iby, and J. S. Lam. 1989. Occurrence of a common lipopolysaccharide antigen in standard and clinical strains of Pseudomonas aeruginosa. J. Clin. Microbiol. 27:962-967. 29. Lightfoot, J., and J. S. Lam. 1991. Molecular cloning of genes involved with expression of A-band lipopolysaccharide, an antigenically conserved form, in Pseudomonas aeruginosa. J.

ANTIMICROB. AGENTS CHEMOTHER.

Bacteriol. 173:5624-5630. 29a.Lightfoot, J., and J. S. Lam. Personal communication. 30. Loh, B., C. Grant, and R. E. W. Hancock. 1984. Use of fluorescent probe 1-N-phenylnaphthlamine to study the interactions of aminoglycoside antibiotics with the outer membrane of Pseudomonas aeruginosa. Antimicrob. Agents Chemother. 26: 546-551. 31. MacArthur, R. F. D., V. Lolans, F. A. Zar, and G. G. Jackson. 1984. Biphasic, concentration-dependent and rate-limited concentration-independent bacterial killing by an aminoglycoside antibiotic. J. Infect. Dis. 150:778-779. 32. Martin, N. L., and T. J. Beveridge. 1986. Gentamicin interaction with Pseudomonas aeruginosa. Antimicrob. Agents Chemother. 29:1079-1087. 33. National Committee for Clinical Laboratory Standards. 1990. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically. Approved standard M7-A2, 2nd ed. National Committee for Clinical Laboratory Standards, Villanova, Pa. 34. Nicas, T. I., and R. E. W. Hancock. 1983. Alteration of susceptibility to EDTA, polymyxin B and gentamicin in Pseudomonas aeruginosa by divalent cation regulation of outer membrane protein Hl. J. Gen. Microbiol. 129:509-517. 35. Nikaido, H., and M. Vaara. 1985. Molecular basis of bacterial outer membrane permeability. Microbiol. Rev. 49:1-32. 36. Parr, T. R., Jr., and A. S. Bayer. 1988. Mechanisms of aminoglycoside resistance in variants of Pseudomonas aeruginosa isolated during treatment of experimental endocarditis in rabbits. J. Infect. Dis. 158:1003-1010. 37. Penketh, A., T. Pitt, D. Roberts, M. E. Hodson, and J. C. Batten. 1983. The relationship of phenotype changes in Pseudomonas aeruginosa to the clinical condition of patients with cystic fibrosis. Am. Rev. Respir. Dis. 127:605-608. 38. Peterson, A. A., S. W. Fesik, and E. J. McGroarty. 1987. Decrease binding of antibiotics to lipopolysaccharides from polymyxin-resistant strains of Escherichia coli and Salmonella typhimurium. Antimicrob. Agents Chemother. 31:230-237. 39. Peterson, A. A., R. E. W. Hancock, and McGroarty. 1985. Binding of polycationic antibiotics and polyamines to lipopolysaccharides of Pseudomonas aeruginosa. J. Bacteriol. 164: 1256-1261. 40. Pitt, T. T., J. MacDougall, A. R. L. Penketh, and E. M. Cooke. 1986. Polyagglutinating and non-typable strains of Pseudomonas aeruginosa in cystic fibrosis. J. Med. Microbiol. 21:179186. 41. Rivera, M., L. E. Bryan, R. E. W. Hancock, and J. E. McGroarty. 1988. Heterogeneity of lipopolysaccharides from Pseudomonas aeruginosa: analysis of lipopolysaccharide chain length. J. Bacteriol. 170:512-521. 42. Rivera, M., T. R. Chivers, J. S. Lam, and E. J. McGroarty. 1992. Common antigen lipopolysaccharide from Pseudomonas aeruginosa AK1401 as a receptor for bacteriophage A7. J. Bacteriol. 174:2407-2411. 43. Rivera, M., R. E. W. Hancock, J. G. Sawyer, A. Haug, and E. J. McGroarty. 1988. Enhanced binding of polycationic antibiotics to lipopolysaccharide from an aminoglycoside-supersusceptible, tolA mutant strain of Pseudomonas aeruginosa. Antimicrob. Agents Chemother 32:649-655. 44. Rivera, M., and E. J. McGroarty. 1989. Analysis of a common antigen lipopolysaccharide from Pseudomonas aeruginosa. J. Bacteriol. 171:2244-2248. 45. Schindler, P. R. G., and M. Teuber. 1975. Action of polymyxin B on bacterial membranes: morphological changes in the cytoplasm and in the outer membrane of Salmonella typhimunum and Eschenchia coli B. Antimicrob. Agents Chemother. 8:95104. 46. Speert, D. P., S. W. Farmer, M. E. Campbell, J. M. Musser, R. K. Selander, and S. Kuo. 1990. Conversion of Pseudomonas aeruginosa to the phenotype characteristic of strains from patients with cystic fibrosis. J. Clin. Microbiol. 28:188-194. 47. Taber, H. W., J. P. Muller, and A. S. Arrow. 1987. Bacterial uptake of aminoglycoside antibiotics. Microbiol. Rev. 51:439457.

VOL. 37, 1993


48. Tsai, C., and C. E. Frasch. 1982. A sensitive silver stain for detecting LPS in polysaccharide gels. Anal. Biochem. 119:115-119. 49. Vaara, M., T. Vaara, M. Jensen, I. Helander, M. Nurminen, E. T. Rietschel, and P. H. Makela. 1981. Characterization of the lipopolysaccharide from the polymyxin-resistant pmrA mutant of Salmonella typhimurium. FEBS Lett. 129:145-149.

721

50. Walker, S. G., and T. J. Beveridge. 1988. Amikacin disrupts the cell envelope of Pseudomonas aeruginosa ATCC 9027. Can. J. Microbiol. 34:12-18. 51. Wilkinson, S. G. 1983. Composition and structure of lipopolysaccharide from Pseudomonas aeruginosa. Rev. Infect. Dis. 5:S941-S949.