influence of far-ultraviolet radiation on the permeability of the outer membrane of ...

Influence of far-ultraviolet radiation on the permeability of the outer membrane of Escherichia coli

RUSTOM MODY, SUDHA KRISHNAMURTHY, AND PRAFULLA DAVE~ Department of Microbiology and Biotechnology Centre, Faculty of Science, M. S. University of Baroda,

Baroda - 390 002, India

Received January 3, 1989

Accepted August 21, 1989

MODY, R., KRISHNAMURTHY, S., and DAVE, P. 1989. Influence of far-ultraviolet radiation on the permeability of the outer membrane of Eschenchia coli. Can. J. Microbiol. 35: 1022 - 1030.

Far-ultraviolet radiation (254 nm) at a dose of 10, 20, and 30 Jim2 was found to disrupt the outer membrane permeability barrier of Escherichia coli to various antibiotics, dyes, and detergents. The degree of sensitization to these agents was propor- tional to the radiation dose. The irradiated cells showed a significant increase in the sensitivity of hydrophilic antibiotics (ampicillin, carbenicillin, penicillin), whereas much less sensitization was found towards hydrophobic probes (kanamycin, erythromycin, rifamycin SV, crystal violet, phenol, novobiocin) and detergents (dodecyl sulfate, bile salt, Triton X-100). The biochemical data and ultrastructural analysis of the outer membrane by freeze-etching have shown that the increase in phospho1ipid:protein ratio after irradiation had changed the architecture of the outer membrane from a highly asymmetric bilayer structure with densely packed lipopolysaccharide-protein particles on the outer half, to one predominantly exhibiting smooth phospholipid bilayer characteristics. The structure, composition, and barrier function of the outer membrane were restored to normal within 3 h of postirradiation incubation in nonproliferative medium. During this period, the acquisition of resistance towards a hydrophilic antibiotic (ampicillin) was faster than that for a hydrophobic agent (phenol).

Key words: far-ultraviolet, outer membrane, permeability, disorganization, recovery.

MODY, R., KRISHNAMURTHY, S., et DAVE, P. 1989. Influence of far-ultraviolet radiation on the permeability of the outer membrane of Escherichia coli. Can. J. Microbiol. 35 : 1022-1030.

L'irradiation a l'ultraviolet lointain (254 nrn) B des doses de 10, 20, et 30 J/m2 s'est avkrk causer une rupture de la bar- r5re de permkabilitk de la membrane externe d'Escherichia coli face a divers antibiotiques, teintures et dktergents. Le degrk de sensibilisation B ces agents a kt6 proportiomel B la dose d'irradiation. La sensibilitk des cellules irradikes s'est accrue de f a~on significative envers les antibiotiques hydrophiles (ampicilline, carbenicilline, pknicilline), tandis qu'elle s'est avkrke beaucoup moindre envers les sondes hydrophobes (kanamycine, Crythromycine, rifamycine SV, violet cristal, phknol, novo- biocine) et les dktergents (dodkcyl sulfate, sels biliaires, Triton X-100). Les donnkes biochimiques et les analyses ultrastruc- turales de membranes externes par dkcapage a froid ont montrk, qu'aprks irradiation, un accroissement du rapport phospho1ipide:protCine avait modifik l'architecture des membranes externes, depuis une structure B deux couches hautement asymktrique avec sur la moitik externe des particules de lipopolysaccharides-protkines disposCes de f a ~ o n compacte B une structure montrant de f a ~ o n prkdominante les caractkristiques d'une double couche lisse de phospholipides. La structure, la composition et la fonction de barrikre de la membrane externe ont Btk rktablies B la normale en dedans de 3 h d'incubation postirradiation dans un milieu de non-prolifkration. Au cours de cette pkriode, l'acquisition d'une rksistance envers un anti- biotique hydrophile (ampicilline) a Ctk plus rapide qu'envers un agent hydrophobe (phknol).

Mots c l h : ultraviolet lointain, membrane externe, permkabilitk, dksorganisation, rCcupkration. [Traduit par la revue]

Introduction The outer membrane of Escherichia coli and other enteric

gram-negative bacteria displays a pronounced structural asym- metry and functional versatility. The phospholipids are con- fined to the inner leaflet, while lipopolysaccharides (LPS) are present on the outer half of the bilayer and provide a strong barrier against hydrophobic solutes (Nikaido 1976). The strong interaction between outer membrane proteins and LPS yields stable LPS -protein aggregates, some of which serve as aqueous channels (Hofstra et al. 1979; Schindler and Rosen- busch 1981; Smit et al. 1975; van Alphen et al. 1979). The discriminatory role of the outer membrane with respect to the permeation of solutes was recognised from the study of mutants defective in one or more outer membrane components, namely, proteins (Bavoil et al. 1977; Lugtenberg et al. 1976; van Alphen et al. 1976; Wilkinson et al. 1972) and LPS (Benson and Decloux 1985; Makela and Stocker 1981; Roantree et al. 1977; Sanderson et al. 1974; Tarnaki et al. 1971), and by the use of membrane-disorganising

'Author to whom correspondence should be addressed. Printed in Canada 1 Imprim6 au Canada

agents, such as EDTA and polycations, whose mode of action is known (Haradaway and Bullet 1979; Leive 1974; Vaara and Vaara 1983a, 19833). In either case, the structural defect in the outer membrane was correlated with changes in the sensitivity of intact cells to various antibiotics and other inhibitory agents. This approach helps to study the details of the molecular organization of the outer membrane as well as the molecular mechanisms of resistance towards compounds with different physical properties. However, the degree of sensitization of cells treated with membrane-disorganising agents may vary under certain conditions, which probably suggests a multiplicity of diffusion pathways across the outer membrane, each operating under a different set of conditions (Nikaido and Nakae 1979; Nikaido and Vaara 1985).

Ionizing and nonionizing radiation are both known to have damaging effects on cell membranes (Kelland et al. 1983; Myers 1970; Wallach 1972). However, evidence in support of radiation-induced biochemical and structural changes in intact cells and isolated membranes is indirect and inconclusive. In the present paper, we report the outer membrane disorganizing action of far-UV radiation (254 nm) on E. coli and corre- late UV-induced structural changes in the membrane with the

Can

. J. M

icro

biol

. 198

9.35

:102

2-10

30.

Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IVE

RSI

TY

OF

MIC

HIG

AN

on

11/1

3/14

. For

per

sona

l use

onl

y.

MODY ET AL. 1023

loss of cell barrier function against various drugs and anti- metabolites.

I Materials and methods Bacterial strain and growth conditions

The strain used was E. coli K-12 (W31 lo), which is thy A36; it was generously donated by Warren Masker, Biology Division, Oak Ridge National Laboratory, TN. The culture was grown in L broth (Miller 1972) at 37°C on a rotary shaker (180 rpm). Cells were harvested in the mid-exponential phase (OD,, ca. 0.3), washed, and resuspended in 0.9% NaCl to give a cell density of 1-2 x los colony- forming units (cfu)/mL. This suspension was maintained on crushed ice prior to and during irradiation.

Irradiation Bacterial suspensions (5 mL) were exposed to 10, 20, and 30 J/m2

of far-UV radiation from a 30 W, low-pressure Philips bactericidal lamp (having about 90% of the emission at 254 nm), at a fluence rate

I of 1.0 J . m-2 . s-'. Immediately after irradiation, cells were sub- I jected to sensitization assay, biochemical analysis, and freeze-

fracture ultramicroscopy.

1 Sensitization assay The minimum inhibitory concentrations (MICs) of various inhibi-

tory agents were determined prior to and after irradiation by a modification of the microtitre plate method (Vaara and Vaara 1983a).

I Aliquots of 1.5 mL from irradiated cell suspensions were diluted in fresh L broth supplemented with 1 % glucose to give a cell density of lo4 cfulmL. Samples (200 pL) from this inoculated medium were dispensed in each well of a microtitre plate. The highest concentration of the test agent was pipetted into one of the wells and then serially diluted out in successive wells (dilution factor, 0.75) with a microtitre pipette. After incubating the plates at 37°C for 18 h, the pH of the medium in each well was checked by adding a drop of 0.1 % bromothymol blue indicator dye. The appearance of yellow color indicated the formation of acid resulting from growth of the culture, while blue color indicated the absence of growth. The lowest concentration of bactericidal agent that completely inhibited growth was interpreted as the MIC. The sensitization factor was determined from the ratio of MIC for unirradiated to that obtained for irradiated culture. These tests were performed three times independently to deter- mine the average MIC value.

UV-induced sensitivity to crystal violet The MIC values were determined by spreading 100 pL of the cell

suspension on L-agar plates containing increasing concentration of crystal violet. The plates were incubated at 37°C for 48 h. The lowest concentration of thk dye which inhibited colony formation was noted as MIC.

UV-induced sensitivity to Triton X-100, sodium deoxycholate, and sodium dodeql sulfate

Compounds such as Triton X-100, sodium deoxycholate, and sodium dodecyl sulphate (SDS), which are bacteriostatic and do not show a clear end point color difference with the microtitre plate assay, even at higher concentrations, were tested by the turbidometric growth determination procedure. Tubes containing increasing concentrations of detergents (from 0 to 20%) in 5 mL of L broth were inoculated with 100 pL of cell suspension and incubated at 37°C in static conditions for 8 h. Growth was monitored by measuring the optical density of the medium at 650 nm with a Kontron 810 spectro- photometer. An OD65o of 1 corresponds to 5 x lo8 cellslmL. The percentage decrease in growth of the irradiated culture in presence of a bacteriostatic agent as compared to that of the unirradiated cells was taken as a measure of the UV-acquired sensitivity of the culture towards these test compounds.

Assessment of recovery from UV radiation induced membrane damage Postirradiation recoverv medium was minimal M9 salts medium

(Miller 1972) supplementkd with 4 mg/mL of glucose, 40 mglmL of casamino acids, and 5 pg/mL of thiamine hydrochloride. This

medium was devoid of thymidine, which is essential for the growth of strain W3 110 (thy -). Aliquots (1 mL) from control and irradiated samples were diluted in M9 salts solution prior to holding in 100 mL of prewarmed recovery medium to give an initial cell concentration of -105 cellslmL. Flasks containing recovery medium were incubated at 37°C on a slow rotary shaker. At regular intervals of 15 min, 5 mL of recovery medium was withdrawn and dispensed into the wells of microtitre plates for determining the MICs of ampicillin and phenol, as described above. Cell viability was also assessed from each aliquot withdrawn.

Analytical procedure The separation of outer and inner membranes was carried out by

density gradient centrifugation, as described by Witholt et al. (1976). Protein was estimated according to the method of Lowry et al. (1951), using bovine serum albumin as standard. Total cell protein was determined after ultrasonic disruption of the cells. Phospholipids were extracted from purified membrane preparations by the method of BIigh and Dyer (1959) and fractionated on silica acid columns to separate 2-keto-3-deoxyoctonic acid lipid A from other phospholipids, as described by Hofstra et al. (1979). Eluants containing phosphatidylethanolamine, phosphatidylglycerol, and cardioiipin were evaporated to dryness and quantitated by assaying for phos- phorous by the method of Bartlett (1959).

Preparation of cell-free culture filtrate (CCF) A log phase culture was suspended in 50 mL of phosphate-buffered

saline (PBS) at a density of 2 X lo8 cfu/mL. This suspension was divided into two fractions of 25 mL each. One of these was irradiated at 30 J/mZ in five batches of 5 rnL each. Both the control and the irradiated fractions were immediately centrifuged at 800 x g and the pellet was resuspended in PBS. The cells were kept in this isotonic surrounding for 10 min at 10°C, and were subsequently removed by centrifugation and filtered through a Sartorius membrane filter @ore size, 0.2 pm. The CCF was then concentrated by lyophilization and analysed for proteins by nondenaturing PAGE.

PAGE The proteins present in the outer membrane fraction and in the CCF

were resolved on a nondenaturing polyacrylamide (7.5%) slab gel in the anodic discontinuous buffer system described by Davis (1964). SDS -PAGE of membrane proteins was carried out according to the method of Ames (1974).

Freeze-fracture electron microscopy Control and irradiated cell suspensions were centrifuged at low

speed (4000 x g for 15 min) and fixed with a mixhrre of 0.5% glutaraldehyde and 2 % formaldehyde in 0. I5 M PBS, pH 7.4. Before quenching, glycerol was added (up to 25%) to prevent freeze damage. The specimens were freeze-fractured at - 100°C in a Balzers BAE301 instrument and shadowed with platinum-carbon replication. Replicas were examined under a Philips 301 electron microscope at 60 kV filament current. The outer membrane particle density was estimated as described by Smit et al. (1975).

Source of chemicals Antibiotics, such as ampicillin, benzyl penicillin (sodium salt),

kanamycin, and tetracycline hydrochloride, were purchased from Alembic Chemical Works, Baroda, India. Chloramphenicol was from Parke-Davis (India) Ltd., Bombay, India; carbenicillin (sodium salt), from Lyka Laboratories, Bombay, India; novobiocin (sodium salt), vancomycin, rifamycin SV, and Triton X-100 were from Sigma Chemical Co., St. Louis, MO. Crystal violet and phenol were from BDH Ltd., Bombay, India. Bile salt was obtained from Kochlight Laboratories, London, U.K., and SDS from Ubichem, London, U.K. Erythromycin was obtained from Alembic Chemical Works, Baroda, India in the form of a solution (1 mL containing 20 mg of erythromycin ethyl succinate). Fresh solutions of all antibiotics were prepared by dissolving them in deionized water. Molecular weight markers used for gel electrophoresis were from Sigma Chemical Co., St. Louis, MO.

Can

. J. M

icro

biol

. 198

9.35

:102

2-10

30.

Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IVE

RSI

TY

OF

MIC

HIG

AN

on

11/1

3/14

. For

per

sona

l use

onl

y.

CAN. J. MICROBIOL. VOL. 35, 1989

TABLE 1. MICs of various inhibitory agents against UV-irradiated E. coli W3110

MICa (ng/rnL) at a UV dose (J/m2) of: Sensitization Partition

Agent 0 10 20 30 factorb coefficientC

Tetracycline 450 250 1.9 0.26 1 400 0.07 Ampicillin 1 236 123 39 12.3 100 0.01 Carbenicillin 23400 2 340 74 1 234 100 0.01 Penicillin G 490 65 20.6 4.9 100 0.02 Kanamy cin 100000 17 798 7 500 3 167 31.5 0.16 Erythromycin lo00 3 16 100 42.2 23.7 0.79 Chloramphenicol 6 250 2 636 1 483 834 7.5 12.4 Crystal violet 31 200 17 550 7404 4 160 7.5 14.4 Phenol 3 900 1645 925 520 7.5 20 Novobiocin 29300 21 975 16480 12 360 2.3 < 20

'The MICs were defined as the lowest concentration of inhibitory agent that prevented growth of bacteria in L broth supplemented with 1 % glucose for 18 h at 3 7 T , starting from an inoculum of lo4 cellsImL. Each MIC value represents the average of three independent determinations. Practically no variation was observed for a defined dose of UV radiation and cell density.

h he ratio of the MIC for unirradiated cells to that obtained after highest dose (30 J I ~ ' ) of UV radiation. 'Values of partition coefficients are according to Coleman and Leive (1979) and Nikaido (1976).

Results Effect of far-UV on sensitivity of E . coli to various anti-

microbial agents The sensitivity of wild-type E. coli W3110 towards various

antibiotics, dyes, and detergents was found to increase when cells were exposed to increasing UV doses. The results of sensitivity tests for a wide range of inhibitory compounds are shown in Table 1. UV-irradiation caused an increase in the sensitivity, particularly to hydrophilic antibiotics such as ampicillin, carbenicillin, and penicillin. The increase in susceptibility to these agents was 100-fold at a dose of 30 J/m2. The decrease in MIC was more significant for lower UV doses (10 and 20 J/m2) as compared with the higher dose of 30 J/m2 (Table 1).

Irradiated cells showed much less sensitization towards hydrophobic probes, which include kanarnycin, erythromycin, rifamycin SV, crystal violet, phenol, and novobiocin. The increase in susceptibility to these probes was strongly dependent on the hydrophobicity of the test compound, thus ranging from 2.3- to 31.5-fold at a UV dose of 30 J/m2. The antibiotic tetracycline does not show the pattern of sensitization based on the hydrophobic index. A drastic increase in the sensitivity to tetracycline by 1400-fold was found at a dose of 30 J/m2 (Table 1). This antibiotic utilizes porin channels for entering through the outer membrane (Nikaido and Nakae 1979). The reason for the remarkably high sensitivity to tetracycline upon irradiation is not clearly deducible.

A log -log plot of sensitization factor versus hydrophobicity (Fig. 1) indicates that the UV-induced permeability has an inverse relationship with hydrophobicity. At UV doses of 10, 20, and 30 J/m2, for each 10-fold increase in the octanol/water partition coefficient of the test agent, there was roughly a 1.5-, 1.8-, and 2.5-fold decrease in its efficacy, respectively.

Normally, E. coli can tolerate high concentration of detergents such as sodium deoxycholate, Triton-100 and SDS. However, upon irradiation, cells showed a distinct sensitivity to these agents (Fig. 2). Subsequent plating of cells on N agar medium from tubes containing 1.5 % deoxycholate, 4 % Triton X-100, and 6% SDS showed no growth, indicating that these concentrations were bactericidal for UV-irradiated cells.

Assay for membrane repair Cells that were sensitized to antibiotics by UV irradiation

FIG. 1. Effect of hydrophobicity on the efficacy of various inhibitory agents against irradiated E. coli W3110. Hydrophobicity of different solutes is expressed in terms of their partition coefficient in 1-octanol/aqueous buffer, pH 7.0, at 24OC (values are according to Coleman and Leive, 1979, and Nikaido, 1976). Partition coefficient of 0.01 corresponds to ampicillin and carbenicillin; 0.02, penicillin G; 0.16, kanamycin; 0.79, erythromycin; 12.4, chloramphenicol; 14.4, crystal violet; and ca. 20 for phenol. Cells were irradiated at 10 (m), 20 (A), and 30 (e) J/m2.

lost their sensitivity upon subsequent incubation in recovery medium, which was nonpermissive for growth. The loss of sensitization to a hydrophilic antibiotic (ampicillin) and a hydrophobic agent (phenol) versus postirradiation incubation time was taken as a measure of membrane recovery from radiation damage. Irradiated cells were kept at 37OC in recovery medium for varying times before testing for sensitivity to ampicillin and phenol. Reacquisition of resistance towards ampicillin was within 2 h of recovery time, while resistance to phenol was acquired in the 3rd hour (Fig. 3). Total recovery was observed at the end of the 3rd hour. The recovery was inhibited when M9 salts solution (without any carbon source) was used, indicating that the membrane repair is essentially an energy-dependent process. This was further confirmed by using various metabolic inhibitors like 2,4-dini- trophenol and sodium azide which inhibit ATP synthesis and also block membrane repair (results not shown).

Can

. J. M

icro

biol

. 198

9.35

:102

2-10

30.

Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IVE

RSI

TY

OF

MIC

HIG

AN

on

11/1

3/14

. For

per

sona

l use

onl

y.

MODY ET AL. 1025

I CONCN. OF DETERGENT (golo)

FIG. 2. Effect of UV irradiation on the growth of E. coli in the presence of increasing concentrations of sodium deoxycholate (0, a), SDS (0, B) and Triton X-100 ( A , A). Open symbols, unirradiated culture; closed symbols, culture irradiated at 30 JlmZ. Growth was monitored by measuring the OD,,, of the medium after 8 h of postirradiation incubation at 37 OC . Detergent concentrations are expressed as % wlv.

UV-induced change in the outer membrane phospho1ipid:protein weight ratio

Isolation and analysis of the outer membrane constituents showed a characteristic increase in the phospholipid to protein weight ratio with increase in the UV dose. The data shown in Table 2 indicate a loss of outer membrane proteins, but no change in the absolute value of phospholipid over the dose range used. At a dose of 30 Jim2, the phospholipid to protein weight ratio increased by nearly twofold. It should be noted that the weight ratio of phospholipid to protein is subject to a considerable variation under different conditions. We do not know the extent of variation that can be attributed to the experimental artifacts of isolation and to what degree the protein content may have been overestimated by the method of Lowry et al. (1951), which is known to yield higher values for the protein content of outer membrane preparations (Schweizer et al. 1976; van Alphen et al. 1977). Therefore, we do not put more emphasis on the absolute values of proteins or phospholipids but, rather, on the relative differences in the phospholipid to protein weight ratio. No difference in the buoyant density of the "H band" was observed in spite of the changes in phospholipid to protein ratio. This could presum- ably be due to the presence of residual peptidoglycan components, which would increase the density of the "H band" in both the control and the irradiated samples.

Gel electrophoresis of outer membrane proteins Figure 4A shows the anodic nondenaturing PAGE profile of

outer membrane proteins of unirradiated and irradiated E. coli W3110. A gradual decrease in the intensity of certain protein bands was observed when cells were subjected to increasing

1 doses of UV radiation. Some proteins appeared to be unaffected I by irradiation. Similarly, SDS -PAGE of control and irradi- I ated (30 Jim2) sample showed a characteristic decrease in the

i intensity of certain bands, indicating the release or breakdown of certain membrane proteins upon irradiation. The release of

1 certain outer membrane proteins upon irradiation was charac-

TIME IN LIQUID MEDIUM (min)

FIG. 3. Time course recovery of irradiated E. coli from UV- induced sensitivity to ampicillin and phenol. Cells were irradiated at 30 Jlm2. Recovery medium consisted of glucose minimal medium supplemented with casamino acids and thiamine. Sensitization factor is the ratio of the MIC for unirradiated to that obtained for irradiated culture. (a) Ampicillin; (0) phenol.

terized by the simultaneous appearance of proteins in the CCF of the irradiated sample (Fig. 5B).

W-induced changes in the ultrastructure of the outer membrane By virtue of the fact that the freeze-etch technique reveals

the molecular details of the apposing two layers of the outer membrane, it was used to study the ultrastructural changes in the architecture of the outer membrane as a consequence of radiation damage. Figure 6 shows the outerJconcave) fracture faces of the outer membrane (0%). The OM of unirradiated cells (Fig. 6A) is completely covered with densely packed particles of 8 to 10 nm in diameter, which represent protein-LPS aggregates (Gilleland et al. 1974; Nikaido and Nake 1979; Osborn and Wu 1980; Verkleij et al. 1977 Ververgaert and Verkleij 1978). The number of these 0% particles were reduced by 50% upon irradiation at a dose of 30 Jim2 (Table 3). Similar membrane damage was also noticeable, but to a lesser extent, even at a UV dose of 10 Jim2 (Fig. 6B). The loss of 0% particles was seen among all the concave fracture faces screened. The simultaneous appearance of smooth areas surrounding the remaining 0% particles (Fig. 6C) suggests a reorientation of phospholipid molecules to fill up the void left by the selective removal of protein-LPS particles. However, no particle aggregation or any other deformity was observed. The outer membrane regained its normal particle density after 3 h of postirradiation incubation in recovery medium (Table 3).

Discussion Low doses of far-UV radiation were found to disrupt the

outer membrane permeability barrier of E. coli towards antibiotics, dyes, and detergents. Interestingly, the UV-induced sensitization showed an inverse correlation with hydrophobicity of the test compounds. We find that the permeability increasing action of far-UV is directed towards both hydrophobic and hydrophilic pathways, although favouring the latter.

Can

. J. M

icro

biol

. 198

9.35

:102

2-10

30.

Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IVE

RSI

TY

OF

MIC

HIG

AN

on

11/1

3/14

. For

per

sona

l use

onl

y.

1026 CAN. J. MICROBIOL. VOL 35, 1989

TABLE 2. Effect of UV radiation on the composition of the outer membrane of E. colia

Component per 1012 cells

Phospholipid Protein Phospholipidlprotein Treatment ( ~ m o l ) ~ (~mol/mg)

Unirradiated 5.25(5.22;5.28) 16.2(16.0;16.4) 0.32

Irradiated 10 J/m2 5.35(5.30;5.40) 12.0(11.3;12.7) 0.44 20 J/m2 5.40(5.33;5.47) 10.1(9.5;10.7) 0.53 30 J/m2 5.42(5.38;5.46) 8.6(8.0;9.2) 0.63

"Values given are means ohtamed from two Independent exper~mental results The numbers in parentheses re resent the values of indlv~dual determinations

'Phosphol~p~d is expressed as pmol of lipid phosphate, excluding KDO lipid A phosphate (see Materials and methods).

'Detenned by the method of Lowry et a1 (1951).

( A ) The UV-induced loss of CM particles and the simultaneous ( ) appearance of smooth interparticulate areas surrounding the remaining particles (Fig. 6) suggest a reorientation of

a b c d e a phospholipid molecules to fill up the void left by the selectee w - 'm removal of protein-LPS complexes that constitute these OM .. .-f particles. This assumption is based on the nonetchability of the - - - 4

150,. interparticulate areas, which many investigators have inter-

# preted to be phospholipid-rich "domains", a hypothesis now

aZ well accepted (Lugtenberg and van Alphen 1983; Nikaido and

+ 108- Nakae 1979; Schweizer et al. 1976; Smit et al. 1975; van

-. @ ?@ Alphen et al. 1977). Outer membranes of this type (i.e., (1* * containing phospholipid bilayer domains) are known to be

* & - permeable to hydrophobic molecules and the permeability is expected to be higher for more hydrophobic compounds

w -,. (Nikaido 1976; Nikaido and Nakae 1979; Nikaido and Vaara 1985). In contrast, the sensitization to hydrophobic agents was not significant, even at a dose of 30 J/m2 (Table 1). The reason for the failure of irradiated cells to show higher levels of sensitization to hydrophobic probes is not known. One pos-

_-L sibility could be that the fraction occupied by such phospho- 23 lipid bilayer domains in the outer membrane may not be large

enough (since there is no net increase in the amount of phospholipid in the outer membrane; Table 2) or that the

I^

-+- bilayer structure is well shielded from the medium. This

14- shielding effect may be caused by poor hydrophobic interactions between lipids and proteins. These assumptions are further supported by the fact that many irradiated cells failed to show significant sensitization to the anionic detergent SDS, which preferentially acts on phospholipid bilayer domains (Lugtenberg and van Alphen 1983).

In view of the fact that outer membrane particles are morphological reflections of hydrophilic pores (van Alphen

FIG. 4. (A) Nondenaturing polyacrylarnide (7%) slab gel electro- et al. 1978), the loss of these particles upon UV irradiation phoresis of outer membrane proteins of control and UV-irradiated (Table 3; Fig. 6) reflects a decrease in the number of porin E. coli W3110 cells. Lanes a to e correspond to the UV fluences of channels. Therefore, a decrease in the sensitivity to hydro- 0, 10, 20, 30, and 60 ~/m', respectively. The amount of protein philic antibiotics like penicillin, ampicillin, and tetracyclin loaded in each lane varied proportionately to the amount recovered was expected after irradiation since these antibiotics use the from a fixed number of cells (10" cells). The total amount of pro- porin pathway for their permeation (Voll and Leive 1970). In tein loaded in lanes a -e was 160, 120, 100, 86, and 80 pg, respectively. (B) SDS-PAGE of outer membrane fractions of E. coli contrast, it was observed that the sensitivity towards these ~ 3 1 1 0 . L~~~ a, nonirradiated (control); lane b, irradiated 30 ~ 1 ~ 2 . antibiotics increased sharply after irradiation. There is one The amount of protein loaded in each lane was 145 pglsample. The plausible mechanism to explain this anomaly the formation of molecular mass standards used were y-globulin (bovine), 150 kilo- aqueous pores in the phospholipid bila~er regions. These daltons (kDa); lypoxidase (soybean), 108 kDa; peroxidase (horse- pores may arise by a spontaneous micellar rearrangement of radish), 40 kDa; chymotrypsinogen A (bovine pancreas), 23 kDa; phospholipids to form a semi-vesiclelike structure, as described and lysozyme (chicken egg white), 14.2 kDa. by Houslay and Stanley (1984). These aqueous openings may

Can

. J. M

icro

biol

. 198

9.35

:102

2-10

30.

Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IVE

RSI

TY

OF

MIC

HIG

AN

on

11/1

3/14

. For

per

sona

l use

onl

y.

MODY ET AL. 1027

0 z LLJ

0.6

0 2 4 G 8 GEL LENGTH (cm)

FIG. 5. Densitometric scan (at 540 nm) of electrophoresed protein of the cell-free culture fdtrate of control (A) and irradiated (B) E. coli cultures on an anodic nondenaturing (7 %) polyacrylamide slab gel after staining with Coomassie blue R. The amount of protein loaded on the gel for both samples was different. It was equivalent to the amount recovered from the culture filtrate of 5 x lo9 cells, which was then concentrated up to 1.0 mL by lyophilization. Molecu- lar mass standards used were bovine serum albumin, 66 kDa; tryp- sinogen (PMSF treated), 24 kDa; trypsin inhibitor (soybean), 20.1 kDa; and a-lactalbumin (bovine milk), 14.2 kDa. Arrows indicate loading point.

allow easy diffusion of hydrophilic molecules, even under I reduced level of porins.

I The loss of outer membrane proteins may result from some modification of their interaction with the surrounding lipids.

I This could result either from direct damage to the proteins, with an ensuing change in their conformation or charge properties, or from photochemical damage to the surrounding lipids, thereby lessening the LPS -protein or phospholipid -

TABLE 3. Surface density of particles on the concave surface of the outer membrane of UV-irradiated E. coli W3110

as observed by freeze-etching electron microscopy

Postirradiation UV dose holding time Number of 0% particles

(JimZ) (min)" * SDb

0 0 8350+48 10 0 6853f51 20 0 5576 *65 30 0 4318f86 30 15 6563f 81 30 60 7096 * 75 30 120 7800 f 82 30 180 8330*90

"Freeze-etching replicas were prepared after holding the irradiated cells in ucose minimal medium (devoid of thymidine) for various time intervals.

$article densities were determined as described by Smit et al. (1975). The number of fractured surfaces scanned for each sample were in the range of 20 to 30 for two independently prepared sets of replicas. Values indicate mean _+ standard deviation.

protein interactions that hold the protein in place. A strong correlation has been previously reported to exist

between the reductioz of outer membrane proteins and LPS and a decrease in OM articles (van Alphen et al. 1978); .R therefore, the loss of 0 particles observed here (Fig. 6) implies LPS is released from the outer membrane when cells are subjected to UV radiation. However, chemical estimation of LPS content was not done since meaningful estimation of LPS content can only be possible if one correlates the number of LPS molecules per unit surface area of the membrane. For this, it is imperative that all LPS molecules must be of pre- cisely the same structure and all cells, of identical surface area. As neither of these restrictions would be met in the case of irradiated cells. estimation of LPS on the basis of KDO content alone would invalidate any of the conclusions drawn. However, in view of the freeze-etch data, the loss of LPS molecules has been considered.

There appears to be some specificity in the action of far-UV light on the components of the outer membrane, some proteins getting released in preference to others (Fig. 4). This apparent specificity could be due to the inherent differences in the UV sensitivity of the proteins or different susceptibilities of the proteins to release from the membrane when surrounding lipids are damaged. It is not known whether the loss of these proteins from the outer membrane is truly complete since their presence below 5% of their normal concentration would no longer be detectable with the methods presently employed (Hofstra et al. 1979). There are several mechanisms that can explain the loss of protein bands on the gel. One possibility is that the protein becomes cross-linked either to another identical protein molecule or to some other protein, thereby changing its electrophoretic mobility. If the newly formed, cross-linked molecules are heterogenous in structure, they would migrate at various rates and go undetected on a polyacrylamide gel. However, there is no supportive evidence to show that UV radiation induces protein -protein cross-links, although DNA - protein cross-links have been observed (Setlow and Carrier 1966). Another possibility is that far-UV light may be causing a degradation of membrane proteins, yielding smaller polypeptides that migrate more rapidly. In that case, distinct bands

Can

. J. M

icro

biol

. 198

9.35

:102

2-10

30.

Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IVE

RSI

TY

OF

MIC

HIG

AN

on

11/1

3/14

. For

per

sona

l use

onl

y.

-

1028 CAN. J. MICROBIOL. VOL. 35, 1989

. .

.";: i.." < : . . , - . . . - - - - - , ? 2 . L U

Can

. J. M

icro

biol

. 198

9.35

:102

2-10

30.

Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IVE

RSI

TY

OF

MIC

HIG

AN

on

11/1

3/14

. For

per

sona

l use

onl

y.

MODY ET AL. 1029

on the gel may not appear if their sizes were heterogenous. However, if a protein species were cleaved at a distinct site or sites, the resultant polypeptides could migrate as distinct bands of low molecular weight. This may be one of the reasons for the origin of the new bands (closed arrows in Fig. 4B). Alter- natively, these high intensity bands may be of membrane proteins that are not affected by UV radiation. Their intensity is

1 likely to be more than in control, because the amount of protein loaded in each lane was same and hence these proteins

I would make up for the lost proteins. If the radiation damage I is in the form of breakage of disulfide bonds of membrane

proteins, which is an important mechanism in UV-induced inactivation of proteins (McLaren 1955; Setlow and Pollard 1962), then the UV-induced monomerization of oligomeric

I proteins cannot be distinguished from the identical effect of the detergent during SDS - PAGE. Therefore, nondenaturing PAGE of outer membrane proteins was performed, which confirmed the results obtained by SDS -PAGE. A dose-dependent loss of certain outer membrane proteins was observed, indicating either a selective release or degradation (breakdown) of membrane proteins. In either case, the proteinic matter would appear in the CCF, either in intact or in degraded form. The CCF of irradiated cells showed a broad (unresolved) peak of low intensity, which reflects a heterogenous mixture of polypeptides ranging from 10 to 70 kilodaltons in molecular mass (Fig. 5) which probably represents the breakdown products of membrane proteins.

It is clear that the action of far-UV on the outer membrane does not reflect mutagenic alteration in the genes governing membrane properties because (i) the effects were observed immediately after irradiation, before any genotypic alteration could be expressed; (ii) after examining a large number of electron micrographs, the membrane damage and its recovery were found to be uniform and gradual, respectively, with all the cells responding simultaneously; and (iii) the extent of radiation injury to the outer membrane in terms of structural and functional changes was the same in a DNA repair pro- ficient strain AB1157 and in DNA repair deficient strains AB1884 (uvrC34), AB1885 (uvrBS), AB1886 (uvrA6), BRc49 (recA56), and AB2480 (recAl3 uvrA6) for a specific dose of UV light (results not shown).

The results of the membrane repair assay (Fig. 3) indicate that the normal levels of resistance to a hydrophilic and a hydrophobic agent were restored in the 2nd and 3rd hour of postirradiation incubation, respectively. This indicates that the membrane assembly process preferentially repairs the hydrophilic pathway, prior to the restoration of the hydrophobic barrier. The recovery observed was not apparently due to growth of the irradiated cells since E. coli have been known to exhibit a growth lag of 1 to 2 h following UV irradiation (Kelner 1953), and also because the recovery medium lacked thymidine, which prevented subsequent growth. The membrane repair process appears to commence immediately following irradiation as some degree of restoration of the barrier function was observed in the first 15 min of recoverv time. Thus. the process that restored impermeability was gradual and energy dependent. The reacquisition of resistance towards

inhibitory agents was found to coincide (with respect to time) with the restoration of normal membrane structure (Fig. 3; Table 3).

We do not have a final answer regarding the molecular basis of UV action on the outer membrane or the mechanism under- lying the recovery of normal structure and function of the membrane. What appears to be conclusive from this study is that UV light induces a spontaneous, nongenotypic alteration in the chemistry and architecture of the outer membrane. Hence, these findings suggest a new approach for studying the details of molecular organization of the outer membrane. The use of UV radiation in the study of nonpermeable substances of microbiological interest, for example, mutagens, antimetab- olites, and inhibitors, should be considered.

Acknowledgements We thank Dr. A. Verkleij of the Department of Microbiol-

ogy and Biological Ultrastructure Research Unit, Utrecht (The Netherlands), for making laboratory space available during the course of one summer and for the technical assistance with freeze-fracture electron microscopy. This work was supported by the Council of Scientific and Industrial Research, New Delhi, by way of a Senior Research fellowship to R.M.

AMES, G. F. 1974. Resolution of bacterial proteins by polyacrylamide gel electrophoresis on slabs. Membrane soluble and periplasmic fraction. J. Biol. Chem. 249: 634-644.

BARTLETT, G. R. 1959. Phosphorus assay in column chromatog- raphy. J. Biol. Chem. 234: 466-468.

BAVOIL, P., NIKAIDO, H., and VON MEYENBURG, K. 1977. Pleiotropic transport mutants of Escherichia coli lack porin, a major outer membrane protein. Mol. Gen. Genet. 158: 23-33.

BENSON, S. A., and DECLOUX, A. 1985. Isolation and characteriza- tion of outer membrane permeability mutants in Escherichia coli K-12. J. Bacteriol. 161: 361-367.

BLIGH, E. G., and DYER, W. J. 1959. A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol. 37: 911 -917.

COLEMAN, W. G., JR., and LEIVE, L. 1979. Two mutations which affect the barrier function of the Escherichia coli K-12 outer membrane. J. Bacteriol. 139: 899 -910.

DAVIS, B. J. 1964. Disc electrophoresis 11. Method and application to human serum. Ann. N.Y. Acad. Sci. 121: 404 -427.

GILLELAND, H. E., STINNETT, J. D., and EAGEN, R. C. 1974. Ultrastructural and chemical alteration of the cell envelope of Pseudomonas aencginosa, associated with resistance to ethylene- diaminetetraacetate resulting from growth in a Mg2+-deficient medium. J. Bacteriol. 117: 302 -31 1.

HARADAWAY, K. L., and BULLET, C. S. 1979. Effect of EDTA on phospholipids and membrane function in Escherichia coli. J. Bac- teriol. 137: 62-68.

HOFSTRA, H., VAN TOLL, M.J.D., and DANKERT, J. 1979. Immuno- fluorescent detection of the major outer membrane protein II* in Escherichia coli O2Sm. FEMS Microbiol. Lett. 6: 147-150.

HOUSLAY, M. D., and STANLEY, K. K. 1984. Dynamic membrane- associated processes. In Dynamics of biological membranes: influence of synthesis, structure and function. Edited by M. D . Houslay and K. K. Stanley. John Wiley and Sons, Chichester, New York. pp. 281 -324.

FIG. 6. Freeze-etch electron micrographs of concave (outer) fracture faces of the outer membrane (0%) of E. coli W3110. (A) Unirradiated; (B) irradiated at 10 Jlm2; (C) irradiated at 30 Jlm2. Samples were quenched at 4OC. Arrows indicate the direction of shadowing. Bars represent 0.1 pm.

Can

. J. M

icro

biol

. 198

9.35

:102

2-10

30.

Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IVE

RSI

TY

OF

MIC

HIG

AN

on

11/1

3/14

. For

per

sona

l use

onl

y.

1030 CAN. J . MICROBI(

KELLAND, L. R., MOSS, S. H., and DAVIES, D. J. G. 1983. An action spectrum for ultraviolet radiation-induced membrane damage in Escherichia coli K-12. Photochem. Photobiol. 37: 301 -306.

KELNER, A. 1953. Growth, respiration and nucleic acid synthesis in ultraviolet-irradiated and in photoreactivated Escherichia coli. J. Bacteriol. 65: 252 -262.

LEIVE, L. 1974. The barrier function of the gram-negative envelope. Ann. N.Y. Acad. Sci. 235: 109-127.

LOWRY, 0. H., ROSENBROUGH, N. J., FARR, A. L., and RANDALL, R. J. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193: 265-275.

LUGTENBERG, B., and VAN ALPHEN, L. 1983. Molecular architecture and functioning of the outer membrane of Escherichia coli and other gram-negative bacteria. Biochim. Biophys. Acta 737: 51-115.

LUGTENBERG, B., PETERS, R., BERNHEIMER, H., and BERENDSEN, W. 1976. Influence of cultural conditions and mutations on the composition of the outer membrane proteins of Escherichia coli. Mol. Gen. Genet. 147: 251 -262.

MAKELA, P. H., and STOCKER, B. A. D. 1981. Genetics of the bacterial cell surface. Symp. Soc. Gen. Microbiol. 31: 219-264.

MCLAREN, A. D. 1955. Relations among cystine content, molecular weights and ultraviolet light sensitivity of proteins. Biochim. Biophys. Acta, 18: 601 -602.

MILLER, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.

MYERS, D. K. 1970. Some aspects of radiation effects on cell membrane. In Biological and medical physics. Vol. 13. Edited by J. H. Lawrence and J. W. Gofman. Academic Press Inc., New York. pp. 219-234.

NIKAIDO, H. 1976. Outer membrane of Salmonella typhimurium: transmembrane diffusion of some hydrophobic substances. Biochim. Biophys. Acta, 433: 118- 132.

NIKAIDO, H., and NAKAE, T. 1979. The outer membrane of gram- negative bacteria. Adv . Microbiol. Phy siol. 20: 163 -250.

NIKAIDO, H . , and VAARA, M. 1985. Molecular basis of bacterial outer membrane permeability. Microbiol. Rev. 49: 1 -32.

OSBORN, M. J., and Wu, H. C. P. 1980. Proteins of the outer membrane of gram-negative bacteria. Annu. Rev. Microbiol. 34: 369 -422.

ROANTREE, R. J . , KUO, T. T., and MACPHEE, D. G. 1977. The effects of defined lipopolysaccharide core defects upon antibiotic resistance of Salmonella typhimurium. J. Gen. Microbiol. 103: 223 -234.

SANDERSON, K . E., MACALISTER, T., and COSTERTON, J. W. 1974. Permeability of lipopolysaccharide-deficient (rough) mutants of Salmonella typhimurium to antibiotics, lysozyme, and other agents. Can. J . Microbiol. 20: 1135 - 1145.

SCHINDLER, H . , and ROSENBUSCH, J. P. 1981. Matrix protein in planar membranes: clusters of channels in a native environment and their functional reassembly. Proc. Natl. Acad. Sci. U.S.A. 78: 2302 -2306.

SCHWEIZER, M., SCHWARZ, H., SONNTAG, I., and HENNING, U. 1976. Mutational change of membrane architecture. Mutants of Escherichia coli K-12 missing major proteins of the outer cell envelope membrane. Biochim. Biophys. Acta, 448: 474 -491.

3L. VOL. 35, 1989

SETWW, R. B., and CARRIER, W. L. 1966. Pyrimidine dimers in ultraviolet-irradiated DNA's. J. Mol. Biol. 17: 237-254.

SETWW, R. B., and POLLARD, E. C. 1962. Action spectra and quan- tum yields. In Molecular biophysics. Edited by R. B. Setlow and E. C. Pollard. Addition-Wesley Publishing Company, Inc., London. pp. 267 - 306.

SMIT, J., KAMIO, Y., and NIKAIDO, H. 1975. Outer membrane of Sal- monella typhirnurium : chemical analysis and freeze fracture studies with lipopolysaccharide mutants. J. Bacteriol. 124: 942-958.

TAMAKI, S., SATO, T., and MATSUHASHI, M. 1971. Role of lipopolysaccharides in antibiotic resistance and bacteriophage adsorption of Escherichia coli K-12. J . Bacteriol. 105: 968-975.

VAARA, M., and VAARA, T. 1983a. Polycations sensitize enteric bacteria to antibiotics. Antimicrob. Agents Chemother. 24: 107- 113.

1983b. Polycations as outer membrane disorganizing agents. Antimicrob. Agents Chemother. 24: 114 - 122.

VAN ALPHEN, L., LUGTENBERG, B., VON BOXTEL, R., and VERHOEF, K . 1977. Architecture of the outer membrane of Escherichia coli K-12. I. Action of phospholipases A2 and C on wild-type strains and outer membrane mutants. Biochim. Biophys. Acta, 466: 257-268.

VAN ALPHEN, L., VERKLEIJ, A., LEUNISSEN-BIJVELT, J., and LUGTEN- BERG, B. 1978. Architecture of outer membrane of Escherichia coli. 111. Protein - lipopoly saccharide complexes in intramem- braneous particles. J . Bacteriol. 134: 1089-1098.

VAN ALPHEN, L., VAN ALPHEN, W., VERKLEIJ, A. J., and LUGTEN- BERG, B. 1979. Architecture of the outer membrane of Escherichia coli K-12. IV. Relationship between outer membrane particles and aqueous pores. Biochim. Biophys. Acta, 556: 233-243.

VAN ALPHEN, W., LUGTENBERG, B., and BERENDSON, W. 1976. Heptose-deficient mutants of Escherichia coli K-12 deficient in up to three major outer membrane proteins. Mol. Gen. Genet. 147: 263 -269.

VERKLEIJ, A., VAN ALPHEN, L., BIJVELT, J., and LUGTENBERG, B. 1977. Architecture of the outer membrane of Escherichia coli K-12. Freeze fracture morphology of wild-type and mutant strains. Biochim. Biophys. Acta, 466: 269-282.

VERVERGAERT, P. H. J. T., and VERKLEIJ, A. J. 1978. A view on intra- membraneous particles. Experientia, 34: 454-455.

VOLL, M. J., and LEIVE, L. 1970. Release of lipopolysaccharide in Escherichia coli resistant to the permeability increase by ethylene- diarninetetraacetate. J. Biol. Chem. 245: 108 - 1 14.

WALLACH, D. F. H. 1972. Membrane effects of radiation. In The plasma membranes. Dynamic perspectives, genetics and pathol- ogy. Edited by D. F. H. Wallach. The English University Press, London. pp. 145 - 155.

WILKINSON, R. G., GEMSKI, P,, and STOCKER, B. A. D. 1972. Non- smooth mutants of Salmonella typhimurium: differentiation by phage sensitivity and genetic mapping. J . Gen. Microbiol. 70: 527-554.

WITHOLT, B., BOCKHOUT, M., BROCK, M., KINGMA, J., VAN HEERIK- HUIZEN, H., and DE LEIJ, L. 1976. An efficient and reproducible procedure for the formation of spheroplasts from variously grown Escherichia coli. Anal. Biochem. 74: 160 - 170.

Can

. J. M

icro

biol

. 198

9.35

:102

2-10

30.

Dow

nloa

ded

from

ww

w.n

rcre

sear

chpr

ess.

com

by

UN

IVE

RSI

TY

OF

MIC

HIG

AN

on

11/1

3/14

. For

per

sona

l use

onl

y.

influence of far-ultraviolet radiation on the permeability of the outer membrane of ...

Documents