Pseudomonas aeruginosa cytotoxin: the influence of sphingomyelin on binding and cation permeability increase in mammalian erythrocytes

Tosiroa Vol . 27, No . S, PP. 531-540, 1989. Printed in Great Britain .

0041-0101/89 53 .00+ .00 ~ 1989 Pergatnoo Prees p~

PSEUDOMONAS AERUGINOSA CYTOTOXIN : THE INFLUENCE

OF SPHINGOMYELIN ON BINDING AND CATION PERMEABILITY INCREASE IN MAMMALIAN ERYTHROCYTES KATHLEEN

M.

CROWELL

and F. LuTZ*

Institut fier Pharmakologie und Toxikologie, Jaslus- Liebig-Universitát Giessen, Frankfurter Str . 107, D-6300 Giessen, F .R.G . (Accepted jor publication 15 November 1988)

K. M. CROWELL and F. LUTZ . Pseudomonas aeruginosa cytotoxin : the influence of sphingomyelin on binding and cation permeability increase in mammalian erythrocytes . Toxicon 27, 531-540, 1989 .-A cytotoxic protein isolated from Pseudomonas aeruginosa damages the plasma membranes of many mammalian cells by forming pores. We studied binding of the ' 25I-cytotoxin and the resulting increase of cation permeability in erythrocytes of various mammalian species. The sensitivity of red blood cells was inversely related to the relative sphingomyelin content in their external surface. Thus, erythrocytes with a sphingomyelin to phosphatidylcholine ratio below 1 (dog, rat, rabbit and man) were sensitive, whereas red blood cells with a ratio above 1 (pig, cattle and sheep) were not attacked even with 100-fold higher cytotoxin concentrations . At 37°C 6.8 f 1 .2 x 103 molecules of 125 I_cytotoxin were bound per rabbit erythrocyte (KD =59 nM), whereas no binding occurred to cattle cells. Cleavage of sphingomyelin by sphingomyelinase C from Bacillus cereus (EC 3.1 .4.12) triggered a dose-dependent enhancement in binding and permeability increase, particularly in red blood cells with a high proportion of sphingomyelin. The KDs for all animal species investigated were 530 nM . Pretreatment with mainly phosphatidylcholine-hydrolyzing phospholipases D from Streptomyces chromofuscus and cabbage (EC 3 .1 .4.4) or phospholipase C from Bacillus cereus (EC 3.1 .4.3) did not influence the cytotoxin effect . The negative correlation between susceptibility and the proportion of sphingomyelin in plasma membranes suggests a binding site close to sphingomyelin . INTRODUCCION

Pseudomonas aeruginosa is an increasingly important pathogen in medical- and veterinaryacquired infections and is associated with a high mortality rate (DnLxor-i:, 1987; Hn~z et al., 1987). Among other substances, P. aeruginosa produces proteins, which are involved in pathogenesis (VASIL, 1986). The pseudomonal cytotoxin (SCHARMANN, 1976; LUTZ, 1979), an acidic protein of approximately 28,000 mol.wt, acts primarily on plasma membranes. The toxin binds to a high affinity binding site on Ehrlich ascites tumor cells (Llrrz, 1986) To whom correspondence should be addressed. 531

532

K. M. CROWELL and F. LUTZ

and forms pores with a diameter of about 2 nm as shown on endothelial cells, rat erythrocytes and Ehrlich cells (Su~rroRP et al., 1985; WE1xER et al., 1985; LUTZ et al ., 1987). Pretreatment of Ehrlich cells with various phospholipases indicated a correlation between phospholipid composition and intoxication (LEWICIü and LUTZ, 1985). An important role was attributed to the sphingomyelin content of the cells. We now have analysed the binding and cytotoxin-induced changes in permeability of plasma membranes with special respect to their sphingomyelin content. As target cells we chose erythrocytes of various mammalian species, because of their varying phospholipid composition .

Materials

MATERIALS AND METHODS

Cytotoxin was prepared from an autolysate of P. aeruginosa strain 158 (Ltrrz, 1979) and stored at - 20°C in phosphate-buffered saline (137 mM NaCI, 2.7 mM KCI, 0.5 mM MgC12, 9.3 mM phosphate, pH 7 .4). The purity of the toxin was about 98% when tested by gel electrophoresis under denaturing and reducing conditions and visualized by silver staining. The preparation was devoid of lipids, phosphate, carbohydrate and of any enzyme activity (Ltrrz, 1979). Iodination was carried out with immobilized lactoperoxidase/glucose oxidase (Bio Rad Laboratories, Richmond, CA, U.S .A .) for lOmin as described by MtYACHI et al. (1973) at a molar ratio of cytotoxin:Na' 2'I (Amersham, Braunschweig, F.R .G .) of 1:0.8 . Iodination did not impair the toxicity of the cytotoxin (Ltrtz et al., 1981) when tested on human granulocytes in the slide adhesion test of Gt .~us~ror~ and vex HEYNINGEN (1957). The specific radioactivity was 0.210 .05 Ci/pmole (n= l2). Ten milliliters of fresh blood (dog, rat, rabbit, human, pig, cow or sheep) were mixed with 1 ml of 10% of sodium citrate, centrifuged at 1000 X g for 5 min and the pellet washed three times with phosphate-buffered saline. The required cell volume was adjusted using haematocrit tubes centrifuged for 4min at 10,000 xg. Phospholipase C from B . cereus (EC 3.1 .4 .3), sphingomyelinase C from B. cereus (EC 3.1 .4.12) and phospholipase D from S. chromofuscus (EC 3.1 .4.4) were purchased from Bcehringer (Mannheim, F.R .G.). Phospholipase D from cabbage (EC 3.1 .4.4) was obtained from Sigma (St Louis, MO, U.S .A .). Sphingomyelinase C pretreatment of erythrocytes was carried out in a buffer solution containing 137 mM NaCI, 2.7 mM KCI, 1 mM MgCI  9.3 mM phosphate, pH 7.4 . For incubation with phospholipase C 1 mM MgClz was replaced by 1 mM ZnClz. Pretreatment with phospholipases D was performed in 110 mM NaCI, 8 mM KCI, 50 mM CaCI 20 mM Tris-HCI . The pH was adjusted to 6.0 for phosphoGpase D from cabbage and to 8.0 for that from S. chromojuscus. Binding assay

The samples (final volume 0.25 ml) wntained phosphate-buffered saline, pH 7.4, 0.2 mg bovine serum albumin, 0.1 mM phenylmethylsulfonyl fluonde, erythrocytes (1 % v/v at 4°C; 3.5% v/v at 37°C) and, if required, unlabeled cytotoxin up to 9 pM . After preincubation (3TC, 30 min), the reaction was started by adding I .2 nM (at 4'C) and 2.2 nM (at 3TC) of "sl~ytotoxin . The reaction was stopped, after 2 hr at 4'C and 30 min at 3TC, by addition of 4 ml of ice-cold buffer. The diluted suspension was filtered immediately under vacuum through membrane filters (Schleicher/Schûll OE 66, 0.2 gym), rapidly washed three times with 3 ml of buffer and counted in a Kontron autogamma system (55% counting efficiency). The non specific adsorption on the filters amounted to 0.8% of the total applied radioactivity. Assay ofpermeability change

Erythrocytes used in these experiments were prepared in either 75 mM NaCI, 1 mM MgCI, and 70 mM potassium phosphate=buffer A (rat, rabbit, man and pig) or 75 mM KCI, 1 mM MgClz and 70mM potassium phosphate=buffer B (dog, cow and sheep), pH 7.4 . The reaction mixture consisted of the respective buffer and cytotoxin in the range of 0.05-40 pM (depending on the species) in a final volume of 0.5 ml . The reaction was started by addition of erythrocytes (3 .5%, v/v) . After 30 min at 37°C the sampleswere centrifuged at 10,000 X gfor 2 min. The pellet was lyophilized, suspended in 0.5 ml distilled water and sonicated. Protein was precipitated by addition to 0.25 ml of 10% tricholoroacetic acid . After centrifugation (2 min, 10,000 x g), Na* and K* were determined in the supernatant by flame photometry (Eppendorf, Hambwg, F.R.G .). Preincubation of erythrocytes with phosphoLpases

Erythrocytes (3.5%, v/v) were incubated at 3TC for 30 min with either phospholipase C, D or sphingomyelinase C. The reaction was terminated by addition of 0.15 M EDTA (ethylenediaminetetraacetic acid; final concentra-

Pseudomonas aeruginosa Cytotoxin

53 3

TABLE 1 . PFIOSPHOLIPASFS, THEIR EFFECT ON THE PHOSPHOLIPIDS OF VARIOUS ERYTHROCYTES AND IT~üt CONCENTRATION REQUIRFD FOR MAXIMAL CLEAVAGE

Agent (source)

Species

n

Concentration required for max. cleavage (LJ/ml)

Sphingomyelinase C (B. eereus)

Dog Rabbit Man Pig Cow$ Cow§ Rabbit Pig Rabbit

7 7 7 7 7 3 3 3 3

0. l 0. l 0 .2 0 .4 1 .0 0.3 4.0 10 .0 80 .0

83 f 2 82 f 2 78 f 5 79 f 6 8l t4 53 t 2

Rabbit

3

5 .0

Phosphohpase D (S. chromofuscus) Phospholipase D (cabbage) Phospholipase C (B . cereus)

% Lipid degraded SM "

PC Mean f S .D .

t

PE

t

t

20 t 2 19 f 2 28 f 4

4t I 8f3 4f2

t

23 t 5

5f3

t t

" SM, sphingomyelin; PC, phosphatidylcholine ; PE, phosphatidylethanolamine . tLess than 2% . $Used in binding experiments . §Used in assays of permeability change .

tion) in the respective incubation buffer devoid of the corresponding activating cation to chelate Cat' (pho¢pholipases D) and Mg" (sphingomyelinase C), or by 10 mM phenanthroline to remove Zn=' (phospholipase C). After a further incubation of 30 min at 3TC the suspension was centrifuged for 5 min at 10,000 x a. The supernatant was collected, boiled for 5 min, centrifuged for 2 min at 10,000 x a and used for the colorimetric determination of the liberated phosphorylcholine/choline as described by IMAMURA and HORnnT (1978) (see below) . The erythrocytes were washed once in phosphato-buffered saline (pH 7.4) for binding assay or in buffer A or buffer B (see above) for experiments on permeability change . Then the required cell volume was adjusted using haematocrit tubes and centrifuged for 4 min at 10,000 X d.. The concentrations of enzyme required for maximal cleavage of the corresponding phospholípids are given in Table 1 . Phospholipid composition and extent ofphospholipid degradation by phospholipases Phospholipase pretreated or untreated erythrocytes were exfractéd áCcording to the procedure of YANO et al. (1984) in chloroform/methanol (2.5:3 ., v/v) . The lipid extract ; was loaded onto a thin-layer plate (Silica Gel 60 F 259, Merck, Darmstadt, F.R .G .) and chromatographed with chloroform :methanol:waterconcentrated NH, (48:40 :7 :5) for the first dimension and chloroform :methanol :formic acid (55 :25 : for the second. The plate was then stained under iodine vapor. The areas of the lipid spots were scraped into glass tubes and phosphate was assayed by a microadaptation of the method of Bartlett as described by MrrCHELL et al. (1986) . Appropriate phospholipid standards (sphingomyelin, phosphatidylcholine, -ethanolamine, -serine) were co-migrated for identification of the lipids. Percentage degradation of glycerophosphoGpids after phospholipase C (B. cereus) or D (cabbage, S. chromojuscus) attack was determined using either sphingomyelin or phosphatidylserine (which are not degraded) as an internal standard . For estimation of sphingomyelin hydrolysis by sphingomyelinase C (B. cereus), phosphatidylethanolamine served as the internal standard. In addition, the extent of sphingomyelin cleavage by sphingomyelirlase was assayed by the method of IMAMURA and HORIUTI (1978) . The reaction mixtures were as follows: 90 U of penoxidase (EC 1 .11 .1 .7), 30 U of choline oxidase (EC 1 .1 .3 .17), 3 mg of 4-aminoantipyrine in 50 mM Tris-HCI, pH 8 .0 (final volume 10 ml). An aliquot of 0.1 ml supernatant (after preincubation), 0.4 ml of the reaction solution and 10 U alkaline phosphatase (EC 3 .1 .3 .1) were incubated for 1 hr at 37°C. The liberated choline was measured at 492 nm, using a choline chloride solution as standard (21+g=0.21 absorbance difference) . Statistics The values are given as arithmetical means f S.D . with n equal to the number of experiments . Student's t-test for paired observations was used for statistical analysis . Kinetic experiments were calculated according to BENNEIT (1978) .

S34

K. M . CROWELL and F . LUTZ

TABLE 2 . PHOBPHOLIPID COMPOSITION OF ERYTHROCYTES AND THE CYTOTOXIN CONCENTRATIONS REQUIRED FOR SO% LOSS OF K' (RAT, RAHBrf, HUMAN, PIG AND SHEEP) OR Na' (DOC AND OOW) WITH OR WITHOUT SPHINGOMYELINASE C PRETREATMENT Ec,°t (pM) Sphingomyelinase pretreatment

Phospholipid composition (%) Species

SM'

PC

PE

PS

SM/PC

-

+

Dog Rat Rabbit Human Pig Cow Sheep

12 .4 13 .8 18 .2 25 .0 28 .1 44 .6 53 .1

48 .7 49.2 37.7 30.9 21 .7 3 .8 <2

21 . S 20 .9 33 .8 28 .4 30 .5 31 .7 27 .6

1 S .6 14 .8 10 .4 14 .7 19 .6 16 .4 17 .8

0.25 0.28 0.48 0.81 1 .29 11 .70 > 26

0 .20 f 0 .02 0 .23 t 0 .02 0 .69 t O .OS 10 .4 t 0 .4 § § §

0.18 t 0.03 ND 0.39 f 0.01$ 2.9 f 0.9 $ 20.0 ±0.4 7.2 tO.S ND

'SM, sphingomyelin; PC, phosphatidylcholine; PE, phosphatidylethanolamine ; PS, phosphatidylserine. effective concentration of cytotoxin required for 50% loss of the main intracellular cation . $Significantly different from erythrocytes without pretreatment (P < O .OS). §No change of the cellular cationic gradient emerged with cytotoxin concentrations up to 401IM . ND, not determined . The Ec,° values are shown as means f S .D.; n = 3. The determination of the phospholipid composition was carried out twice for each animal species . tEC,° ,

RESULTS

Phospholipid composition The phospholipid composition in erythrocyte membranes of various animal species is given in Table 2. The results are in fair agreement with data published previously (JOKINEN and GAi-nHSaRG, 1979; BARSxxol.z and TxoMPSOx, 1980; KUYPERS et al., 1985) . Spec city oj'~sl-cytotoxin binding Binding of 'ssI-cytotoxin was studied with rabbit, human and bovine erythrocytes (Fig. 2). Saturation of the binding and the time and temperature dependence revealed its specificity . Figure 1 presents the'z'I-cytotoxin binding to rabbit red blood cells at 37°C. At 37°C specific binding was in equilibrium after 25 min, at 4°C within 2 hr (data not shown) and lasted for several hr. The binding was largely reversible (95% at 4° C, 92% at 37°C) by unlabeled cytotoxin in a monophasic manner . ' z sI-cytotoxin was bound with a half time of < 1 min and less than 5 nM cytotoxin were required to obtain a measuuable binding (not shown). From a Scatchard plot (inset Fig .2), only one class of binding sites was ascertained . As shown in Table 3, at 4°C, the binding capacity of rabbit erythrocytes was seven-fold higher as compared to 37°C, and even human cells bound ' z s I-cytotoxin at 4°C . Cattle red blood cells displayed no measurable binding either at 37°C or 4°C. Hence, the less sphingomyelin located on the membrane exterior, the more cytotoxin is bound . In addition, the KDS for all animal species investigated were similar (530 nM) (Table 3) at both of the temperatures. Nonspecific binding did not vary between 4°C (3.8±2 fmole/ 9 x 10' cells) and 37°C (3.3 f 1 .6 fmole/9 x 10' cells), and changed to a small extent only depending on time (Fig. 1). Binding of l zsi-cytotoxin to sphingomyelinase Cpretreated erythrocytes Sphingomyelinase C (B. cereus) cleaves phosphorylcholine from sphingomyelin of the red blood cell membrane and leaves N-acyl-sphingosine (ceramide) in the bilayer (BAREN-

Pseudomonas aeruginosa Cytotoxin

535

FAG. 1. Tt~-couRSE oF'~'I~rTaroxiN BnvnnaG To RABBrr aerrHAOCVTFS . Red blood cells were incubated at 37°C with '=°I-cytotoxin (2 .2 nM) in the absence (p) and presence (D) of SpM unlabeled cytotoxin. Specific binding (O) was defined as the difference between these two values. Dissociation was determined by addition of unlabeled cytotoxin (arrow) at the indicated times ( "), means t S.D., n=4. The association velocity constant was 0.005/nM/min, the dissociation velocity constant 0.082/nM/ min and the K° 16 nM . xoLZ and Txol~soN,1980;

hc>EZnwn et al., 1986). As mentioned above (Table 1) up to 83% (depending on the species) of the outer leaflet sphingomyelin was hydrolysed. Figure 3 shows, that sphingomyelinase preincubation promoted the cytotoxin effect. Compared to untreated cells the binding increases for all mammalian blood cells investigated (see Table 3). Only a slight enhancement was achieved with rabbit erythrocytes, while with human and especially with cattle cells a distinct increase occurred.

0.01 FIG .

2.

0.1

1

NM

10

cytotoxin

CO~E'171TON OA 1 ZSI-CYT070XIN WIl'H NATIVE CYT01nXIN FOR HIGH AFFINITY BINDING SrrE3 ON ERYTHROCYTES AT 3iC.

Rabbit (O) n=7, human (p) n=3 and cow (p) n=4 all suspensions (3.5% v/v) were used. The values are given as means t S.D . the inset shows Scatchard plot analysis . For details see Methods.

536

K. M . CROWELL and F. LUTZ

TABLE

3.

TFn: INFLUENCE OF SPHINGOMYELW CLEAVAGE BY SPHINGOMYELINASE C FROM B . CereuS ON TEn; SPECIFIC

izsl_~.LOXIN

HINDING TO ERYTFmOCYTES

Specific' 2°I-cytotoxin binding Species

SM': PC ratio

Rabbit

0.48

Human

0.81

Cow

Molecules bound/cell x 10'

K° (nM)

Sphingomyelinase pretreatment

n

37°C

4°C

37°C

4°C

+ + +

7 7 3 3 4 4

0.68f0.12 0.84t0.1Ot NM O.OSf0.02 NM O.ISf0.05

4.4 t 0.5 4.8t0 .4 1.2 f 0.4 2.1f0 .2$ NM ND

59 f6 58±3 NM 60±3 NM 53t5

54 t8 58t5 58 t3 SSf7 NM ND

11 .7

'SM, sphingomyelin; PC, phosphatidylcholine. tSignificantly different from erythrocytes without pretreatment (P < 0.05). NM, bound radioactivity was less than two-fold of the background (about 300 cpm) corresponding to l0 molecules bound per cell. ND, not determined, because haemolysis occurred when brought to 4°C. All results are shown as meansfS.D .

Cytotoxin-induced permeability increase and the influence of sphingomyelinase C pretreatment The loss of 50% of the main intracellular cation was reached within 1 min after intoxication . As shown in Fig. 4 the sensitivity of various mammalian erythrocytes to the cytotoxin decreases in the order dog> rabbit > man. No permeability increase was obtained

E

0.08

N O

\

ui N

ö ' E

FiG.

3.

w

\~ ~ 0.5 1 1.5 7 \i ~~ mol bound/9x10 ~~\~i cells ~T~

SPECQ~7C BINDING OF ' 2s I-CYI'OTOXIN TO SPFIINGOMYELINASE P Rr~F~ TED ERYTHItOCYTI~ AT

37°C. Rabbit (~) n=7, human (/) n=3 and ww (~) n=4, means f S.D . The inset shows Scatchard plot analyses .

Pseudomonas aeruginosa

Cytotoxin

53 7

loss of the main intracellular monov4lent cation k 100 0

c0 ca

ó 50~

0 0.01

Q1

1

10

100

NM gtotoxin

FIG . 4. EFPE(T OF PRETREATMENT WITH SPHINOOMYELINASE C FROM B. CCrCYS ON TED: CYT010XINIIdDUCED BREAKDOWN OF THE CELLULAR CATIONIC GRADIINT OF VARIOUS MAMMALIAN ERYTHROCYTES.

After incubation of 30 min at 37°C with cytotoxin concentrations in the range of O.OS-40 pM (depending on the animal species) the reaction was stopped and the intracellular K' (rabbit, man, pig) or Na' (dog, cattle) concentration was determined. The concentration of themain monovalent inside cation achieved in non-intoxicated cells was taken as 100% . The first of the paired symbols below represents untreated erythrocytes, the second symbol of the pair represents sphingomyelinase C-treated red blood cells; dog(p, ~), rabbit (Q, ~), human (p, /), cow (6, ®) and pig (p, LQ). The S.E . values were less than 12% of We mean. n=3.

with pig, cattle and sheep red blood cells, even with cytotoxin concentrations as high as 40 ~M. Rabbit cells were three times and human cells 15 times less sensitive as compared to those of dog and rat (Table 2), as regards the cytotoxin concentration required for 50% loss of the intracellular rations. While dog, rat and rabbit erythrocytes were completely permeable for rations, human cells did not release more than about 62% of their intracellular K+ . Preincubation with sphingomyelinase C enhanced the cytotoxin-induced permeability for small rations (Fig. 4). Only a slight increase of susceptibility was observed with dog and rabbit erythrocytes, whereas other red blood cells richer in sphingomyelin displayed a clear distinction (Table 2). Lack of in~fuence by phospholipase C and phospholipase D pretreatment on the cytotoxic attack

The intrinsic specificity of the phospholipases C and D is mainly towards glycerophospholipids (MÓLLBY, 1978 ; EThMADi, 1980). As shown in Table 1, they cleaved 19-28% of the erythrocyte phosphatidylcholine and 4-8% ofits phosphatidylethanolamine. However, preincubation of rabbit erythrocytes with one of these phospholipases influenced neither at 4°C nor at 37°C the specific 125 I-cytotoxin binding (not shown) . The cytotoxin-induced permeability increase was also not altered after pretreatment with phospholipase D (S. chromofuscus) . The cytotoxin concentration required for 50% loss of the cellular K+ for rabbit erythrocytes remained the same (0.7 f 0.03 pM) as compared to untreated cells (0.72f 0.04 ~M; n = 3) and those of pig were not attacked at all.

53 8

K. M.

CROWELL

and F. LUTZ

DISCUSSION

The activity of cytotoxic agents of bacteria, which enhance or inhibit the actions of each other have been studied for many bacterial products, using erythrocytes as targets (B>rltNr-IEI~R et al., 1975; LnvDm, 1984) . Evidence is here presented, that erythrocyte membranes with low sphingomyelin to phosphatidylcholine ratios bind the P, aeruginosa cytotoxin and the degree of binding is correlated to the loss of intracellular cations . By contrast, erythrocytes with a high sphingomyelin to phosphatidylcholine ratio (cattle and sheep) were resistant, if not pretreated with sphingomyelinase C from B. cereus. An exceptional position has to be attributed to pig cells, which are similar to those of man as far as their amount ofsphingomyelin (28 .1 % pig, 25.0% human, Table 1) is concerned, but are not attacked by the cytotoxin . However, the sphingomyelin to phosphatidylcholine ratio is 1 .29 for pig and 0.81 for human erythrocytes . Moreover, pig cells have a smaller surface radius (6.1 gym) as compared to that of man (7.9 gym) (VnN DEENEN and DE GER, 1964). Thus, pig erythrocytes have a higher density of sphingomyelin on their membrane exterior, which may result in a topographically different and for the cytotoxic attack disadvantageous composition of the membrane surface . Even after cleavage of sphingomyelin pig and human red blood cells differed in their sensitivity to the toxin . The reason for this discrepancy is not obvious . Cleavage of sphingomyelin by sphingomyelinase C from B. cereus promoted the cytotoxic effect exerted on the erythrocytes of all animals investigated . The enzymatic removal of phosphorylcholine headgroups leaves an altered but functioning cell membrane (ZWAAL et al., 1973; BARENHOLZ and THOMPSON, 1980). The residual ceramide is still able to interact with membrane sterols to maintain the cell's integrity . Below 10°C, however, the stability is reduced (IKEZAWA et al., 1978; TAGUCHI et al., 1983) . Accordingly we found haemolysis of `sphingomyelin-depleted' cattle erythrocytes after cooling to 0°C. This reduction of stability implies an interference in the cytotoxin-induced permeability increase . Nevertheless, the ' z SI-cytotoxin binding process remained unaffected. The following findings corroborate this conclusion. Firstly, neither intact (Table 3) nor haemolysed (not shown) cattle erythrocytes were able to bind cytotoxin, if not pretreated with sphingomyelinase . Secondly, the Kps achieved with untreated or sphingomyelinase-pretreated erythrocytes of all animal species investigated were the same. The cytotoxin binding and action on erythrocytes was similar to that on Ehrlich cells (LEWICKI and LUTZ, 1985; LuTZ, 1986). For both cell types, binding leads to a loss of intracellular cations, and there are similarities in: (i) time, (ü) dose dependence, and (iü) the enhanced effect after sphingomyelin cleavage . (i) At 30°C for Ehrlich cells and 37°C for rabbit erythrocytes, ' 2 s I-cytotoxin was bound with a half time of < 1 min. The loss of 50% of the intracellular cations occurred within 1 min for both cell types. (ü) About 2 nM 'ssI-cytotoxin were required to obtain a measurable binding to Ehrlich cells and less than 5 nM for rabbit erythrocytes . The minimal concentration provoking a loss of intracellular cations was 100 nm for Ehrlich cells and 200 nM for rabbit erythrocytes. This discrepancy in cytotoxin concentrations required for binding as compared to permeability increase can not be explained, as no molecular events between binding and cytolytic effect are known. (iii) Among the phospholipases investigated only sphingomyelinase C promoted binding of the cytotoxin as well as an increase in permeability. Phosphatidylcholine active phospholipases like phospholipase D (cabbage) did not influence the 'zSI-cytotoxin

Pseudomonas aeruginosa Cytotoxin

539

(Llrrz, 1986). Incubation of Ehrlich B. cereus (which hydrolyses phosphatidylcholine-PC, phosphatidylethanolamine-PE, and phosphatidylserine-PS; OTIVAPSS et al., 1977) or phospholipase D from cabbage (which cleaves PC, PE, PS and lyso PC; DAWSON and HEMIIdGTON, 1967) or phospholipase D from S. chromofuscus (which is active on lyso PC, PC, sphingomyelin and PE ; IMAMURA and HORIUTI, 1979) did not alter the Na+/K+ gradient . The presence of sphingomyelin or ceramide liposomes, tested in binding binding to Ehrlich ascites plasma membrane vesicles cells or erythrocytes with phospholipase

C

from

studies (with rabbit erythrocytes) and assays on permeability increase (with E111tL1cH cells), did not change the cytotoxic effect

(LvTZ et al., 1988), which excludes them as probable or of the r~'I-cytotoxin binding after sphin-

additional acceptors. Hence, the enhancement

gomyelin cleavage may reflect a binding site in close proximity to sphingomyelin. The quantity

of available binding sites increases with

pretreatment, but is not accompanied by

changes in the cytotoxin affinity to the receptor. Similar synergisms were found for

C and insulin (C~JATRECASAS, 1971 ; CLARK et al., 1980; TURYN et al., 1985) Staphylococcus aureus ß-toxin (sphingomyelinase C) and the B-protein of Streptococcus agalactiae, known as the CAMP factor (CHRISTIE et al., 1944; BEItNI-n:Ih¢R et al.; 1979). Cooperative reactions of P. aeruginosa and other bacteria were elaborated by FRASER (1964), who found lytic effects enhanced by diffusible substances of S. agalactiae, Corynebacterium equi and C. renale . The basic reaction mechanism was not elucidated . Our data strongly support the idea that membrane sphingomyelin is of essential importance to the action of the pseudomonal cytotoxin. Cleavage of sphingomyelin phospholipase and for the

exposes more binding sites for the toxin, suggesting an acceptor in close proximity and partially masked by sphingomyelin.

Acknowledgements-We are indebted to REGtx~ L»o~ for her excellent purification of the pseudomonal cytotoxin, and to the Deutsche Forschungsgemeinschaft for financial support. REFERENCES BARENHOLZ, Y. and THOAfP30N, T. E. (1980) Sphingomyelins in bilayets and biological membranes. Biochim . biophys . Acta 604, 129. BENNETT, J. P . Jr (1978) Methods in binding studies . In: Neurotransmitter Receptor Binding, p . 57 (YAMAMURA, H . L, Errxn, S. J. and KurtAx, M. J., Eds) . New York : Raven Press. Interaction between aerolysin, erythrocytes, and BpRrrm~my A. W., AvtGen, L. S. and Avion, G. (197 erythrocyte membranes. Infect . Immun . 11, 1312 . Bswvrn~~, A. W., LrxnEre, R. and Av~GAU, L. S. (1979) Nature and mechanism of action of CAMP protein of group B streptococci . Infect . Immun . 23, 838. Cxiers~rn:, R., Amarra, N. E. and MUNCH-Per~N, E. (1944) A note on a lyric phenomenon shown by group B Streptococci. J. exp. Bio( . 22, 197. Ct.Axtc, S., Lt~Rtcnvs, R. G., De LUISE, M. and MSaCx, R. A. (1980) The effects of trypsin and phospholipase C on insulin binding and action in the isolated adipocyte. Biochem . J. 186, 535. C~A~raEC~s, P. (1971) Unmasking of insulin receptors in fat cells and fat cell membranes. J. biol. Chem . 246, 6532 . DALHO~, A. (1987) Opportunistic infections caused by Pseudomonas aeruginosa. Infection 15, S 60. DAVNSON, R. M. C. and HEA~ING1'ON, N . (1967) Some properties of purified phospholipase D and especially the effect of amphipatic substances . Biochem . J. 102, 76. E~resunr, A. H. (1980) Membrane asymetry . A survey and critical appraisal of the methology-II . Methods for assessing the unequal distribution of lipids. Biochim. biophys. Acta 604, 423 . FIe~Ie, G. (1964) The effect on animal erythrocytes of wmbinations of diffusible substances produced by bacteria. J. Path . Bacteriot. 88, 43 . GLADS'I'ONE, G. P . and VAN HEYNINGEN, W . E. (1957) Staphylococcal leucocidins. Br. J. exp. Path . 38, 123. HAFEZ, H . M ., WoExNLE, H . and HEU., G. (1987) Pseudomonas-aeruginosa-Infektionen bei Putenkûken and Behandlungsversuche mit Apramycin. Bert. Mûnch . Tierdrztt. Wschr. 100, 48 . IxEZAwA, H., MoRI, M., OHYAHU, T . and TAC3UCHI, R. (1978) Studies on sphingomyelinase of Bacillus cereus : I . Purification and properties . Biochim. biophys. Acta 528, 247.

540

K. M. CROWELL and F. LUTZ

IKEZAWw, H., MATSUSHITA, M., Tosu'I'w, M. and Twauctm, R. (1986) Effects of metal ions on sphingomyelinase activity of Bacillus cereus. Arch . Biochem. Biophys. 249, 588. IstwMUnw, S. and Hoxlun, Y. (1978) Enzymatic determination of phospholipase D activity with choline oxidase. J. Biochem. 83, 677. IstwMUxw, S. and HoxIUTI, Y. (1979) Purification of Streptomyces chromojuscus phospholipase D by hydrophobic affinity chromatography on palmitoyl cellulose. J. Biochem. 85, 79 . JoKINEN, M. and GwI~EaG, C. G. (1979) Phospholipid composition and external labeling of aminophospholipids of human En(a-) erythrocyte membranes which lack the major sialoglycoprotein (glycophorin A) . Biochim. biophys. Acta 554, 114. KurPEIes, F. A., EasTON, E. W., VAN DEN HovEN, R., WP.NSING, T., RoeLOI~err, B., OP uEN Kwntr, J. A. F. and VwN DEENEN, L. L. M. (1985) Survival of rabbit and horse erythrocytes in viva after changing the fatty acyl composition of their phosphatidylcholine. Biochim. biophys . Acta 819, 170. LEWICKI, N. and Lurz, F. (1985) Role of sphingomyelin in the plasma membrane damage due to a cytotoxin from Pseudomonas aeruginosa . Toxicon 23, 587. LIrroEa, R. (1984) Alteration of mammalian membranes by cooperative and antagonistic actions of bacterial proteins. Biochim. biophys. Acta 779, 423 . Lurz, F. (1979) Purification of a cytotoxic protein from Pseudomonas aeruginosa. Toxicon 17, 467. LuTZ, F. (1986) Interaction of Pseudomonas aeruginosa cytotoxin with plasma membranes from Ehrlich ascites tumor cells. Natmyt :-Schmiedebergs Arch. Pharmac. 332, 103. LuTZ, F., GxiESxw~tt, S. and KwurEx, I. (1981) Cytotoxin from Pseudomonas aeruginosa in mice : distribution related to pathology. Toxicon 19, 763. LuTZ, F., MwtrxElt, M. SIId FAILING, K. (1987) Cytotoxic protein from Pseudomonas aeruginosa : formation of hydrophilic pores in Ehrlich ascites tumor cells and effect on cell viability. Toxicon 25, 293. LuTZ, F., CxowELL, K., LEWICKI, N. and CONRATH, R. (1988) The role of lipids in the action of Pseudomonas aeruginosa cytotoxin on mammalian cells. Zbl. Bakt 17 (Suppl .), 95 . MITCHELL, K. T., F~tAFr.r . , J. E. Jx and HUESns, W. H. (1986) Separation of phosphoinositides and other phospholipids by two~imensional thin-layer chromatography . Analyr. Biochem. 158, 447. MIYwCHI, Y., CHRAMBACH, A., MECKLENBURG, R. and LlrsE-rr, M. B. (1973) Preparation and properties of'z'ILH-RH. Endocrinology 92, 1725 . MOLLBY, R. (1978) Bacterial phospholipases . In : Bacterial Toxins and Cell Membranes, p. 367 (JEWASZEWICZ, J. and WwnsTxóM, T., Eds) . London : Academic Press. Orxw~, A. B., LITTLE, C., SLECIEN, K., WALLIN, R., JoHNSeN, S., FLENQ9RUD, R. and PRYDZ, H . (1977) Some characteristics of phospholipase C from Bacillus cereus . Ew . J. Biochem. 79, 459. SCFIARMANN, W. (1976) Purification and characterization of leucocidin from Psetuiomonas aeraginosa. J. gen. Microbial. 93, 292. $UTTORP, N., SEDCiEA, W., UHL, J., Lurz, F. and Roxw, L. (1985) Pseudomonas aeruginosa cytotoxin stimulates prostacyclin production in cultured pulmonary artery endothelial cells: membrane attack and calcium influx. J. cell. Physiol. 123, 64 . TwGUCHI, R., Mlztnvo M., INOUE, M. and IKEZAWw, H. (1983) Increase in osmotic fragility of bovine erythrocytes induced by bacterial phospholipases C. J. Biochem. 93, 403. TuxvN, D., SCwCCHI, G. E. and DELLACHA, J. M. (1985) Unmasking of insulin receptors in rat submaxillary gland microsomes: effect of high ionic strength, phospholipase C and S-adenosyl-L-methionine . Biochim. biophys. Acta 845, 333. Vwx DEENEN, L. L. M. and DE Gmr, l . (1964) Chemical composition and metabolism of lipids in red cells of various animal species . In : The Red Blood Cell, p. 243 (B>SI-IOP, C. and SuxcENOR, D. M., Eds) . New York : Academic Press. VASIL, M. L. (1986) Pseudomonas aeruginosa: biology, mechanisms of virulence, epidemiology . J. Pediat . 108, 800. WEU1Ett, R. N., $CFtNE1DER, E., HwesT, C. W. M., DEUTICKE, B., BENZ, R. and Fxls~mt, M. (1985) Properties of the leak permeability induced by a cytotoxic protein from Pseudomonas aeruginosa (PACI) in rat erythrocytes and black lipid membranes. Biochim. biophys. Acta 820, 173. YANG, K., HIGASi1mA, H., INOUE, R. and Nozwww, Y. (1984) Bradykinin-induced rapid breakdown of phosphatidylinositol 4,5-biphosphate in neuroplastoma X glioma hybrid NG108-15 cells. J. biol. Chem 259, 10201 . ZWML, R. F. A., ROEIAFSEN, B. and COLLEY, C. M. (1973) Localization of red cell membrane constituents . Biochim. biophys. Acta 300, 159.