Oligomerization of 3H-labelled staphylococcal alpha-toxin and fragments on adrenocortical Y1 tumour cells

Microbial Pathogenesis 1988 ; 4 : 223-229

Oligomerization of 3H-labelled staphylococcal alpha-toxin and fragments on adrenocortical Y1 tumour cells Lennart Blomgvist* and Monica Thelestam Department of Bacteriology, Karolinska Institutet, Box 60400, S-104 01 Stockholm, Sweden (Received September 30, 1987 ; accepted January 11, 1988)

Blomqvist L . (Dept of Bacteriology, Karolinska Institutet, Box 60400, S-104 01 Stockholm, Sweden) and M . Thelestam . Oligomerization of 3 H-labelled staphylococeal alpha-toxin and fragments on adrenocortical Y1 tumour cells . Microbial Pathogenesis 1988 ; 4 : 223-229 . Staphylococcus aureus a-toxin has previously been shown to bind to erythrocyte membranes and the isolated membranes contain the toxin in both monomeric and hexameric form . The hexamers are believed to form the ring-shaped structures observed by electron microscopy on toxin-treated erythrocytes . It has not previously been shown that hexamers are formed also on nucleated mammalian cells although it has been assumed that hexamers in both systems create transmembrane channels, responsible for the toxin- induced membrane damage . Here we demonstrate by autoradiography that 3 H -z-toxin bound to and formed high molecular weight complexes-presumably hexamers-on cultured adrenocortical Y1 tumour cells . The binding kinetics suggested a non-specific association of a-toxin with the membrane, rather than specific receptor-binding . The pH during toxin binding did not influence the subsequently induced membrane damage . Non-membrane damaging a-toxin fragment preparations also bound firmly to the cell membranes . Upon contact with Y1 cells the fragments formed complexes of the same apparent molecular size as those generated from intact a-toxin . Two interpretations are possible : either the fragment oligomers are somehow defective i .e . not able to form transmembrane structures or the functional relevance of toxin oligomerization for a-toxininduced membrane damage must be questioned . Key words: S . aureus; a-toxin ; toxin fragments ; cultured Y1 cells ; membrane damage ; oligomers .

Introduction Alpha-toxin from Staphylococcus aureus is a cytotoxic exoprotein which is secreted as a 3 S monomer (M, 33000), but under certain conditions oligomerizes to a 12 S high molecular weight complex . Tryptic toxin fragments of different sizes have been described .' Native alpha-toxin exhibits a broad spectrum of biological activities, including lethality, dermo-necrosis, haemolysis and cell membrane-damage (for review, see") . Alpha-toxin appears to be a major virulence factor in staphylococcal infection .' The molecular cloning of the a-toxin gene and subsequent elucidation of the amino acid sequence revealed three major hydrophobic regions .' A two-domain structure of the molecule has been proposed,' with the membrane binding region contained in the C-terminal domain .' Author to whom all correspondence should be addressed . 0882-4010/88/030223+07 $03 .00/0

© 1988 Academic Press Limited



2 24

L. Blomgvist and M . Thelestam

Secreted a-toxin is highly surface active . It penetrates mixed lipid monolayers, 10 forms a protein film at an air-water interface," and permeabilizes liposomes as measured by release of marker molecules . t0 Electron microscopy revealed the presence of ring-shaped structures on both natural (erythrocyte) and artificial membranes treated with Lx-toxin .` The rings appeared identical to the 12 S form of the toxin . Partial membrane penetration of a-toxin has been implicated by the demonstration of toxin induced changes in the hydrophobic fracture plane 12 and by photolabelling studies ." The oligomerization of hydrophilic monomeric toxin molecules to an amphiphilic hexamer and its partial embedment within the membrane was proposed to generate a transmembrane channel ." More recently, the ionic conductance of the toxin channels in artificial membranes has been thoroughly studied . 15-16 Thus a-toxin was the first example of a so-called channel-forming protein of procaryotic origin, with properties resembling those of the C5b-9 complement factors and the lymphocytic perforins ." Cultured mammalian cells of different origin exhibit varying susceptibility to a-toxins' $ We recently demonstrated that the mouse adrenocortical Y1 tumour cell is irreversibly intoxicated after only brief exposure to a-toxin ." This cell was used as a model system in studies of the early events in %-toxin interaction with cultured mammalian cel ls . 20 In the present study we investigated binding of 3 H-labelled 06-toxin to Y1 cells and oligomerization of the toxin and of different toxin fragments on the Y1 cell membrane . Binding did occur, but the data do not support specific receptor-ligand interaction . However, both intact toxin and toxin fragments were shown to form high molecular weight complexes (oligomers) on the Y1 cell membrane .

Results and discussion In the present study a fully active 3 H-labelled a-toxin was found to associate with cultured Y1 cells . The amount of radioactivity associated with the Y1 cells was <10% of the total . This figure was not significantly changed by pretreatment of the cells with unlabelled toxin or by increasing the cell density . The kinetics suggested non-specific toxin interaction with the cells rather than specific receptor binding . The fact that cellassociated toxin remained with the membranes after the extensive isolation procedure, however, suggests that the toxin was firmly bound to the membranes . Although binding to Y1 cells was suggested by recent indirect binding experiments' 9-20 this is the first direct demonstration of Lx-toxin binding to cultured cells . Different authors have reported that erythrocytes from various species had an increased susceptibility to a-toxin at a low pH . 2 ' ,22 It was speculated that the low pH caused charge alterations both on the toxin and on the cell surface, thus facilitating the toxin-membrane interaction . In contrast, the present study suggested that binding of a-toxin to the Yl cell membrane was not promoted by lowering the pH, since toxintreated cells released essentially the same amount of nucleotide marker independent of the pH during toxin exposure (Fig . 1 ) . Alpha-toxin has been shown to interact with Y1 cells in at least three separable steps : (i) binding ; (ii) a putative conformational change, and (iii) membrane damage . 20 Those findings were consistent with hexamer formation in the second step, but it was not possible then to demonstrate hexamers directly . In the present study intact a-toxin and fragment preparations (Fig . 2) were 3 H-labelled and toxin-membrane interaction was studied by autoradiography . High molecular weight complexes were visualized after isolation of plasma membranes from toxin-treated Y1 cells (Fig . 3) . The migration



225

a-toxin oligomerization on Y7 cells

100

80

C. 20

0 0

1

2

3

4

5

alpha-toxin concentration µg/m1 Fig . 1 . Nucleotide labelled Y1 cells were exposed for 15 min at 0°C to a-toxin diluted in phosphatebuffered saline at various pHs and post-incubated 24 h in toxin-free growth medium . Membrane damage was measured as release of cytoplasmic nucleotide marker from the labelled Y1 cells . Symbols: L = pH5; A = pH6 ; ∎ = pH7 ; A = pH8 .

23z4 a-toxin hexamers, support the notion that these complexes are actually hexamers and that they consist of toxin only . The fact that a-toxin induces cell membrane lesions is well established . Recently Bhakdi and coworkers hypothesized that a-toxin-induced membrane damage results from the transition of hydrophilic toxin monomers to amphiphilic membrane-penetrating hexamers which form protein-lined transmembrane channels .' 4.23 That both monomeric and hexameric toxin was found on erythrocyte membranes was explained as the result of partial hexamer dissociation upon the boiling in SDS for 2 sec . 21 Also on the Y1 cell membrane monomeric toxin was detected autoradiographically (Fig . 3) . Since boiling was totally avoided, and since a-toxin hexamers are stable in SDS at room temperature 4,24 we consider a partial dissociation of the oligomers less likely in this case . Possibly, not all of the initially associated monomeric toxin has been converted to oligomeric form in the membrane . It was also directly demonstrated here that both the a-toxin fragment preparations bound firmly to the Yl cell membrane, as already implicated from previous indirect binding studies .' 2' However, it is surprising that also the fragments, which are nontoxic and non-membrane damaging, were able to form oligomers of the same apparent molecular size as those formed from intact a-toxin (Fig . 3) Thus membrane insertion and oligomerization of a foreign protein does not automatically lead to membrane damage, at least not in a nucleated cell . This aspect deserves further investigation . In summary : we have shown by autoradiography that a-toxin binds to the Y1 cell membrane, presumably by non-specific association . As expected it oligomerizes, but oligomerization takes place also with non-membrane damaging toxin fragements . Two interpretations are possible : (i) either the fragment oligomers, although firmly bound, are somehow unable to fully penetrate the membrane or (ii) the functional relevance of oligomerization for toxin-induced membrane damage has to be questioned .



2 26

L . Blomqvist and M . Thelestam

Fig . 2 . SDS-PAGE of a-toxin and fragments . Lane A : a-toxin ; lane B : natural fragments : lane C : tryptic fragments ; lane D : high molecular weight marker proteins (Pharmacia, Uppsala, Sweden) . Numbers indicate molecular weight in kilodaltons .

Materials and methods Chemicals . Trypsin, sera and cell culture media were obtained from Flow Laboratories, Irvine, Scotland . Hanks balanced salt solution (HBSS) was prepared at this department . [5- 3 H]uridine (specific activity 26 .7 Ci/mmol) and Aquasol TM Universal Cocktail were from Du Pont, NEN Research Products, Boston, U .S .A ., Bolton-Hunter reagent (N-succinimidyl [2,3 -3 H] propionate] was purchased from Amersham International, Amersham, UK . Other chemicals were from E . Merck AG, Darmstadt, FRG and KEBO, Stockholm, Sweden . All chemicals were of analytical grade .



a-toxin oligomerization on Y1 cells

227

220

36

Fig . 3 . Autoradiography of Y1 cell membrane proteins isolated from cells exposed to approximately 6 µg/ml 3 H-a-toxin and 3 H-a-toxin fragments . Cell membranes were prepared and processed as detailed in Materials and Methods . The exposure time was 33 weeks . Lane A : 3 H-a-toxin/Y1 cell membrane unboiled ; lane B : 3 H-a-toxin natural fragment/Y1 cell membrane unboiled ; lane C : 3 H-a-toxin tryptic fragments/Y1 cell membrane unboiled ; lane D : 3 H-a-toxin/Y1 cell membrane boiled in SDS ; lane E : 3 H-a-toxin natural fragment/Y1 cell membrane boiled in SIDS ; lane F : 3 H-a-toxin tryptic fragments/Y1 cell membrane boiled in SDS . = monomeric a-toxin ; y = oligomeric a-toxin .

Alpha-toxin and fragment preparation . Alpha-toxin was produced from S. aureus strain Wood 46 and purified as described ." A naturally occurring fragment preparation, 25 and fragments generated by tryptic digestion of a-toxin have been well characterized .' The natural fragment preparation contained a major fragment (M,- 18 .5 kD) and small amounts of a minor fragment (M,-14 kD) . The tryptic fragments had a Mr of 18 kD and 17 kD respectively (Fig . 2) . Both fragment preparations lacked the lethal and cytotoxic activities of native a-toxin . The natural fragment was neutralized by polyclonal anti-a-toxin and retained the membrane-binding region, since it inhibited the effect of intact a-toxin .25 3H-labelling of a-toxin and fragments. Alpha-toxin and the fragment preparations were labelled with Bolton-Hunter reagent (N-succinimidyl [2,3 3H] propionate) as earlier described . 27 Briefly, 50 µg of the different preparations were incubated with 0 .2 mCi of the reagent and after addition of 0 .1 M borate buffer incubated overnight at 4°C . After the reaction was terminated the labelled products were separated from unreacted reagent on a PD 10 (Sepharose G50, Pharmacia, Uppsala, Sweden) column . The specific radioactivities of the labelled preparations were calculated as previously described (A . E. Bolton, Radioiodination techniques, review 18, Amersham International, Amersham, United Kingdom, 1977), and ranged from approximately 1 .0 to 1 .2 µCi/µg protein .



228

L. Blomqvist and M . Thelestam

Binding of 3H-labelled a-toxin to Y1 cells . Attempts to determine 3 H-a-toxin binding to cells growing attached to culture wells were unsuccessful due to a high unspecific association of the toxin with the plastic . Therefore, suspended cells were used in a method similar as described for a-toxin binding to rabbit erythrocytes . 28 Trypsinized Y1 cells were suspended in Eagles medium and left for 2 h at 4°C . Equal numbers of cells (10 6 ) were then carefully distributed in polypropylene test tubes, and incubated with radiolabelled a-toxin for 30 min at 0°C or 37°C . The cells were then centrifugated (3000xg for 5 min) and washed twice with tris-buffered saline (pH 7 .4), whereafter the pellets were solubilized in 0 .5 ml 0.1 M NaOH . The radioactivity was determined by liquid scintillation (LKB-Wallac, Bromma, Sweden) in 10 pl samples of total cell suspensions, supernatants and pellet fractions . In these experiments the influence of atoxin concentration (0 .7-2 .8 pg/ml), cell number (0 .25 to 4 .0x106 cells/ml) and the presence of increasing amounts of unlabelled a-toxin (0 .5 to 8 .0 pg/ml) was investigated . Y1 cell membrane isolation and detection of a-toxin by autoradiography . Y1 cells were cultivated in 6-well plates (A/S Nunc, Roskilde, Denmark), exposed to 3 H-a-toxin in eluation buffer (PBS + 0 .25% albumin + 0 .25% gelatin, pH 7,4) for 15 min at 0°C and further incubated at 37°C for 45 min . The cells were rinsed twice with HBSS, collected from the wells, centrifugated 3000 xg for 5 min, and suspended in sucrose medium (0 .25 M sucrose, 0 .2 mm Mg, and 5 mm tris/HC1, pH 7 .4) . After 10 min at 22°C, the lysed cells were homogenized at 4°C whereafter EDTA was added to a final concentration of 1 mm . The homogenate was centrifuged (1700xg, 10 min, 22°C) and the resulting supernatant centrifugated at 33000xg for 1 h . The pellet was suspended in 1 ml of 1 mm tris/HC1 + 1 MM MgSO 4 (pH8 .6) and centrifugated at 33000xg for 1 h . The final pellet was suspended in 100 yl of sodium dodecylsulphate (SDS) preparation buffer 29 and the radioactivity in 10 pl was determined by liquid scintillation . Half of the remaining sample was boiled for 5 min, before sodium dodecylsulphate polyacrylamide gel electrophoresis (SDS-PAGE) according to Laemmli . 29 Gels were stained with Coomassie Blue R-250 . The volumes of the samples applied on the same gel were adjusted to contain approximately equal amounts of radioactivity . As enhancer 1 M sodium salicylate (+0 .5% glycerol) was used during the film exposure, which was carried out at -70°C for varying periods of time . Assay of membrane damage . Mouse adrenocortical Y1 tumour cells were cultivated in HAM F10 medium to near confluency in 24 well polystyrene trays (Linbro Scientific, Irvine, Scotland) . Plasma membrane damage was measured as release of radioactive cytoplasmic nucleotides from toxin-treated cells which had been pre-labelled with [5- 3 H]uridine . 19 The nucleotide release was expressed as percentage of the maximal release ." All experiments were conducted twice with duplicate samples .

We thank Kerstin Andreasson and Lena Norenius for excellent technical assistance . This work was sponsored by the Swedish Medical Research Council (grant 16X-2562) and the A . 0 . Sward Foundation .

References 1 . Blomqvist L, Bergman T, Thelestam M, Jbrnvall H . Characterization of domain borders and of a naturally occurring major fragment of staphylococcal a-toxin . FEBS Lett . 1987 ; 221 : 127-132 . 2 . Kato I, Watanabe M . Chemical studies on staphylococcal alpha-toxin and its fragments . Toxicon 1980 ; 18 :361-365 . 3 . Watanabe M, Kato I . Purification and some properties of a lethal toxic fragment of staphylococcal a-toxin by tryptic digestion . Biochim . Biophys Acta 1978 ; 535 : 388-400 . 4 . Tobkes N, Wallace BA, Bayley H . Secondary structure and assembly mechanism of an oligomeric channel protein . Biochemistry 1985; 24 : 1915-1920 . 5 . Freer JH, Arbuthnott JP . Toxins of Staphylococcus aureus. In : Dorner F, Draws J, eds. Pharmacology of Bacterial Toxins . Oxford : Pergamon Press, 1986 : 581-633 . 6 . Thelestam M, Blomqvist L . Staphylococcal alpha toxin ; recent advances . Toxicon 1988; 26 : 51-65 . 7 . Kinsman 0, Jonsson P, Haraldsson I, Lindberg M, Arbuthnott JP, Wadstri m T . Decreased virulence of a-haemolysin negative and coagulase negative mutants of Staphylococcus aureus infections in mice . In : Jeljaszewicz J, ed . Staphylococci and staphylococcal infections . Stuttgart: Gustav Fischer Verlag, 1981 ; 652-659. 8. O'Reilly M, de Azavedo JCS, Kennedy S, Foster TJ . Inactivation of the alpha-haemolysin gene of Staphylococcus aureus 8325-4 by site-directed mutagenesis and studies on the expression of its haemolysins . Microbial Pathogenesis 1986 ; 1 : 125-138.



a-toxin oligomerization on Y1 cells

229

9 . Gray GS, Kehoe M . Primary sequence of the a-toxin gene from Staphylococcus aureus Wood 46 . Infect Immun 1984; 46 :615-618 . 10 . Freer JH, Arbuthnott JP, Bernheimer AW . Interaction of staphylococcal a-toxin with artificial and natural membranes . J Bacteriol 1968 ; 95 : 1153-1168 . 11 . Buckelew AR, Colacicco G . Lipid monolayers . Interactions with staphylococcal a-toxin . Biochim Biophys Acta 1971 ; 233 : 7-16 . 12 . Freer JH, Arbuthnott JP, Billcliffe B . Effects of staphylococcal a-toxin on the structure of erythrocyte membranes : a biochemical and freeze-etching study . J Gen Microbiol 1973; 75 : 321-332 . 13 . Thelestam M, Jolivet-Reynaud C, Alouf JE . Photolabelling of staphylococcal a-toxin from within rabbit erythrocyte membrane . Biochem Biophys Res Commun 1983 ; 111 : 444-449 . 14 . Fi ssle R . Bhakdi S, Sziegoleit A, Tranum-Jensen J, Kranz T, Wellensiek H-J . On the mechanism of membrane damage by Staphylococcus aureus a-toxin . J Cell Biol 1981 ; 91 : 83-94 . 15 . Belmonte G . Cescatti L, Ferrari B, Nicolussi T, Ropele M, Menestrina G . Pore formation by Staphylococcus aureus alpha-toxin in lipid bilayers . Dependence upon temperature and toxin concentration . Eur Biophys J 1987 ; 14 : 349-358 . 16 . Menestrina G . Ionic channels formed by Staphylococcus aureus alpha-toxin ; voltage dependent inhibition by divalent and trivalent ions . J Membr Biol 1986 ; 90 : 177-190 . 17 . Bhakdi, S, Tranum-Jensen J . Damage to mammalian cells by proteins that form transmembrane pores . Rev Physiol Biochem Pharmacol 1987 ; 107 : 147-223 . 18 . Thelestam M . Membrane damage by staphylococcal a-toxin to different types of cultured mammalian cell . Biochim Biophys Acta 1983 ; 762 : 481-488 19 . Thelestam M, Blomqvist L . Intoxication of cultured mouse adrenocortical tumour cells by staphylococcal a-toxin . In : Alouf JE, Fehrenbach FJ, Freer JH, Jeljaszewicz J, eds . Bacterial Protein Toxins . London : Academic Press, 1984 ; 279-285 . 20 . Blomqvist L, Thelestam M . Early events in the action of staphylococcal alpha-toxin on the plasma membrane of adrenocortical Y1 tumor cells . Infect Immun 1986 ; 53 : 636-640 . 21 . Bhakdi S, Muhly M, Fi ssle R . Correlation between toxin binding and hemolytic activity in membrane damage by staphylococcal a-toxin . Infect Immum 1984 ; 46 : 318-323 . 22 . lkigai H, Nakae T . The rate assay of alpha-toxin assembly in membrane . FEMS Lett 1984; 24 : 319-322 . 23 . Bhakdi S, Fi ssle R, Tranum-Jensen . Staphylococcal a-toxin : Oligomerization of hydrophilic monomers to form amphiphilic hexamers induced through contact with deoxycholate detergent micelles . Proc Natl Acad Sci 1981 ; 78 : 5475-5479 . 24 . Cassidy P, Harshman S . Characterization of detergent-solubilized iodine-125-labelled a-toxin bound to rabbit erythrocytes and mouse diaphragm muscle . Biochemistry 1979 ; 18 : 232-236 . 25 . Blomqvist L, Thelestam M . A staphylococcal a-toxin fragment ; its characterization and use for mapping biologically-active regions of a-toxin . Acta Pathol Microbiol Immun Scand Sect B 1986 ; 94 : 277-283 . 26 . Wadstrom T . Studies on extracellular proteins from Staphylococcus aureus IV . Separation of )r-toxin by isoelectric focusing . Biochim Biophys Acta 1968 ; 168: 228-242 . 27 . Blomqvist L, Appelgren L-E, Thelestam M . Distribution of 3 H-labelled staphylococcal alpha-toxin and a toxin fragment in mice . Infect Immun 1987; 55 : 1906-1913 . 28 . Phimister GM, Freer JH . Binding of 1251-alpha-toxin of Staphylococcus aureus to erythrocytes . J Med Microbiol 1984 ; 18 :197-204. 29 . Laemmli UK . Cleavage of structural proteins during the assembly of the head of bacteriophage T4 . Nature 1970 ; 227 : 680-685 . 30 . Thelestam M, MbIlby R . Determination of toxin-induced leakage of different-size nucleotides through the plasma membrane of human diploid fibroblasts . Infect Immun 1975 ; 11 : 640-648 .