Detection of Clostridium novyi type B K toxinby cell culture systems

Erika Borrmann *, Frank SchulzeFederal Institute for Health Protection of Consumers and Veterinary Medicine, Division 4, Jena, Germany

Received 17 November 1998; accepted 20 January 1999

Abstract

Ten permanent cell lines were examined for their reaction to the Clostridium novyi K toxin. The action of the toxin wasdetermined after 3 days by microscopic examination and the MTT assay. The K toxin exhibited the strongest effect on ESH-Lcells rather than other cell lines. Vero and SFT-R cells reacted in a comparable way, but less sensitively. We were able to showthat the cytopathic effect on the three types of cells was neutralised by the international standard for gas gangrene antitoxin (C.novyi) but in no case by heterologous antisera. Our results have shown that the three cell lines were specific indicators for thedetection of the cytopathic effect of K toxin. The cytopathic effect can be measured reproducibly by the cell culture assay used.These results are suitable as the starting point for the development of the neutralisation test using cell cultures. ß 1999Published by Elsevier Science B.V. All rights reserved.

Keywords: Clostridium novyi ; Alpha toxin; Cell culture; Detection

1. Introduction

Clostridium novyi is classi¢ed into four types des-ignated A, B, C and D based on the production ofeight di¡erent soluble antigens and, to some extent,on the diseases which they produce in men and ani-mals. Type B is the causative agent of infectiousnecrotic hepatitis (black disease, deutscher Bradsot),especially in sheep but seldom in other animals. Themajor pathological determinant is the K toxin whichis an oedema-inducing and lethal exotoxin [1].Although not precisely measured, ¢nancial losses to

the sheep farms are high when vaccination is notcarried out. Economic losses result from deathsand depreciation of infected stock and from the costsof preventive programmes and sanitary disposal ofdead sheep a¡ected with the disease [2]. An earlyprophylactic immunisation with C. novyi toxoid vac-cines is recommended in endangered areas, becausethe disease can be e¡ectively controlled by vaccina-tion of animals [3].

According to the European Pharmacopoeia [4] thepotency testing of C. novyi (type B) vaccines for vet-erinary use requires vaccination of laboratory rabbitsfollowed by quantitative determination of the anti-bodies against C. novyi K toxin in the rabbit sera bythe toxin neutralisation test in mice (TNT). Sincethis test is cumbersome, time-consuming and causessevere distress and su¡ering to animals, an in vitro

0928-8244 / 99 / $20.00 ß 1999 Published by Elsevier Science B.V. All rights reserved.PII: S 0 9 2 8 - 8 2 4 4 ( 9 9 ) 0 0 0 4 2 - 5

* Corresponding author. Tel. : +49 (3641) 804-251;Fax: +49 (3641) 804-251; E-mail: e.borrmann@bgvv.de

FEMSIM 1031 8-6-99

FEMS Immunology and Medical Microbiology 24 (1999) 275^280

method using cell cultures should be developed as apossible alternative to the TNT.

As antibody titres in sera determined by neutrali-sation are based on the detection of the non-neutral-ised toxins, the ¢rst step taken was to test the cyto-toxic/cytopathic e¡ect of K toxin on cell cultures andto examine the speci¢city of the reaction.

Several reports have described the cytopathic ef-fect of the K toxin on di¡erent cell cultures: Vero,CHO, HT 29 [5], L-929, PC-12 [2], WI-38 [6] andendothelial cells [7]. The quantitative determinationof the antibodies against K toxin, however, has notyet been performed successfully.

2. Materials and methods

2.1. Toxin

C. novyi type B toxin (IRP 425), used as standardfor the C. novyi K TNT in mice, was kindly providedby Dr Hauer (Animal and Plant Health InspectionService, Centre for Veterinary Biologics, Ames, IA,USA). A volume of 0.8 ml of this toxin and 0.2 mldiluent is equivalent to 0.1 l� dose [8]. In order toremove any glycerol the K toxin was dialyzed over-night at 4³C against phosphate bu¡er, pH 7.2. Thetoxin was characterised by means of SDS-PAGE andimmunoblotting. A monoclonal antibody kindly pre-sented by Dr Pietrzykowski was taken as primaryantibody for the detection of the toxin pattern inimmunoblotting.

2.2. Cell lines

For the detection of cytotoxicity/cytopathicity, thefollowing cell lines were used: embryonal bovinelung cells (EBL), foetal bovine tracheal cells(FBTR), kidney cells of the African green monkey(Vero), rat intestinal cells from the small intestine(IEC-6), hamster ovary cells (CHO-K1), embryonalkidney cells of the rhesus monkey (MA 104), dogkidney cells (MDCK), embryonal skin cells of sheep(ESH-L), Syrian hamster kidney cells (BHK-21) andthymus cells of sheep (SFT-R).

Vero, ESH-L, EBL and FBTR cells were culti-vated in minimal essential medium (Eagle) contain-ing Earle's salts (MEM). MA 104 cells grew in Dul-

becco's minimal essential medium (DMEM) and theIEC-6 cells grew in DMEM with the addition ofHEPES and insulin. MDCK cells were cultivatedin MEM containing non-essential amino acids(NEAA). MEM containing NEAA and 120 mg l31

sodium pyruvate was used for the cultivation ofBHK-21 and SFT-R cells. CHO-K1 cells were culti-vated in MEM containing 150 Wg ml31 proline. Foe-tal calf serum (FCS, 10%), 2 mM glutamine and 50Wg ml31 gentamicin were added to the media. Theuse of inactivated FCS was necessary for the culti-vation of CHO-K1. All cells were maintained at37³C in a 5% CO2 atmosphere.

2.3. Antisera

The following antisera were used:

b International standard for gas gangrene antitoxin(C. novyi, equine, 3rd standard, 1966, Statens Se-ruminstitut, Copenhagen, Denmark), dissolved in1 ml phosphate bu¡er, pH 7.2, containing 1100international units.

b International C. perfringens K antitoxin (5thstandard, 1963, Central Veterinary Laboratory,Weybridge, UK), dissolved in 5 ml phosphatebu¡er, pH 7.2, containing 270 international units.

b International standard for Clostridium welchii(perfringens) type D antitoxin (1st standard,1954, Central Veterinary Laboratory, Weybridge,UK), dissolved in 5 ml phosphate bu¡er, pH 7.2,containing 250 international units.

b International standard for Clostridium welchii(perfringens) type B antitoxin (1st standard,1954, Central Veterinary Laboratory, Weybridge,UK), dissolved in 5 ml phosphate bu¡er, pH 7.2,containing 1000 international units.

2.4. Cell culture assay

The action of the K toxin on cells was determinedby a cell test which has been described previously [9].The cell culture assay was performed in 96-wellplates (Costar Europe Ltd., Badhoevedorp, TheNetherlands). In each well of a 96-well plate alreadycontaining 50 Wl of the medium used for cell culti-vation (5% FCS), 50 Wl of the toxin solutions were

FEMSIM 1031 8-6-99

E. Borrmann, F. Schulze / FEMS Immunology and Medical Microbiology 24 (1999) 275^280276

added in their respective dilution steps, as well as 50Wl cell suspension. Each of the dilution steps wasmeasured in duplicate. For the cell control (cellswithout contact of toxin) 50 Wl of medium wasused instead of toxin solution. On each plate, thecell control was measured six times. The plateswere incubated at 37³C under 5% CO2 for 3 days.

The cell suspensions were prepared by the removalof con£uent growth from the cell culture £asks usingtrypsin, followed by resuspension in the medium (5%FCS). The same media were used for preparation ofthe toxin dilution series.

The neutralisation test used to obtain evidence ofspeci¢city was performed as described earlier [10]. Ineach well of a 96-well plate, the toxin and the anti-serum in serial dilution were mixed and incubated atroom temperature for 1 h. The cell suspension wasthen added. The plates were incubated at 37³C under5% CO2 for 3 days. The cell suspension and the toxinsolution were used as controls.

2.5. Evaluation

The determination of viable cells was carried outwith the 3-(4,5-dimethylthiazo-2-yl)-2,5-diphenyltet-razolium bromide (MTT) test [11] and microscopic

examination. The optical density (OD) was measuredat 550 nm using an ELISA reader (SLT GmbH,Crailsheim, Germany).

The number of cells which remained viable afterexposure to the toxin was calculated by means ofthe evaluation programme from SLT relating torespective cell controls (% viable cells = (ODtest/ODcell control)U100).

3. Results

The comparison of the cytopathic e¡ect of C. no-vyi type B K toxin on di¡erent cell lines is shown inFig. 1. The results are means of triplicate determina-tions. The calculated standard deviations varied be-tween 3% (minimum) and 25% (maximum). The cellswhich remained viable after toxin contact were de-termined depending on the toxin concentration.

The K toxin exhibited the strongest e¡ect on theESH-L cells as compared to the other cell lines. Ex-posure of ESH-L cells to highly concentrated K toxinresulted in rounded cells (cytopathic e¡ect) besidelysed cells (cytotoxic e¡ect). The proliferation ofthe cells was clearly reduced. Higher dilution of thetoxin resulted in spindle-like beside rounded cells

Fig. 1. Comparison of the cytopathic e¡ect (% viable cells) of C. novyi type B K toxin on di¡erent cell lines. The cytopathic e¡ect isshown as a function of the toxin concentration.

FEMSIM 1031 8-6-99

E. Borrmann, F. Schulze / FEMS Immunology and Medical Microbiology 24 (1999) 275^280 277

Fig. 2. Morphological alterations of the cells exposed to K toxin for 3 days in microtiter plates. Light microscopy, magni¢cation U90. A:ESH-L cells without toxin contact (cell control), 1.25U104 cells per well, the cells are con£uent and show a ¢broblast-like morphology.B: ESH-L cells after contact with K toxin (titre 1:100), complete rounding. C: Vero cells without toxin contact (cell control), 1.0U104

cells per well, the cells are con£uent and show a ¢broblast-like morphology. D: Vero cells after contact with K toxin (titre 1:100), com-plete rounding. E: SFT-R cells without toxin contact (cell control), 1.0U104 cells per well, the cells are con£uent and show an epithelial-like morphology. F: SFT-R cells after contact with K toxin (titre 1:100), rounding and retracting cells.

FEMSIM 1031 8-6-99

E. Borrmann, F. Schulze / FEMS Immunology and Medical Microbiology 24 (1999) 275^280278

(Fig. 2A,B). All investigated cells exhibited the sametype of rounding e¡ect (Fig. 2C^F), but the ESH-Lcells were more sensitive than the other cells. Ap-proximately 50% viable cells were measured at a tox-in titre of 1:10 000 on ESH-L, 1:2000 on Vero and1:3000 on SFT-R cells.

The toxin test was also used for the determinationand standardisation of the neutralisation test condi-tions. The parameters using the ESH-L cells were thefollowing:

b Test media: all dilutions were made with the cellculture medium (5% FCS);

b Cell number: 1.25U104 cells per well ;b Comparison of the cell passages: 13th^33rd pas-

sages;b Age of the cells : 4 days after passage;b Evaluation of viable cells : MTT test.

The speci¢city of the ESH-L cell line as well as theVero and SFT-R cell lines for the detection of the Ktoxin was determined in the neutralisation test. Theinternational standard for gas gangrene antitoxinwas able to neutralise the action of the toxin up toa 1:10 000 dilution of the antiserum. The heterolo-gous antisera were unable to neutralise the cyto-pathic e¡ect in any case.

4. Discussion

The reproducible and speci¢c detection of the cy-topathic e¡ect of the C. novyi K toxin on ESH-L cellsis a prerequisite for the development of a neutralisa-tion test using cell cultures as an alternative to theTNT. In comparison to other cell lines, the ESH-Lcells were found to react most sensitively to the tox-in. The cell shape alterations were characterised byretraction and rounding when the toxin was highlyconcentrated. The use of the toxin at a dilutions 1:2000 resulted in spindle-like cells side by sidewith rounded cells. The e¡ects observed were inagreement with reports in the literature. The C. novyiK toxin belongs to the so-called large clostridial cy-totoxins (LCT) with reference to the high molecularmass of the single-chain molecules, their in vitro ac-tion on the micro¢lament system of cultured cellsand a series of in vivo actions [12]. LCTs added to

cells in culture are taken up by receptor-mediatedendocytosis into cytosol. The contact of culturedcells with the K toxin results in irreversible changesin cell shape based on the action on the actin cyto-skeleton and by minor e¡ects at the vimentin andtubulin systems [13,14]. The K toxin has been foundto have glycosyltransferase activities that use acti-vated sugars to modify di¡erent small GTP-bindingproteins [15].

Cell cultures can be e¡ective indicators of free tox-in in the TNT, but it is necessary to demonstrate thatthe cytopathic e¡ects have been caused by the toxinand not by other components such as preservatives,detoxifying agents, non-immunogenic bacterial tox-ins and other excipients [6]. For this reason, it isnecessary to ¢nd sensitive cells in order to removeinterfering components by dilution. Therefore, theESH-L cells would be the most suitable from therange of cells investigated. The ESH-L cells, how-ever, were growth-limited cells (so-called ¢nite cellline) and it was not possible to use cells after the33rd passage because of the morphological altera-tions of the cells. The Vero and SFT-R cells wereless sensitive than the ESH-L cells, but both cell linesare permanently growing cells.

Clostridial vaccines are multicomponent vaccinesand beside C. novyi K toxoid also contain C. perfrin-gens O and L toxoids. It was very important to dem-onstrate that the cytopathic e¡ect on the cells hadbeen caused by the K toxin and not by other com-ponents. The neutralisation tests using both homol-ogous antiserum and heterologous antisera demon-strated the speci¢city of the reaction of K toxin onthe ESH-L, as well as the Vero and the SFT-R cells.In addition, the calculated standard deviations dem-onstrate that the cytopathic e¡ect on the three celllines could be measured reproducibly. Our resultshave shown that the three cell lines are suitable in-dicators of excess toxicity of C. novyi K toxin in theTNT using cell cultures. These results are the start-ing point for the development of the neutralisationtest using cell cultures.

Acknowledgements

This work was supported by the Federal Ministryof Education, Science, Research and Technology,

FEMSIM 1031 8-6-99

E. Borrmann, F. Schulze / FEMS Immunology and Medical Microbiology 24 (1999) 275^280 279

Germany. Mrs C. Muselmann and Mrs A. Dram-burg carried out the laboratory work. Dr Hauer pro-vided the K toxin and Dr Pietrzykowski provided themonoclonal antibody against K toxin.

References

[1] Hatheway, C.L. (1990) Toxigenic clostridia. Clin. Microbiol.Rev. 3, 66^98.

[2] Bette, P., Frevert, J., Mauler, F., Suttorp, N. and Habermann,E. (1989) Pharmacological and biochemical studies of cytotox-icity of Clostridium novyi type A alpha-toxin. Infect. Immun.57, 2507^2513.

[3] Schranner, I., Erhard, M.H., Kaltner, H. and Lo«sch, U.(1992) Isolation of immunogenic and lethal peptides of K-tox-in from Clostridium novyi type B. Toxicon 30, 653^668.

[4] European Pharmacopoeia, 2nd edn., Part II-8, No. 362 (1984)Vaccinum Clostridii Novyi B ad usum veterinarium.

[5] Ball, D.W., Tassell, R.L.V., Roberts, H.D., Hahn, P.E., Ly-erly, D.M. and Wilkins, T.D. (1993) Puri¢cation and charac-terization of alpha-toxin produced by Clostridium novyi typeA. Infect. Immun. 61, 2912^2918.

[6] Knight, P.A., Burnett, C., Whitaker, A.M. and Queminet, J.(1986) The titration of clostridial toxoids and antisera. Dev.Biol. Stand. 64, 129^136.

[7] Mu«ller, H., von Eichel-Streiber, Ch. and Habermann, E.(1992) Morphological changes of cultured endothelial cells

after microinjection of toxins that act on the cytoskeleton.Infect. Immun. 60, 2007^3010.

[8] Product sheet, Animal and Plant Health Inspection Service,Center for Veterinary Biologics, Ames, IA.

[9] Borrmann, E. and Schulze, F. (1997) Use of cell culture assaysfor potency testing of C. perfringens immunobiologicals as analternative to mouse toxin neutralisation test. Pharmeuropa 1,(Special issue BIO) 30^39.

[10] Borrmann, E. and Schulze, F. (1998) Nachweis von Clostri-dium novyi type B alpha toxin mittels Zellkulturen. ALTEX98, (Suppl.) 53^56.

[11] Mossmann, T. (1983) Rapid colorimetric assay for cellulargrowth and survival : application to proliferation and cytotox-icity assays. J. Immunol. Methods 65, 55^63.

[12] Hofmann, F., Herrmann, A., Habermann, E. and von Eichel-Streiber, C. (1995) Sequencing and analysis of the gene encod-ing the alpha toxin of Clostridium novyi proves its homologyto toxins A and B of Clostridium di¤cile. Mol. Gen. Genet.247, 670^679.

[13] Donelli, G. and Fiorentini, C. (1994) Bacterial protein toxinsacting on the cell cytoskeleton. Microbiologica 17, 345^362.

[14] Oksche, A., Nakov, R. and Habermann, E. (1992) Morpho-logical and biochemical study of cytoskeletal changes in cul-tured cells after extracellular application of Clostridium novyialpha-toxin. Infect. Immun. 60, 3002^3006.

[15] Selzer, J., Hofmann, F., Rex, G., Wilm, M., Mann, M., Just,I. and Aktories, K. (1996) Clostridium novyi K-toxin-catalyzedincorporation of GlcNAc into RHO subfamily proteins.J. Biol. Chem. 271, 25173^25177.

FEMSIM 1031 8-6-99

E. Borrmann, F. Schulze / FEMS Immunology and Medical Microbiology 24 (1999) 275^280280