Microcystis Aeruginosa Toxin: Cell Culture Toxicity, Hemolysis, and Mutagenicity Assays W

APPLIED AND ENVIRONMENTAL MICROBIOLOGY. June 1982, p. 1425-1433 Vol. 43, No. 6 0099-2240/82/061425-09$02.00/0

Microcystis aeruginosa Toxin: Cell Culture Toxicity, Hemolysis, and Mutagenicity Assays W. 0. K. GRABOW,l* W. C. Du RANDT,1 O. W. PROZESKY,2 AND W. E. SCOTT1 National Institute for Water Research, Council for Scientific and Industrial Research, P.O. Box 395, Pretoria 0001,1 and National Institute for Virology, Johannesburg,2 South Africa Received 9 November 1981/Accepted 22 February 1982

Crude toxin was prepared by lyophilization and extraction of toxic Microcystis aeruginosa from four natural sources and a unicellular laboratory culture. The responses of cultures of liver (Mahlavu and PLC/PRF/5), lung (MRC-5), cervix (HeLa), ovary (CHO-Kl), and kidney (BGM, MA-104, and Vero) cell lines to these preparations did not differ significantly from one another, indicating that toxicity was not specific for liver cells. The results of a trypan blue staining test showed that the toxin disrupted cell membrane permeability within a few minutes. Human, mouse, rat, sheep, and Muscovy duck erythrocytes were also lysed within a few minutes. Hemolysis was temperature dependent, and the reaction seemed to follow first-order kinetics. Escherichia coli, Streptococcus faecalis, and Tetrahymena pyriformis were not significantly affected by the toxin. The toxin yielded negative results in Ames/Salmonella mutagenicity assays. Micro- titer cell culture, trypan blue, and hemolysis assays for Microcvstis toxin are described. The effect of the toxin on mammalian cell cultures was characterized by extensive disintegration of cells and was distinguishable from the effects of E. coli enterotoxin, toxic chemicals, and pesticides. A possible reason for the acute lethal effect of Microcystis toxin, based on cytolytic activity, is discussed.

Some cellular components of various species number of disturbances in enzymatic systems of algae have toxic, carcinogenic, or mutagenic and metabolic processes have been observed, effects (7, 11, 23, 31, 32, 34, 37). Neither the and it has been concluded that the toxins of blue- chemical nor the toxicological properties of green algae cause degenerative changes in par- these compounds have been fully established (6, enchymatous organs and brain cells (22). Toxins 7, 11). Certain strains of the blue-green alga of blue-green algae have been reported to agglu- (cyanobacterium) Microcystis aeruginosa Kutz tinate erythrocytes (6) and to be cytopathogenic emend Elenkin produce toxins which may con- for cultures of mammalian cells (22). Primary sist of different combinations of a number of fibroblast cultures of human and rat embryo peptide or peptide-containing toxins of unde- cells have proved particularly sensitive, fol- fined structure with hepatotoxic or neurotoxic lowed, in decreasing order of sensitivity, by a rat activities (7, 11, 22). The M. aeruginosa toxin fibroblast cell line and the HeLa, A-8, and HEp- known as microcystin or fast-death factor, 2 cell lines (22). which kills white mice in 1 to 3 h, may consist of This study deals with the response of previ- more than one type of cyclic polypeptide which ously untested cells to Microcystis toxin, the may contain up to 16 amino acids and have a development and evaluation of biological assays molecular weight ranging from 654 to 19,400 (7, for Microcystis toxin, and the testing of toxin for 11, 28, 37). Death of mice and vervet monkeys mutagenic activity by means of the Ames/Sal- injected with Microcystis toxin has been as- monella microsome mutagenicity assay (Ames cribed to circulatory failure as an indirect conse- test). The results contribute to the characteriza- quence of extensive liver damage (10, 37, 38). tion of Microcystis toxin, the establishment of However, the short survival time suggests that practical methods for research on the toxin, and liver damage may not be the primary cause of the understanding of its mechanism of action. death, and indications that neurotoxic activity located in the same fraction as the hepatotoxin MATERIALS AND METHODS or in a different fraction is the cause of death Microcystis toxin. Harvests consisting predominantly have been observed (11). In white rat and mouse of toxic M. aeruginosa were collected in South Africa toxicological studies in which algal toxin prepa- from the Hartbeespoort Dam (near Pretoria) in 1978, rations that probably included microcystin but the Vaal Dam (near Johannesburg) in 1977 and 1980, possibly also other toxins have been used, a and the Roodeplaat Dam (near Pretoria) in 1978 (35, 1425 1426 GRABOW ET AL. APPL. ENVIRON. MICROBIOL. 37). Similar harvests of nontoxic Microcystis orga- of streptomycin and the other containing (per millili- nisms collected at the Hartbeespoort Dam in 1974 and ter) 1 mg each of EDTA (Titriplex III; E. Merck) and 1979 and the Rietvlei Dam (near Pretoria) in 1975 glucose in sodium chloride (8 mg/ml), 0.2 mg of served as negative controls. A unicellular laboratory potassium chloride, 2.89 mg of disodium hydrogen culture of a toxic strain of M. aeruginosa designated phosphate 12 hydrate, and 0.2 mg of potassium hydro- WR70 (35) was propagated as described elsewhere (35; gen phosphate (calcium- and magnesium-free phos- W. E. Scott, D. J. Barlow, and J. H. Hauman, Pro- phate-buffered saline), were mixed together immedi- ceedings ofthe International Conference on the Water ately before being used for the trypsinization of the Environment: Algal Toxins and Health, in press). cells. Freeze-dried samples were thoroughly ground with a Microtiter cell culture assay. The mammalian cell pestle and mortar, suspended in 0.1 M ammonium cultures were trypsinized, and the cells were suspend- bicarbonate (1 g/100 ml), stirred overnight at 10°C, and ed in appropriate volumes of growth medium supple- centrifuged (10,000 x g for 30 min); the toxin-contain- mented with 2 to 10% serum (the harvest from conflu- ing supernatant was then freeze-dried. This material ent growth in 25-cm2 flasks was suspended in 10 to 25 was suspended in distilled water (4 mg/ml) and then ml of medium), and 0.1-ml volumes of these suspen- filtered (Sartorius SM/S13400 [diameter, 47 mm; pore sions were seeded into wells of microtiter plates size, 5 ,m] followed by Sartorius SM/N11306 [diame- (Titertek plates with 96 flat-bottom wells; Flow Labo- ter, 25 mm; pore size, 0.45 pLm]) for testing. ratories). Dilutions of test suspensions (0.1 ml) were Escherichia coli enterotoxin. E. coli strains 18, B7A, added immediately or after 24 h. Sterile distilled water and H10407, which all produce both heat-labile and and similar preparations of nontoxic strains of M. heat-stable enterotoxins (8, 9, 15, 33), were supplied aeruginosa were used in negative controls. After incu- by A. S. Greeff (Institute of Pathology, Pretoria). bation at 37°C in an atmosphere of 5% C02, the cells Cultures were grown in Oxoid Nutrient Broth no. 2 for were studied for cytopathogenic effects. 48 h at 35°C without shaking and centrifuged (10,000 x Hemagglutination and hemolysis tests. Sheep, Mus- g for 30 min); then the supernatant was filtered covy duck, mouse, rat, and human (group A, RH+) through a membrane (pore size, 0.45 pm) for toxicity erythrocytes were washed three times with 0.85% testing. sodium chloride and resuspended in sodium chloride. Cell cultures. The human hepatoma cell lines PLC/ In hemagglutination tests, 1% suspensions (0.1 ml) of PRF/5 (hepatitis B surface antigen positive) (passage erythrocytes were seeded into wells of microtiter 91) and Mahlavu (hepatitis B surface antigen negative) plates (Titertek plates with 96 V-shaped wells; Flow (passage 102) (1, 29, 30) were obtained from J. J. Laboratories), and dilutions of the test suspensions Alexander, R. Saunders, and J. A. Pienaar (National (0.1 ml) were added immediately. Saline and similar Institute for Virology, Johannesburg). The Chinese preparations of nontoxic strains of M. aeruginosa hamster ovary cell line CHO-Kl (passage 14) was were used in negative controls. The plates were cov- purchased from Flow Laboratories, Inc., and the fetal ered and allowed to stand for 3 to 4 h at room rhesus monkey kidney cell line MA-104 and the human temperature. Results were recorded as ++ (no cells embryo lung cell line MRC-5 (passage 27) were pur- settling at the bottom, erythrocyte agglutination), + - chased from M. A. Bioproducts. Cultures of the Vero (partially positive: about half of the cells settling at the (African green monkey kidney) and HeLa (human bottom, the rest remaining diffused), or - - (negative: cervical carcinoma) cell lines were provided by E. M. well-defined spot of erythrocytes settled at the bottom Bey (National Institute for Virology) and J. A. Pien- of the well and clear solution at the top), as described aar. The BGM (African green monkey kidney cell elsewhere (6). In hemolysis tests, equal volumes of lines) and primary vervet kidney cells have been erythrocytes in saline (10% [vol/vol]) and test material described elsewhere (14). Hepatoma, HeLa, MA-104, suspended in saline were mixed and kept at 35°C for 60 Vero, BGM, and primary vervet kidney cells were min. Reaction mixtures were then macrohematocrit grown in Eagle minimal essential medium with Earle centrifuged in Wintrobe tubes (3,000 x g for 3 min at salts EMEM (Auto-Pow; Flow Laboratories) supple- room temperature or 2,260 x g for 30 min at 5°C) (41) mented (per milliliter) with 100 U of penicillin, 100 pg or microhematocrit centrifuged by the method de- of streptomycin, 50 ptg of neomycin, 292 pg of L- scribed previously (41) or in 1.5-ml Eppendorff tubes glutamine, and 2 mg of sodium bicarbonate. Gentamy- with a Heraeus-Christ type 00912 microhematocrit cin (50 ,ug/ml) and tylosin (8 p.g/ml) were occasionally centrifuge for 5 min at room temperature. Saline or added to control mycoplasma contamination in PLC/ saline preparations of nontoxic strains of M. aerugin- PRF/5 cultures. CHO-Kl cells were cultured in Nutri- osa were mixed with erythrocyte suspensions in nega- ent Mixture F12 Ham (Bio-Cult; GIBCO Laboratories) tive controls. Test and negative-control suspensions containing (per milliliter) penicillin (100 U), streptomy- were centrifuged simultaneously. After centrifugation, cin (100 ,ug), and sodium bicarbonate (1,176 ,ug). E199 the supernatants were assayed, and any reduction in medium with L-glutamine and Earle salts (Flow Labo- packed-erythrocyte volume and concomitant increase ratories), supplemented (per milliliter) with penicillin in visible hemoglobin concentration relative to the (100 U), streptomycin (100 p.g), and sodium bicarbon- negative-control levels was recorded as a positive ate (2.2 mg), was used for MRC-5 cell cultivation. hemolysis reaction (41). Each medium was supplemented with 10% heat-inacti- Tetrahymena pyriformis. A culture of T. pyriformis W vated fetal calf serum (State Vaccine Institute, Cape was supplied by J. L. Slabbert (National Institute for Town, South Africa) for the propagation of cells at Water Research, Pretoria). Propagation of the orga- 37°C and with 2% serum for the maintenance of nisms and preparation of washed cell suspensions in cultures at 33°C (in 75- or 25-cm2 plastic tissue culture growth medium or balanced Osterhout salt solution flasks [Costar]). Equal volumes of two solutions, one were done as described elsewhere (J. L. Slabbert and containing (per milliliter) 2.5 mg of Trypsin 1:250 W. S. G. Morgan, Water Res., in press). Appropriate (Difco Laboratories), 100 pg of penicillin, and 100 jig mixtures of cell suspensions and suspensions of test VOL. 43, 1982 M. AERUGINOSA TOXIN 1427 material in water were incubated at 27°C. Osterhout TA100 were used were performed as described previ- solution, saline, or distilled water was used in negative ously (3, 13). Tests were done on the toxin prepara- controls. Reaction mixtures contained about 3 x 104 tions described above as well as on slurries of toxic M. cells per ml. The effect of test suspensions on the aeruginosa in dimethyl sulfoxide (10 mg/ml), which motility and morphology of organisms was evaluated were tested without filtration. with a hemacytometer by comparison with the effect Chemical toxicants and pesticides. Solutions of anhy- of negative controls. Cell membrane permeability was drous copper sulfate, mercuric chloride, ammonium studied by using the trypan blue test. sulfate, lead chloride, cadmium chloride, and phenol Mouse toxicity test. White mice weighing about 20 g were prepared as described previously (21). Ethyl each were inoculated intraperitoneally with the test alcohol was used to facilitate the solution of the suspensions (7, 18, 28, 37). pesticides parathion, chlordane, and pentachlorophe- Test for E. coli heat-labile enterotoxin. Suspensions nol (26) in water. All solutions were decontaminated of CHO-Kl cells were subcultured in 0.1-ml quantities by membrane filtration. into wells of flat-bottom microtiter plates, and dilutions of enterotoxin preparations (0.1 ml) were added immediately. Similar preparations of non-enterotoxi- RESULTS genic strains of E. coli were used in negative controls. Response of mammalian cell cultures to Micro- After incubation for 24 h at 37°C in an atmosphere of cystis toxin. Cells of all cultures were severely 5% C02, the morphology of the cells was investigated. damaged or disintegrated after overnight incuba- The percentage of elongated cells (clearly bipolar, tion in the presence of toxin (Fig. 1C and G). All length three times greater than diameter) was deter- cell lines underwent the same type of cell disin- mined (15, 33), and the results were recorded as + + (more than 15% elongated cells), + - (ca. 15% elongat- tegration, which resulted in cell debris of differed cells), or - - (less than 15% elongated cells). This ent sizes and granular material (Fig. 1C and G). procedure was used in all studies of the effect of E. coli The disintegration of cells was not preceded by enterotoxin and Microcystis toxin on the morphology elongation or associated morphological changes. of mammalian cells. Cytopathogenic effects were detectable within Trypan blue test of cell membrane permeability. The about 6 h in cells seeded in the presence of confluent growth of cell cultures in 25-cm2 flasks was undiluted toxin preparations and about 8 h after trypsinized, and the cells were suspended in appropri- the preparations were added to established 24-h ate volumes (5 to 10 ml) of growth medium supple- cultures. Results of toxin titrations were usually mented with 10% serum. A 1.0-ml volume of each cell to 24 but suspension was mixed with 1.0 ml of test suspension read after overnight incubation (16 h), and incubated at 35°C in an atmosphere of 5% CO2- In the lowest concentrations of toxin which exerted negative controls, the test suspension was replaced by a cytopathogenic effect were generally detect- phosphate-buffered saline and similar preparations of able only after 2 to 5 days of incubation. Serum nontoxic strains of M. aeruginosa. After 60 min at had no effect on the action of toxin: serum added 35°C, 0.1 ml each was removed from the test and concomitantly with toxin or 4 h after toxin was control suspensions and mixed with an equal volume added to washed cells in serum-free growth of membrane-filtered 0.5% trypan blue solution in medium did not detectably alter the cytopatho- phosphate-buffered saline (17), and the percentage of genic effect. inactivated cells was evaluated with a hemacytometer a toxin titration (improved Neubauer, Spencer Bright-line; American The results of typical (Table 1) Optical Corp.) as described by Hoskins (17). illustrate that the sensitivities of the cell lines did Bacteria. Strains of E. coli and Streptococcus faeca- not differ significantly from one another. Similar lis were isolated from wastewater and identified by results were recorded for all toxin preparations. Gram staining and API 20E and API 10 Strep test CHO-Kl and Mahlavu were marginally more systems (12). Inactivation of stationary-phase bacteria sensitive than were the other cell lines, and was evaluated by adding 0.1-ml overnight broth cul- BGM and Vero cell lines were the least sensi- ture (35°C) samples diluted to 10-2 in saline to 4 ml of tive. Primary vervet kidney cells were slightly saline containing 2 mg of lyophilized Microcystis ex- more sensitive than were the cell lines. One tract per ml and counting viable bacteria before and disadvantage of primary cells for toxicity testing after incubation for 5 h at 35°C. The effect of toxin on for the multiplication of bacteria was evaluated by inocu- was that they had to be incubated relatively lating 0.1-ml overnight broth culture samples diluted long periods before yielding established cell 10' in saline into 4 ml of nutrient broth (Difco) growth suitable for testing. The optimum nutri- containing 2 mg of lyophilized Microcystis extract per ent concentration in the growth medium and ml and counting viable bacteria before and after incu- number of cells seeded in the wells had to be bation for 24 h at 35°C. In negative controls, toxin was established for each cell line. Tests in which replaced by similar extracts of nontoxic Microcystis cells were seeded in the presence of toxin were harvests. We counted bacteria by membrane filtration, generally more sensitive than were tests in using m-Endo LES agar for E. coli and m-Enterococ- which toxin was added to established cell layers. cus agar for S. faecalis (12). Lower counts in test cells to be the mixtures than in negative controls were taken as CHO-Kl and Mahlavu proved evidence of bacterial inactivation or inhibition of mul- most practical for the detection of toxicity by a tiplication. microtiter assay in which cells are seeded in the Ames tests. Plate incorporation assays and spot tests presence of test suspensions. These cell lines in which tester Salmonella sp. strains TA98 and were easy to keep for instant use over several 1428 GRABOW ET AL. APPL. ENVIRON. MICROBIOL.

En-_

FIG. 1. Mammalian cells in microtiter cultures after overnight incubation (x234). CHO-Kl cells: (A) control, (B) with E. coli enterotoxin, (C) with Microcystis toxin, (D) with 100 p.g of Cd21 per ml. Mahlavu cells: (E) control, (F) with E. coli enterotoxin, (G) with Microcystis toxin, (H) with 100 [Lg Of CU21 per ml. VOL. 43, 1982 M. AERUGINOSA TOXIN 1429 TABLE 1. Microtiter cell culture, mouse, purification works in Pretoria (12), water from hemolysis, and trypan blue tests for toxin derived the Hartbeespoort Dam, and solutions of toxic from a bloom of toxic M. aeruginosa in compounds in distilled water. The results of a Hartbeespoort Dam typical comparative experiment show that raw Response to indicated concn (mg/ml) of and treated wastewater and river and dam water Test toxin' lacking an algal bloom yielded negative results in 1.0 0.5 0.25 0.125 0.063 0.032 cell culture assays in which Mahlavu or CHO- Cell culture Kl cells were used (Table 2). Similar results were recorded for samples collected CHO-Kl ++ ++ ++ + .-- __ on 8 differ- Mahlavu + + + + + + - - - ent days and assays in which BGM, Vero, MA- HeLa . + + + - + - . _ _ _ _ 104, HeLa, or PLC/PRF/5 cells were used. Even PLC/PRF/5 ++ +. membrane-filtered samples of water with a MRC-5 .... ++ +- bloom of toxic M. aeruginosa or laboratory MA-104 .... ++ + cultures of M. aeruginosa WR70 generally yield- Vero .+-. +. ed negative results. Cell toxicity was detectable BGM +.- +- only after the release of toxin from the toxic algae by lyophilization or sonication. The cyto- Mouse ....+.+ ++ + pathogenic effect of the chemicals and pesticides Trypan blue .. ++ _ (Table 2) was characterized by the rounding of cells and by death without instant disintegration Hemolysis ++ ++ ++ I_++ ++ ++ (Fig. 1D and H) and was distinguishable from the effect of Microcystis toxin. a Toxin preparation and test procedures are described in text. The cell culture results after 24 h are Response of cell cultures to E. coli enterotoxin. expressed as follows: + +, distinct cytopathogenic In microtiter cell culture assays, CHO-Kl cells effect; +-, indications of cytopathogenic effect; --, exposed to E. coli enterotoxin displayed the no effect. The mouse test results after 5 h are ex- typical cell elongation (Fig. 1B) described previ- pressed as follows: ++, dead; +-, convulsions, ab- ously (15). Mahlavu cells elongated similarly normal breathing, pallor; - -, no effect. The results of (Fig. 1F), but PLC/PRF/5 cells did not. The cell the trypan blue test of Mahlavu cells after 60 min at elongation induced by the E. coli enterotoxin 35°C are expressed as follows: + +, percentage of was distinguishable from the cytopathogenic ef- stained cells higher than that in negative controls; - -, fect percentage of stained cells not higher than that in of Microcystis toxin (Fig. 1). negative controls. The results of the hemolysis test Mouse toxicity test. The mouse toxicity test after 60 min at 35°C are expressed as follows: + +, was less sensitive than microtiter cell culture hematocrit volume of packed erythrocytes smaller assays to all Microcystis toxin preparations. than that in negative controls. Mice did not die within 5 h after the inoculation of wastewater, river and dam water, E. coli enterotoxin, or relatively high concentrations of potentially toxic compounds (Table 2). The weeks simply by incubation at 33°C in mainte- mouse toxicity test has the advantage of giving nance medium that was changed twice a week. results within a few hours, and death is a well- The microtiter cell culture assay yielded repro- defined endpoint for titration purposes. ducible results, had a fairly distinct endpoint, Trypan blue test. Disruption of membrane and proved suitable for titrating toxin derived permeability was clearly visible in mammalian from a variety of sources. cells about 5 min after the addition of Microcys- In the case of the toxin obtained from Hart- tis toxin. Control (unaffected) cells were round, beespoort Dam, cell culture assays proved to be shining, and colorless, whereas affected cells more sensitive than the mouse toxicity test, and were flat, dull, and clearly blue as a result of the the hemolysis test was the most sensitive (Table penetrating stain. The responses of CHO-Kl, 1). In similar comparative tests of M. aerlugin- Mahlavu, PLC/PRF/5, HeLa, and monkey kid- osa WR70 toxin and the toxins obtained from ney cells to Microcystis toxin did not differ Vaal and Roodeplaat Dams, the cell culture significantly from one another. As an assay for assays were also more sensitive than the mouse the toxin, the trypan blue test was not particular- toxicity test; in contrast, the hemolysis test was ly sensitive (Table 1). Tests in which CHO-Kl less sensitive than both. None of the tests re- cells were used and incubation was for 60 min at sponded to negative controls. 35°C proved practical and had a well-defined and To evaluate possible interference by other reproducible endpoint. The assay was not affect- compounds which may occur in water and have ed by wastewater, river and dam water, E. coli cytopathogenic effects, we performed microtiter enterotoxin, or relatively high concentrations of cell culture assays of membrane-filtered samples toxic compounds (Table 2). of wastewater collected at the Daspoort sewage Hemagglutination and hemolysis tests. Micro- 1430 GRABOW ET AL. APPL. ENVIRON. MICROBIOL. TABLE 2. Microtiter cell culture, mouse, trypan blue, and hemolysis tests for various water samples and toxic compound solutions" Sample

Test StlStld Activated ApiesrHabeespoortHatesortE|E co/i LT Cu2+ at (igIml): Hg>+ at (pg/ml): Ammoniaat (as N) sludge River Dam water enterotoxinboiL - (~LgIml): sewage effluent water 100 10 1 100 10 1 1,000 100 10 Cell culture CHO-Kl ______E ++ ++ -- ++ ++ ++ +++- Mahlavu - E ++ ----+++++- ++--- Mouse +- + + + +- +- ND ND +- ND ND +- ND ND Trypan blue -- ++

Hemolysis __ _ _ _-_- _- _- __ _ ++_ _-__ ++. a E, More than 15% of the cells were elongated; ND, not done; -- (hemolysis test), hematocrit volume of packed erythrocytes not smaller than that in negative controls. The other symbols are explained in footnote a of Table 1. b LT, Heat labile. titer hemagglutination tests of Microcystis toxin tometer or ordinary microscope slide, a reaction yielded results resembling the agglutination of mixture consisting of 0.1 ml of erythrocyte sus- erythrocytes. However, hemolysis tests showed pension (50 x 105 erythrocytes per ml) and 0.1 that the apparent diffusion of erythrocytes and ml of test suspension was sufficient and was lack of settling were mainly, probably exclusive- conveniently prepared in wells of microtiter ly, due to the lysis of cells. In tests of dilutions of plates. The hemolysis tests were simple and toxin preparations, no evidence was found that economical and yielded results with a clearly agglutination occurred in addition to lysis. Both defined endpoint within a short period. For reactions were temperature dependent. When instant availability, erythrocyte suspensions moderate concentrations of toxin were used, were stored at 4°C for several weeks. Decon- both tests yielded negative results at 4°C, even tamination of test suspensions was generally not after 16 h. At 10°C, both tests yielded marginally necessary, but it was important to have an positive results within 5 h, whereas at 35°C, isotonic reaction mixture to avoid the autolysis reactions had reached an endpoint within 5 h. of erythrocytes. Tests of ordinary samples of Reactions at different temperatures also failed to water therefore required the addition of sodium indicate agglutination in addition to lysis. No chloride to a final concentration of 0.85%. Un- difference was detected among the responses of der these conditions, negative results were re- the five types of erythrocytes. The results of corded for wastewater, river and dam water, E. titrations indicated that the lysis of erythrocytes coli enterotoxin, and relatively high concentra- by Microcystis toxin was a first-order kinetic tions of toxic compounds (Table 2). reaction. Similar results were recorded for all Response of T. pyriformis to Microcystis toxin. toxin preparations. No significant difference between suspensions Even though the sensitivities of the hemolysis with and without toxin in viability, motility, or test to various toxin preparations differed from morphology of T. pyriformis was detected, even those of microtiter cell culture and mouse toxici- after 16 h at 27°C. As far as could be established ty tests, no difference was detected among the by trypan blue tests, the toxin also had no effect hemolytic reactions to the toxin preparations. In on the membrane permeability. These results tests of 4-mg/ml samples of Hartbeespoort Dam show that toxin in concentrations highly cytolyt- Microcystis toxin, microscopic observation re- ic to mammalian cells and lethal to mice (Table vealed that the lysis of erythrocytes started 1) had no significant effect on T. pyriformis. immediately and that more than 95% of the cells Response of bacteria to Microcystis toxin. In were lysed within 2 to 8 min. The detection of saline suspensions containing toxin, the average low concentrations of toxin required prolonged count of E. coli cells increased from 38 x 103 to incubation. Mixtures of erythrocyte and test 71 x 106/ml in 5 h at 35°C, and in saline suspensions were successfully incubated in a suspensions without toxin, the count declined variety of tubes or vials and centrifuged directly from 35 x 103 to 23 x 103/ml. In similar tests on for screening purposes or analyzed accurately S. faecalis, the count in saline suspensions with after hematocrit centrifugation. For microscopic toxin increased from 20 x 103 to 85 x 105/ml, detection of erythrocyte lysis with a hemacy- and in negative controls, the average count VOL. 43, 1982 M. AERUGINOSA TOXIN 1431

TABLE 2-Continued Sample Pb2+ at (,ug/ml): Cd2+ at (pg/ml): Phenol at (pg/ml): Parathion at Chlordane at (±g/ Pentachlorophenol (pg/mi): ml): at (,ug/ml): 100 10 1 100 10 1 1,000 100 10 20 10 1 100 10 1 100 10 1

+- ND ND +- ND ND +- ND ND +-ND ND +- ND ND ND +- ND

declined from 21 x 103 to 9 x 103/ml. These The results of the trypan blue and hemolysis results show that both E. coli and S. faecalis tests show that one effect of the toxin was to actually thrived on the toxin preparations. The inactivate the cell membrane permeability barri- toxin had no detectable effect on the multiplica- er almost instantaneously. As in the case of the tion of the bacteria in broth for 24 h at 35°C. In sensitivities ofthe cell cultures, there was hardly broth with or without toxin, the average count of any difference among the sensitivities of the E. coli increased from 4 x 102 to 20 x 108/ml, different cells to this effect. Damage to cell and that of S. faecalis increased from 2.8 x 102 membranes substantiates the observation that to 90 x 107/ml. Similar results were recorded in destruction of liver cells is the earliest histologi- experiments in which preparations of toxic M. cal sign of a toxic effect on animals. Liver aeruginosa were replaced by homologous prepa- damage is characterized by the breakdown of rations of nontoxic strains. These results show sinusoidal endothelia, the disappearance of the that toxin in concentrations highly cytolytic to spaces of Disse between hepatocytes and sinu- mammalian cells and lethal to mice (Table 1) had soids, irregular-sized inter- and intralobular con- no detectable effect on E. coli and S. faecalis. nective tissue focus proliferation, leukocyte in- Ames tests. No mutagenic activity was detect- filtration, and progressive disruption of tissue ed in assays of any of the test strains in the structure (6, 10, 38). In vivo liver cell destruc- presence or absence of activating liver enzymes. tion, the damage to in vitro cell cultures characterized by almost complete cell disintegration (Fig. 1C and G), and the penetration of trypan DISCUSSION blue into cells that had been exposed for a few The absence of a significant difference among minutes to toxin in suspension cultures indicate the sensitivities of liver, lung, cervix, ovary, and that cell membrane destruction is probably the kidney cells shows that the Microcystis toxin did primary, possibly only, mechanism of action of not specifically attack liver cells. Histological Microcystis toxin. Destruction of the liver cell damage in animals is possibly first detected in permeability barrier would explain the early the liver (6, 10, 38) because toxins administered increase in serum levels of liver enzymes (10). orally or intraperitoneally are directly transport- The failure of disrupted hepatocytes and the ed to the liver, and, as a result of their metabolic liver structure to dispose of the influx of circu- activity, liver cells may absorb most of the toxic lating blood may account for the toxin poisoning effect. Epithelial cells (CHO-Kl, Mahlavu, symptoms of accumulation of blood in the liver HeLa, PLC/PRF/5) are possibly slightly more (10, 18, 37), pallor (18, 37), and death as a sensitive than fibroblast cells (MRC-5, Vero, consequence of circulatory collapse. The instan- MA-104, BGM) (Table 1). However, it should be taneous disruption of membrane permeability kept in mind that these tests were done in vitro and hemolysis indicate that protein synthesis is on transformed cells and that untransformed not involved in the cytolytic action of the toxin, cells under in vivo conditions may behave differ- which may explain the sudden onset of symp- ently. This is illustrated by the primary vervet toms of toxin poisoning in vivo and demonstrate kidney cells, which were more sensitive than that the toxin has no cytotoxic effect. transformed kidney cells or any of the other The absence of cell elongation or related mor- transformed cells. The observation that untrans- phological changes in all of the cell types after formed cells are more sensitive than trans- exposure to toxin shows that the mechanism of formed cells has been noted before (22). action of the toxin differs from that of E. coli 1432 GRABOW ET AL. APPL. ENVIRON. MICROBIOL. heat-labile enterotoxin. The latter induces typi- duced is not the same for all strains of toxic M. cal cell elongation in CHO-Kl (Fig. 1B), Mah- aeruginosa. The presence of various compounds lavu (Fig. 1F), and various other cell lines such with different molecular weights, physical prop- as HeLa and Vero by stimulation of the adenyl erties, and amino acid compositions in Micro- cyclase enzyme (15). The typical pattern of rapid cystis extracts has been established (11, 28, 37), and extensive cellular disintegration by Micro- and the data on the hemolytic activities of differ- cystis toxin (Fig. 1C and G) may prove useful in ent preparations are in agreement with the hem- the identification and characterization of the agglutination test results which did not correlate toxin: the pattern was also distinguishable from quantitatively with data on the lethal doses of that of heavy metal and pesticide toxin effects different toxin preparations for mice (6). (Table 2), which included the gradual rounding The in vitro procedures described in this study of cells, death, and no early disintegration (Fig. make it possible to investigate the effect of 1D and H). toxins on the morphology and chemical compo- The failure of Microcystis toxin to significant- sition of membranes of selected cells under ly affect the protozoan T. pyriformis, the gram- controlled conditions and to distinguish between negative bacterium E. coli, and the gram-posi- cytolytic and cytotoxic activities. The negative tive bacterium S. faecalis indicates that the results recorded in multiple tests of settled and toxin may specifically attack cell membranes of treated wastewater, river and dam water lacking warm-blooded animals. This is substantiated by a toxic algal bloom, and solutions of toxic chem- the temperature dependence of erythrocyte ly- icals and pesticides (Table 2) in concentrations sis. The resistance of T. pyriformis shows that which are rarely exceeded in natural bodies of the toxin may be similar to toxins of blue-green water (16) indicate that false-positive results of algae such as Anabaena flos-aquae, which do the cell assays are unlikely to be due to the not affect the protozoan Paramecium caudatum, conventional pollutants in the test samples. In and may differ from toxins such as those of terms of analytical methods, the microtiter cell Fischerella epiphytica and Gloeotricha echinu- culture assay was more sensitive than the con- lata, which inactivate P. caudatum (31). ventional mouse toxicity test to all preparations, The erythrocyte lysis results and the lack of and the hemolysis test was the most sensitive to evidence of agglutination suggest that unless the Hartbeespoort Dam toxin. The in vitro cell different compounds were involved, the hemag- assays may also prove useful for the general glutination reported in an earlier study (6) may screening of water quality. Although they are have been mistaken for hemolysis. This indi- less sensitive to toxicants than more sophisticat- cates that the mechanism of action of Microcys- ed biological toxicity assays (21, 26, 27), they tis toxin differs from that of hemagglutinating may be used for the detection of shock loads of a toxic lectins of higher plants (6, 19, 20, 36) and variety of potentially acute toxicants because of hemagglutinating compounds of certain marine being simple, quick, and economical. algae (5, 6). The molecular weights of plant Even though the results of our studies of toxic lectins (in the range of 100,000) are also much M. aeruginosa from five sources were in close higher than that of Microcystis toxin (6, 11), and agreement with one another, comparison of our the survival time of 2 to 4 days after acute findings with those of other studies should be dosages of toxic lectins is much longer than that done with caution because M. aeruginosa, as observed after acute dosages of Microcystis well as other blue-green algae, produces a vari- toxin. It appears that the mechanism of action of ety of toxic compounds with different proper- Microcystis toxin may be more similar to that of ties, toxic effects, and mechanisms of action (6, bacterial cytolytic toxins (2, 24) or even snake 7, 11, 18, 22, 28, 34, 37). The toxicity noted in venom cardiotoxins (25). The negative Ames this study may be the result of combined or test results show that the toxin differs from synergistic effects of different compounds (2, mutagenic and carcinogenic compounds of cer- 24), and the effects of other compounds may tain marine algae (23), fungi (4, 39), and higher have been antagonized or obscured. Similar plants (40). studies on chemically purified and characterized The cell assays described in this study should fractions are therefore in progress. prove useful in research on the characterization and mechanism of action of Microcystis toxin ACKNOWLEDGMENTS and related toxins. The observation that the We thank D. P. Botes and J. Alexander for advice and J. A. sensitivity of hemolysis tests to preparations of Pienaar and H. Kruger for technical assistance. different harvests of toxic M. aeruginosa did not correlate with those of mouse toxicity and cell LITERATURE CITED 1. Alexander, J., G. Macnab, and R. Saunders. 1978. Studies culture tests suggests that these three tests mea- on in vitro production of hepatitis B surface antigen by a sure activities of different components and that human hepatoma cell line. Perspect. Virol. 10:103-120. the ratio in which these components are pro- 2. Alouf, J. E. 1977. Cell membranes and cytolytic bacterial VOL. 43, 1982 M. AERUGINOSA TOXIN 1433

toxins, p. 219-270. In P. Cuatrecasas (ed.), The specificity substances in water by means of a mammalian cell culture and action of animal, bacterial and plant toxins. Chapman technique. Water Res. 15:553-559. and Hall, London. 22. Kirpenko, Y. A., V. V. Stankevich, V. M. Orlovskiy, N. I. 3. Ames, B. N., J. McCann, and E. Yamasaki. 1975. Methods Kirpenko, A. V. Bokov, and T. F. Karpenko. 1979. A for detecting carcinogens and mutagens with the Salmio- comparative assessment of the toxic effect of biologically nella/mammalian-microsome mutagenicity test. Mutat. active substances of blue-green algae at the cellular and Res. 31:347-364. organismic levels. Hydrobiol. J. 15:83-86. 4. Bjeldanes, L. F., G. W. Chang, and S. V. Thomson. 1978. 23. Leary, J. V., R. Kfir, J. J. Sims, and D. W. Fulbright. Detection of mutagens produced by fungi with the Salmo- 1979. The mutagenicity of natural products from marine nella typhimuriium assay. Appl. Environ. Microbiol. algae. Mutat. Res. 68:301-305. 35:1150-1154. 24. Le Vine, H., and P. Cuatrecasas. 1981. An overview of 5. Blunden, G., D. J. Rogers, and W. F. Farnham. 1975. toxin-receptor interactions. Pharmacol. Ther. 12:167-207. Survey of British seaweeds for hemagglutinins. Lloydia 25. Louw, A. I., and L. Visser. 1977. Kinetics of erythrocyte 38:162-168. lysis by snake venom cardiotoxins. Biochim. Biophys. 6. Carmichael, W. W., and P. E. Bent. 1981. Hemagglutina- Acta 498:143-153. tion method for detection of freshwater cyanobacteria 26. Morgan, W. S. G. 1975. Monitoring pesticides by means (blue-green algae) toxins. Appl. Environ. Microbiol. of changes in electric potential caused by fish opercular 41:1383-1388. rhythms. Prog. Water Technol. 7:33-40. 7. Collins, M. 1978. Algal toxins. Microbiol. Rev. 42:725- 27. Morgan, W. S. G. 1977. An electronic system to monitor 746. the effects of changes in water quality on fish opercular 8. DuPont, H. L., S. B. Formal, R. B. Hornick, M. J. Sny- rhythms, p. 38-55. In J. Cairns, Jr., K. L. Dickson, and der, J. P. Libonati, D. G. Sheahan, E. H. LaBrec, and G. F. Westlake (ed.), Biological monitoring of water and J. P. Kalas. 1971. Pathogenesis of Escherichia coli diar- effluent quality, ASTM STP 607. American Society for rhea. N. Engl. J. Med. 285:1-9. Testing and Materials, Philadelphia. 9. Evans, D. J., Jr., D. G. Evans, and S. L. Gorbach. 1973. 28. Murthy, J. R., and J. B. Capindale. 1970. A new isolation Production of vascular permeability factor by enterotoxi- and structure for the endotoxin from Microcvstis aeriugin- genic Escherichia coli isolated from man. Infect. Immun. osa NRC-1. Can. J. Biochem. 48:508-510. 8:725-730. 29. Oefinger, P. E., D. L. Bronson, and G. R. Dreesman. 10. Falconer, I. R., A. R. B. Jackson, J. Langley, and M. T. 1981. Induction of hepatitis B surface antigen in human Runnegar. 1981. Liver pathology in mice in poisoning by hepatoma-derived cell lines. J. Gen. Virol. 53:105-113. the blue-green alga Mic rocvstis aeru4ginosa. Aust. J. Biol. 30. Prozesky, 0. W., C. Brits, and W. 0. K. Grabow. 1973. In Sci. 34:179-187. vitro culture of cell lines from Australia antigen positive 11. Gorham, P. R., and W. W. Carmichael. 1979. Phycotox- and negative hepatoma patients, p. 358-360. In S. J. ins from blue-green algae. Pure Appl. Chem. 52:165-174. Saunders and J. Terblanche (ed.), Liver. Pitman Medical 12. Grabow, W. 0. K., J. S. Burger, and E. M. Nupen. 1980. Books, London. Evaluation of acid-fast bacteria, Candida albicans, enter- 31. Ransom, R. E., T. A. Nerad, and P. G. Meier. 1978. ic viruses and conventional indicators for monitoring Acute toxicity of some bluegreen algae to the protozoan wastewater reclamation systems. Prog. Water Technol. caudatum. J. Phycol. 14:114-116. 12:803-817. Paramecilum 13. Grabow, W. 0. K., R. Denkhaus, and P. G. van Rossum. 32. Reynolds, C. S. 1980. Cattle deaths and blue-green algae: 1980. Detection of mutagens in wastewater, a polluted a possible instance from Cheshire, England. J. Inst. Water river and drinking-water by means of the Ames Salmonel- Eng. Sci. 34:74-76. la/microsome assay. S. Afr. J. Sci. 76:118-123. 33. Schoub, B. D., A. S. Greeff, G. Lecatsas, 0. W. Prozesky, 14. Grabow, W. 0. K., and E. M. Nupen. 1981. Comparison I. T. Hay, J. G. Prinsloo, and R. C. Ballard. 1977. A of primary kidney cells with the BGM cell line for the microbiological investigation of acute summer gastroen- enumeration of enteric viruses in water by means of a tube teritis in black South African infants. J. Hyg. 78:377-385. dilution technique, p. 253-256. In M. Goddard and M. 34. Schwimmer, M., and D. Schwimmer. 1968. Medical as- Butler (ed.), Viruses and wastewater treatment. Perga- pects of phycology, p. 279-358. In D. F. Jackson (ed.), mon Press, Oxford, England. Algae, man and the environment. Syracuse University 15. Guerrant, R. L., L. L. Brunton, T. C. Schnaitman, L. I. Press, Syracuse. Rubhun, and A. G. Gilman. 1974. Cyclic adenosine mono- 35. Scott, W. E. 1974. The isolation of Microcystis. S. Afr. J. phosphate and alteration of chinese hamster ovary cell Sci. 70:179. morphology: a rapid, sensitive in vitro assay for the 36. Sharon, N., and H. Lis. 1972. Lectins: cell-agglutinating enterotoxins of Vibrio cholerae and Escherichia coli. and sugar-specific proteins. Science 177:949-959. Infect. Immun. 10:320-327. 37. Toerien, D. F., W. E. Scott, and M. J. Pitout. 1976. Mi- 16. Hattingh, W. H. J. 1977. Reclaimed water: a health haz- crocvstis toxins: isolation, identification, implications. ard? Water S A (Pretoria) 3:104-112. Water S A (Pretoria) 2:160-162. 17. Hoskins, J. M. 1967. Virological procedures, p. 25-319. 38. Tustin, R. C., S. J. van Rensburg, and J. N. Eloff. 1973. Butterworths Publishers, Inc., London. Hepatic damage in the primate following ingestion of toxic 18. Hughes, E. O., P. R. Gorham, and A. Zehnder. 1958. algae, p. 383-385. In S. J. Saunders and J. Terbianche Toxicity of a unialgal culture of Microcvstis aeruiginosa. Liver. Pitman Medical London. Can. J. Microbiol. 4:225-236. (ed.), Books, 19. Ikegwuonu, F. I., and 0. Bassir. 1977. Effects of phytohe- 39. Wehner, F. C., P. G. Thiel, and R. Vleggaar. 1979. Mu- magglutinins from immature legume seeds on the function tagenicity of 5,6-dimethoxysterigmatocystin, a metabolite and enzyme activities of the liver, and on the histopatho- from Aspergillus multicolor, in the Salmonella/microsome logical changes of some organs of the rat. Toxicol. Appl. system. Appl. Environ. Microbiol. 38:1015-1017. Pharmacol. 40:217-226. 40. Weisburger, E. K. 1979. Natural carcinogenic products. 20. Kakade, M. L., K. K. Keahey, C. K. Whitehair, and R. J. Environ. Sci. Technol. 13:278-281. Evans. 1965. Morphological changes in rats fed Navy 41. Wintrobe, M. M., G. R. Lee, D. R. Boggs, T. C. Bithell, beans. Proc. Soc. Exp. Biol. Med. 119:934-937. J. W. Athens, and J. Foerster. 1974. Clinical hematology, 21. Kfir, R., and 0. W. Prozesky. 1981. Detection of toxic 7th ed., p. 110-117. Lea & Febiger, Philadelphia.