Vol. 55, No. 1APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Jan. 1989, p. 190-1970099-2240/89/010190-08$02.00/0Copyright © 1989, American Society for Microbiology

Biotransformation and Detoxification of T-2 Toxin by Soil andFreshwater Bacteria

STEVEN BEETON AND ALAN T. BULL*

Biological Laboratory, University of Kent at Canterbury, Canterbury, Kent CT2 7NJ, United Kingdom

Received 12 July 1988/Accepted 26 October 1988

Bacterial communities isolated from 17 of 20 samples of soils and waters with widely diverse geographicalorigins utilized T-2 toxin as a sole source of carbon and energy for growth. These isolates readily detoxified T-2toxin as assessed by a Rhodotorula rubra bioassay. The major degradation pathway of T-2 toxin in the majorityof isolates involved side chain cleavage of acetyl moieties to produce HT-2 toxin and T-2 triol. A minordegradation pathway of T-2 toxin that involved conversion to neosolaniol and thence to 4-deacetyl neosolaniolwas also detected. Some bacterial communities had the capacity to further degrade the T-2 triol or 4-deacetylneosolaniol to T-2 tetraol. Two communities, TS4 and KS10, degraded the trichothecene nucleus within 24 to48 h. These bacterial communities comprised 9 distinct species each. Community KS10 contained 3 primarytransformers which were able to cleave acetate from T-2 toxin but which could not assimilate the side chainproducts, whereas community TS4 contained 3 primary transformers which were able to grow on the cleavageproducts, acetate and isovalerate. A third community, AS1, was much simpler in structure and contained onlytwo bacterial species, one of which transformed T-2 toxin to T-2 triol in monoculture. In all cases, the completecommunities were more active against T-2 toxin in terms of rates of degradation than any single bacterialcomponent. Cometabolic interactions between species is suggested as a significant factor in T-2 toxindegradation.

The trichothecenes are a group of toxic secondary metab-olites produced by several genera of fungi, such as Fusar-ium, Myrothecium, Trichothecium, and Stachybotrys (Table1) (1, 15, 21). Ingestion of mouldy cereal crops contaminatedby these toxigenic fungi has resulted in serious outbreaks ofmycotoxicoses in humans and domesticated animals (4, 9).The toxicological effects of these mycotoxins include cyto-toxicity (7) and potent inhibition of eucaryotic protein syn-thesis (21).

Studies on trichothecene transformations have centeredon in vivo and in vitro mammalian systems, and in general,these studies have revealed the production of other less toxictrichothecenes (5, 13, 17, 23). Mammalian transformations oftrichothecenes appear to be catalyzed by nonspecific car-boxylesterases located in the microsomal fraction of liver (5,10). Early research also showed that a number of microor-ganisms were able to transform but not to degrade tricho-thecenes (3, 26, 27).

Interest in microbial transformations was stimulated bythe discovery of a detoxification pathway for T-2 toxin[4, - 15 - diacetoxy - 8a -(3 - methylbutyryl - oxy) - 12,13 - epoxy -trichothec-9-ene] and deoxynivalenol [3a,7a-15-trihydroxy-12,13-epoxytrichothec-9-ene-8-oneJ, two naturally occurringtrichothecenes found in mouldy cereals, to their respectivenontoxic de-epoxy forms (11, 12). The possible role of soilbacteria in trichothecene transformation was suggested byUeno et al. (22). Aerobic axenic cultures of a Curtobacte-rium species strain 114-2, isolated from soil, transformed T-2toxin to T-2 triol [3a,43-15-trihydroxy-8-(3-methylbutyry-loxy)-12,13-epoxytrichothec-9-ene] via HT-2 toxin [15-ace-toxy - 8a - (3 - methylbutyryloxy) - 3a,4I3 - dihydroxy - 12,13 -epoxy-trichothec-9-ene], two trichothecenes belonging tothe T-2 toxin series. T-2 toxin could be utilized by thisbacterium as a sole source of carbon and energy, and theauthors reported that prolonged incubation resulted in the

* Corresponding author.

disappearance of T-2 triol without the further production ofany related trichothecene structures (16, 22). In commonwith in vitro transformations by mammalian systems (6, 10),the initial deacetylation reactions of the Curtobacteriumisolate were catalyzed by soluble carboxylesterases.We have investigated the role of natural bacterial commu-

nities and monocultures as agents for the detoxification andbiodegradation of T-2 toxin and related trichothecenes.Bacterial communities capable of detoxification and biodeg-radation were enriched from soil and freshwater samples,and their detoxification capacities were screened by a re-cently described bioassay (19) based on the sensitivity of theyeast Rhodotorula rubra. The efficacy of bacterial commu-nities from widely diverse geographical locations in thedetoxification of T-2 toxin is discussed, together with somefeatures of the community structures.

MATERIALS AND METHODSChemicals. T-2 toxin, deoxynivalenol, and neosolaniol

[4,-15-diacetoxy-3co,8a-dihydroxy-12,13-epoxytrichothec-9-ene] were gifts from the Chemical Defence Establishment,Porton Down, Wiltshire, England. Standard trichothecenekits containing T-2 toxin, HT-2 toxin, T-2 triol, T-2 tetraol[3a,4,B,8a - 15 - tetrahydroxy - 12,13 -epoxytrichothec - 9 - ene],verrucarol [4,B-15-dihydroxy-12,13-epoxytrichothec-9-ene],and diacetoxyscirpenol [3-hydroxy-4,-15-diacetoxy-12,13-epoxytrichothec-9-ene] were purchased from Sigma LondonChemical Company, Pool, Dorset, England. Tetraethylenepentamine, 4-dimethylamino pyridine, 3-nitrobenzoyl chlo-ride, and 4-nitrobenzyl pyridine were purchased from Fluka,Buch, Switzerland. All other chemicals or solvents wereeither chemically pure or of analytical reagent grade.

Environmental samples. Environmental samples for en-richment isolation of T-2 toxin-detoxifying microorganismswere collected from the following sites (bacterial communitydesignations in parentheses): leaf litter and soils, Canter-bury, Kent, England (KS); the River Stour, Kent, England

190

on May 22, 2021 by guest

http://aem.asm

.org/D

ownloaded from

TRANSFORMATION OF T-2 TOXIN 191

TABLE 1. Chemical structures and chromatographic data forT-2 toxin and other trichothecenes

1 2 3 A

T-2trolOH H0129.0 2.1

CH2R 2~~

4-Deactheneosolanio OH OAOH N LN

Retenionime (m) o p-ntrobetuoedervatves

(TS);hoteffuncaermsrenea pape manufacture HPChin

result'1 2 3 A B

T-2 toxin OAc OAc V 0.53 0.59 9.45HT-2 toxin OH OAc V 0.23 0.21 18.40Neosolaniol OAc OAc OH 0.32 0.30 8.35T-2 tdiol OH OH V 0.12 0.09 26.174-Deacetylneosolaniol OH OAc OH NDod ND NDT-2 tetraol OH OH OH 0.05 0.02 20.54

a See diagram above. OAc, Acetate; V, isovalerate.lDetermined by TLC (see Materials and Methods).oRetention time (min) of p-nitrobenzoyl derivatives.reND, Not determined.

(KW); soils from deciduous forest, Chiang Mai, Thailand(TS); effluent water from screen paper manufacture, ChiangMai, Thailand (TW1); soils from Signy Island, Antarctica(AS); the River Rhine, Cologne, Federal Republic of Ger-many (GW and GU). Soil samples (ca. 1.0 g) were shakenwith distilled water (5.0 ml), and water samples (1.0 ml) wereused directly as inocula for enrichment cultures.Enrichment isolation. Soil and water inocula were used to

set up enrichument cultures in a basal medium (8) supple-mented with the following trace elements (grams liter-L):H3BO3, 0.05; MnSO4 -4H20, 0.04; (NH4)6Mo7024, 0-02;KI, 0.01; CUS04 - H2O, 0.004; CoCl6 -6H20, 0.00. Theinitial pH of the medium was adjusted to 7.0 with NaOHsolution (1 M). Initial experiments were based on a T-2 toxinconcentration of 1 mg ml-a, but at this concentration, T-2toxin was insoluble, and therefore the concentration wasreduced to 1 mM (0.046 mgml-t), at which T-2 toxin wassoluble. Incubations (10 ml) were made in 100-ml Erlen-meyer flasks at 30C and 200 irpm. Isolates AS1 and AS2were incubated at 4°C without shaking. After 7 days ofincubation, samples (0.1 ml) of the enrichment cultures wereremoved aseptically and inoculated into 10 ml of fresh T-2toxin medium. Daily samples (0.46 ml) were removed asep-tically for analysis. The course of T-2 toxin biotransforma-tion was followed by thin-layer chromatography (TLC) andhigh-performance liquid chromatography (HPLC) of culturesupernatants. Supernatants were shaken three times withequal volumes of ethyl acetate (0.46 ml), and the organicphase was analyzed. The efficiency of this extraction proce-dure has been demonstrated by Yagen et al. (25), whoreported recoveries of T-2 toxin and HT-2 toxin from plasmaof 95 and 102%, respectively, by using a single ethyl acetateextraction. We have obtained similar recoveries from culturesupernatants. The culture supernatants were also subjectedto a bioassay to determine residual trichothecene toxicity.Bacteriological analyses were made as described below.

Assimilation of other trichothecenes by bacterial communi-ties. To basal medium (10 ml), separate additions of the

following trichothecenes were made: HT-2 toxin, 0.02 mM;neosolaniol, 1.0 mM; diacetoxyscirpenol, 1.0 mM; deoxyni-valenol, 1.0 mM; roridin A, 0.05 mM; verrucarin A, 0.5 mM;verrucarol, 0.1 mM. Incubations, sampling, and analyseswere made as for T-2 toxin enrichment experiments.Chemical analysis of trichothecenes. In TLC analysis, two

solvent systems were used: chloroform-acetone (3:2, vol/vol) (system A) and chloroform-methanol (95:5, vol/vol)(system B). T-2 toxin and trichothecene metabolites werevisualized either by spraying with 10% H2SO4 and thenheating or by treatment with 4-(p-nitrobenzyl) pyridine andthen with tetraethylenepentamine (20, 21). Photographicnegatives of developed TLC plates were analyzed for spotdensity on a transmittance densitometer (Bio-Rad Laborato-ries). After calibration to zero by measuring an area of thenegative with no spots, the negative was scanned and thepeaks were recorded. The area under each peak was cut outand weighed. The weight of the T-2 toxin peak in zero-timesamples was assigned as 100% of trichothecenes present.The other peaks were compared with this T-2 toxin valueand recorded as a percentage of the trichothecenes present.Standard trichothecenes (1 ,ug per spot) reacted differently tosulfuric acid development, resulting in a different spot inten-sity. Experimental samples were corrected for this differ-ence by multiplying them by the ratio of spot intensity of T-2toxin (1 ,ug per spot) to spot intensity of appropriate standardtrichothecenes. HPLC analysis of trichothecenes was madeby the method of Maycock and Utley (14). Trichothecenestandards for HPLC analysis (T-2 toxin, neosolaniol, HT-2toxin, T-2 triol, and T-2 tetraol) were derivatized separatelyand spiked into derivatized test samples to make identifica-tion of metabolites by cochromatography possible. For TLCanalysis, the underivatized standards described above (1 ,ugper spot) were chromatographed alongside experimentalsamples (Table 1).

Isolation of bacterial communities and individual species.Culture samples (0.1 ml) were diluted serially in sterile basalmedium, and selected dilutions (0.1 ml) were spread ontonutrient agar and the following minimal agar media (1 mM):sodium acetate, isovaleric acid, and T-2 toxin. Plates wereincubated at 30°C (4°C for isolate AS1) until the appearanceof CFU. CFU were counted and distinguished initially bycolony morphology. Putative monospecies were stored onnutrient agar slants and subcultured weekly. Bacterial com-munities and individual species were freeze-dried and storedat 4°C. Individual species were identified, to genus levelwhen possible, by the National Collection of Industrial andMarine Bacteria Ltd. (Torry Research Station, Aberdeen,Scotland).Metabolism of T-2 toxin by individual bacterial species.

Individual species from three T-2 toxin-degrading bacterialcommunities (AS1, KS10, and TS4) were precultured innutrient broth and were subcultured 18 h later (0.1 ml) in 10ml of T-2 toxin (1 mM) basal salts medium. After 5 days ofcultivation, trichothecene metabolites were searched for byTLC.

Bioassay for T-2 toxin. The toxicity of culture media andother trichothecene-containing samples was determined byan R. rubra bioassay based on the method of Stone et al.(19). R. rubra NRRL Y-7222 (mutant UV1/11) was sensitiveto T-2 toxin in the range of 2 to 10 ,ug ml- 1. Filter paper disks(6-mm diameter; 3MM; Whatman Inc., Clifton, N.J.) were

soaked with 10 RI of enrichment culture, air dried, and thentreated with 10 pul of chloroform to lyse bacteria present inthe sample.

VOL. 55, 1989

on May 22, 2021 by guest

http://aem.asm

.org/D

ownloaded from

192 BEETON AND BULL

TABLE 2. Abilities of bacterial communities to detoxify andgrow on T-2 toxin as the sole source of carbon and energy

Time forEnrichment source disappearanceb Growth onand isolate numbera toxin(h T-2 toxin

Antarctica (SignyIsland)AS1 48 +AS2 >120 -

England (Kent)KS1 72 +KS2 48 +KS3 48 +KS4 48 +KS5 48 +KS6 48 +KS7 48 +KS10 24 +KW3 36 +KW4 >320 -

Federal Republic ofGermany (Cologne)

GUl >120 -GW1 36 +

Thailand (Chiang Mai)TW1 48 +TS2 96 +TS3 72 +TS4 48 +TS5 72 +TS6 72 +

aAbbreviations (second letter of each isolate designation): S, soil; W,water; U, sewage sludge.

b As determined by bioassay.

RESULTS

Screening for T-2 toxin detoxifying activity. Of the 20environmental samples, 17 produced growth after incubationin T-2 toxin basal salts medium. During and after batchenrichment isolation, the toxicity of the culture medium wasdetermined by bioassay. R. rubra was sensitive to concen-trations of T-2 toxin above 2.5 pg mnl1, with a lower limit ofsensitivity between 1.0 and 2.5 ,ug ml-'; however, this yeastshowed no sensitivity to HT-2 toxin, T-2 triol, T-2 tetraol, orneosolaniol in the range of 2.5 to 10.0 tig ml-'. A linearresponse between the diameter of the zone of inhibition andT-2 toxin concentration was found between 2.5 and 15 ,ugml-'. Table 2 summarizes data derived from the first sub-culturing of the bacterial communities obtained from theprimary enrichment isolations. Each culture was analyzedfor growth and T-2 detoxification as a function of theincubation time. Enrichment cultures showing increasedturbidity became nontoxic to R. rubra at between 24 and 96h, depending upon the particular enrichment source (Table2). The time taken for the disappearance of toxicity to R.rubra decreased with subsequent subculturing of the originalenrichment culture. For example, isolates KS10 (deciduousleaf litter) and TS4 originally detoxified T-2 toxin within 48and 72 h, but after a single subculture in T-2 basal medium,detoxification had occurred by 24 and 48 h, respectively.These results strongly suggest adaptation of the communitiesto growth on T-2 toxin.

Biotransformation profiles of T-2 toxin. T-2 toxin transfor-mation products from selected time course experiments

100190

80

70'cnL 601zLJo 50.LU

0

40.

0 3 4 5 6 7

TIME (DAYS)FIG. 1. Transformation of T-2 toxin by community TW1. Sym-

bols: 0, T-2 toxin; 0, HT-2 toxin; A, T-2 triol.

were analyzed by TLC. Experimental samples were runalongside standards, and Rf values were calculated andcompared (Table 1). Two of the isolated communities (TW1;and AS1, undisturbed pasture) produced metabolites of T-2toxin which cochromatographed with HT-2 toxin and T-2triol. In both cases, the HT-2 toxin concentration increasedto a maximum after 2 to 3 days of incubation as T-2 toxin wastransformed; thereafter, as HT-2 toxin was transformed, T-2triol increased to a maximum at 3 to 5 days (Fig. 1). Withcommunities KW3 and TS4, a further metabolite which hadRf values of 0.05 and 0.02 in solvent systems A and B,respectively, and which cochromatographed with T-2 tetraolwas observed. The pattern of trichothecene metabolism bycommunity KW3 (Fig. 2) showed that HT-2 toxin wastransformed and that a lag period of 2 days followed beforethe T-2 triol was converted to T-2 tetraol. Subsequently, theT-2 tetraol was rapidly metabolized without the furtherproduction of other trichothecene-related compounds (Fig.2). Community TS4 transformed T-2 toxin in a sequencesimilar to that of community KW3, with the production ofT-2 tetraol from T-2 triol, but TS4 was more active thanisolate KW3, since transformed products showed little ten-

1001

90

80'

s2 70'

L 60zLL( 50LLI 400O 3Q.Cr 20-

10

0

10 1 2 3 4 5 6 7

TIME (DAYS)FIG. 2. Transformation of T-2 toxin by community KW3. Sym-

bols: 0, T-2 toxin; 0, HT-2 toxin; A, T-2 triol; A, T-2 tetraol.

APPL. ENVIRON. MICROBIOL.

on May 22, 2021 by guest

http://aem.asm

.org/D

ownloaded from

TRANSFORMATION OF T-2 TOXIN 193

0 24 4 8 72 96 120 144

TIME ( h)FIG. 5. Batch growth of communities KS10 and TS4. KS10

grown on T-2 toxin (0) or on acetate plus isovalerate (A); TS4grown on T-2 toxin (0) or on acetate plus isovalerate (A).

less neosolaniol produced (17%). As neosolaniol was gradu-ally transformed, an unknown metabolite (previously de-tected in GW1) appeared in the culture supernatant.To substantiate the identity of putative T-2 transformation

products, two communities (KS10 and TS4) were resusci-tated from the freeze-dried state and grown in T-2 toxinmedium. During 7 days of incubation, aliquots (0.429 ml,equivalent to 200 jig of T-2 toxin at time zero) were extractedand analyzed by HPLC. The retention times of referencetrichothecenes are shown in Table 1. Trichothecenes in theculture samples were separated by HPLC, and after theaddition of known quantities of trichothecene standards, acorresponding increase in peak area was observed in theexperimental samples. The unknown metabolite producedby communities GW1 and KS10 had a retention time of 15.45min and was eluted between T-2 toxin and HT-2 toxin. Thegeneral profile of T-2 toxin transformation as revealed byHPLC was consistent with that obtained by TLC (Fig. 6).TLC analyses of the communities (Fig. 1-4) revealed that a100% recovery of trichothecenes was not obtained. Forexample, in Fig. 1, the sum of the percentages of HT-2 toxinand T-2 triol after 3 days does not equal the initial percentageof T-2 toxin. It is unclear whether this is due to theinefficiency of the extraction procedure, which may favorextraction of nonpolar trichothecenes, or to another path-way, such as de-epoxidation operating in parallel with theside chain cleavage reactions, which removes epoxy-tricho-thecenes from the culture.

Microbial analysis of bacterial communities. Two of theisolated communities, KS10 and TS4, were resuscitatedfrom the freeze-dried state by incubation in T-2 toxin mini-mal salts medium at 30°C. After 7 days of growth, aliquots(0.1 ml) were subcultured further into fresh T-2 toxin mini-mal salts medium or into sodium acetate and isovalerate saltsmedium. The community growth profiles of KS10 and TS4grown on T-2 toxin and the acetate-isovalerate mixture areshown in Fig. 5. Both the communities had similar maximumspecific growth rates on T-2 toxin and on a mixture of acetateand isovalerate (TS4: T-2 toxin, 0.10 h-1; acetate andisovalerate, 0.12 h-1; KS10: T-2 toxin, 0.10 h-1; acetate and

-1*0

-2-0

Ec00

UJz

mC -4-0-0co

C - 5- 0

0 1 2 3

TIME (DAYS)4 5 6 7

-6 0

FIG. 3. Transformation of T-2 toxin by community TS4. Sym-bols: 0, T-2 toxin; 0, HT-2 toxin; A, T-2 triol; A, T-2 tetraol.

dency to accumulate in the culture (Fig. 3). Further analysisof community TS4 by TLC and color development with4-(p-nitrobenzyl) pyridine and tetraethylpentamine indicatedthat no trichothecenes remained in the culture medium after5 days.Communities GW1 and KS10 produced a trichothecene

metabolite, in addition to HT-2 toxin, T-2 triol, and T-2tetraol, that had Rf values of 0.32 and 0.30 in solvent systemsA and B, respectively. This metabolite cochromatographedwith neosolaniol. Neosolaniol and HT-2 toxin (each ca. 35%)reached maximum concentrations at 2 days in cultures ofisolate GW1 (Fig. 4); thereafter, HT-2 toxin was transformedwith a corresponding increase in T-2 triol and T-2 tetraol. Anunknown metabolite (Rfs, 0.08 and 0.06 in solvent systems Aand B, respectively) had a mobility between those of T-2triol and T-2 tetraol and began to increase in concentration inGW1 culture supernatants as neosolaniol reached a maxi-mum. With prolonged culture (data not shown) of up to 14days, only T-2 tetraol remained in the culture medium. Thepattern of T-2 toxin degradation by community KS10 wassimilar to that of community GW1 but with proportionally

TIME (DAYS)

FIG. 4. Transformation of T-2 toxin by community GW1. Sym-bols: 0, T-2 toxin; 0, HT-2 toxin; A, T-2 triol; A, T-2 tetraol; OI,neosolaniol; *, unidentified trichothecene.

100

90

80

aQ 70V)zwZ 60llJ1E40-I-0I 30-0tr 20

10

100.

90.

80

aS 70

W 60zo 50LIIE 40

0

z 20-

10

0'

0 1 3 4

VOL. 55, 1989

on May 22, 2021 by guest

http://aem.asm

.org/D

ownloaded from

194 BEETON AND BULL

TABLE 3. Identification of components of two T-2toxin-transforming communities isolated from soil

Community Identityaand isolate

KS101.Pseudomonas-Alcaligenes-

Agrobacterium sp.2.Blastobacter natatorius3.Pseudomonas-Xanthomonas-

Alcaligenes sp.4.Pseudomonas-Xanthomonas-

Alcaligenes sp.5.Pseudomonas-Alcaligenes sp.6.Memberof the family Rhizobiaceae7.Not determinable8.Atypical Arthrobacter sp.9.Atypical Arthrobacter sp.

TS41.Arthrobacter-Brevibacterium sp.2.Pseudomonas sp.3.Pseudomonas sp.4.Pseudomonas sp.5.Pseudomonas sp.6.Arthrobacter-Brevibacterium-

Microbacterium sp.7.Arthrobacter sp.8.Pseudomonas sp.9.Pseudomonas-Alcaligenes sp.

a Analyses made by the National Collection of Industrial and MarineBacteria Ltd., Aberdeen, Scotland.

isovalerate, 0.12 h-1). Initially, the growth of the communi-ties on T-2 toxin was probably limited by the rate of sidechain cleavage. Community TS4 produced a higher finalculture density than community KS10 when grown on T-2toxin, presumably because of assimilation of the products oftrichothecene ring degradation. These results are consistentwith the patterns of T-2 toxin degradation revealed bychromatographic analyses.

Microbiological analysis revealed that communities KS10and TS4 each were composed of nine distinct bacteria. Afterpurification of the isolates on nutrient agar, they wereidentified as far as possible by biochemical tests to assignthem to genera and species (Table 3). Both communitieswere composed of gram-negative and gram-positive bacteriacommonly found in soil and water environments. Only oneisolate, KS10-2, was identified to the species level and wasfound to be Blastobacter natatorius, an unusual gram-

TABLE 5. Transformation of T-2 toxin and growth on side chaincleavage products by individual species from bacterial

communities KS10 and TS4

Community Transformation Growth on:and isolate of T-2 toxina Acetate Isovalerate

KS101 - + +2 + _3 - +4 - +S - + +6 - _7 + +8 + _9 + _

TS41 + +2 + + +3 + + +4 - + +S + + +6 - +7 - +8 - + +9 - + +

a Transformation of T-2 toxin to T-2 triol or further within 5 days. ±, Traceamount of neosolaniol only.

negative rod previously isolated from swimming pools (18).Both communities, when growing on T-2 toxin, were domi-nated by a single species, designated KS10-5 and TS4-4 intheir respective communities (Table 4). The majority ofspecies in these communities made up less than 10% of thetotal population.

In contrast to these complex bacterial communities, iso-late AS1 was composed of only two Arthrobacter species.This community revealed a transformation sequence similarto that observed with other communities.Growth of individual species on T-2 toxin and putative side

chain cleavage products. Individual species isolated fromcommunities KS10 and TS4 were tested for the ability totransform T-2 toxin and to utilize acetate or isovalerate(Table 5). Community KS10 contained two primary utilizersin the sense of Bull and Slater (2), KS10-2 and KS10-9,which converted T-2 toxin to T-2 triol within 5 days ofincubation, and one primary utilizer, KS10-8, which con-verted T-2 toxin to T-2 triol together with trace amounts of

TABLE 4. Temporal variation in the species composition of bacterial communities KS10 and TS4 growing on T-2 toxin

% Isolate in community (on isolation day)a:

Isolate KS10b TS4

0 1 2 3 4 5 0 1 2 3 4 5

1 6.5 0.7 0.9 2.1 4.2 3.3 9.0 28.2 7.6 5.4 35.8 2.12 1.9 1.9 0.5 0.9 3.0 2.4 4.5 12.9 4.1 3.3 2.2 0.33 0.8 0.4 0.5 0.6 0.6 0 13.7 2.9 1.4 1.7 1.9 0.34 3.7 2.4 9.3 0.9 4.0 12.9 12.3 10.0 31.1 34.5 30.9 90.25 56.9 91.6 84.1 92.3 80.9 65.8 18.4 7.6 6.4 4.4 11.0 1.66 25.0 2.9 4.2 2.7 6.6 14.7 19.8 7.6 36.4 2.1 0.4 07 0 0.1 0 0.9 0 0.2 6.9 1.8 0.5 1.3 2.9 0.38 0.4 0 0 0 0.2 2.5 6.9 3.5 9.9 23.1 6.7 3.09 8.5 8.4 1.2 7.2 8.3 2.5a The zero-time sample represents the composition of the inoculum which was grown in T-2 toxin medium.b Colony morphologies of isolates KS10-4 and KS10-9 were indistinguishable; therefore, the data are recorded collectively under isolate KS10-4.

APPL. ENVIRON. MICROBIOL.

on May 22, 2021 by guest

http://aem.asm

.org/D

ownloaded from

TRANSFORMATION OF T-2 TOXIN 195

2001

TIME (DAYS)FIG. 6. Batch culture of community KS10 on T-2 toxin: HPLC analysis of trichothecenes. Symbols: 0, T-2 toxin; 0, HT-2 toxin; A, T-2

triol; A, T-2 tetraol.

T-2 tetraol. These three primary utilizers were unable toutilize acetate or isovalerate as a carbon source and werepresent as low percentages of the total viable count through-out the incubation period (Table 4). One organism, KS10-7,produced a trace amount of neosolaniol from T-2 toxin, butT-2 toxin remained largely unchanged after 7 days of incu-bation. Bacterial community TS4 also contained three pri-mary utilizers, TS4-1, TS4-2, and TS4-3, but, in contrast totheir counterparts in community KS10, they also utilizedacetate and isovalerate as growth substrates. In conse-quence, these three bacterial species were present in muchhigher proportions of the total community than the primaryutilizers in community KS10. Bacterial community AS1contained one primary utilizer (AS1-1) which was capable oftransforming T-2 toxin to T-2 triol, via HT-2 toxin, within 7days. The complete bacterial communities (AS1, KS10, andTS4) were more active, in terms of the rate of transformationof T-2 toxin, than any single primary utilizer of T-2 toxin.

Assimilation of trichothecenes other than T-2 toxin bybacterial communities. Bacterial communities KS10 and TS4were able to transform and utilize for growth HT-2 toxin,neosolaniol, and diacetoxyscirpenol. HT-2 toxin was quicklytransformed to T-2 triol and T-2 tetraol after 7 days; neoso-laniol was transformed more slowly after 5 days to anunknown metabolite with an Rf value of 0.08 (solvent systemA). Diacetoxyscirpenol was rapidly transformed by TS4 andKS10 within 1 and 2 days, respectively, to a metabolite (Rf0.35; solvent system A) that did not comigrate on TLC plateswith any available reference trichothecene. This metabolitewas subsequently transformed to a metabolite with an Rf of0.08. All of the trichothecenes described above supportedgrowth of KS10 and TS4 communities. In contrast, neithercommunity had the capacity to transform roridin A, verru-

carin A, or deoxynivalenol. However, community TS4 was

able to transform verrucarol but only after prolonged incu-bation (17 days).

DISCUSSION

It was reported recently (22) that T-2 toxin could betransformed by axenic cultures of 12 gram-positive aerobicbacteria enriched from soil environments, although no de-tails were provided of the soil types and origins or of the

number of positive enrichments obtained. The experimentsreported here show that 85% of the soil and water isolatesscreened were able to utilize T-2 toxin as a sole source ofcarbon and energy and that this activity was associated withthe removal of T-2 toxicity. Screening for T-2 toxin detoxi-fication was facilitated by the R. rubra bioassay, becauseresulting trichothecene transformation products were non-toxic to this yeast, thereby allowing a selective assay for T-2toxin. The wide range of environmental samples used in ourstudy originated from four different geographical areas, andwithin these areas a variety of soil and water types werescreened. All samples originating from Thailand were posi-tive, and only one sample from the other three geographicalareas showed a negative result for detoxifying activity.

Biotransformation of T-2 toxin to HT-2 toxin has beenreported for fungal systems (26), soil bacteria (22), andruminal microorganisms (11, 24); further transformation ofHT-2 toxin to T-2 triol (22, 24) has also been observed. Themajor transformation pathway for T-2 toxin in the majorityof our enrichment samples was to T-2 triol via HT-2 toxin;however, more active communities were able to transformthe T-2 triol to T-2 tetraol. This transformation sequenceresulted in progressively less toxic trichothecenes; for ex-ample, T-2 triol and T-2 tetraol are approximately 20-foldless toxic than T-2 toxin when administered to mice (28).Another minor pathway involving neosolaniol was suggestedin some of the enrichments. Community GW3 showed thatthis pathway to neosolaniol was actively producing compa-rable amounts of neosolaniol and HT-2 toxin as the firstmajor metabolites of T-2 toxin transformation (Fig. 4). Anunidentified transformation product was detected only whenneosolaniol was produced, and we suggest that the formermay be 4-diacetyl neosolaniol, which is further transformedto T-2 tetraol. 4-Diacetylneosolaniol has been found previ-ously as a transformation product of HT-2 toxin in rat liver(29), and it undergoes further transformation to T-2 tetraol.Although we have not substantiated the identity of diace-toxyscirpenol transformation products by isolates KS10 andTS4, it is possible that selective deacetylation of the 4-acetylmoiety of diacetoxyscirpenol would result in monoacetoxy-scirpenol [15-acetoxy-4p-dihydroxy-12,13-epoxytrichothec-9-ene] followed by deacetylation of the C15 acetyl moiety to

VOL. 55, 1989

on May 22, 2021 by guest

http://aem.asm

.org/D

ownloaded from

196 BEETON AND BULL

scirpentriol [3a,4,3-15-trihydroxy-12,13-epoxytrichothec-9-ene]. Ueno et al. (22) revealed that the transformationproduct of diacetoxyscirpenol, mediated by a Curtobacte-rium species, was scirpentriol, although no mention ofmonoacetoxyscirpenol as an intermediate was made. Mono-acetoxyscirpenol may have been overlooked in the Curto-bacterium system because of the long time of initial sampling(3 days). Results shown here indicate that the putativetransformation product monoacetoxyscirpenol is trans-formed within 3 days.

In the study presented here, only ethyl acetate-extractablemetabolites were chromatographed. It has been suggested(29) that extraction with XAD-2 resin is more efficient forpolar trichothecenes; however, reported TLC and HPLCdata have revealed that T-2 tetraol was extracted by ethylacetate, suggesting that other workers extracting transfor-mation products of T-2 toxin should have detected T-2tetraol if it was present.Many natural and xenobiotic compounds have been

shown to be degraded by microbial communities (2). Al-though axenic cultures of soil organisms that transform T-2toxin to T-2 triol have been isolated (22), the complexbacterial communities KS10 and TS4 were always more

active against T-2 toxin than were pure cultures of theconstituent bacteria. These two communities showed dif-ferent interactive relationships. The biotransformationmechanism of community KS10 involved cometabolism inthat the three primary utilizers KS10-2, KS10-8, and KS10-9were unable to utilize the initial side chain cleavage productsresulting from their transformation of T-2 toxin. Theseorganisms were found at low levels in the community (Table4), presumably surviving within the community by utilizingexcretion products from the other bacteria present. In com-

munity TS4, the primary utilizers TS4-1, TS4-2, and TS4-3possessed both transformation activity and side chain prod-uct assimilation. These primary utilizers were present athigher levels in the community than the correspondingspecies were in community KS10. Both communities con-

tained gram-negative soil bacteria able to transform T-2toxin, and this is the first reported case of the involvement ofgram-negative bacteria with such activity.We conclude that relatively high concentrations of T-2

toxin can be quickly transformed in various natural andman-made environments representative of very wide geo-

graphical locations and that this transformation results inprogressively less toxic trichothecenes and mineralization.

ACKNOWLEDGMENTS

This work was supported by Ministry of Defence contract 2099/016/CDE.We thank Richard Maycock for advice on HPLC analysis of

trichothecenes and Tim Rubidge and Ken Macdonald for helpfuldiscussions and encouragement throughout the project.

LITERATURE CITED

1. Bamburg, J. R., W. F. 0. Marasas, N. V. Riggs, E. B. Smalley,and F. M. Strong. 1968. Toxic spiroepoxy compounds fromfusaria and other hypomycetes. Biotechnol. Bioeng. 10:445-455.

2. Bull, A. T., and J. H. Slater. 1982. Microbial interactions andcommunity structure, p. 13-44. In A. T. Bull and J. H. Slater(ed.), Microbial interactions and communities, vol. 1. AcademicPress, Inc. (London), Ltd., London.

3. Claridge, C. A., and H. Schmitz. 1977. Microbial and chemical

transformations of some 12,13-epoxytrichothec-9,10-enes.Appl. Environ. Microbiol. 36:63-67.

4. Cole, R. J., and R. H. Cox. 1981. Handbook of toxic fungalmetabolites. Academic Press, Inc., New York.

5. Ellison, R. A., and F. N. Kotsonis. 1974. In vitro metabolism ofT-2 toxin. Appl. Microbiol. 27:423-424.

6. Fonnum, F., S. H. Sterri, P. Aas, and H. Johnsen. 1985.Carboxylesterases, importance for detoxification of organo-phosphorus anticholinesterases and trichothecenes. Fundam.Appl. Toxicol. 5:529-538.

7. Grove, J. F. 1969. The cytotoxicity of some transformationproducts of diacetoxyscirpenol. Chem. Commun. (J. Chem.Soc. Sect. D) 1969:1473-1478.

8. Jigami, Y., T. Omori, Y. Minoda, and K. Yamada. 1974.Screening of n-alkylbenzene assimilating yeasts and identifica-tion of oxidation products from n-alkylbenzenes. Agric. Biol.Chem. 38:401-408.

9. Joffe, A. S. 1971. Alimentary toxic aleukia, p. 139-189. In S.Kadis, A. Ciegler, and S. J. Ajl (ed.), Microbial toxins, vol. 7.Academic Press, Inc., New York.

10. Johnsen, H., E. Odden, 0. Lie, B. A. Johnsen, and F. Fonnum.1986. Metabolism of T-2 toxin by rat liver carboxylesterase.Biochem. Pharmacol. 35:1469-1473.

11. Kiessling, K.-H., H. Pettersson, K. Sandholm, and M. Olsen.1984. Metabolism of aflatoxin, ochratoxin, zearalenone, andthree trichothecenes by intact rumen fluid, rumen protozoa, andrumen bacteria. Appl. Environ. Microbiol. 47:1070-1073.

12. King, R. R., R. E. McQueen, D. Levesque, and R. Greehalgh.1984. Transformation of deoxynivalenol (vomitoxin) by rumenmicroorganisms. J. Agric. Food Chem. 32:1181-1183.

13. Matsumota, H., T. Ito, and Y. Ueno. 1978. Toxicological ap-proaches to the metabolites of fusaria. XII. Fate and distribu-tion of T-2 toxin in mice. Jpn. J. Exp. Med. 48:393-399.

14. Maycock, R., and D. Utley. 1985. Analysis of some tricho-thecene mycotoxins by liquid chromatography. J. Chromatogr.347:429-433.

15. Mirocha, C. J., S. V. Pathre, B. Schauerhamer, and C. M.Christensen. 1976. Natural occurrence of Fusarium toxins infeedstuff. Appl. Environ. Microbiol. 32:553-556.

16. Nakayama, K., A. Kato, Y. Ueno, Y. Minoda, T. Omori, and K.Komagata. 1980. Studies on the metabolism of trichothecenemycotoxins. II. Metabolism of T-2 toxin with the soil bacteria.Proc. Jpn. Assoc. Mycotoxicol. 12:30-32.

17. Ohta, M., K. Ishii, and Y. Ueno. 1977. Metabolism of tricho-thecene mycotoxins. I. Microsomal deacetylation ofT-2 toxin inanimal tissues. J. Biochem. 82:1591-1598.

18. Sly, L. I., and M. H. Hargreaves. 1984. Two unusual buddingbacteria isolated from a swimming pool. J. Appl. Bact. 56:479-486.

19. Stone, T. E., T. Rubidge, and K. D. Macdonald. 1986. R. rubrabio-assay: increase in sensitivity and wider applications. J.Microbiol. Methods 5:59-63.

20. Takitani, S., Y. Asabe, T. Kato, M. Suzuki, and Y. Ueno. 1979.Spectrodensitometric determination of trichothecene mycotox-ins with 4-(p-nitrobenzyl) pyridine on silica gel thin-layer chro-matography. J. Chromatogr. 172:335-342.

21. Ueno, Y. 1984. General toxicology, p. 135-146. In Y. Ueno(ed.), Trichothecenes: chemical, biological and toxicologicalaspects. Elsevier Scientific Publishing Co., New York.

22. Ueno, Y., K. Nakayama, K. Ishii, F. Tashiro, Y. Minoda, T.Omori, and K. Komagata. 1983. Metabolism of T-2 toxin inCurtobacterium sp. strain 114-2. AppI. Environ. Microbiol. 46:120-127.

23. Visconti, A., and C. J. Mirocha. 1985. Identification of variousT-2 toxin metabolites in chicken excreta and tissues. Appl.Environ. Microbiol. 49:1246-1250.

24. Westlake, K., R. I. Mackie, and M. F. Dutton. 1987. T-2 toxinmetabolism by ruminal bacteria and its effects on their growth.Appl. Environ. Microbiol. 53:587-592.

25. Yagen, B. B., M. Bialer, and A. Sintov. 1985. Gas chromato-graphic assay with pharmacokinetic applications for monitoringT-2 and HT-2 toxins in plasma. J. Chromatogr. 343:67-75.

26. Yoshizawa, T., and N. Morooka. 1975. Biological modification of

APPL. ENVIRON. MICROBIOL.

on May 22, 2021 by guest

http://aem.asm

.org/D

ownloaded from

TRANSFORMATION OF T-2 TOXIN

trichothecene mycotoxins: acetylation and deacetylation of de-oxynivalenols by Fusarium spp. Appl. Microbiol. 29:54-58.

27. Yoshizawa, T., C. Onomoto, and N. Morooka. 1980. Microbialacetyl conjugation of T-2 toxin and its derivatives. Appl. Envi-ron. Microbiol. 39:%2-966.

28. Yoshizawa, T., T. Sakamoto, and K. Okamoto. 1980. In vitro

formation of 3'-hydroxy T-2 and 3'-hydroxy HT-2 toxins fromT-2 toxin by liver homogenates from mice and monkeys. Appl.Environ. Microbiol. 47:130-134.

29. Yoshizawa, T., S. P. Swanson, and C. J. Mirocha. 1980. In vitrometabolism of T-2 toxin in rats. Appl. Environ. Microbiol. 40:901-906.

VOL. 55, 1989 197

on May 22, 2021 by guest

http://aem.asm

.org/D

ownloaded from