Folia Microbiol. 47 (5),467-472 (2002)

http: //www. biomed, cas. c z/tabu / folia /

Biotransformation of Trichloroethene by Pure Bacterial Cultures J. R0~.I~KAa, J. MOLLERa, D. VITa, V. HLrI~KA a, J. HOFFMANNa, H. DA'I'KOVAb, M. NI~MECe aDepartrnent01'Environmental Technology and Chemistry, Tomdg Bat'a University, 762 72 Zlfn.,Czechia blnstitute for Testing and Certification, 764 21 Zlfn-Louky, Czechia CDepartmentof Microbiology, Masaryk University, 602 O0Brno, Czechia Received 19 September 2001 Revised version 6 March 2002

ABSTRACT. From natural samples 11 isolates able to remove trichloroethene (CC12CHCI) from an aqueous environment were obtained which were capable of cometabolic degradation of CC12CHCI by an enzyme system for phenol degradation. At an initial CC12CHC1 concentration of 1 mg/L, the resting cells of particular cultures degraded 33-94 % CC12CHC1 during 1 d and their transformation capacity ranged from 0.3 to 3.1 mg CCI2CHC1 per g organic fraction. An analysis of a mixed phenol-fed culture with an excellent trichloroethene-degrading ability found a markedly minority isolate represented in the consortium to be responsible for this property. This culture degraded CCIzCHCI even at a low inoculum concentration and attained a transformation capacity of 14.7 mg CCI2CHC1 per g. The increase in chloride concentration after degradation was quantitative when compared with the decrease in organically bound chlorine. The degree of CCI2CHC1 degradation was affected by Me2S2; this substance can significantly reduce the degrading ability of some tested cultures (> 60 %); however, it does not cause this inhibition with others.

Trichloroethene (CChCHC1) is a typical representative of unsaturated chlorinated hydrocarbons which have been and still are used to a considerable extent in various industrial fields where, together with noncombustibility, their remarkable degreasing and dissolving qualities are particularly appreciated. At the same time, CCIzCHCI represents a significant contaminant of groundwaters and soils in industrial and military zones in a number of localities and countries. Its persistence in the environment is multiplied by the fact that under anaerobic conditions its microbial dehalogenation leads to lower chlorinated alkenes (cis-cfichloroethene, trans-dichloroethene, 1,1-dichloroethene) and subsequently to carcinogenic chloroethene (vinylchloride; Vogel et al. x987). Trichloroethene has been regarded for years as a microbially nondegradable substance. The first data on CC12CHC1 biotrausformation appeared in mid-1980s with a description of CC12CHC1 degradation in soil based on cometabolism (Wilson and Wilson I985). The capability of cometabolic CC12CHC1 degradation was later proved for bacteria of several genera growing on methane, toluene, phenol, propane (Chang and Alvarez-Cohen I995), propene (Ensign et al. 1992; Saeki et al. I999), ammonia (Arciero et al. I989), cumene (isopropylbenzene; Dabrock et al. 1992) or dimethylsulfide (Takami et al. I999). These substances induce in bacterial cells the formation of oxygenases capable of catalyzing the transformation of the pollutant. Shortly after discovery of these processes efforts appeared to apply bacteria for bioremediation of contaminated groundwaters. Microbial degradation of CCI2CHC1 can be performed by means of mixed microbial cultures (e.g. with activated sludge adapted to phenol or other substrate), utilizing pure bacterial cultures of natural origin and also applying genetically engineered bacteria capable of constitutive oxygenase production. Even though some aspects of CCI2CHCI bacterial degradation are still the object of research, field tests of groundwater decontamination have already been carried out (Hopkins et al. I993; Steffan et al. 1999). It was proved, e.g., that phenol injected into contaminated groundwaters together with oxygen can bring about a marked decrease in CCI2CHC1 concentration. With an initial CCI2CHCI concentration <500 Ixg/L, its drop down to 87-89 % after 30 h was achieved by an injection of 12.5 mg/L phenol and 35 mg/L oxygen (Hopkins et al. x993). Degradation of dichloroethenes was later proved as well, and hydrogen peroxide was also employed a s a n oxygen source (Hopkins and McCarty 1995). A pilot-plant test of remediating groundwaters contaminated with a mixture of chlorinated hydrocarbons, using a strain of Burkholderia cepacia constitutively expressing toluene-o-monooxygenase, is also described, where a decrease in volatile organic substances from a value of approximately 2200 to 250 I.tg/L appeared in the most conclusive case, and locally even below 50 ~g/L (Steffan et al. I999).

468 J. R0~IC*~A et al.

Vol.

47

The objective of our work was to obtain pure natural cultures able to remove CCI2CHCI from an aqueous environment and to study their properties, which are of key importance for CC12CHCI biotransformation.

MATERIALS

AND

METHODS

Isolation of phenol utilizin9 bacteria. Samples of soils, sediments or wastewaters were enriched 2 per d in a liquid medium with phenol (composition in g/L): NH4CI 1.1, K2HPO4 1.0, NaC1 0.5, phenol 0.3, MgSO4-7H20 0.2, FeSO4-7H20 0.01, CaC12 0.01, trace dement solution 1 mL (in rag/L): H3BO3 57, MnSO4"5H20 43, ZnSO4"TH20 43, CuSO4"5H20 40, (NH4)6Mo7024"4H20 37, Co(NO3)2"6H20 25. One loop was inoculated onto solid media (of the same composition, with added agar 18 g/L and bromothymol blue 60 rag/L) and cultivated at 25 *C. Colonies thus obtained were transferred to fresh solid media with phenol and subsequently to universal media. Purity of the cultures was checked microscopically by Gram staining and isolates were characterized by the OF-test, oxidase and catalase tests. Significant cultures were biochemically identified in advance by Nefermtest 24 (Lachema, Czechia). Determination of the CCI2CHCI transformation capacity of the isolates. Cells of each culture were enriched in a liquid medium with phenol 300 mg/L during 1 d, and an additional induction of cells was performed with phenol 200 mg/L for 2 h. The cells were harvested by centrifugation (21000 g, 10 rain) or by filtration and washed several times with physiological solution. Following further retention, the cells were resuspended in a liquid mineral medium (g/L: NaC1 0.5, K2HPO4 0.5, NH4C1 0.3, MgSO4"7H20 0.1, FeSO4"7HzO 0.01, CaCI2 0.01; trace element solution 1 mL) so that inoculum concentration reached an A620 of 0.7-0.8 (approximately corresponding to 300 mg biomass per L). Inoculum biomass was characterized by organic fraction: dry mass of 5 mL sample was burned at 550 *C for 3 h and weighed. Inoculum (20 mL) for CC12CHCI degradation was placed in 40-mL gastight sample vials (Supelco, USA) enclosed with Teflon-coated silicon septa. Immediately before closure, determined volumes of CCI2CHC1 stock solution (approximately i g/L) were dosed into the sample vials by means of a gast-ight syringe (Hamilton, USA) so as to attain a CC12CHC1 initial concentration of 750-850 Ixg/L in the aqueous phase. The residual concentrations of CCI2CHC1 were measured after i d of shaking at 25 *C. The control and abiotic tests were always get up in parallel with biotransformation tests. For determining 1-d transformation capacity, CClzCHC1 decrease in the aqueous pt~ase was related to 1 g of inoculum organic fraction. All the tests were performed in duplicate. Determination of CCI2CHCI concentration. The quantity of CCI2CHCI in the liquid phase was assayed by gas chromatography (Hewlett-Packard 5890) after concentrating 5 mL of sample by the Purge and Trap method with a Tekmar 2000 concentrator. Separation of volatile substances was done in a Quadrex column -- 30 m x 0.53 ram, film 3 ~m, detection by flame-ionizing detector (FID). CCI2CHCI quantification was carried out to an external standard by the calibration curve method using a methanol solution of CC12CHC1 (2 mg/mL) diluted to the necessary concentration. Determination of chloride ions released during biotransformation. Cultivation, enrichment, harvesting and washing of the cells were performed in the same way as in transformation capacity determination, but the cells were resuspended in a mineral medium free from chlorides (composition in g/L): K2HPO4 0.5, (NI-/4)2SO4 0.37, MgSO4"7H20 0.1, NaNO3 0.05, Ca(NO3)2"4H20 0.016; trace elements solution 1 mL. After 1-d degradation at 25 ~ the quantity of C1- ions in samples was determined by spectrophotometry (Iwasaki I952) and corrected for the blank tests. The quantity of CI- ions was related by percentage to the quantity of C1- ions contained in eliminated CC12CHCI. In order to calculate the CCI2CHC1 content in the gaseous phase of the tests, the dimensionless Henry constant for CC12CHC1 at 25 *C, H = 0.4, was used (Folsom et al. x99o).

2002

BIOTRANSFORMATION OF TRICHLOROETHENE BY PURE BACTERIAL CULTURES

469

RESULTS A N D DISCUSSION

Ability of bacterial isolates to degrade CCI2CHCI Various samples of industrial wastewater treatment sludges, and soils and sediments contaminated with phenolic compounds yielded 11 cultures of Gram-negative bacteria utilizing phenol as the only carbon and energy source. It was possible to divide the isolates morphologically into two groups, rods and larger cocci. As was found in preliminary identification, the rods were predominantly Pseudomonas putida representatives, while cocci represented cultures resembling biochemically the genus Acinetobacter. All the cultures were successfully subjected to CC12CHCI biotransformation tests (Table I). The most efficient isolates were capable of removing 88-94 % CClzCHC1 under the given condiTable I. Transformation capacity (TC) of isolates in tions and attained a 1-d transformation capacity of I d (rag transformed CCI2CHCI per g biomass) 2.6-3.1 mg CCIzCHC1 per g biomass. On the opposite, the least active ones degraded 2 0 - 4 5 % Isolate TC Isolate TC CC12CHC1 and their transformation capacity was in the range of 0.3-0.4 mg CC12CHC1 per g biomass, c1 3.11 A109 2.03 Different degradation ability of various phenol utilizRz 2.66 C3 1.21 VT1 2.58 R1 0.77 ing bacteria was also described by other authors. For Bll0 2.23 CA 0.40 instance, Shih et al. 0996) isolated a total of 47 strains L1 2.05 C.5 0.25 of bacteria from 4 reactors with mixed cultures F3 2.O4 growing on phenol but only 13 utilized phenol in a pure culture. Ten strains degraded CCIzCHCI but only three exhibited CCI2CHC1 elimination to a marked extent. The other strains merely exhibited a mild, weak or even no ability to degrade CCIzCHCI. Similar conclusions were published by Fries et al. (1997) who tested 60 straln~ of phenol growing bacteria isolated from trichloroethene-contaminated groundwaters. Among these, 36 were able to remove 80-100 % CCI2CHCI within 15 d (the initial concentration 1 pg/L). The others exhibited a highly varied CClzCHCl-degrading ability (0-60 %). An explanation of the varying degradation ability of bacteria was presented by Futamata et al. (I998) who proved that the variability of pure cultures in their activity to degrade CClzCHCI is explicable by the various amino acid sequences of parts of their phenol hydroxylases responsible for transformation of CC12CHCI. These, as well as our results, thus indicate the need to select cultures for CClzCHCI degradation carefully, and substantiate the disadvantage of applying mixed undefined microbial cultures growing on a primary substrate. Under such conditions a selection of the culture(s) most efficient against a contaminant hardly takes place.

Dimethyl disulfide production by the isolates A peculiarity, noted in all CCI2CHCI transformation tests, was a regular appearance of an unknown substance showing up as a peak in the chromatographic record after biotransformation, behind the peak of residual CCI2CHCI (Fig. 1). The magnitude of the peak varied with culture and did not correspond to the quantity of transformed CC12CHCI. In view of the fact that it could be a metabolite of CCI2CHCI degradation, this substance was identified by GC-MS. This identification revealed an unexpected mass spectrum. At first sight, it was quite different from the mass spectra of chlorinated compounds and comparison with spectrum library showed a high similarity to the mass spectrum of Me2S2 (Fig. 2). In the mass range 29-150 the line of oxygen molecule (mz 32) naturally present in headspace sample represented the only difference between the above spectra, while the other characteristics were identical. Consequently, comparison of the retention times of the substance and standard compound in GC determination proved that the substance was dimethyI disulfide.

470

Vol. 47

J. ROffgJ~KA a aL

ID

v4

,

O l",,

Fig. 1. G C r e c o r d a l t e r CCI2CHCI d e g r a d a t i o n ; t -- time, R -- d e t e c t o r response; 1 - - CCI2CHCI, 2 - - u n k n o w n c o m p o u n d . !

m~

It was not probable that a subIoo I 941 I l~S I stance of this structure could be % a metabolite of CCI2CHC1 biotrans75 formation; nevertheless, tests were '79 done to follow the formation of MezS2 50 32 with selected cultures in the absence of CCIzCHCI. These tests showed that 25 61 production of Me2S2 is fully indepen0 ,..I ,I dent of CCI2CHCI transformation, and ....... I Ill 100 94 the substance, therefore, accumulates in a closed system as a natural volatile 75 45 79 metabolite of the used bacterial cells. Similar conclusions can also be found 50 in other studies; e.g. Tomita et al. (z987) isolated a number of bacterial 25 15 1 61 strains producing Me2S2 from actiI, ff :Ill vated sludge. The formation of diI J' I methyl oligosultides in bacterial degra20 ~0 60 80 100 120 m z dation of organic matter in water sources has also been described (GinzFig. 2. Mass spectrum (mz) of an unknownsubstance (top) and of burg et al. I998, I999). Me2S2 was Me2S2 (bottom); % -- detectorresponse. observed as a dominant volatile metabolite in some species of the Pseudom o n a s genus when grown on organic acid salts (Sch611er et al. z997).

[L

Influence o f dimethyl disulfide on CCI2CHCI biotransformation

Since Me2S2 is a compound displaying a mild antibacterial effect (Kyung and Fleming I997), CCI2CHC1 biotransformation tests were conducted in the presence of Me2S2 at a concentration of 500 rtg/L. We were motivated also by the fact that some studies aimed at CCI2CHCI biotransformation have described the occurrence of toxic metabolites during its course (Shurtliff et al. 1996; Sun and Wood t998). Four cultures were selected for these tests and their CCIzCHC1 transformation capacity was determined in the presence of Me2S2 (Table II).

2002

BIOTRANSFORMATION OF TRICHLOROETHENE BY PURE BACrERIAL CULTURES 471

Table II.

Influence of Me2S2 (500lag/L) on CCI2CHCI transformation capacity (%) by some isolates (means •

Isolate C1 RF2 L1 A109

withoutMe2S2 100 • 100 • 100 • 100 •

0.5 16.0 1.8 4.1

with Me2S2 95A • 92.6 • 63.1 • 30.5 •

0.3 15.8 2.5 2.6

A significant inhibition of CC12CHCI biotransformation by Me2S2 occurred with cultures A109 and L1, while the impact of Me2S2 on cultures C1 and RF2 was slight or insignificant. Hence it can be assumed that production of Me2S2 and possibly other metabolites can significantly affect the degradation ability of mixed bacterial cultures in particular, where various interactions between the microbes can take place and that the question of producing such substances has to be taken into account when choosing a strain for field bioremediation.

Suitable culturefor CCI2CHCI transformation Together with determination of the Me2S2 influence on the transformation capacity of the cultures, a part of this work was devoted to microbiological analysis of a mixed microbial culture with excellent degradation ability (activated sludge adapted for 2 weeks to phenol at a concentration of 300 rag/L). This mixed inoculum was capable of removing 97 % CC12CHC1 in 22 h, at a biomass concentration near the level of pure cultures. The point in question was to find out whether this activity of the mixed culture toward CCI2CHCI was caused by a combined action of several degrading microorganisms or by the presence of a microbe with superior efficiency. Five microbial representatives were isolated from the mixed culture on universal media: entirely dominant Pseudomonas putida (10 CFU/nL), also Corynebacterium sp. (1 CFU/nL), two isolates of yeasts (0.1-1 CFU/nL) and a minor bacterial isolate RF2 (10 CFU/gL). All these microbes were able to utilize phenol, as was demonstrated by growth on phenol agars. CC12CHC1 biotransformation tests were carried out according to the above-mentioned procedure, differing in a lower inoculum concentration of some isolates due to their slower growth on phenol (Table III). Table III. Degradationabilityof the microbesisolatedfroma mixedculture

Isolate

P.putida RF1 RF2 Yeast K'V3 Yeast KV4

Coomebacteriumsp. RF5

Inoculum biomass mg/L 301 68 439 84 166

CCI2CHCI removal % 51.4 100 0 0 0

TCt 1.34 11.1 0 0 0

tTransformationcapacity(1 d, mg CCI2CHCIper g biomass). Neither yeasts (isolates KV3, KV4) nor Corynebacterium sp. RF5 were capable of CC12CHC1 degradation. P. putida RF1, dominant in the consortium, displayed a mild degradation ability in tests -- its transformation capacity reached only 1.34 mg CC12CHC1 per g biomass during 1 d. However, the isolate RF2 was able to degrade all of CCI2CHC1. In a repeated determination of its transformation capacity, which was performed at an even lower inoculum concentration (43 mg biomass per L), it even reached 14.7 mg CCI2CHC1 per g which is a value several times higher than that achieved with pure cultures. Apart from its excellent degradation ability, the RF2 culture was also significant for a very low Me2S2 production and it was even capable of partly removing Me2S2 from an aqueous environment

(data not shown). Degree of CCI2CHCImineralization with the RF2 isolate Measuring chlorides after performed biotransformation is a suitable method for evaluating the biodegradation of chlorinated compounds and a scale of mineralization of the most problematic part of the pollutant molecule. Two independent measurements (each in duplicate) of released CI- ions after CC12CHC1 degradation by the RF2 culture, yielding values of 103 and 107 % chlorine quantity of trans-

472

J. R[~I~--,ICV"KAet al.

Vol. 47

formed CC12CHCI. The high extent of the mineralization of organically bound chlorine by RF2 permits its comparison with the properties of other cultures. The extents of CC12CHCI mineralization after its degradation by several pure cultures was measured by Sun and Wood (I996). Only one strain, Methylosinus trichosporium OB3b, mineralized all degraded CCI2CHC1 (measured mineralization efficiency 109 %), while others showed a lower efficiency, viz~ 77 % for Burkholderia cepacia G4PR1, 62 % for B. cepacia G4, 85 % for Pseudomonas mendocina KR1 and only 51% for P. putida F1. On the other hand, Chang and Alvarez-Cohen (I995) observed that after CC12CHCI degradation by mixed microbial cultures growing on methane, phenol, toluene or propane more than 95 % chlorine of degraded CCI2CHCI was transformed into chlorides. The same results were obtained by Ishida and Nakamura (2ooo) for CCI2CHCI mineralization by Ralstonia sp. KNI-10A, constitutively expressing phenol hydroxylase -- more than 95 % of chlorine in CC12CHCI was released as chloride ions during prolonged degradation. Although the data show the need of greater precision when measuring the released chlorides, they show the necessity of such determinations in every concrete case of biodegradation of the chlorinated pollutants. The RF2 isolate thus appears to be a prospective culture for further study and also for potential application to bioremediation of groundwaters contaminated with CC12CHCI. REFERENCES ARCIERO D., VANNELLIT., I_XX;ANM., HOOPER A.B.: Degradation of trichloroethylene by the ammonia-oxidizing bacterium Nitrosomonas europaea.Biochem.Biophys.Res.Commun. 159, 640-643 (1989). Crtm~o H.L., ALVAREz-COHE'~L.: Transformation capacities of chlorinated organics by mixed cultures enriched on methane, propane, toluene or phenol. BioteclmoLBioeng. 45, 440-449 (1995). DABROOr B., RAVEL J., BERTRAMJ., Go'rrscrtALK G.: Isopropylbenzene (cumene) -- a new substrate for the isolation of trichloroethene-degrading bacteria. Arch.Microbiol. 158, 9-13 (1992). ENSIGNS.A., H'OaA~ M.R., Am' DJ.: Cometabolic degradation of chlorinated alkenes by alkene monooxygenase in a propylene-grown Xanthobacter strain. Appl.Environ.Microbiol. 58, 3038-3046 (1992). FOLSOMB.IL, CHAPMANPJ., PRrrcrlARD P.H.: Phenol and trichloroethylene degradation byPseudomonas cepacia G4: kinetics and interaction between substrates. AppLEnviron.Microbiol. 56, 1279-1285 (1990). FRIESM.IL, FORNEYLJ., TmDJE J.M.: Phenol- and toluene-degrading microbial populations from an aquifer in which successful trichloroethene cometabolism occurred. ApptEnviron.Microbiol. 63, 1523-1530 (1997). FtrrAMATAH., WATANABEIC, I-IARAYAMAS.: Relationships between the triehloroethylene-degrading activities and the amino acid sequences of phenol hydroxylases in phenol-degrading bacteria. Battelle 1st Internat. Conf. on Remediation of Chlorinated and Recalcitrant Compounds, Monterey (CA) 1998. GIN7_.BUROB., CttALIFAI., HADASO., DOR I., LEV O.: Formation of dimethyloligosulfides in lake Kinneret. WaterSci.TechnoL 40, 73-78 (1999). GINZBUROB., ~ F ^ I., ZOHAm T., HAt)AS O., DOR I., LEV O.: Identification of oligosulfide odorous compounds and their source in the sea of Galilee. WaterRes. 32, 1789-1800 (1998). HOPKINS G.D., MuNAr~T^ J., SEMPmNIL., McCARTY P.L.: Trichloroethylene concentration effects on pilot field-scale in situ groundwater bioremediation by phenol-oxidizing microorganisms. Environ.Sci.Technol. 27, 2542-2547 (1993). HOPKINSG.D., MCCARTYP.L.: Field evaluation of in situ aerobic cometabolism of trichloroethylene and three dichloroethylene isomers using phenol and toluene as the primary substrates. Environ.Sci.Technol. 29, 1628-1637 (1995). ISHIDAH., NAKAMURAK.: Trichloroethylene degradation by Ralstonia sp. KNI-10A constitutively expressing phenol hydroxylase: transformation products, NADH limitation, and product toxicity. ZBiosci.Bioeng. 89, 438-445 (2000). IWASAmI., UTSUMIS., OZAWAT.: New colorimetrie determination of chloride using mercuric thiocyanate and ferric ions. Bull. Chem.Soc.Japan 25, 226 (1952). KYUNOK.H., FLEMINGH.P.: Antimicrobial activity of sulphur compounds derived from cabbage. ZFood Prot. 60, 67-71 (1997). SAF.gJH., ASaRA M., FURUrtASHXK., AVERHOFFB., Go'rrSCnALK G.: Degradation of trichloroethylene by a linear-plasmid-encoded alkene monooxygenase in Rhodococcus corallinus (Nocardia corallina) B-276. Microbiology 145, 1721-1730 (1999). SCHOLt.ER C., Moran S., WILKINS K.: Volatile metabolites from some gramnegative bacteria. Chemospl~,re 35, 1487-1495 (1997). SmH C., DAVL~'M.E., ZHOUJ., "I~EDJEJ.M., CmDDLEC.S.: Effects of phenol feeding pattern on microbial community structure and cometabolism of trichloroethylene.Appl.Environ.Microbiol. 62, 2953-2960 (1996). SHURTHFF M.M., PAR~N G.F., WEATHm~SLJ., GlasoN D.T.: Biotransformation of trichloroethylene by a phenol-induced mixed culture. ZEnviron.Eng. 122, 581-589 (1996). STEFFANR.J., SPERRYK.L., WALSHM.T., VAINBEROS., CONDEEC.W.: Field-scale evaluation of in situ bioaugmentation for remediation of chlorinated solvents in groundwater. Environ.Sci.Teclmol. 33, 2771-2781 (1999). SUN A.K., HONOJ., WOOD T.K.: Modeling trichloroethylene degradation by a recombinant pseudomonad expressing toluene ortho-monooxygenase in a fixed-film bioreactor. BioteclmoLBioeng. 59, 40-51 (1998). SuN A.K., WOOD T.IC: Trichloroethylene degradation and mineralization by pseudomonads and Methylosinus trichosporium OB3b. AppLMicrobiol.Biotechnol. 45, 248-256 (1996). TAKAMIW., HORINOUCrllM., NoJIm H., YAXlANEH., OMOm T.: Evaluation of trichloroethylene degradation by E. coli transformed with dimethylsulphide monooxygenase genes and/or cumene dioxygenase genes. BiotechnoLLett. 21, 259264 (1999). TOMITAB., INOUEH., CHAYAK., NAKAMURAA., HAMAMURAN., UENO K., WATANABEIC, Os~- Y.: Identification of dimethyl disulfide-forming bacteria isolated from activated sludge. AppLEnviron.MicrobioL 53, 1541-1547 (1987). VOGEL T.M., CRIDDLEC.S., McCARTY P.L.: Transformations of halogenated aliphatic compounds. Environ.Sci.Technol. 21, 722-736 (1987). WILSONJ.T., WILSONB.H.: Biotransformation of trichloroethylene in soil. Appl.Environ.Microbiol. 49, 242-243 (1985).