INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Although many efforts have been made to reduce the emission of polychlorinated dibenzofurans and dioxins during combustion of dust or bleaching of pulp in paper mills, these compounds remain widespread environmental contaminants. This problem has given rise to an intensive search for microorganisms capable of metabolizing cyclic biaryl ether structures in order to evaluate their degradation potential in nature. Recently, some bacterial strains able to cometabolize (4, 13) or to grow with dibenzofuran (7, 16, 20) have been isolated and the mechanisms of oxygenation and ring cleavage have been described in detail (23). Degradation of halogenated dibenzofurans and dibenzo-p-dioxins by aerobic bacteria has so far been shown only for derivatives with up to two chlorine substituents (24). The mineralization potential for higher chlorinated compounds might be rather low. The biodegradation capacity of aerobic microorganisms could be important in the destruction of nonhalogenated or slightly halogenated intermediates arising from anaerobic dehalogenation of dibenzofurans, e.g., in sediments (1) or soils.

Although fungi constitute the majority of microbial biomass in soil, data on fungal catabolism of biaryl ether compounds are limited and restricted to hydroxylation processes (4, 11). Our objectives were (i) to identify fungal strains that could degrade dibenzofuran and (ii) to determine the pathway and the chemical intermediates through which the breakdown occurs.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Isolation procedure. We isolated yeast strains from a dioxin-contaminated soil sample from a copper works under selective conditions in a mineral salts medium (MM) (14) containing, per liter, 5 g of NH₄H₂PO₄, 2.5 g of KH₂PO₄, 1 g of MgSO₄ · 7H₂O, 2 mg of FeCl₃ · 6H₂O, 20 mg of Ca(NO₃)₂ · 4H₂O, and 10 ml of a trace element solution (containing, per liter, 50 mg of H₃BO₃, 40 mg of MgSO₄ · 4H₂O, 40 mg of ZnSO₄ · 7H₂O, 20 mg of Na₂MoO₄, 10 mg of CuSO₄ · 5H₂O, 10 mg of CoCl₂, 10 mg of potassium iodide). Soil (2 g) was suspended in 100 ml of MM and supplemented with dibenzofuran (1 g/liter) as the sole source of carbon and chloramphenicol (30 µg/liter) to inhibit bacterial growth. After 7 days of shaking cultures at 30°C and 180 rpm, 10 ml was transferred to 90 ml of fresh medium and incubated under the same conditions. Identical transfers were performed 7 and 14 days later. Afterwards, microorganisms were obtained by plating 0.1 ml of the cultures on malt agar plates. Pure cultures were maintained on malt agar and stored at 4°C.

Identification and characterization of strain SBUG 801. The yeast used was assigned to the genus Trichosporon by the urea hydrolysis test, sugar utilization (IC32; Biomerieux, Nürtingen, Germany), and morphology. Further identification to the species level using physiological characteristics, coenzyme Q type, and the D2 domain of 26S rRNA sequences was performed by the method of Gueho et al. (9, 10). Strain SBUG 801 has been deposited as strain DSM 12017 in the Deutsche Sammlung für Mikroorganismen, Braunschweig, Germany.

Growth and incubation conditions. Cells were grown in 500-ml shake flasks with 100 ml of MM supplemented with 1 or 2% glucose and 1 ml of a vitamin solution (21). After incubation for 16 h at 30°C and 180 rpm on a rotary shaker, cells were harvested by centrifugation (3,100 × g, 5 min) and washed twice with sterile MM. The cell pellet was resuspended in MM to an optical density (600 nm) of 6. Dibenzofuran and 2- or 3-hydroxydibenzofuran were added in suitable concentrations of about 10 to 500 µg/ml (see Results) to the cell suspensions, and the cultures were incubated as described above. Cells in MM without substrate and dibenzofuran derivatives in MM without cells were used as controls.

Kinetics of dibenzofuran degradation and isolation of metabolites. Cultures were incubated in 100-ml Erlenmeyer flasks containing 20 ml of the cell suspension in MM and 160 µg of dibenzofuran per ml at 30°C with agitation at 180 rpm. Flasks with autoclaved cells and dibenzofuran were used as additional controls. At each sampling period, cultures were removed and stored at 20°C. At the end of the experiment all samples and controls were thawed and extracted with 4 ml of trichloromethane for 5 min. The organic layer was separated from the aqueous phase by centrifugation in glass tubes (3,100 × g, 5 min). The contents of dibenzofuran and metabolites were estimated by gas-chromatographic analysis of 1 µl of the trichloromethane phase and HPLC analysis of 100 µl of the aqueous supernatant. The data are reported as means for two separate experiments with replicated batch cultures. Standard deviation was no more than 8%.

Chemical analysis and identification of metabolites. Gas chromatography was carried out on a Shimadzu model GC-14A gas chromatograph with a flame ionization detector at 310°C equipped with a Permabond SE-54-DF-0.35 capillary column (50 m by 0.25 mm [inside diameter]; Macherey-Nagel, Düren, Germany). Helium was used as the carrier gas (180 kPa). The column temperature was programmed from 150 to 300°C at 10°C/min.

Chemicals. Dibenzofuran and 2-hydroxydibenzofuran were purchased from Aldrich-Chemie. 3-Hydroxydibenzofuran was synthesized from 3-aminodibenzofuran by the method of Wilkes (22). All chemicals and solvents were of the highest purity available.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

From enrichment cultures with dibenzofuran various yeast strains were isolated, but none could grow with the biaryl compound as the sole source of carbon. Screening of the yeasts obtained for their potential to metabolize dibenzofuran following growth on glucose identified a strain that produced a broad pattern of degradation products when the culture supernatant was analyzed by HPLC.

Taxonomic characterization of the dibenzofuran-metabolizing yeast. Microscopic observation showed formation of arthroconidia. This finding together with the results of the carbon utilization spectrum (IC32) and the hydrolysis of urea led us to classify the organism as a yeast belonging to the genus Trichosporon. This genus can be separated into two groups, one consisting of 15 species containing coenzyme Q-9 ubiquinone systems, and the other consisting of four species containing coenzyme Q-10 ubiquinone systems (10). Since this isolate had coenzyme Q-10, subsequent studies were confined to Trichosporon mucoides, T. moniliforme, T. cutaneum, and T. jirovecii. The physiological features of the isolate pointed to T. mucoides: assimilation of specific carbon sources (galactitol, glucono--lactone, L-arabinitol, melibiose, and erythritol), no growth with nitrate or creatine, and growth at 37°C. However, all four species have very similar carbon and nitrogen compound assimilation patterns (2, 10). For final identification, we sequenced the taxonomically most informative 26S rRNA region from strain SBUG 801 and compared the data obtained with data for the type strains of the four species obtained by Gueho et al. (10). The sequence of strain SBUG 801 was the same as that of the CBS 7625 type culture of T. mucoides.

Metabolism of dibenzofuran. Glucose-grown cells of T. mucoides SBUG 801 metabolized dibenzofuran very rapidly. More than 50% of the substrate added was degraded within 8 h (Fig. 1) at a degradation rate of about 15 µg of dibenzofuran ml¹ h¹. After that time the degradation of the biaryl compound dropped drastically. The kinetics seem to be time dependent, because a similar decrease in the degradation rate occurred after 8 to 10 h when the initial dibenzofuran concentrations were 100 and 200 µg/ml. Nonsignificant change in the dibenzofuran concentration was observed in heat-killed cultures.

View larger version (13K): [in this window] [in a new window]   FIG. 1.   Degradation of dibenzofuran (160 µg/ml) by T. mucoides SBUG 801.

Detection and identification of intermediates. Dibenzofuran degradation was accompanied by the formation of at least six water-soluble intermediates. The major metabolites (designated I to VI) had retention times of 10 to 12 min (I to IV), 9.3 min (V), and 7.6 min (VI). The HPLC peak at 13.8 min was dibenzofuran (Fig. 2). Metabolites I to V accumulated over a period of about 10 h and seemed to be metabolized during subsequent incubation. The concentration of metabolite VI increased continuously (Fig. 3). Metabolites were characterized by their UV-visible spectra during detection with a diode array detector. None of these substances accumulated in the control flasks.

View larger version (20K): [in this window] [in a new window]   FIG. 2.   HPLC elution profile of an aqueous culture supernatant after incubation of T. mucoides SBUG 801 with dibenzofuran (500 µg/ml) for 12 h. At least six dibenzofuran (DBF) degradation products (I to VI) were detectable after separation on a reversed-phase column. AU, absorbance units.

View larger version (22K): [in this window] [in a new window]   FIG. 3.   Product accumulation and decomposition during incubation of glucose-grown cells of T. mucoides SBUG 801 with dibenzofuran (160 µg/ml). Amounts of intermediates were calculated from their absorbance at 220 nm.

Monohydroxylated compounds. Compounds I to IV were identified as the four isomers of monohydroxylated dibenzofurans by comparing UV and mass spectra with those of available standards (2- and 3-hydroxydibenzofuran), with data from the literature (1-hydroxydibenzofuran [4]), or with our own data (4-hydroxydibenzofuran [11]). 2-Hydroxydibenzofuran was always the most common isomer (maximum concentration, approximately 20 µg/ml). In contrast, 3- and 4-hydroxydibenzofuran were detected at about 2 to 5 µg/ml, and 1-hydroxydibenzofuran was detected in much smaller quantities (no more than 0.5 µg/ml).

Metabolites V and VI. Metabolites V and VI were rapidly released into the medium. Compound V reached its maximum concentration (6 µg/ml) at about 8 h, while compound VI accumulated throughout the entire incubation up to 35 µg/ml (Fig. 3). Incubation experiments with glucose-grown cells of T. mucoides SBUG 801 with 125 µg of 2-hydroxydibenzofuran or 3-hydroxydibenzofuran per ml as the substrate also resulted in the accumulation of metabolites V and VI in the culture fluid. Thus, 2- and 3-hydroxydibenzofuran were formed as intermediates during dibenzofuran degradation.

View larger version (15K): [in this window] [in a new window]   FIG. 4.   1H NMR spectrum of metabolite V, 2,3-dihydroxydibenzofuran (300 Mhz; CD3OD). Trimethylsilane was the internal standard.

View larger version (25K): [in this window] [in a new window]   FIG. 5.   Mass spectrum of the methyl derivative (A) and 1H NMR spectrum of metabolite VI, 2-(1-carboxy methylidene)-2,3-dihydrobenzo[b]furanylidene glycolic acid, a ring cleavage product formed during incubation of T. mucoides SBUG 801 with dibenzofuran (B).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

We investigated the ability of a yeast to biodegrade dibenzofuran, a model compound for intermediates arising from highly persistent environmental pollution with biaryl ether structures. These structures may be nonhalogenated and slightly halogenated compounds which result from anaerobic or chemical dehalogenation processes. Screening of soil samples for suitable organisms revealed a strain identified as T. mucoides which had been isolated only from human tissue previously. This strain was able to grow with aromatic compounds like phenol or benzoic acid and transformed dibenzofuran as well. The degradation of monocyclic aromatic compounds is common in yeasts (12, 15), and phenol degradation by T. cutaneum has been investigated in detail (8, 17). However, information on the ability of yeasts to cleave the aromatic ring in biaryl or polycyclic compounds is still lacking.

We suggest that T. mucoides SBUG 801 degrades dibenzofuran via 2,3-hydroxydibenzofuran (metabolite V) since ring cleavage of this intermediate leads to the formation of 2-(1-carboxy methylidene)-2,3-dihydrobenzo[b]-furanylidene glycolic acid (metabolite VI [Fig. 6]). The accumulation of high levels of monohydroxylated dibenzofurans and the formation of the dihydroxylated derivative from both 2-hydroxydibenzofuran and 3-hydroxydibenzofuran indicated the action of monooxygenases during the first degradation steps. Hydroxylation of biaryl compounds by monooxygenases in fungi has been frequently described (3, 4, 6, 11, 19); however oxidation of the resulting hydroxylated derivatives has not been well documented. The only previous report of this activity in yeasts showed that diphenyl ether could be transformed via hydroxylated derivatives to a lactone by Trichosporon sp. strain SBUG 752 (18).

View larger version (22K): [in this window] [in a new window]   FIG. 6.   Proposed pathway for the first steps of dibenzofuran degradation by T. mucoides SBUG 801.

Monohydroxylated and dihydroxylated dibenzofuran derivatives were metabolized very rapidly by cells of T. mucoides SBUG 801, leading to the formation of a derivative of hydroxymuconic acid with a benzofuran substructure. Cleavage of the dihydroxylated ring presumably utilizes a mechanism similar to that described for the degradation of diphenyl ether by Trichosporon sp. strain SBUG 752 (18). In both species a third hydroxyl group must be introduced into dihydroxylated intermediates before ortho cleavage of the aromatic structure can take place. In contrast to T. mucoides (coenzyme Q-10), strain SBUG 752 belongs to the group of coenzyme Q-9-containing Trichosporon species.

This is the first report of a fungus capable of cleaving the aromatic structure of dibenzofuran. The results together with those for strain SBUG 752 show that Trichosporon strains not only degrade monoaromatic compounds but also can degrade biaryl ether substrates. Further studies, particularly with chlorinated dibenzofurans, will be required to determine if these yeasts can effectively degrade halogenated environmental pollutants.