HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by Rapp, P. Articles by Gabriel-Jürgens, L. H. E. Search for Related Content PubMed PubMed Citation Articles by Rapp, P. Articles by Gabriel-Jürgens, L. H. E. Agricola Articles by Rapp, P. Articles by Gabriel-Jürgens, L. H. E.

Degradation of alkanes and highly chlorinated benzenes, and production of biosurfactants, by a psychrophilic Rhodococcus sp. and genetic characterization of its chlorobenzene dioxygenase

GBF-National Research Centre for Biotechnology, Division of Microbiology, Mascheroderweg 1, D-38124 Braunschweig, Germany

Correspondence
Peter Rapp
pra{at}gbf.de

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Rhodococcus sp. strain MS11 was isolated from a mixed culture. It displays a diverse range of metabolic capabilities. During growth on 1,2,4-trichlorobenzene, 1,2,4,5-tetrachlorobenzene (1,2,4,5-TeCB) and 3-chlorobenzoate stoichiometric amounts of chloride were released. It also utilized all three isomeric dichlorobenzenes and 1,2,3-trichlorobenzene as the sole carbon and energy source. Furthermore, the bacterium grew well on a great number of n-alkanes ranging from n-heptane to n-triacontane and on the branched alkane 2,6,10,14-tetramethylpentadecane (pristane) and slowly on n-hexane and n-pentatriacontane. It was able to grow at temperatures from 5 to 30 °C, with optimal growth at 20 °C, and could tolerate 6 % NaCl in mineral salts medium. Genes encoding the initial chlorobenzene dioxygenase were detected by using a primer pair that was designed against the -subunit (TecA1) of the chlorobenzene dioxygenase of Ralstonia (formerly Burkholderia) sp. strain PS12. The amino acid sequence of the amplified part of the -subunit of the chlorobenzene dioxygenase of Rhodococcus sp. strain MS11 showed >99 % identity to the -subunit of the chlorobenzene dioxygenase from Ralstonia sp. strain PS12 and the parts of both -subunits responsible for substrate specificity were identical. The subsequent enzymes dihydrodiol dehydrogenase and chlorocatechol 1,2-dioxygenase were induced in cells grown on 1,2,4,5-TeCB. During cultivation on medium-chain-length n-alkanes ranging from n-decane to n-heptadecane, including 1-hexadecene, and on the branched alkane pristane, strain MS11 produced biosurfactants lowering the surface tension of the cultures from 72 to 29 mN m^-1. Glycolipids were extracted from the supernatant of a culture grown on n-hexadecane and characterized by ¹H- and ¹³C-NMR-spectroscopy and mass spectrometry. The two major components consisted of ,-trehalose esterified at C-2 or C-4 with a succinic acid and at C-2' with a decanoic acid. They differed from one another in that one 2,3,4,2'-trehalosetetraester, found in higher concentration, was esterified at C-2, C-3 or C-4 with one octanoic and one decanoic acid and the other one, of lower concentration, with two octanoic acids. The results demonstrate that Rhodococcus sp. strain MS11 may be well suited for bioremediation of soils and sediments contaminated for a long time with di-, tri- and tetrachlorobenzenes as well as alkanes.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Numerous Gram-negative bacteria are known to utilize halogenated benzenes as sole carbon source (de Bont et al., 1986; Schraa et al., 1986; Spain & Nishino, 1987; van der Meer et al., 1987; Haigler et al., 1988; Oltmanns et al., 1988; Oldenhuis et al., 1989; Sander et al., 1991; Brunsbach & Reineke, 1994; Spiess et al., 1995; Potrawfke et al., 1998). However, only very few Gram-positive bacteria, mainly rhodococci, have been described as having this capability (Goulding et al., 1988; Stoecker et al., 1994; Zaitsev et al., 1995). Among these few Gram-positive bacteria utilizing chlorinated benzenes, none have hitherto been found able to metabolize tri- and tetrachlorobenzenes. Most of the key enzymes of the Gram-negative bacteria involved in the degradation pathways, as well as the coding genes and regulatory elements, have been described (van der Meer, 1997; Sander et al., 1991; van der Meer et al., 1991a, b, c; Beil et al., 1997, 1998, 1999). It was shown that degradation of chlorobenzenes in Pseudomonas sp. strain P51 and Ralstonia sp. strain PS12 is initiated by the non-haem-iron-containing TcbA and TecA dioxygenases (Werlen et al., 1996; Beil et al., 1997). These class IIB chlorobenzene dioxygenases are multicomponent enzyme systems consisting of a reductase and ferredoxin and a catalytic terminal dioxygenase composed of a large - and a small -subunit, which are possibly in an ₃₃ configuration (Kauppi et al., 1998). The -subunit can be divided into two distinct domains: a Rieske domain which contains the [2Fe–2S] centre, and the catalytic domain containing the mononuclear Fe(II) centre. Beil et al. (1998) showed that the -subunit of the chlorobenzene dioxygenase from Ralstonia sp. strain PS12 is responsible for its substrate specificity. In contrast to these data for the enzymes of Gram-negative bacteria involved in the degradation of chlorinated benzenes, no corresponding data have been reported up to now from Gram-positive bacteria.

A well-known property of Gram-positive bacteria, however, especially the rhodococci, is the capability to degrade alkanes (Rehm & Reiff, 1981; Finnerty, 1992). This property is often accompanied, especially among rhodococci, by the ability to produce biosurfactants (Hommel, 1990; Finnerty, 1992). Among these biosurfactants, glycolipids are the most important group (Hisatsuka et al., 1971; Rapp et al., 1979; Inoue & Ito, 1982; Ristau & Wagner, 1983; Passeri et al., 1992). They show a high surface or interfacial activity as well as pH and temperature stability, low toxicity and good biodegradability.

Chlorinated benzenes are poorly soluble in water, and therefore less available to micro-organisms. Particularly in old polluted sites, a considerable fraction of the contaminants appears to be unavailable for biodegradation. This decrease of bioavailability of pollutants in the course of time is often designated as ‘ageing’. It may result from slow diffusion into very small pores, where micro-organisms are absent, and/or from absorption into organic matter of soils and sediments (Bosma et al., 1997). This decrease of bioavailability of contaminants in the course of time by ageing may be overcome to some extent by applying biosurfactants. These have the ability to desorb and disperse or dissolve such very hydrophobic compounds, i.e. alkanes or chlorinated benzenes. However, the bioremediation of contaminated soils and sediments by the addition of biosurfactants is restricted by their high production costs. Current efforts are therefore directed towards the isolation or design of micro-organisms that exhibit the desired degradative capabilities together with the simultaneous production of suitable biosurfactants. Gallardo et al. (1997) described the construction of such a recombinant micro-organism. They combined the production of biosurfactants with enhanced biodesulfurization of dibenzothiophene in a recombinant Pseudomonas strain. To our knowledge, however, no naturally occurring micro-organism has been isolated which combines the two properties, i.e. the degradation of hydrophobic chlorinated aromatic compounds and the production of suitable biosurfactants.

In the present study, the isolation of a psychrophilic, slightly halotolerant Rhodococcus sp. from a mixed culture is reported. The mineralization of tri- and tetrachlorobenzenes by this bacterium is described, as well as its ability to degrade alkanes of variable chain length coupled with the production of biosurfactants. Finally, its chlorobenzene dioxygenase is genetically compared with related benzene, chlorobenzene, toluene and biphenyl dioxygenases.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Isolation and growth of bacteria.
Strain MS11 was isolated as a contaminant from a liquid batch culture of Burkholderia sp. strain PS14 (Rapp & Timmis, 1999). Pure cultures of strain MS11 were obtained by serial dilutions on nutrient agar (Difco) and retransfer of single colonies onto plates consisting of mineral salts medium (Rapp & Timmis, 1999) and 12 g l^-1 of agar no. 1 (Oxoid) with 1,2,4-trichlorobenzene (1,2,4-TCB) provided from the headspace. Liquid cultures were grown in 500 ml Erlenmeyer flasks containing 100 ml mineral salts medium on a rotary shaker (120 r.p.m.) at 20 °C. 1,2,4,5-tetrachlorobenzene (1,2,4,5-TeCB) and other solid substrates were added as finely mortar-ground powder after sterilization to a concentration of 5 mM. 1,2- and 1,3-dichlorobenzene (DCB), 1,2,4-TCB and short-chain n-alkanes ranging from n-hexane to n-nonane as well as other volatile toxic substrates were added via the vapour phase also to a concentration of 5 mM (from a tube with two small openings inside the shake flask, which was mounted in the seal of its screw cap). Less volatile medium- and long-chain alkanes were added directly to the mineral salts medium to a concentration of 1 % (w/v). For growth on agar plates, solid substrates were placed in the lids of inverted Petri dishes and liquid volatile substrates were provided via the vapour phase.

Characterization of strain MS11.
The pure culture was characterized by 16S rRNA sequence analysis (Woese et al., 1983; Gutell et al., 1985). It was classified additionally on the basis of Bergey's Manual of Systematic Bacteriology (Goodfellow, 1989). Tests were performed as described in the manual or by Smibert & Krieg (1981).

Preparation of resting cells and cell extracts.
Cells grown for 48 h on 1,2,4-TCB, 1,2,4,5-TeCB or glucose were centrifuged at 12 000 g for 20 min at 4 °C and the pellets were washed twice with 33 mM Tris/HCl buffer, pH 8·0, and resuspended in 1 ml of the same buffer. Crude cell extracts were prepared by passing cell suspensions twice through a chilled French pressure cell (Aminco) at 90 MPa. After centrifugation at 25 000 g for 30 min at 4 °C, the clear supernatant was used as cell extract. Soluble protein concentrations were determined by the method of Bradford (1976) with bovine serum albumin as the standard.

Determination of cell growth with alkanes as carbon sources.
A 10 ml volume of culture, grown on alkanes, was mixed with 10 ml ethanol/butan-1-ol/chloroform (10 : 10 : 1, by vol.) and centrifuged at 10 000 g for 30 min; the pellet was washed with 10 ml water and dried at 60 °C to a constant weight.

Oxygen uptake determinations.
Assays for determination of the oxygen uptake rate were carried out with resting cells suspended in 54 mM phosphate buffer, pH 7·4, to an OD₅₇₈ of 1·0 in a volume of 1 ml. The oxygen uptake rate was measured polarographically as described previously (Potrawfke et al., 1998). Substrates were dissolved in dimethyl sulfoxide and their final concentration in the assay mixture was 0·1 mM. Protein concentrations of cell suspensions were determined by the method of Schmidt et al. (1963).

Chloride release.
Chloride concentrations were determined by the mercury(II) thiocyanate method (Bergmann & Sanik, 1957; Chauhan et al., 1998) and protein concentrations of cell suspensions by the method of Schmidt et al. (1963). Specific rates of chloride release were determined by using washed cells suspended in 54 mM phosphate buffer, pH 7·4. Chloride concentration after degradation of 1,2,4,5-TeCB, 1,2,4-TCB and 3-chlorobenzoate of increasing concentration was determined in cultures with mineral salts medium, in which chloride salts had been replaced by the corresponding sulfate salts. Standards were prepared by dissolving appropriate amounts of sodium chloride in this chloride-free mineral salts medium.

Nucleic acid extraction.
Cells grown to an OD₅₇₈ of about 1 were centrifuged and the pellet was suspended in phospate buffer supplied with the components of the FastDNA Spin kit for soil (Bio 101 Systems). The subsequent extraction was carried out by the manufacturer's protocol.

Oligonucleotides.
The designation, sequence (5'–3') and priming direction of the oligonucleotide primers used for amplification of DNA fragments by PCR (Saiki et al., 1988) are as follows: prLG-Tec-FI (forward), CGTGAGGAGACAATGAATCACA; prLG-Tec-RI (reverse), CCAGTCAGACGAGGTCATCATC. These primers were designed against TecA1, the -subunit of the chlorobenzene dioxygenase from Ralstonia sp. strain PS12 [pr-LG-Tec-FI, against nucleotides of positions 1347–1368; pr-LG-Tec-RI, against nucleotides of positions 2662–2684; nucleic acid numbering according to the database numbering of the chlorobenzene dioxygenase from Ralstonia sp. strain PS12 (accession no. U78099)] (Beil et al., 1997). The fragment obtained by PCR amplification comprised approximately 1330 bp. The 20 µl reaction mixture contained the following ingredients: 1 µl template DNA, 0·4 µl 1·25 mM dNTPs, 2 µl reaction buffer, 10 pM each primer, forward and reverse, 0·3 µl Taq polymerase (4 U µl^-1) (Promega), and sterile water up to 20 µl. The reaction mixture was heated to 94 °C for 2 min. Thirty-five cycles were carried out in a Perkin Elmer thermal cycler at 94 °C for 30 s, at 55 °C for 30 s, and at 72 °C for 1 min, followed by an elongation at 72 °C for 10 min.

Cloning of the genes encoding the chlorobenzene dioxygenase.
Prior to cloning, PCR products were purified using QIAquick PCR purification columns (Qiagen). Ligation into the pGEM-T vector (Promega) was carried out as recommended by the manufacturer. Ligation products were transformed into Escherichia coli JM109 competent cells (Promega) and plated onto Luria–Bertani (LB) agar supplemented with 1 mM IPTG, 40 mg X-Gal l^-1 and ampicillin according to the manufacturer's (Promega) instructions. Transformants containing an insert were identified by blue/white screening and transferred onto a new LB plate, prepared as described above, with a numbered grid for subsequent identification of the clones. White colonies were checked for inserts using primers as described above. Positive amplicons were purified using QIAquick PCR purification columns (Qiagen).

Sequence analysis.
Sequences were assembled using Sequencher version 4.0.5. They were submitted unaligned to the European Molecular Biology Laboratory (EMBL: http://dove.embl-heidelberg.de/Blast2) to determine the most similar sequence.

Enzyme assays.
Dihydrodiol dehydrogenase activity was determined by monitoring NAD⁺ reduction at 340 nm (Reineke & Knackmus, 1984). Catechol 1,2-dioxygenase activity was quantified by measuring the formation of cis,cis-muconic acids at 260 nm (Dorn & Knackmuss, 1978a). Interfering catechol 2,3-dioxgenase activity was eliminated by preincubation of the assay mixture with 0·01 % (v/v) H₂O₂ for 10 min as described by Nakazawa & Yokota (1973). Catechol 2,3-dioxygenase (EC 1.13.11.2) activity was determined by monitoring the formation of 2-hydroxymuconic semialdehyde at 375 nm (Nozaki, 1970). Absorption coefficients for chloromuconates were reported by Dorn & Knackmuss (1978b) and Sander et al. (1991).

Determination of 1,2,4,5-TeCB concentration.
Two parallel 100 ml shake flask cultures with 1,2,4,5-TeCB as carbon source were extracted with 100 ml n-hexane. The extracts were dried over anhydrous sodium sulfate, and concentrated by evaporation to 1 ml at 1·4x10⁴ Pa at 30 °C. 1,2,4,5-TeCB concentration was analysed by GC as described previously (Rapp & Timmis, 1999). Peaks were identified and quantified by comparing injections with authentic external standards, prepared by dissolving defined amounts of 1,2,4,5-TeCB in mineral salts medium. These aqueous solutions were extracted and the extracts concentrated in the same way as the samples to be analysed.

Analysis of fatty acids in the culture broth.
A 100 ml shake flask culture grown on n-dodecane was acidified with conc. HCl to a pH of approximately 2·0 and extracted with 50 ml ethyl acetate. The extract was dried over anhydrous sodium sulfate and evaporated to dryness under reduced pressure. The residue was dissolved in 5 ml CH₂Cl₂/CH₃OH/conc. HCl (10 : 1 : 1, by vol.) and heated for 90 min at 100 °C. After cooling, the fatty acid methyl esters were recovered in CH₂Cl₂, dried under nitrogen and dissolved in 1 ml n-octane for analysis by GC and GC/MS. Capillary GC was performed on an HP5890 Series II gas chromatograph with a fused-silica capillary column (50 mx0·11 µm) with cross-linked 5 % phenylmethyl silicone (film thickness 0·11 µm). The instrument was equipped with a flame ionization detector and H₂ was used as carrier gas. The operating temperatures of the injector and detector were 250 and 300 °C, respectively. The oven temperature programme was as follows: 100 °C for 2 min, increase to 290 °C at a rate of 4 °C min^-1, with an isothermal period of 14 min at the end. The GC/MS analyses were carried out on a similar chromatograph as described above (He was the carrier gas), connected to an HP 5989 A quadrupole mass spectrometer. The electron-impact ion source was maintained at 200 °C, while the quadrupole temperature was 100 °C. The electron energy was set at 70 eV.

Determination of surface tension.
The surface tension of cultures was determined with a Du Noüy-ring tensiometer (K6, Krüss, Hamburg, Germany) at room temperature (Weser, 1980).

Thin-layer chromatography (TLC).
This was performed on silica gel 60 F₂₅₄ nano-plates (0·2 mm) (Macheray-Nagel). Trehalose lipids were chromatographed with the solvent system CHCl₃/CH₃OH/H₂O (65 : 15 : 2, by vol.) and mono-and disaccharides with the solvent system propan-2-ol/H₂O (85 : 15, v/v). Sugars and trehalose lipids were detected by spraying with anisaldehyde/sulfuric acid reagent (Krebs et al., 1967).

Isolation and partial characterization of trehaloselipids.
One litre of mineral salts medium with 1 % (w/v) n-hexadecane as sole carbon source in a 3 l shake flask was inoculated with strain MS11, grown on 1,2,4-TCB. The culture was grown at room temperature at 120 r.p.m. for 168 h. After centrifugation at 12 000 g at 4 °C, the supernatant was extracted with CH₂Cl₂/CH₃OH (2 : 1, v/v). The extract was concentrated under reduced pressure and applied to a silica gel 60 column (70–230 mesh, ASTM, Merck). Residual n-hexadecane and other compounds less polar than the trehalose lipids were eluted with CHCl₃/CH₃OH/HCl (5 : 1 : 0·01, by vol.). Trehalose lipids were eluted with CHCl₃/CH₃OH (5 : 1·5, v/v) followed by CHCl₃/CH₃OH (5 : 2, v/v). Elution was followed by TLC with the solvent system CHCl₃/CH₃OH/H₂O (65 : 15 : 2, by vol.). Fractions containing trehalose lipids were further purified by thick-layer chromatography on silica gel 60 F₂₅₄ (2 mm) using the solvent system CHCl₃/CH₃OH/H₂O (65 : 15 : 2, by vol.). Trehalose lipids were obtained as a white and wax-like substance and their examination by TLC on silica gel showed a single spot with an R_F value of 0·41 in the solvent system CHCl₃/CH₃OH/H₂O (65 : 15 : 2, by vol.). This R_F value and the green colour after spraying with anisaldehyde/sulfuric acid reagent (Krebs et al., 1967) suggested that the biosurfactants were glycolipids. They were saponified and the water-soluble fraction was analysed for reducing sugars by the 3,5-dinitrosalicylic acid method as described by Rapp et al. (1979). Acidic hydrolysis of the non-reducing carbohydrate moiety (Rapp et al., 1979) delivered D-glucose as demonstrated by TLC and enzymic analysis (Boehringer Mannheim Biochemica, 1994), suggesting trehalose to be the sugar moiety of the biosurfactants.

NMR spectroscopy.
NMR 1D (¹H and ¹³C), 2D ¹H COSY (correlated spectroscopy) and 2D ¹H, ¹³C COSY spectra were recorded at 300 K on a Bruker ARX 400 NMR spectrometer locked to the major deuterium resonance of CD₃OD in the mixed CDCl₃/CD₃OH (7 : 3, v/v) solvent. All chemical shifts are given in p.p.m. relative to trimethylsilane and coupling constants in Hz.

¹H NMR.
Trehalose system =5·44 [H-1, J(1–2) 3·6]; =4·9 [H-2, J(2–3) 10·2]; =5·54 [H-3, J(3–4) 9·8]; =5·08 [H-4, J(4–5) 9·8]; =3·75 [H-5, J(5–6A) 2·2]; =3·58 [H-6A, J(5–6B) 4·8]; =3·5 [H-6B, J(6A–6B) 12·4]; =5·28 [H-1'J(1'–2') 3·6], =4·73 [H-2', J(2'–3') 9·9]; =4·0 [H-3', J(3'–4') 9·5]; =3·36 [H-4', J(4'–5') 9·2]; =3·68 [H-5']; =3·83–3·69 [H-6'A]; =3·83–3·69 [H-6'B]; fatty acids =0·89 [terminal CH₃'s (R_2', R₃, R₄)]; =1·28 [(CH₂)_n (R_2', R₃, R₄)]; =2·7–2·45 [CH₂() (R₂ or R₄)]; =1·57 [CH₂() (R₃, R₂ or R₄)], =1·65 [CH₂() (R_2')]; =2·7–2·45 [CH₂() (R₂ or R₄)]; =2·27 [CH₂() (R₃, R₂ or R₄); =2·5–2·34 [CH₂() (R_2')].

¹³C NMR.
Trehalose system =90·7 (C-1); =91·2 (C-1'); =71·3* (C-2); =73·0^# (C-2'); =70·2 (C-3); =71·2* (C-3'); =68·9 (C-4); =71·4* (C-4'); =70·7 (C-5); =73·1^# (C-5'); =60·8 (C-6); =61·8 (C-6') (* ^# these assignments are interchangeable); fatty acids =174·3 [CO (R_2')]; =173·5 [CO (R₃)]; =173·2 [CO (R₂ or R₄)]; =34·4 [C() (R_2', R₃, R₂ or R₄)]; =32·2/32·0 [CH₂.CH₂.CH₃ (R_2', R₃, R₂ or R₄)]; =30·0 [C() (R₂ or R₄)]; =29·9 [C() (R₂ or R₄)]; =29·3–29·7 [CH₂-rest (R_2', R₃, R₂ or R₄)]; =25·2/25·1 [C() (R₃, R₂ or R₄)]; =25·2 [C() (R_2')]; =22·9 [CH₂.CH_2.CH₃ (R_2', R₃, R₂ or R₄); =14·2 [terminal CH₃ (R₃, R₂ or R₄, R₂')].

Electrospray ionization tandem mass spectrometry (ESI-MS/MS).
A quadrupole time-of-flight mass spectrometer (Micromass) equipped with a nanospray ion source was used and a voltage of approximately 1000 V was applied. For collision-induced dissociations, parent ions were selectively transmitted from the quadrupole mass analyser into the collision cell with argon as the collision gas. The kinetic energy was set at approximately -30 eV. The resulting daughter ions were separated by an orthogonal time-of-flight mass analyser.

Trehalose tetraester M₁.
m/z=899·4, I_rel=100 %, (M₁+Na)⁺; m/z=799·41, I_rel=25·0 %, (M₁+Na-HOOCCH₂CH=C=O)⁺; m/z=781·39, I_rel = 36·3 %, (M₁+Na-succinic acid)⁺; m/z=755·3, I_rel=22·5 %, (M₁+Na-octanoic acid)⁺; m/z=727·28, I_rel=16·3 %, (M₁+Na-decanoic acid)⁺; m/z=655·29, I_rel=13·4 %, (M₁+Na-octanoic acid-HOOCCH₂CH=C=O)⁺; m/z=627·28, I_rel=6·3 %, (M₁+Na-decanoic acid-HOOCCH₂CH=C=O)⁺; m/z=583·25, I_rel=89·1 %, (M₁+Na-anhydroglucose decanoate)⁺; m/z=565·25, I_rel=48·4 %, (M₁+Na-anhydroglucose decanoate-H₂O)⁺; m/z=483·2, I_rel=16·9 %, (M₁+Na-anhydroglucose decanoate-HOOCCH₂CH=C=O)⁺; m/z=465·24, I_rel=47·5 %, (M₁+Na-anhydroglucose decanoate-succinic acid)⁺; m/z=439·16, I_rel=23·1 %, (M₁+Na-anhydroglucose decanoate-octanoic acid)⁺; m/z=411·13, I_rel=35·0 %, (M₁+Na-anhydroglucose decanoate-decanoic acid)⁺; m/z=357·16, I_rel=23·1 %, (M₁+Na-anhydroglucose decanoate-HOOCCH₂CH=C=O-CH₃(CH₂)₅CH=C=O)⁺; m/z=339·16, I_rel=35·9 %, (M₁+Na-anhydroglucose decanoate-HOOCCH₂CH=C=O-octanoic acid)⁺; m/z=321·14, I_rel=26·3 %, (M₁+Na-anhydroglucose decanoate-succinic acid-octanoic acid)⁺; m/z=293·12, I_rel=21·3 %, (M₁+Na-anhydroglucose decanoate-succinic acid-decanoic acid)⁺; m/z=267·03, I_rel=40·0 %, (M₁+Na-anhydroglucose decanoate-decanoic acid-octanoic acid)⁺; m/z=167·1, I_rel=9·7 %, (M₁+Na-anhydroglucose decanoate-octanoic acid-decanoic acid-HOOCCH₂CH=C=O)⁺.

Trehalose tetraester M₂.
m/z=871·38, I_rel=100 %, (M₂+Na)⁺; m/z=771·39, I_rel=29·7 %, (M₂+Na-HOOCCH₂CH=C=O)⁺; m/z=753·38, I_rel=42·5 %, (M₂+Na-succinic acid)⁺; m/z=727·29, I_rel=39·4 %, (M₂+Na-octanoic acid)⁺; m/z=627·27, I_rel=26·3 %, (M₂+Na-succinic acid-CH₃(CH₂)₅CH=C=O)⁺; m/z=583·26, I_rel=31·3 %, (M₂+Na-2 octanoic acids)⁺; m/z=555·24, I_rel=88·8 %, (M₂+Na-anhydroglucose decanoate)⁺; m/z=537·24, I_rel=46·9 %, (M₂+Na-anhydroglucose decanoate-H₂O)⁺; m/z=465·25, I_rel=18·1 %, (M₂+Na-2 octanoic acids-succinic acid)⁺; m/z=437·22, I_rel=58·4 %, (M₂+Na-anhydroglucose decanoate-succinic acid)⁺; m/z=411·14, I_rel=83·8 %, (M₂+Na-anhydroglucose decanoate-octanoic acid)⁺; m/z=393·13, I_rel=39·1 %, (M₂+Na-anhydroglucose decanoate-H₂O-octanoic acid)⁺, m/z=311·13, I_rel=23·8 %, (M₂+Na-anhydroglucose decanoate-HOOCCH₂CH=C=O-octanoic acid)⁺; m/z=293·13, I_rel=63·8 %, (M₂+Na-anhydroglucose decanoate-succinic acid-octanoic acid)⁺; m/z=267·04, I_rel=83·7 %, (M₂+Na-anhydroglucose decanoate-2 octanoic acids)⁺; m/z=167·1, I_rel=19·4 %, (M₂+Na-anhydroglucose decanoate-2 octanoic acids-HOOCCH₂CH=C=O)⁺.

Chemicals.
1,2-, 1,3- and 1,4-DCB were obtained from Merck. 1,4-Difluorobenzene, 1,2-, 1,3- and 1,4-dibromobenzene were from Fluka. All other substituted benzenes were purchased from Aldrich. 1,2,4,5-TeCB, 1,2,4-TCB and the three isomeric dichlorobenzenes were analysed by GC as described by Rapp & Timmis (1999) and no structurally similar contaminants could be detected in significant amounts. Chlorobenzoic acids and D- and L-2-chlorosuccinic acid, all of the highest purity, were obtained from Fluka. Alkanes (purity of 99 %) were from Sigma. All other chemicals were of analytical grade from commercial sources.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Isolation and characterization of strain MS11
Strain MS11 was obtained from a mixed culture utilizing 1,2,4,5-TeCB and 1,2,4-TCB as the sole carbon and energy source, by multiple streaking on nutrient agar. The isolate was aerobic, Gram-positive and non-motile. Its morphology changed from cocci to rods, which could elongate to longer ones or to branched filaments. It formed no aerial hyphae. Strain MS11 was catalase- and urease-positive and oxidase- and arylsulfatase-negative. It was not acid-fast, reduced nitrate to nitrite and its growth was not inhibited by 6 % (w/v) NaCl. Strain MS11 grew at temperatures from 5 to 30 °C; optimum growth was observed at 20 °C. It grew well at 5 °C on glucose and, after adaptation, also on the much less water-soluble n-hexadecane as the sole carbon source, thus demonstrating its psychrophilic character. Growth of strain MS11 on mineral salts medium with 5 g l^-1 of universal peptone (Merck) was not inhibited by penicillin G (10 µg ml^-1), while mitomycin (2 µg ml^-1), gentamicin (25 µg ml^-1), chloramphenicol (5 µg ml^-1) and rifampicin (25 µg ml^-1) inhibited its growth. Strain MS11 utilized the following compounds as sole carbon and energy sources: D-glucose, D-fructose, D-mannose, D-ribose, D-xylose, maltose, sucrose, cellobiose, starch, sorbitol, mannitol, myo-inositol, methanol, ethanol, n-butanol, butan-2,3-ol, glycerol, acetate, succinate, pimelate, citrate, L-glutamate, L-leucine, DL-norleucine, L-histidine, DL-phenylalanine, -alanine, L-proline and L-asparagine. It did not utilize the following compounds as sole carbon and energy sources: D-galactose, L-arabinose, lactose, raffinose, gelatin, L-lactate, DL-mandelate, D-tartrate and DL-methionine. This nutrional versatility and antibiotic susceptibility of strain MS11 is typical of the genus Rhodococcus (Goodfellow, 1989; Zaitsev et al., 1995; Bizet et al., 1997; Goodfellow et al., 1998).

16S rRNA of strain MS11 was partially sequenced and compared with known sequences in the database of the Gesellschaft für Biotechnologische Forschung, Division of Microbiology (Maidak et al., 1997; Stoesser et al., 1997). Its partial sequence (1390 bp) was 99·8 % similar to that of type strain Rhodococcus erythropolis DSM 43066^T.

On the basis of biological and biochemical characteristics as well as 16S rRNA sequence analysis, strain MS11 was classified as a Rhodococcus sp.

Growth on alkanes and on halogenated and nonhalogenated aromatic compounds
Rhodococcus sp. strain MS11 was able to utilize many alkanes as sole carbon and energy source (Table 1). It grew well on n-alkanes ranging from n-heptane to n-triacontane and slowly on n-hexane and on n-pentatriacontane. The products of the terminal oxidation of n-alkanes, in the form of 1-hexadecene, 1-hexadecanol and 1-hexadecanoic acid, as well as the branched alkane 2,6,10,14-tetramethylpentadecane (pristane) were also utilized by strain MS11 as sole carbon source. Fig. 1 illustrates the growth of strain MS11 in shake-flask cultures on alkanes: in this case with 1 % (w/v) n-tetradecane as carbon source. With this substrate, a maximum specific growth rate of 0·042 h^-1 was determined. The values of µ_max determined in shake-flask cultures with other n-alkanes as carbon sources ranged from 0·03 h^-1 for n-decane to 0·08 h^-1 for n-hexadecane and from 0·06 h^-1 for n-heptadecane to 0·03 h^-1 for n-nonadecane.

View larger version (21K): [in this window] [in a new window]   Fig. 1. Growth () and biosurfactant formation (; measured as the surface tension of the culture) of Rhodococcus sp. strain MS11 on 1 % (w/v) n-tetradecane as the sole source of carbon. , Surface tension of the abiotic control. Data are shown as the mean±ranges of two independent cultures.

View larger version (21K): [in this window] [in a new window]   Fig. 2. Growth () and chloride release () of Rhodococcus sp. strain MS11 on 1,2,4,5-TeCB. Parallel 100 ml cultures initially contained 1·72 mM 1,2,4,5-TeCB. , 1,2,4,5-TeCB concentration of the culture; , 1,2,4,5-TeCB concentration of the abiotic control. Data are shown as the mean±SD of four independent cultures.

View larger version (17K): [in this window] [in a new window]   Fig. 3. Chloride concentration after degradation of 1,2,4,5-TeCB (),1,2,4-TCB () and 3-chlorobenzoate () of increasing concentration by Rhodococcus sp. strain MS11. Data are shown as the mean±ranges of two independent cultures.

View larger version (34K): [in this window] [in a new window]   Fig. 4. Phylogenetic tree based on nucleotide sequences of the -subunit of initial dioxygenases from different bacteria. The names of bacterial strains and the DDBJ, EMBL, GenBank, SWISS-PROT and PIR accession numbers are indicated in parentheses. The distance scale indicates 10 % substitutions and the numbers shown at the nodes indicate bootstrap percentages. Only bootstrap values >80 % are shown. The tree was resampled 100 times. It was constructed using -subunit sequences of group IV described by Nam et al. (2001).

Using n-alkanes ranging from n-decane to n-heptadecane,1-hexadecene or pristane as sole carbon and energy sources, the surface tension of the cultures always strongly decreased. Fig. 1 illustrates the biosurfactant formation, measured by means of the decrease of surface tension, during cultivation of MS11 on n-tetradecane as sole carbon source. It was lowered from 72 to 29 mN m^-1. This decrease took place rapidly at the beginning of cultivation, when the cell mass was still low. However, the surface tension of the cultures was not lowered markedly when strain MS11 was grown at room temperature on n-alkanes with chain lengths of n18 and n9 or on 1-hexadecanol and hexadecanoic acid as sole carbon source (data not shown).

Characterization of trehaloselipids
The structure of the glycolipids was established from the combined ¹H- and ¹³C- NMR, and MS data (see Methods). The -glucopyranose moieties in the glycolipids were identified from the correlations in the 2D COSY spectrum and from the magnitude of the chemical shifts and vicinal coupling constants of the 1D ¹H spectrum. In the ¹H detected ¹³C–¹H correlated spectrum recorded by the HMBC (heteronuclear multiple bond correlation) procedure, the anomeric proton of the glucopyranose moieties showed both a residual one-bond and three-bond correlations to the anomeric carbon, thus indicating an ,-trehalose system. The 2D ¹H COSY spectrum of the trehalose lipids indicated by the low-field shifts of the corresponding glucose ring protons next to the esterified hydroxyl groups, in the region of 4·7–5·5 ppm, that the hydroxyl groups at positions 2, 3, 4 and 2' were esterfied with fatty acids. The ¹H NMR data showed additionally that the fatty acids consisted of three medium-chain-length acids and one succinic acid. Long-range ¹³C–¹H correlations in the 2D ¹H, ¹³C COSY spectrum of the trehalose tetraesters between the trehalose ring and methylene protons of the fatty acids with their carbonyl carbons showed that the fatty acids at C-3 and C-2' were of medium chain length and that a succinic acid residue was at C-2 or C-4.

The ESI-mass spectrum of the 2,3,4,2'-trehalose tetraesters showed, besides some very small signals, two prominent (M+Na)⁺ ion peaks at m/z 899·4 and 871·38 (see Methods), the first with a more than twofold higher relative abundance than the latter. The next highest peaks after the molecular ion peaks were those at m/z 583·25 and 555·24. Their appearance resulting from the removal of an anhydroglucose moiety esterified with a decanoic acid demonstrated that the medium-chain-length fatty acid at C-2' of both trehalose tetraesters was a decanoic acid. The fragmentation patterns of both molecular ions confirmed additionally that trehalose was, as already indicated by the NMR spectra, esterified at C-2 or C-4 with a succinic acid. However, the two fragmentation pathways also differed from one another. While an octanoic and a decanoic acid were split off from one glucose moiety of the higher molecular mass trehalose tetraester, two octanoic acids were removed from one glucose residue of the lower molecular mass trehalose lipid. NMR data and fragmentation patterns of both 2,3,4,2'-trehalose tetraesters led to the structure shown in Fig. 5. In both cases, ,-trehalose was found to be esterified at C-2 or C-4 with a succinic and at C-2' with a decanoic acid. Furthermore, the trehalose lipid of molecular mass 876 was found to be linked at C-2, C-3 or C-4 with an octanoic and a decanoic acid and that with the lower molecular mass of 848 with two octanoic acids.

View larger version (8K): [in this window] [in a new window]   Fig. 5. Structure of 2,3,4,2'-trehalose tetraester of molecular mass 848: R2 or R4, succinic acid; R2, R3 or R4, two octanoic acids. Structure of 2,3,4,2'-trehalose tetraester of molecular mass 876: R2 or R4, succinic acid; R2, R3 or R4, octanoic and decanoic acid.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The chlorobenzene dioxygenase of strain MS11 can be grouped most probably into the class IIB dioxygenase subfamily. These Rieske non-haem-iron oxygenases are now classified into groups based on their substrate specificity and the sequence of their -subunits (Gibson & Parales, 2000; Nam et al., 2001). The sequence of the -subunit of the chlorobenzene dioxygenase of strain MS11 was >99 % identical with that of the chlorobenzene dioxygenase from Ralstonia sp. strain PS12 (Beil et al., 1997) and 98 % identical with that of the chlorobenzene dioxygenase from Pseudomonas sp. strain P51 (Werlen et al., 1996). These sequences cluster very closely together with sequences of the -subunits of toluene, benzene and biphenyl dioxygenases assembled in group IV of the classification scheme of Nam et al. (2001) (Fig. 4). Additionally, the region of the amino acid sequence of the -subunit of the two chlorobenzene dioxygenases from strains MS11 and PS12 responsible for substrate specificity (Beil et al., 1998) showed 100 % identity. It is well known that many catabolic genes are associated with transposons (Wyndham et al., 1994) and that transposition is a major mechanism for the acquisition of catabolic genes by bacterial genomes. However, such genes can also be acquired by site-specific integration (Poelarends et al., 2000). Ralstonia (formerly Burkholderia) sp. strain PS12 was isolated together with Burkholderia sp. strain PS14 from a soil sample of a waste disposal site (Sander et al., 1991). Both strains are able to degrade the same chlorinated benzenes, ranging from 1,2,4,5-TeCB to monochlorobenzene (Sander et al., 1991; Beil et al., 1997), and have plasmids of the same size (Sander, 1991). These correspondences of the two strains suggest that they have the same gene encoding chlorobenzene dioxygenase. Taking this into consideration as well as the similarity of >99 % among the sequences of the chlorobenzene dioxygenases of strains PS12 and MS11, one can possibly conclude that the gene encoding chlorobenzene dioxygenase of the Gram-positive strain MS11 may have been acquired from the Gram-negative strain PS14 by horizontal gene transfer.

While the gene encoding the -subunit of the chlorobenzene dioxygenase and possibly also the other upper pathway genes of Rhodococcus sp. strain MS11 exhibit high sequence similarities to those of Ralstonia sp. strain PS12 and Pseudomonas sp. strain P51, such a correspondence between the lower pathway genes encoding chlorocatechol 1,2-dioxygenase, chloromuconate cycloisomerase and dienolactone hydrolase seems to be doubtful and has to be examined. The reason for this assertion is that the chlorocatechol-degrading enzymes and especially the chloromuconate cycloisomerase of another Rhodococcus, R. opacus 1CP, have unusual biochemical properties and sequences with relatively little similarity to those of proteobacterial chlorocatechol-degrading enzymes, thus indicating a different origin (Schlömann, 2002).

The most striking difference, however, between the four above-mentioned Gram-negative, tri- and/or tetrachlorobenzene-degrading proteobacteria and the Gram-positive, tri- and tetrachlorobenzene-degrading Rhodococcus sp. strain MS11 is the ability of the latter to utilize an unusually broad range of alkanes as sole carbon and energy source. Strain MS11 was able to grow on straight-chain alkanes consisting of 7–30 carbon atoms, on its possible metabolites as demostrated by means of 1-hexadecene, 1-hexadecanol and 1-hexadecanoic acid, and on the recalcitrant branched alkane 2,6,10,14-tetramethylpentadecane (pristane). It grew slowly on n-pentatriacontane and on n-hexane. The metabolism of alkanes is a not uncommon feature of rhodococci (Finnerty, 1992). There are strains able to grow on n-decane to n-triacontane (Milekhina et al., 1998), on n-dodecane to n-dotriacontane (Whyte et al., 1998) or on n-hexane to n-eicosane and on pristane (Bej et al., 2000). The rhodococci, however, also include strains metabolizing only a rather narrow spectrum of n-alkanes ranging from n-undecane to n-heptadecane (Lofgren et al., 1995) or from n-dodecane to n-eicosane (Sorkoh et al., 1990). Like some other Rhodococcus species (Bej et al., 2000; Whyte et al., 1998), strain MS11 was psychrotolerant, since it could grow at 5 °C, while its optimum temperature was 20 °C.

It has been observed that strain MS11 degraded n-alkanes with chain lengths of C₁₀–C₁₇ as readily as or even better than those of shorter chain length, although the water solubilities of the latter are higher (McAuliffe, 1969; Zhang & Miller, 1995). This may partly depend on the toxicity of n-alkanes with chain lengths of n9. However, it may also indicate that an additional mechanism to the mere dissolution in water is responsible for the bioavailability of alkanes ranging from n-decane to n-heptadecane. It is well known that many hydrocarbon-degrading bacteria optimize the uptake of medium-chain-length alkanes by producing biosurfactants (Hommel, 1990). Among them, rhodococci are the most prominent group (Finnerty, 1992). Therefore, it was not surprising that Rhodococcus sp. strain MS11 also produced biosurfactants, lowering the surface tension of cultures from 72 mN m^-1 to 29 mN m^-1 during growth on n-alkanes ranging from n-decane to n-heptadecane, as well as on 1-hexadecene and the branched alkane pristane. The biosurfactants were extracted and purified from the supernatant of a culture grown on n-hexadecane as sole carbon source. ¹H and ¹³C NMR spectroscopy as well as ESI-MS/MS showed that the two main biosurfactants were ,-trehalose tetraesters linked at C-2 or C-4 with a succinic and at C-2' with a decanoic acid. They differed in that one tetraester was esterified at C-2, C-3 or C-4 with an octanoic and a decanoic acid and the other one at the same positions with two octanoic acids. Similar 2,3,4,2'-trehaloselipids, the main components of which were esterified with two decanoic, one octanoic and one succinic acid, were found in cultures of R. erythropolis (DSM 43215) grown on n-tetra- and n-pentadecane under nitrogen-limiting conditions. However, the exact linkages between trehalose and the four fatty acids could not be determined (Kim et al., 1990; Ristau & Wagner, 1983). These anionic 2,3,4,2'-trehalose tetraesters were found to be powerful biosurfactants, lowering the interfacial tension between water and n-hexadecane to less than 1 mN m^-1 at a critical micelle concentration of 15 mg l^-1 (Kim et al., 1990). Another 2,3,4,2'-trehalose tetraester obtained from a culture of Rhodococcus sp. strain 51T7 grown on waste lubricant oil also contained succinic acid and fatty acids ranging from hepta- to undecanoic acid (Espuny et al., 1995). Fatty acids with chain lengths of C₈–C₁₄ were the lipophilic part of a 2,3,4,2'-trehalose tetraester isolated from the marine bacterium Arthrobacter sp. strain EK1. In this tetraester, trehalose was found to be esterified at C-2 position with succinate (Passeri et al., 1991). R. erythropolis strain SD74 grown on n-hexadecane excreted two other trehalose tetraesters, 2,3,4,2'-di-O-succinoyl-di-O-alkanoyl-,-trehalose and 2,3,4-mono-O-succinoyl-di-O-alkanoyl-,-trehalose. The fatty acids of both trehaloselipids were tetradecanoic and hexadecanoic acid (Uchida et al., 1989). Finally, Bartrakov et al. (1981) extracted from Mycobacterium paraffinicum, grown on n-hexadecane, 2-O-octanoyl-3,2'-di-O-decanoyl-6-O-succinoyl-,-trehalose together with four other trehaloselipids.

Rhodococcus sp. strain MS11 is, to our knowledge, the first micro-organism described to degrade di-, tri- and tetrachlorobenzenes and to produce biosurfactants during growth on a wide range of alkanes. Alkanes and chlorinated benzenes, especially the higher substituted ones, are hydrophobic and tend to sorb onto soils and sediments. Over a longer contact time they slowly diffuse into their inorganic or organic matrix. Furthermore, accumulation of these contaminants into fissures, cavities and pores also renders them inaccessible to micro-organisms (Bouwer & Zehnder, 1993). The extracellular surface- and interfacial-active trehalose tetraesters excreted by Rhodococcus sp. strain MS11 may be well suited for bioremediation of sites polluted in such a way. Moreover, strain MS11 was found to be psychrophilic and, according to Larsen's classification (Larsen, 1986), also slightly halotolerant. This wide range of capabilities together with the known environmental persistence of rhodococci (Warhurst & Fewson, 1994) makes Rhodococcus sp. strain MS11 a good candidate for bioremediation of soils and sediments, contaminated for years or even decades with these recalcitrant compounds.

	ACKNOWLEDGEMENTS

 
We thank A. Krüger and C. Strömpel for generous sequencing support and M. Nimtz and V. Wray for very helpful discussions of the NMR and mass spectra. The assistance of M. Sylla, B. Jaschok-Kentner, C. Kakoschke and P. Wolff is gratefully acknowledged.

	REFERENCES

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Bartrakov, S. G., Rozynov, B. V., Koronelli, T. V. & Bergelson, L. D. (1981). Two novel types of trehalose lipids. Chem Phys Lipids 29, 241–266.

Beil, S., Happe, B., Timmis, K. N. & Pieper, D. H. (1997). Genetic and biochemical characterization of the broad spectrum chlorobenzene dioxygenase from Burkholderia sp. strain PS12. Dechlorination of 1,2,4,5-tetrachlorobenzene. Eur J Biochem 247, 190–199.[Medline]

Beil, S., Mason, J. R., Timmis, K. N. & Pieper, D. H. (1998). Identification of chlorobenzene dioxygenase sequence elements involved in dechlorination of 1,2,4,5-tetrachlorobenzene. J Bacteriol 180, 5520–5528.[Abstract/Free Full Text]

Beil, S., Timmis, K. N. & Pieper, D. H. (1999). Genetic and biochemical analyses of the tec operon suggest a route for evolution of chlorobenzene degradation genes. J Bacteriol 181, 341–346.[Abstract/Free Full Text]

Bej, A. K., Saul, D. & Aislabie, J. (2000). Cold-tolerant alkane-degrading Rhocococcus species from Antarctica. Polar Biol 23, 100–105.[CrossRef]

Bergmann, J. G. & Sanik, J. (1957). Determination of trace amounts of chlorine in naphtha. Anal Chem 29, 241–243.[CrossRef]

Bizet, C., Barreau, C., Harant, C., Nowakowski, M. & Pietfroid, A. (1997). Identification of Rhodococcus, Gordona and Dietzia species using carbon source utilization tests ("biotype-100" strips). Res Microbiol 148, 799–809.[Medline]

Boehringer Mannheim Biochemica (1994). Methoden der enzymatischen Bioanalytik und Lebensmittelanalytik. Mannheim, Germany: Boehringer Mannheim.

Bosma, T. N., Middeldorp, P. I. M., Schraa, G. & Zehnder, A. J. B. (1997). Mass transfer limitation of biotransformation: quantifying bioavailability. Environ Sci Technol 31, 248–252.[CrossRef]

Bouwer, E. J. & Zehnder, A. J. B. (1993). Bioremediation of organic compounds – putting microbial metabolism to work. TIBTECH 11, 360–367.

Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72, 248–254.[CrossRef][Medline]

Brunsbach, F. R. & Reineke, W. (1994). Degradation of chlorobenzenes in soil slurry by a specialized organism. Appl Microbiol Biotechnol 42, 415–420.[Medline]

Chauhan, S., Barbieri, P. & Wood, T. K. (1998). Oxidation of trichloroethylene, 1,1-dichloroethylene, and chloroform by toluene/o-xylene monooxygenase from Pseudomonas stutzeri OX1. Appl Environ Microbiol 64, 3023–3024.[Abstract/Free Full Text]

De Bont, J. A. M., Vorage, M. J. A. W., Hartmans, S. & van den Tweel, W. J. J. (1986). Microbial degradation of 1,3-dichlorobenzene. Appl Environ Microbiol 52, 677–680.[Abstract/Free Full Text]

Dorn, E. & Knackmuss, H.-J. (1978a). Chemical structure and biodegradability of halogenated aromatic compounds. Two catechol 1,2-dioxygenases from a 3-chlorobenzoate-grown pseudomonad. Biochem J 174, 73–84.[Medline]

Dorn, E. & Knackmuss, H.-J. (1978b). Chemical structure and biodegradability of halogenated aromatic compounds. Substituent effects on 1,2-dioxygenation of catechol. Biochem J 174, 85–94.[Medline]

Espuny, M. J., Egjido, S., Mercade, M. E. & Manresa, A. (1995). Characterization of trehalose tetraester produced by a waste lube oil degrader Rhodococcus sp. 51T7. Toxicol Environ Chem 48, 83–88.

Finnerty, W. R. (1992). The biology and genetics of the genus Rhodococcus. Annu Rev Microbiol 46, 193–218.[CrossRef][Medline]

Gallardo, M. E., Fernandez, A., de Lorenzo, V., Garcia, J. L. & Diaz, E. (1997). Designing recombinant Pseudomonas strains to enhance biodesulfurization. J Bacteriol 179, 7156–7160.