The ability of sediment bacteria to utilize polycyclic aromatic hydrocarbons (PAHs) when present as components of mixtures was investigated. One strain, identified as Mycobacterium flavescens, could utilize fluoranthene in the presence of pyrene, although utilization of pyrene was slower in the presence of fluoranthene than in its absence. The second strain, a Rhodococcus species, could utilize fluoranthene in the presence of anthracene, although the presence of fluoranthene slowed the rate of utilization of anthracene. Cometabolism of fluoranthene in these strains was confirmed by the isolation of metabolites of fluoranthene and by kinetic analysis of the rate of utilization of the growth substrate in the presence of fluoranthene. In both strains, metabolism of fluoranthene occurred on the fused ring of the fluoranthene molecule, producing 9-fluorenone-1-carboxylic acid. In the Rhodococcus sp., a second metabolite, a-(carboxymethylene)fluorene-1-carboxylic acid, was identified, indicating that this strain has the capacity to metabolize fluoranthene via ortho as well as meta cleavage. The presence of PAHs in a mixture produces interactive effects which can either increase or decrease the rate of utilization of individual PAHs, results which need to be taken into account when estimating rates of degradation in contaminated environments.
Polycyclic aromatic hydrocarbons (PAHs) are widespread industrial pollutants that are released into the environment from coking of coal, distillation of wood, operation of gas works and oil refineries, runoff from asphalt pavements, and combustion processes. Their physico-chemical properties, which include low water solubility and high adsorption coefficient, combine to make soils and sediments environmental sinks for PAHs. Concentrations of 3037 mg kg−1 and 334 mg kg−1 for anthracene are typical of soils from wood-preserving and creosote production facilities . Concentrations of total PAHs in Mediterranean sediments varied between 1 and 20 500 ng g−1 dry weight , while in Penobscot Bay, concentrations ranged from 286 to 8794 ng g−1. In both cases, the highest concentrations corresponded to the most heavily industrialized areas.
PAHs are of concern because some of them have been identified as genotoxicants in short-term mutagenicity tests such as the Ames test and as animal carcinogens in long-term carcinogenicity bioassays [4,5]. Their mutagenicity varies with the number of aromatic rings. The two-ringed naphthalene and three-ringed compounds phenanthrene and anthracene do not demonstrate genotoxicity and carcinogenicity while some four-ringed compounds such as pyrene are genotoxic as measured in the Ames test yet are noncarcinogenic as determined in rodent bioassays [4,5]. Other four-ringed PAHs as well as five-ringed compounds such as benzanthracene and benzo[a]pyrene are both genotoxic and carcinogenic [4,5].
One of the most promising methods of removal of PAHs from contaminated environments is bioremediation [1,6]. A wide array of bacteria isolated from contaminated sites have been shown to be capable of degrading a range of PAHs when exposed to individual PAHs [7,8]. Studies using single PAHs, however, do not reflect the true complexity of PAH degradation in natural environments where the compounds are present in multicomponent mixtures. When present in mixtures, PAHs have the capacity to influence the rate and extent of biodegradation of other components of the mixture. In some cases, these interactions may be positive, resulting in an increase in biodegradation of one or more components [9–12], while in other cases negative effects have been observed [11,13,14]. To elucidate the extent of such interactions, the biodegradation of PAHs alone and in two-component and three-component mixtures was studied using bacteria isolated from contaminated sediments.
2Materials and methods
2.1Organisms and growth conditions
Bacterial strains were isolated by enrichment culture using PAHs as the sole source of carbon and energy. One bacterium, isolated from sediments of the Grand Calumet River using pyrene as sole source of carbon and energy, has been identified as a Mycobacterium flavescens. A second bacterium was isolated on anthracene as sole source of carbon and energy utilizing sediment from a lagoon at the headwaters of the Grand Calumet River in northwestern Indiana as the inoculum. The bacterial strain was identified as a Rhodococcus sp. based on Gram reaction, morphology and fatty acid methyl ester profile (Microbial ID, Newark, DE, USA). A mineral salts medium  was used in enrichment experiments and subsequent growth experiments. Solid PAH was added prior to autoclaving. After several transfers, the enrichment cultures were plated on nutrient agar and sprayed with an acetone solution of PAH to produce a solid film of PAH on the plate; PAH-utilizing cultures were identified by the presence of a clear zone around the colony . Metabolism was confirmed by means of a mineralization screening test .
2.2Utilization of mixtures of PAHs
To determine utilization of PAH alone or in mixtures, sufficient PAH was added to mineral salts medium to provide the target concentration of 50 μg ml−1 as a solution in dimethyl sulfoxide. Under these conditions, the insoluble PAH forms a suspension which can be reproducibly sampled. Duplicate flasks were inoculated with the respective bacterial strains and incubated at 24°C on a gyratory shaker (150 rpm) in the dark. At intervals, 10-ml samples were removed for extraction to determine the amount of PAH remaining in suspension .
2.3Isolation of fluoranthene metabolites
Metabolism of fluoranthene was verified and metabolites were identified by utilizing a dual substrate system in which the growth PAH was unlabeled and fluoranthene was radiolabeled. In experiments with either pyrene–fluoranthene or anthracene–fluoranthene combinations, both PAHs were supplied to the culture flasks of M. flavescens or the Rhodococcus sp., respectively, in mineral salts medium at a concentration of 50 mg/l. After appropriate time intervals, the contents of the flasks were subjected to two extractions using three aliquots of 50 ml ethyl acetate, one on the neutral suspension and a second after acidification with 6 M HCl to achieve a pH of 2. In each case, the solvent was evaporated and the residue dissolved in 100 μl methanol and analyzed for the presence of metabolites as described below.
2.4Measurement of PAH utilization rates
Utilization rates of PAHs were determined using a mineralization assay. For these experiments, resting cells were prepared by centrifugation of cultures of the bacterial strains grown on the respective PAH, washing the pellet with 0.01 M phosphate buffer (pH 7.2), followed by centrifugation and resuspension in buffer to yield a suspension with an optical density of 0.150. The cell suspension (1 ml) was added to serum vials containing 10 ml of phosphate buffer and fitted with a wick containing 0.2 ml of 1 M KOH. The reaction was initiated by the addition of a solution of either anthracene to give a concentration of 0.75–3 μM or pyrene to give a concentration of 0.2–0.5 μM in the serum vial in addition to radiolabeled PAH to provide 20 000 dpm per vial. At each time interval, the reaction in triplicate vials was stopped by the addition of 1 ml of 1 M H2SO4. Wicks were removed and counted by liquid scintillation counting. Data were plotted to determine the initial rate of reaction. In order to determine the effect of fluoranthene on the degradation of the growth PAH, a second series of vials were prepared containing fluoranthene at a concentration of 1.0 μM. To verify that the observed mineralization was biological and to correct for PAH volatilization, killed-cell controls were prepared in an identical manner except that acid was added prior to addition of the radiolabeled PAH. Data from the experimental flasks were corrected by subtraction of dpm from the killed-cell controls for each combination of PAH concentration and organism used. Protein was determined by the bicinchoninic acid method .
For the determination of PAH utilization, 10 ml of culture medium was extracted with three 10-ml portions of methylene chloride. The portions were pooled, the methylene chloride extract was dried over sodium sulfate and evaporated using a rotary evaporator. The methylene chloride-extractable residue was dissolved in 100 μl of methylene chloride supplemented with eicosane as internal standard for subsequent chemical analysis. Recovery of pyrene under these extraction conditions was 89.4%; slightly higher recoveries of phenanthrene and anthracene (92–93%) have been obtained using the same procedure.
Quantitation of PAHs was performed by GC using a Model 8500 gas chromatograph (Perkin-Elmer, Norwalk, CT, USA) equipped with a flame ionization detector using a method developed by Mueller et al. . The column was a fused silica capillary column, 30 m long with an internal diameter of 0.24 mm, coated with 1.0 μm of a bonded and cross-linked stationary phase consisting of 5% phenyl-substituted polymethylsiloxane (DB-5, J&W Scientific, Rancho Cordova, CA, USA). Initial column temperature 150°C; temperature was increased to 275°C at a rate of 5°C min−1, and was held at the final temperature for 5 min. The injector and detector temperatures were 250°C. Using this method, the limit of detection was determined to be 2 μg ml−1 of the original mineral salts medium.
PAH metabolites were separated by high-performance liquid chromatography (HPLC) using a Hewlett-Packard model 1050 pump system (Hewlett-Packard, Palo Alto, CA, USA) with a Hewlett-Packard photo-diode array model 1040A detector at 254 nm and a Radiomatic A-500 radiochromatography detector connected in series. The compounds were eluted using a linear gradient of 40–95% methanol/water over 40 min at 1 ml min−1 with a 4.6×250 mm 5 μm C18 Inertsil ODS-3 column (MetaChem Technologies, Torrance, CA, USA). UV absorbance spectra were acquired on line.
Electron ionization mass spectrometry analysis was performed using a TSQ 700 triple quadrupole mass spectrometer (Finnigan, San Jose, CA, USA) equipped with a direct exposure probe. Liquid chromatography/mass spectrometry (LC/MS) analysis was performed using an HP 5989B mass spectrometer equipped with an HP 1090L/M HPLC (Hewlett-Packard). The mass spectrometer was operated in the negative electrospray ionization mode with the capillary exit voltage at −100 V. Full scans were acquired from m/z 50–650 at 1.28 scans s−1. HPLC was performed with a Prodigy ODS(3) 2.0×250 mm 5μ 100A HPLC column (Phenomenex, Torrance, CA, USA). The mobile phase delivered at 0.2 ml min−1 with a linear gradient from 5% acetonitrile/95% water to 95% acetonitrile/5% water in 30 min with no buffer.
1H Nuclear magnetic resonance (NMR) spectra were obtained at 500 MHz with a Bruker AM500 spectrometer (Bruker Instruments, Billerica, MA, USA) at 28°C with the following parameters: sweep width, 7K; datum size, 32K; acquisition time, 2.33 s; flip angle, 90°; relaxation time, 0 s. A 4-s saturation time was used for the nuclear Overhauser effect (NOE) experiments.
Unlabeled PAHs were obtained from Aldrich Chemical Company (Milwaukee, WI, USA) and were all greater than 98% (w/w) purity. Several radiolabeled chemicals were obtained from Chemsyn Science Laboratories (Lenexa, KS, USA): 4,5,9,10-[14C]pyrene (55 mCi mmol−1); 9,10-[14C]anthracene (58 mCi mmol−1); 5,6,11,12-[14C]chrysene (54.4 mCi mmol−1); and 3-[14C]fluoranthene (50 mCi mmol−1). The following PAHs were supplied by Sigma (St. Louis, MO, USA): 1-[14C]naphthalene (10.3 mCi mmol−1) and 9-[14C]phenanthrene (13.1 mCi mmol−1).
3.1Isolation and characterization of PAH-degrading bacteria
Isolation and characterization of the pyrene-degrading strain as M. flavescens has been previously described . Using anthracene as enrichment substrate and an inoculum consisting of sediment from lagoons at the headwaters of the Grand Calumet River, a pure culture was obtained which was capable of growth on anthracene as sole source of carbon and energy. Based on morphology, Gram reaction and results of physiological tests as well as the fatty acid profile, the culture was identified as a Rhodococcus sp. The Rhodococcus sp. was tested for the ability to mineralize several commonly occurring PAHs. It could not utilize naphthalene or chrysene to a significant extent, but could mineralize phenanthrene (31.0%), pyrene (13.6%), anthracene (53.0%) and fluoranthene (4.7%) over a 2-week period.
3.2Utilization of mixtures of PAHs
Experiments were performed to measure the effect of fluoranthene on the utilization of pyrene by M. flavescens (Fig. 1). Fluoranthene was utilized slowly when present alone (0.610 nmol h−1 mg protein−1). When fluoranthene and pyrene were provided in combination, the rate of utilization of fluoranthene increased to 1.28 nmol h−1 mg protein−1. In contrast, an inhibitory effect on pyrene degradation was observed in comparison to utilization of pyrene when present alone, the rate of utilization decreasing from 2.55 nmol h−1 mg protein−1 to 1.42 nmol h−1 mg protein−1.
When a similar experiment was performed with the Rhodococcus sp., a similar effect was observed. Fluoranthene inhibited the utilization of anthracene (Fig. 2), decreasing the utilization rate from 1.65 nmol h−1 mg protein−1 to 0.362 nmol h−1 mg protein−1. The utilization of fluoranthene in the mixture (0.93 nmol h−1 mg protein−1) was stimulated in comparison to its utilization when alone (0.21 nmol h−1 mg protein−1).
3.3Competitive inhibition of PAH utilization
Kinetic experiments were undertaken to determine whether the observed inhibition by fluoranthene of the utilization rate of the primary substrate was competitive. Analysis of the initial reaction rates of PAH mineralization gave estimates of 0.044 mg l−1 for the Ks for pyrene mineralization by M. flavescens and 0.470 μg l−1 for the Ks for anthracene mineralization by Rhodococcus sp. Transformation of the initial rate data by Lineweaver–Burk analysis is given in Figs. 3 and 4. For both organisms, double reciprocal plots of the utilization of the growth substrate in the presence and absence of fluoranthene result in the same intercepts on the y-axis, suggesting that fluoranthene is a competitive inhibitor .
3.4Identification of fluoranthene metabolites
Analysis of the extracts of suspensions of the M. flavescens and Rhodococcus when exposed to pyrene–fluoranthene and anthracene-fluoranthene revealed the presence of one and two, respectively, radiolabeled metabolites (Figs. 5 and 6). Fluoranthene metabolite (I) was isolated from incubations with both organisms. It was shown to have a retention time of 23.14 min. The UV spectrum, with absorbance maxima at 207 and 258 nm, was identical to that of 9-fluorenone-1-carboxylic acid. This identification was confirmed by mass spectrometry. The spectrum gave a molecular ion peak (M+) with fragment ions at m/z 180 and 152, corresponding to losses of CO2 and CO2–CO, respectively, a spectrum which is consistent with a compound having a molecular mass of 224 and a single carboxylic acid group. The 1H NMR assignments and coupling constants in deuterated methanol were 7.21 (H2, J2,3=7.5 Hz, J2,4=0.9 Hz), 7.48 (H3, J3,4=7.5 Hz), 7.56 (H4), 7.64 (H5, J5,6=7.3 Hz), 7.52 (H6, J6,7=7.5 Hz, J6,8=1.1 Hz), 7.31 (H7, J7,8=7.5 Hz), 7.58 (H8). Decoupling experiments indicate the presence of two aromatic rings. Selective saturation of H5 produced NOEs to both H4 and H6. Homonuclear decoupling experiments were used to determine the site of substitution as C-1. Based on the mass spectral data, the substituted group was determined to be a carboxylic acid and this compound was identified as 9-fluorenone-1-carboxylic acid.
The second fluoranthene metabolite (II) obtained from the anthracene–fluoranthene combination when analyzed by LC/MS gave a mass spectrum with prominent ions at 265, 221 and 177. The ions observed for this compound were ions with a hydrogen removed at [M−H]−; molecular masses were one higher than the ions. The molecular mass of the compound was therefore 266, with peaks corresponding to consecutive loss of two carboxyl moieties [M−H−CO2]− and [M−H−2CO2]−. The 1H NMR assignments and coupling constants in D2O were 6.83 (H1), 7.77 (H4, J4,5=7.5 Hz, J4,6=1.1 Hz), 7.37 (H5, J5,6=7.7 Hz), 7.44 (H6), 7.73 (H7, J7,8=7.5 Hz, J7,9=1.1 Hz), 7.36 (H8, J8,10=1.1 Hz, J9,10=7.8 Hz), 7.29 (H9), and 7.73 (H10). The observed coupling patterns were consistent with the three rings of fluoranthene: one ring having four proton resonances, one ring with three and a third ring with one. The singlet at 6.83 ppm (H1), when irradiated, produced a NOE at a resonance (H10) in the ring consisting of four proton resonances. The singlet had no long-range couplings to any other resonances, so the carboxyl groups were determined to be at C2 and C3. Based on the mass spectrum, the ring was broken between C2 and C3. The metabolite was identified as 9-(carboxymethylene)fluorene-1-carboxylic acid.
The ability of bacteria to utilize PAHs as growth substrates has been documented by extensive studies over the past few decades [7,8,20]. These studies share a common approach: that of isolating bacteria from the environment and using the isolated strains as pure cultures in order to establish the metabolic pathways by which the bacterial strains metabolize individual PAHs. While necessary to characterize the bacterial metabolism of these compounds, such studies do not reflect the true complexity of PAH degradation in natural environments where the compounds are present in multicomponent mixtures.
Results of the relatively few studies which have focused on the degradation of PAHs as components of mixtures indicate that simultaneous utilization of PAHs is common when pure cultures of PAH-degrading strains are supplied with mixtures of PAHs. Pure cultures, including M. flavescens and strains of Pseudomonas, have been found to be capable of utilizing mixtures of PAHs simultaneously [21,22]. Similar results have been obtained with bacterial communities from soil and sediment [9,23]. The present study confirms the ability of bacterial strains to utilize PAHs simultaneously.
Frequently, interactive effects such as inhibition and cometabolism are observed in addition to simultaneous utilization. Cometabolism as defined by Criddle  consists of the transformation of a nongrowth substrate by growing cells in the presence of a growth substrate, as well as transformation by resting cells in the absence of growth substrate and resting cells in the presence of an energy substrate. Enhancement in the rate of utilization of one PAH in the presence of a growth substrate has been frequently observed and attributed to cometabolism [10–12,25,26].
In the present study, the utilization of fluoranthene in the presence of the respective growth substrates was scrutinized in order to verify that cometabolism was indeed occurring. Metabolites of fluoranthene were isolated and identified (Fig. 7). With both bacterial strains, metabolism was consistent with cleavage of one of the fused aromatic rings of the fluoranthene molecule. A similar metabolic pathway has been noted in a fluoranthene-degrading Mycobacterium[27,28]. These researchers also identified one of the metabolites of fluoranthene as 9-fluorenone-1-carboxylic acid. Evidence for the fission of the fused ring by Rhodococcus sp., resulting in a dicarboxylic acid, is consistent with initial attack on the carbons of that ring followed by ortho ring cleavage. This Rhodococcus strain has been shown to attack anthracene by an analogous dual cleavage pathway . The presence of an ortho cleavage enzyme has also been identified in mycobacteria [30,31].
In contrast to the enhanced biodegradation rates observed in the case of cometabolism, inhibitory effects were observed in the present study, for example in the case of anthracene utilization decelerated in the presence of fluoranthene by Rhodococcus sp., and pyrene utilization also decelerated in the presence of fluoranthene by M. flavescens. Similar inhibitory effects have been noted by other researchers [11,12,21]. Stringfellow and Aitken  used degradation kinetics to elucidate the mechanism of inhibition of PAH utilization by two strains of Pseudomonas isolated from creosote-contaminated soil on phenanthrene as growth substrate. Kinetics consistent with competitive inhibition were observed when the degradation of phenanthrene was studied in combination with one of the following PAHs: naphthalene, 1-methylnaphthalene, 2-methylnaphthalene and fluorene. The analysis of Stringfellow and Aitken, along with the current data, suggests that PAH-degrading bacterial strains utilize a common enzyme for the degradation of two or more PAHs. The extent to which common enzymes are utilized for the degradation of PAHs within a single bacterial strain is not well documented. Enzymes of naphthalene degradation encoded on the NAH plasmid were found to be involved in the degradation of phenanthrene and anthracene in Pseudomonas[32,33].
Bioremediation has been frequently proposed as a method of removing PAHs from contaminated sediments and soils [1,34–36]. To be useful for bioremediation, bacterial strains must be capable of utilizing target PAHs in the presence of other PAHs as well as other contaminants such as phenols and heavy metals. While the present study has confirmed the phenomenon of simultaneous utilization of PAHs as components of mixtures, the results reported here also provide evidence of interactive effects which have significant implications for the biological removal of PAHs from contaminated soils and sediments. While cometabolic effects have the capacity to increase biodegradation rates of otherwise recalcitrant PAHs, inhibitory effects need to be taken into consideration, as they may reduce the effectiveness of bacterial strains in mitigating pollution.
This work was partially supported by a grant from the National Oceanographic and Atmospheric Administration through the Illinois–Indiana Sea Grant College program. The senior author thanks Kathy Myers, Denise Quance, and Kym Crandell for technical assistance. Additional support was provided by the Oak Ridge Institute for Science and Education Faculty Research Program at the National Center for Toxicological Research, Jefferson, AR, USA. The authors thank James. P. Freeman and Thomas M. Heinze for valuable advice and assistance.