Polycyclic aromatic hydrocarbons (PAHs) are ubiquitous environmental toxicants with various emission sources [1
]. Alkylated PAHs are more abundant than their unsubstituted forms in crude oil and refined petroleum products and are thought to be major contributors to the overall toxicity of spilled oil [4
]. However, the environmental fate and transformation of alkylated PAHs after spills or other contamination requires further elucidation.
In the environment, oils undergo various weathering processes (e.g., spreading, dispersion, dissolution, evaporation, emulsification, sinking, biodegradation, and photodegradation), resulting in changes of chemical composition. Biodegradation plays a very important role in the complete mineralization of hydrocarbons. According to calculations of the remaining oil in the Gulf of Mexico after the Deepwater Horizon oil spill, for example, it was estimated that hydrocarbon-degrading bacteria removed up to approximately 50% of the hydrocarbons [6
It is well-known that many aerobic bacteria degrade aromatic hydrocarbons as substrates [9
]. Many bacterial strains have been evaluated for their ability to degrade aromatic hydrocarbons including PAHs via one of the three methods: mineralization, co-metabolic transformation, and non-specific oxidation [10
]. Among these, Sphingomonas
sp. strains can degrade PAHs with 2–5 rings, which are major PAHs of toxicological concerns after spills, via non-specific oxidation utilizing enzymes such as dioxygenase or other oxygenase [11
]. Although the biodegradation of unsubstituted PAHs by microorganisms has been studied for decades [9
], aerobic biodegradation of alkylated derivatives has not been widely examined and studies have mainly focused on methylated naphthalenes [16
]. Because alkylated phenanthrenes and pyrenes are important toxicants in oil contamination, studies are needed to identify their biodegradation and transformation products during weathering processes. In a few studies, the biodegradation kinetics of alkylated PAHs were investigated together with their unsubstituted forms [23
]. Zhong et al. showed that microbial transformation of 1-methylphenanthrene by Sphingobium
sp. MP9-4 occurs by the simultaneous monooxygenation of the methyl group and dioxygenation on unsubstituted benzene ring [22
]. However, metabolic transformation studies have been conducted for only a few alkylated PAHs, whereas there are many other alkylated PAHs present in spilled oils [28
]. Further studies are needed to understand whether oxidation pathways in aerobic bacteria such as Sphingobium
sp. can be generalized to various alkylated PAHs.
It is generally accepted that oxygenated PAHs (oxy-PAHs) are generated by degradation of PAHs in aerobic environments and that they are more polar than unsubstituted PAHs. Oxy-PAHs have a higher water solubility and thereby higher bioavailability compared to unsubstituted PAHs [31
]. Although alkylated PAHs have very low water solubility, the transformation products are expected to be more soluble, resulting in increased exposure of organisms if water solubility limits chemical exposure. Moreover, the carcinogenicity of PAHs is mediated by their reactive transformation products. As concerns regarding environmental effects have increased, the ecotoxicity of oxy-PAHs including hydroxylated PAHs, aromatic acids, and ketone- and Quinone-substituted PAHs has been investigated [33
]. Fallahtafti et al. examined the chronic toxicity of 1-methylphenanthrene and its hydroxylated derivatives to Japanese medaka (Oryzias latipes
), showing that ring hydroxylation enhanced the toxicity of PAHs [35
]. Similar results were observed in developmental toxicity tests of zebrafish embryo by chrysene and its hydroxylated derivatives, revealing that hydroxylated PAHs are more toxic than unsubstituted PAHs [37
The main goals of this study were to (1) determine and compare the biodegradation kinetics of alkylated phenanthrenes and pyrenes with those of unsubstituted forms, (2) identify major biodegradation metabolites, and (3) evaluate the toxic potency and efficacy of identified metabolites. Two alkylated phenanthrenes, 3-methylphenanthrene (3MPhe) and 3,6-dimethylphenanthrene (36DMPhe), and 1-methylpyrene (1MP) were chosen as model alkylated PAHs and the bacterium Sphingobium quisquiliarum EPA505 was used to degrade PAHs. The biodegradation kinetics of alkylated PAHs and their parent PAHs was assessed in batch experiments. The main metabolites of selected alkylated PAHs were enriched by high-performance liquid chromatography (HPLC) fractionation and identified by gas chromatography-mass spectrometry (GC-MS). The toxicity of metabolic fractions was also measured by determining the luminescence inhibition of Aliivibrio fischeri as an ecotoxicity endpoint and compared to that of parent compounds.
2. Materials and Methods
GC-grade 3-methylphenanthrene (98%), 3,6-dimethylphenanthrene (95%), and 1-methylpyrene (94%) were purchased from Tokyo Chemical Industry Co. (Tokyo, Japan). Analytical-grade phenanthrene (97%), pyrene (99%), fluoranthene (98%), dimethyl sulfoxide (DMSO; 99.5%) and a solution of N-tert-butyldimethylsilyl-N-methyltrifluoroacetamide with 1% tert-butyldimethylchlorosilane (MTBSTFA, ≥95%) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Bioreagent-grade nitrilotriacetic acid (≥99.0%), potassium phosphate dibasic (≥99.0%), sodium phosphate dibasic (≥99.0%), calcium chloride dihydrate (≥99.0%), iron(ΙΙ) sulfate heptahydrate (≥99.0%), magnesium sulfate heptahydrate (≥99.0%), ammonium molybdate tetrahydrate (≥99.0%), zinc sulfate heptahydrate (≥99.0%), copper(ΙΙ) chloride dihydrate (≥99.0%), Ethylenediaminetetraacetic acid disodium salt dihydrate (98.5–101.5%), cobalt(ΙΙ) chloride hexahydrate (98–102%), sodium borohydride (99%), manganese(ΙΙ) sulfate hydrate (≥99.0%) and ammonium sulfate (≥99.0%) were purchased from Sigma-Aldrich (St. Louis, MO, USA). The medical-grade polydimethylsiloxane (PDMS) sheet (1 mm thickness, density of 1170 kg m−3) was purchased from Specialty Silicone Products, Inc. (Ballston Spa, NY, USA). Methanol (HPLC-grade) was purchased from Honeywell Burdick & Jackson (Ulsan, Korea). Acetonitrile (HPLC ultra-gradient solvent) was purchased from Avantor Performance Materials, Inc. (Center Valley, PA, USA).
The minimal salts medium was prepared by adding following three solutions per liter: (1) 40 mL of buffer solution containing 134 g of Na2
and 68 g of KH2
per liter (pH adjusted to 7.25 using KOH), (2) 10 mL of mineral base containing 10 g of nitrilotriacetic acid, 14.45 g of MgSO4
O, 3.33 g of CaCl2
O, 0.00925 g of (NH4
O, 0.099 g of FeSO4
O, and 50 mL of Metal “44” solution per liter (pH adjusted to 7.25 using KOH), and (3) 10 mL of 100 g L−1
2.2. Culturing of Bacteria
DSM 7526 (strain EPA505) was selected as the test species for biodegradation tests and was purchased from DSMZ (Braunschweig, Germany). A custom-cut PDMS sheet (5 × 5 cm) was used to supply fluoranthene to the bacterial growth medium by passive dosing [42
]. To saturate the PDMS with fluoranthene, PDMS sheets were submerged in methanol solution containing excess crystals of fluoranthene in a glass bottle. The bottle was placed on a horizontal shaker at 150 rpm and 25 °C for 24 h. After shaking, the PDMS sheets were removed from the methanol solution and air-dried in a clean bench. A frozen supply of S. quisquiliarum
was plated onto LB agar plates and incubated at 28 °C. After incubation, a single colony was picked and inoculated into a glass bottle containing 50 mL of minimal salt medium and a fluoranthene-loaded PDMS sheet. The bottle was closed loosely with a cap and incubated at 25 °C and 120 rpm for 4 days. After incubation, 0.8-mL aliquots of medium containing the grown biomass were dispended into a sterile external cryovial (Thermo Fischer Scientific, Waltham, MA, USA). The medium was then mixed with 0.8 mL of glycerol and stored at −80 °C until use.
2.3. Biodegradation Kinetics
Thirty milliliters of minimal salts medium were added to a glass bottle containing fluoranthene-saturated PDMS and inoculated with 200 μL of a thawed aliquot of strain EPA505. The bottle was loosely closed with a cap and placed on a horizontal shaker at 25 °C and 120 rpm. After 4 days, the grown biomass suspension was diluted to 1/100 with minimal salts medium. This suspension was also serially diluted, plated on LB agar plates, and incubated for 3 days for colony counting.
For biodegradation tests, an individual compound dissolved in DMSO was spiked into 1 L of the strain EPA 505 culture (2.54 × 107
and 5.78 × 107
for phenanthrenes and pyrenes, respectively) at the initial concentration close to its aqueous solubility limit (Table 1
). The DMSO content in the medium was 0.1% by volume. After mixing the spiked medium, 5 mL of the sample was collected and added to a 20 mL glass vial with a Teflon-coated screw cap. The biodegradation test was initiated by placing the test vials on a rotary shaker (200 rpm) at 25 °C in the dark. All samples were prepared in triplicate, and the negative control prepared with sterilized cells was included in all tests. Three vials were used to detect the concentrations of remaining compounds at 0, 1.5, 3, 4.5, and 6 h after initiating the biodegradation test for phenanthrene, 3-methylphenanthrene, and 3,6-dimethylphenanthrene, and at 0, 9, 12, 15, 18, and 24 h after initiation for pyrene and 1-methylpyrene. Bacterial biomass at the end of the tests was counted to assure that bacterial activity was not altered during the tests. At each sampling time, the whole medium was extracted using 3 × 20 mL hexane:ethylacetate (7:3, v
). The extract was then passed through a sodium sulfate layer and regenerated cellulose membrane filter (pore size 0.2 µm, diameter 47 mm, chmlab group, Barcelona, Spain). The filtered extract was evaporated using a rotary evaporator to approximately 2 mL and the remaining solution was evaporated under a gentle nitrogen gas stream to complete dryness. The dried residue was re-dissolved in 1 mL methanol for ultra-performance liquid chromatography (UPLC) analysis.
The concentrations of PAHs and alkylated PAHs was quantified using an Acquity UPLC® (Waters, Milford, MA, USA) equipped with a Sample manager-FTN, Quaternary Solvent Manager, and FLR Detector. The mobile phase in isocratic mode was 70% acetonitrile and 30% water at a flow rate of 0.2 mL min−1 at ambient temperature. The target compounds were separated on an Acquity UPLC® BEH C18 column (2.1 × 50 mm, 1.7 µm particle size; Waters) at 40 °C. PAHs and alkylated PAHs were detected using the fluorescence detector with excitation (λex) and emission wavelengths (λem) of 260 and 352 nm for phenanthrene, 3MPhe, and 3,6DMPhe and 260 and 420 nm for pyrene and 1MP.
In quantitative analysis of target compounds, the standard and blank samples were analyzed every 15 samples for quality control. No target peaks were observed in all blank samples and the relative standard deviations of peak areas for chemical standards during the analysis were 4.42%, 4.18%, 6.33%, 3.11% and 4.47% for phenanthrene, 3MPhe, 3,6DMPhe, pyrene, and 1MP, respectively.
The extraction recoveries at solubility level were 82.0 ± 1.4%, 90.3 ± 3.5%, 83.7 ± 1.5%, 94.3 ± 2.5%, and 88.3 ± 1.9% for phenanthrene, 3MPhe, 3,6DMPhe, pyrene, and 1MP, respectively.
2.4. Fractionation of Transformation Products
Following the biodegradation kinetic tests, the metabolic products of 3,6-dimethylphenanthrene and 1-methylpyrene were qualitatively identified in the UPLC chromatograms. To evaluate the toxicity of metabolic products based on the luminescence inhibition of A. fischeri
, the column effluent containing metabolites was enriched. A 1-L bottle containing 800 mL medium and S. quisquiliarum
spiked with 3,6-dimethylphenanthrene or 1-methylpyrene at their aqueous solubility levels (Table 2
) was placed on a rotary shaker (200 rpm) at 25 °C for at least for 24 h. After incubation, the solution was passed through an Oasis®
HLB (200 mg) cartridge (Waters). The loaded cartridge was eluted with 4 mL methanol for metabolites from 3,6-dimethylphenanthrene or 4 mL dichloromethane for metabolites from 1-methylpyrene. The eluent was filtered through a 0.2-µm regenerated cellulose membrane filter (chmlab group), and then re-dissolved in 4 mL methanol after evaporation of the extraction solvent using a rotary evaporator and nitrogen gas stream for 3D-scanning (fixed λex
at 260 nm) with a fluorescence detector (FLD). Based on the peaks identified by 3D-scanning, the eluent was further fractionated using a Waters HPLC system equipped with a Waters 717+ autosampler, Waters 2475 multi λ fluorescence detector, and two Waters 515 HPLC pumps. Fifty microliters of eluent were injected into the HPLC system, and the metabolites were separated on a Thermo C18 column (4.6 × 150 mm, 5 μm particle size; Thermo Fisher Scientific). The mobile phase in isocratic mode was 90% acetonitrile and 10% water (v/v
) at a flow rate of 1 mL min−1
at ambient temperature.
The effluent fraction containing metabolite peaks under FLD was collected. The fractionation procedure was repeated 40 times to obtain a sufficient mass of metabolites for chemical identification. The combined column effluent was passed through a sodium sulfate layer, followed by evaporation of the solvent using a rotary evaporator and a gentle nitrogen stream, and finally re-dissolved in 2 mL ethyl acetate for GC-MS analysis.
2.5. Identification of Metabolic Products of Alkylated PAHs
To identify metabolites of 3,6-dimethylphenanthrene and 1-methylpyrene, the samples were derivatized with MTBSTFA, which is frequently used to detect phenolic metabolites. One milliliter of ethyl acetate solution containing the metabolites was collected into a 2-mL vial. After evaporation of the ethyl acetate under a gentle nitrogen stream, the residue was reacted by spiking 20 µL MTBSTFA solution in a water bath at 60 °C for 60 min. After the reaction was complete, 480 μL ethyl acetate was added for GC-MS identification. Metabolites of alkylated PAHs were identified using an Agilent Technologies model 7890 gas chromatograph (GC) with a HP-5MS capillary column (30 m, 0.25 mm, 0.25 µm; Agilent Technologies), coupled with a model 5975 mass spectrometer (MS). The GC column temperature was held at 100 °C (1 min), increased to 160 °C at a rate of 15 °C min−1, increased to 300 °C at a rate of 4 °C min−1, and then held for 2 min. The inlet was held at 280 °C and used in splitless mode. Helium was used as a carrier gas at a rate of 1 mL min−1. The GC-MS interface temperature was maintained at 280 °C. The MS was used in electron impact mode (70 eV), and scans ranged from 50 to 500 m/z. The ion source and mass filter temperatures were held at 230 °C and 150 °C, respectively. Alkylated PAHs metabolites were identified by comparison with authentic and/or general standards and the literature. The metabolites identified in the fractionated samples were not detected in controls.
2.6. Luminescence Inhibition Assay
A Microtox® Model 500 system (Strategic Diagnostics, Inc., Newark, DE, USA) was used to measure the luminescence inhibition of A. fischeri by detecting the metabolic products of 3,6-dimethylphenanthrene and 1-methylpyrene. Freeze-dried A. fischeri and saline diluent (2% NaCl) were purchased from Strategic Diagnostics, Inc. Two hundred microliters of ethyl acetate concentrate were added to a 4-mL glass vial, and the solvent was evaporated under a gentle nitrogen stream. The dried residue was dissolved in 2 mL of HPLC-grade water and this dissolved mixture solution was used for the Microtox® assay without further treatment. Luminescence inhibition after 15 min of exposure was measured according to the manufacturer’s protocol. Dose-response curves were derived using Microtox® Omni software (Strategic Diagnostics, Inc.).
The biodegradation rate constants of (alkylated) phenanthrenes were much faster than (alkylated) pyrenes. This tendency was also observed in a previous study comparing the biodegradation of phenanthrene, 1-methylphenanthrene, 2-methylphenanthrene, 3,6DMPhe, pyrene, and 1MP by S. quisquiliarum
(DSM 7526), showing a higher biodegradation rate in phenanthrene and their alkylated forms than that for pyrenes [24
Earlier studies showed conflicting effects of alkylation of phenanthrene and pyrene on the biodegradation rate constants. Whereas the biodegradation rate of 1-methylphenanthrene by Sphingobium
sp. MP9-4 was slower than that of phenanthrene [22
], the rate constants were increased by methylation for 1-methylphenanthrene (3.24 ± 1.44 L mgprotein−1
) and 2-methylphenanthrene (3.73 ± 1.56 L mgprotein−1
) and decreased for 36DMPhe (1.56 ± 0.36 L mgprotein−1
) compared to unsubstituted phenanthrene (2.09 ± 0.79 L mgprotein−1
For pyrene and 1-methylpyrene, the literature values are 1.10 ± 0.31 L mgprotein−1
and 0.82 ± 0.50 L mgprotein−1
, respectively [24
], indicating no significant difference between the biodegradation rate constants of these compounds. The biodegradation rate constant decreased with alkylation for phenanthrenes whereas the rate constant for 1-methylpyrene was greater than that of pyrene in this study. It has been suggested that the main site for dioxygenation of phenanthrene by Sphingomonas
sp. is the 5,6-C site (or 3,4-C site which are equivalent; Figure S8a, Supplementary Material
]. In 3-methylphenanthrene, however, the 3,4-C site is hindered by methylation, resulting in inhibition of dioxygenation-derived metabolism. In 3,6-dimethylphenanthrene, both dioxygenation sites are hindered by methyl groups, resulting in a significant decrease in the biodegradation rate constants compared to the unsubstituted form. The observed trend in the biodegradation kinetics for phenanthrene, 3-methylphenanthrene, and 3,6-dimethylphenanthrene agrees well with this hypothesis. In case of pyrene, no site preference was observed in the metabolism because of its high symmetry (Figure S8b, Supplementary Material
). Diverse metabolic pathways by methylation of pyrene may result in a higher biodegradation rate constant of 1MP than that of pyrene. The identification of the biotransformation products of 36DMPhe and 1MP also supported this hypothesis as explained below. However, it is also well-known that co-factors such as metal ions included in the culture medium influence on the observed bacterial or enzymatic degradation rate and transformation pathways [55
]. Further investigations on the mechanism and pathways of the transformation of alkylated PAHs would explain the differences in biodegradation rates among PAHs and their alkylated homologues.
Two isolated fractions for the metabolic products of 36DMPhe were identified as monooxygenated form of 36DMPhe (36DMPhe_F2) and monooxygenated form of 36DMP_F2 (36DMPhe_F1), respectively. This indicates that monooxygenation of the methyl group could initiate the degradation of 36DMPhe by S
. In bacterial degradation of 1MP by the strain MP9–4, monooxygenation of the methyl group occurred with dioxygenation on carbon atoms in benzene ring simultaneously [22
]. It was observed that unsubstituted positions in benzene ring was preferentially attacked in bacterial degradation of 1-methylphenanthrene [22
]. For 36DMPhe in this study, however, monooxygenation of the methyl group appeared to be the major biodegradation pathway because the 5,6-C site (or equivalent 3,4-C site), which is known as the main site for dioxygenation by Sphingobium
], is hindered by methylation. The isolated fraction of the metabolic products of 1MP, 1MP_F, was also identified as the monooxygenated form of 1MP, confirming that monooxygenation of the methyl group also occurred during bacterial degradation of 1MP. This indicates that the alkyl-substitution on pyrene could serve multiple degradation pathways and make the biodegradation faster than unsubstituted pyrene.
The luminescence inhibitions of A. fischeri
by the isolated fractions from biodegradation products of 36DMPhe and 1MP were detectable despite non-observable toxicity for their original compounds, indicating that the ecotoxicity of alkylated PAHs could be enhanced via biodegradation processes. This increase in toxicity can be explained by the increased solubility of the compounds via biotransformation or by intrinsic toxicity of transformation products. The median effect concentrations (EC50
) of the identified transformation products were derived using quantitative structure-activity relationships developed for predicting the baseline toxicity of petroleum hydrocarbons in our previous study [57
] with log Kow
values estimated by EPISuiteTM
. These values were 0.39, 0.49 and 1.5 μM for 36DMPhe_F1, 36DMPhe_F2, and 1MP_F, respectively. The toxicity ratios (TR) were calculated by dividing the EC50
values derived from the quantitative structure-activity relationships prediction by the measured EC50
in this study. These values were 0.47, 4.61, and 1.23 for 36DMPhe_F1, 36DMPhe_F2, and 1MP_F, respectively. It is generally accepted that chemicals with a TR < 5 cause narcosis [58
]. Although all tested fractions could be classified as baseline toxicants, the fraction 36DMPhe_F2 may have a specific toxic mode-of-action. The luminescence inhibition of A. fischeri
by aromatic acids showed that the toxicity increased with increasing hydrophobicity, which is the general tendency observed for baseline toxicants [33
]. Previous studies compared the toxicity of PAHs and their oxygenated derivatives [35
]. Enhanced chronic toxicity in the early life stage of Oryzias latipes
was observed for 1-methylphenanthrene by hydroxylation, as hydroxylated 1-methylphenanthrene derivatives were 4-fold more toxic than 1-methylphenanthrene [35
]. The developmental toxicity of chrysene and its hydroxylated derivatives tested using zebrafish embryos showed similar results, indicating that the photochemical or microbial transformation of PAHs enhances the toxicity of PAHs [36
]. This suggests that both the toxic potency and efficacy of metabolic products of alkylated PAHs could be greater than their parent compounds because of the greater inherent toxicity of metabolites and/or enhanced water solubility.