Environ Sci Pollut Res

DOI 10.1007/s11356-014-2890-z

PAHS AND FISH – EXPOSURE MONITORING AND ADVERSE EFFECTS – FROM MOLECULAR TO INDIVIDUAL LEVEL

Effects of an oil spill in a harbor assessed using biomarkers

of exposure in eelpout
Joachim Sturve & Lennart Balk & Birgitta Liewenborg &
Margaretha Adolfsson-Erici & Lars Förlin &
Bethanie Carney Almroth

Received: 18 September 2013 / Accepted: 4 April 2014

# The Author(s) 2014. This article is published with open access at Springerlink.com

Abstract Oil spills occur commonly, and chemical com- Introduction

pounds originating from oil spills are widespread in the aquat-
ic environment. In order to monitor effects of a bunker oil spill Oil spills occur commonly, and chemical compounds origi-
on the aquatic environment, biomarker responses were mea- nating from oil spills, such as polycyclic aromatic hydrocar-
sured in eelpout (Zoarces viviparus) sampled along a gradient bons (PAHs), are widespread in the aquatic environment
in Göteborg harbor where the oil spill occurred and at a (Beyer et al. 2010; de Hoop et al. 2011). Aquatic organisms
reference site, 2 weeks after the oil spill. Eelpout were also protect themselves against the harmful effects of exposure to
exposed to the bunker oil in a laboratory study to validate field these and other xenobiotics via molecular and cellular defense
data. The results show that eelpout from the Göteborg harbor systems, such as detoxifying enzymes and molecules, metal-
are influenced by contaminants, especially polycyclic aromat- binding proteins, and trapping of foreign toxic compounds by
ic hydrocarbons (PAHs), also during “normal” conditions. lysosomes. These responses, as well as any cellular or molec-
The bunker oil spill strongly enhanced the biomarker re- ular damage that may occur as a result of exposure, are often
sponses. Results show elevated ethoxyresorufin-O-deethylase used in monitoring and assessment programs addressing the
(EROD) activities in all exposed sites, but, closest to the oil environmental impact of pollutants (van der Oost et al. 2003).
spill, the EROD activity was partly inhibited, possibly by Interactions between pollutants and biochemical and physio-
PAHs. Elevated DNA adduct levels were also observed after logical functions in fish, detected as subcellular, cellular, or
the bunker oil spill. Chemical analyses of bile revealed high organ disturbances, are referred to as biomarkers and can
concentrations of PAH metabolites in the eelpout exposed to serve as early warning signs, indicating possible disturbances
the oil, and the same PAH metabolite profile was evident both in reproduction success, growth, or survival of the fish (Forlin
in eelpout sampled in the harbor and in the eelpout exposed to et al. 1986; Haux and Forlin 1988).
the bunker oil in the laboratory study. Many of the compounds found in oil, for example PAHs,
are oxidatively metabolized in fish by detoxifying enzymes in
the phase I cytochrome P450 system, also called the CYP
Keywords Oil spill . Polycyclic aromatic hydrocarbons . system. The CYP1A subfamily belongs to the superfamily of
Eelpout . Biomarkers . EROD . DNA adducts CYPs, and is important in the phase I detoxification reactions
in fish. The phase I metabolites (e.g., hydroxylated PAHs) are
further metabolized by phase II systems to form water-soluble
conjugates that are excreted into the bile (Leonard and Hellou
Responsible editor: Philippe Garrigues 2001). The metabolites can be detected by high-performance
J. Sturve (*) : L. Förlin : B. Carney Almroth
liquid chromatography (HPLC)/fluorescence or by gas
Department of Biological and Environmental Sciences, University of chromatography/mass spectrometry (GC/MS) after hydrolysis
Gothenburg, Box 463, SE-405 30 Göteborg, Sweden of the conjugates. The bioconcentration factor bile/water can
e-mail: joachim.sturve@bioenv.gu.se be up to 106 in fish, which means that low concentrations of a
L. Balk : B. Liewenborg : M. Adolfsson-Erici
contaminant in the aquatic environment can be detected via
ITM, Department of Applied Environmental Science, Stockholm metabolites in the bile. Identification of PAH metabolites in
University, SE-106 91 Stockholm, Sweden the bile can be used to distinguish between petrogenic and
 Environ Sci Pollut Res

pyrogenic PAH exposure. Petroleum PAH are often dominat- of the Gothenburg archipelago. Eelpout were sampled before,
ed by alkylated two- and three-ringed aromatics, while pyro- and 2 weeks and 5 months after the oil spill and levels of
genic PAH are dominated by four- and five-ringed aromatics CYP1A protein, hepatic EROD activity and DNA adducts
(Anderson and Lee 2006). Many PAHs, or their metabolites, were measured. We also measured the amounts of PAH me-
are known to be toxic and/or carcinogenic (Aas et al. 2001), tabolites in bile as their hydroxylated transformation products
and PAHs are considered the most toxic of all petroleum as well as bile fluorescent aromatic compounds (FAC), and
compounds (Yanik et al. 2003). The induction of CYP1A in verified their use as biomarkers of exposure.
fish, usually measured as increased activity of
ethoxyresorufin-O-deethylase (EROD), is considered to be
the most responsive and consistent biomarker for aryl hydro- Material and methods
carbon (AH)–receptor ligands, such as PAHs and planar PCBs
(Goksøyr and Förlin 1992). However, the relationship be- Chemicals
tween exposure to PAHs and consequent induction of
EROD activity is not always positively correlated. High levels Sodium acetate, acetic acid, and potassium carbonate, all of
of PAHs have been shown in some cases to inhibit EROD analytical grade (Merck, Darmstadt, Germany), n-hexane
activity (Schiedek et al. 2006). PAHs are also known to be “LiChrosolve” (Merck, Darmstadt, Germany), acetic anhy-
genotoxic, and they can be bioactivated via the cytochrome dride of analytical grade (Riedel-de Haen, Seelze, Germany),
P450 system, resulting in more toxic intermediate metabolites and β-glucuronidase (102 100 units/mL of β-glucuronidase
(Baird et al. 2005). These metabolites are toxic and effects and 290 units/mL of sulfatase) (Sigma-Aldrich Chemie
include DNA adducts which can be quantified (Lyons et al. Gmbh, Steinheim, Germany), methyl-t-butyl ether of HPLC
2004; Amat et al. 2006; Malmström et al. 2009). PAHs are grade (Rathburn Chemicals Ltd, Walkerburn, UK). Analytical
hydrophobic and semivolatile, factors that contribute to their standards 2-OH-naphthalene, 1-OH-fluorene, 1-OH-
accumulation in sediments and biological tissues and to their phenanthrene, 1-OH-pyrene (Sigma-Aldrich Chemie Gmbh,
persistency in the environment (Yanik et al. 2003), and the Steinheim, Germany), and β,β′-binaphthyl were purchased
major degradation pathways, which are microbial, are highly from Larodan Fine Chemicals AB, Malmö, Sweden. Ethanol,
dependent upon environmental conditions (Haritash and acetone, and peanut oil were all of analytical grade.
Kaushik 2009).
Previous studies have addressed the effects of an oil spill on Field samplings
eelpout (Frenzilli et al. 2004; Carney Almroth et al. 2005), and
have shown oxidative stress and DNA damage as a result. The Eelpout were captured with fyke nets by local fishermen at
eelpout (Zoarces viviparus) is a viviparous species of blenny that four sites around the River Göta älv estuary, Göteborg. The
is commonly used in European environmental biomonitoring sites were Nordre älv, a local reference site; Hjuvik in the outer
campaigns due to the fact that its range reaches from Norway to harbor; Skalkorgarna, the site located within the oil harbor;
northern France including also the Baltic Sea (Whitehead et al. and the inner harbor site, Aspholmarna. Fish were also taken
1984). Several additional characteristics of this species make it from the national reference site, Fjällbacka, 150 km north of
suitable for field studies, e.g., the fish are relatively stationary, the River Göta älv estuary. Figure 1 shows a map displaying
making it possible to correlate exposure conditions to physio- the sampling sites. Fish were captured at three time points.
logical responses in wild fish, and the viviparous nature of The first sampling was prior to the spill in May when samples
their reproduction allows for studies addressing reproduction were taken as part of a regional biomonitoring campaign.
and maternal effects. This species has been used in Swedish Biomarker results from this sampling campaign have already
biomonitoring programs for more than 20 years as a sentinel been presented in Sturve et al. (2005). Samples were also
species for environmental pollution (Vetemaa et al. 1997; collected in July (2 weeks after the oil spill) and in
Ronisz et al. 1998; Larsson et al. 2000; Frenzilli et al. 2004; November, approximately 5 months after the oil spill. In
Carney Almroth et al. 2005; Ronisz et al. 2005; Sturve et al. July, sampling had to be repeated after 4–5 days at the
2005; Gercken et al. 2006), resulting in considerable amounts Nordre älv and Aspholmarna sites due to low catches at the
of data concerning its basic physiology, responses to pollutant first fishing time point. Due to practical reasons, it was not
exposure, and consistent information describing the reference possible to take samples from the national reference site,
sites. Fjällbacka, in July.
In this study, we have addressed the effects of expo-
sure to bunker oil on feral eelpout. In June 2003, Laboratory studies
between 10 and 100 t of bunker oil containing around 25 %
PAHs were accidentally released from a storage facility in Two laboratory studies were performed with bunker oil from
Göteborg harbor in Sweden and spread into the outer parts the batch that was spilled into the harbor. The oil was a gift
Environ Sci Pollut Res

Fig. 1 Map of field sampling sites, showing two reference sites (1 Fjällbacka, national reference site; 2 Nordre Älv, local reference site) and three sites
within the harbor (3 Hjuvik, 4 Skalkorgarna, 5 Aspholmarna)

from the Västra Götaland County administration. Eelpout for The second exposure study aimed to study DNA adducts
the laboratory studies were collected from Grundsund, formation and was therefore longer. Ten female eelpout were
Sweden, with the help of local fishermen. This site is located injected intraperitoneally (i.p.) with 100 mg bunker oil/kg fish.
65 km north of Göteborg city and is considered to be an After 10 days, five of the fish were re-injected with the same,
unpolluted site. The fish were transported to the Department resulting in two exposure groups (five fish per group), injected
of Biological and Environmental Sciences, Gothenburg once (Inj 1) and injected twice (Inj 2). The control group (five
University, and kept in aerated and filtered sea water (32 ‰) fish) was injected with the carrier alone (peanut oil). All of the
at 12 °C and 12:12 light cycle for up to 2 weeks for fish were sampled after 21 days. Five fish were sampled
acclimatization. immediately when arriving to the laboratory, thus forming
In the first exposure study, eelpout was exposed to the oil an extra control group (0 sample).
through water exposure for 96 h. Female eelpout (n=36) were
divided into four 600-L tanks (giving 1 g of fish per liter of Sampling
water) with aerated seawater at approximately 12 °C and
exposed to the oil in a static system. Besides the control group, Eelpout were killed by a sharp blow to the head and the weight
the fish were exposed to three different doses of the oil (10, and length was recorded. The fish were cut open, bile was
100, and 1,000 μg/L), dissolved in 60 mL of acetone. The collected with a syringe, and the liver excised, weighed, and
control group was exposed to 60 mL of acetone alone. frozen in liquid nitrogen. Before freezing, the liver was
 Environ Sci Pollut Res

divided into two pieces; one for enzymatic analysis and one Bile fluorescent aromatic compounds
for DNA adduct analysis. Bile samples from female fish were
pooled (five fish per pool) in order to obtain sufficient vol- Synchronous fluorescence spectrometry of PAH metabolites
umes. The pooled bile samples were put in dark glass bottles, in bile was performed following the protocol described by Aas
frozen on dry ice, and kept in −20 °C until analysis. et al. (2000). Individual bile samples were diluted in 1:800 in
48 % ethanol with further dilution (1:1,600, 1:3,200, 1:6,400,
or 1:12,800) when necessary and submitted to synchronous
Biochemical analysis
fluorescence scan with Δλ=42 nm. The peak corresponding
to pyrene-like compounds (emission wavelength 383) was
The microsomal fraction was obtained following the protocol
quantified by integrating and calculating the peak area be-
described by Forlin et al. (1984). Livers were homogenized
tween the emission wavelengths 365 to 400. PAH metabolite
using glass/Teflon® for 3×5 s in four volumes of homogeni-
levels are expressed as arbitrary fluorescence.
zation buffer (0.1 M Na+/K+-phosphate buffer (pH 7.4) con-
taining 0.15 M KCl). The homogenate was centrifuged in two
Speciation of PAH metabolites in bile
steps, first 10,000×g for 20 min at 4 °C followed by 105,000×
g for 60 min at 4 °C. The supernatant (cytosol fraction) was
Approximately 200 mg bile was enzymatically hydrolyzed by
collected and the pellet (microsomal fraction) resuspended in
adding 20 μL β-glucuronidase (102,100 units/mL) in 1 mL
homogenization buffer containing 20 % glycerol and stored at
0.2 M acetic acid buffer, pH 5, and incubated for 2 h at 37 °C.
−80 °C until analysis.
Free compounds were extracted with 2×3 mL hexane/methyl-
EROD activity was measured in the microsomal fraction
t-butyl-ether (1:1) after addition of 2 mL water and 0.5 g
according to a spectrofluorometric method described by
NaCl. The combined organic phases were evaporated to dry-
Forlin et al. (1984) using rhodamine as standard.
ness; the residue was dissolved in 2 mL hexane. For acetyla-
CYP1A levels were determined in the microsomal fraction
tion of the hydroxyl groups, 100 μL acetic acid anhydride/
with an enzyme-linked immunosorbent assay (ELISA) ac-
pyridine (1:1) was added, and the mixture was heated at 60 °C
cording to Ronisz and Förlin (1998). Due to the lack of
for 30 min. The combined organic phases were concentrated
standard CYP1A protein, the results are presented as absor-
and analyzed by GC/MS-selected ion monitoring after addi-
bance and not as absolute CYP1A levels, giving relative
tion of internal standard (100 μL β,β′-binaphthyl). For quan-
differences between groups.
titation purposes, the following ions were used: 115, 144 (2-
Protein levels in the microsomal fractions were determined
OH-naphthalene), 165, 194 (1-OH-phenanthrene), 189, 218
according to the method described by Lowry et al. (1951)
(1-OH-pyrene), and 182 (2-OH-fluorene). The acetylated ex-
using bovine serum albumin as standard.
tracts were also analyzed by full-scan gas chromatography/
mass spectrometry for identification of other hydroxylated
DNA adduct measurement compounds. By interpretation of mass spectra, and by com-
paring retention times in the GC/MS chromatogram, 10 meth-
Deep-frozen liver tissue from blenny were semithawed, and the ylated hydroxylated PAH metabolites were tentatively identi-
DNA extracted and purified according to previous reports (Dunn fied. An estimation of their concentrations were made using
et al. 1987; Reichert and French 1994; Ericson and Balk 2000) the following ions: 158, 200 (M+) for CH3-OH-naphthalenes,
and slightly modified as described previously (Ericson et al. 172, 214 (M+) for (CH3)2-OH-naphthalenes, 186, 228 (M+)
1998; 2000). DNA adducts were enriched using the Nuclease for (CH3)3-OH-naphthalenes, 196, 238 (M+) for CH3-OH-
P1 method, 0.8 μg nuclease P1/μg DNA, and a 45 min incuba- fluorenes, 210,252 (M+) for (CH3)2-OH-fluorenes, 208, 250
tion period (Reddy and Randerath 1986; Beach and Gupta (M+) for CH3-OH-phenanthrenes, 222, 264 (M+) for (CH3)2-
1992). Finally, the DNA adducts were radiolabelled using OH-phenanthrenes, 232, 274 (M+) for CH3-OH-pyrenes, and
5′-[γ-32P]triphosphate([γ-32P]ATP) and T4 polynucleotide ki- 246, 288 (M+) for (CH3)2-OH-pyrenes. The extracts were
nase (Aas et al. 2000). Separation and clean-up of adducts was analyzed in splitless mode on a Hewlett-Packard 5890 series
performed by multidirectional thin layer chromatography (TLC) II GC (Avondale, PA, USA) coupled with a JEOL low-
on laboratory produced polyethyleneimine cellulose sheets, de- resolution automass with electron ionization, 70 eV
scribed as suitable for adducts formed from large hydrophobic (Stoughton, WI, USA). The ion source was 200 °C, and the
xenobiotics, such as four to six ring, PAHs (Reichert and French interface was 250 °C. Helium was used as carrier gas. The
1994; Ericson et al. 1998, 1999). column (30 m×0.25 mm MSDB5 with a phase thickness of
In addition, several quality control experiments were per- 0.25 μm, J&W Scientific, Folsom, CA, USA,) was held at
formed in parallel to the analysis of the eelpout samples. All 90 °C for 1 min, then quickly raised to 200 °C followed by an
these quality assurance experiments strongly suggested a increase of 10 °C/min up to 300 °C. The injector temperature
faultless assay for the DNA adduct measurements. was 275 °C. Two pools per site were analyzed for the field
Environ Sci Pollut Res

samples while only one pool per exposure condition was

analyzed in the laboratory exposure study. For the Nordre
älv and Aspholmarna sites, one pool per sampling time point
was selected for analysis.

Statistical analyses

Data was analyzed with one-way analysis of variance

(ANOVA) followed by Student–Newman–Keuls test using
the software SPSS® version 18 for Windows. Data not
displaying homogeneity of variance (Leven’s test) were log
transformed prior to testing. Data are presented as mean±
standard error (SEM), and the significance level was set at
p<0.05.

Results

Laboratory studies

EROD activity

EROD activities showed a dose-dependent elevation after

exposure to the crude oil in the laboratory, all doses resulting
in significantly elevated EROD activities compared to the
control. The low dose resulted in 3 times elevation, the middle
dose 18 times elevation, and the high dose 72 times elevation
in EROD activity (Fig. 2a).

CYP1A levels

Hepatic CYP1A levels in eelpout were significantly elevated

in the middle and the high dose compared to the control group
after oil exposure in the laboratory (Fig. 2b).

DNA adduct levels

DNA adduct levels in the livers were analyzed in eelpout

exposed to the oil for 3 weeks through IP injections. Eelpout Fig. 2 Results from the 4-day laboratory water exposure study using
from both control groups, 0 sample and carrier control, were bunker oil. EROD activities (a), levels of CYP1A protein (b), and PAH
metabolites (FACs) in bile (c). Fish were exposed in four groups, control,
found to have similar levels of DNA adducts (2.6±1.1 and 2.9 low dose (10 μg/L), medium dose (100 μg/L), and high dose
±0.8, respectively). Also, eelpout receiving a single or two (1,000 μg/L). Results are shown as mean±standard error, n=8. Letters
doses of bunker oil were found to have similar levels of DNA (a, b, c) indicate statistical differences, p<0.05
adducts (16.4±9.3 and 15.2±4.9, respectively). Both groups
exposed to bunker oil showed statistically significant eleva-
tion in DNA adduct levels compared to both control groups high exposures resulted in significantly elevated levels of
(data not shown). Units for the DNA adducts are nmol FAC metabolites in bile compared to the control (Fig. 2c).
adducts/mol normal nucleotides.
Hydroxylated PAH metabolites
FACs
The concentration of hydroxylated PAH metabolites was an-
FAC levels in bile of eelpout exposed to oil showed similar alyzed in bile from the laboratory exposure, and the results are
dose-dependent pattern as EROD activity. The middle and displayed in Table 1. After the 3-week long exposure study via
 Environ Sci Pollut Res

i.p. injection, the total amount was approximately 35 times levels were measured in fish from Nordre älv. All sites had
higher in the exposed fish compared to the control. Results significantly higher FAC levels in July compared to the same
from the water exposure study show a dose–response relation- sites in May (Fig. 3c), and levels at Aspholmarna were
ship with approximately 8 times higher levels of hydroxylated highest. Four months after the spill, the amounts of FACs
PAH metabolites in the fish exposed to the low dose, approx- were similar to those present prior to the spill.
imately 97 times higher in the middle dose and approximately
600 times higher in the high dose.
Hydroxylated PAH metabolites
Field study
Results from the chemical analysis of hydroxylated naphtha-
EROD activity lenes, pyrenes, fluorenes, and phenanthrenes of different de-
gree of methylation in bile from eelpout captured at the
EROD activities showed significant differences between the sampling sites after the oil spill (July) are presented in
control sites Fjällbacka and Nordre älv and Aspholmarna, Table 2. The sum of identified chemicals show that eelpout
within the harbor, in the May sampling (Fig. 3a). In July, the from the Aspholmarna site contained most hydroxylated
EROD activities were significantly higher in the Nordre älv, PAHs followed by the Skalkorgarna, Hjuvik, and finally
Hjuvik, and Skalkorgarna sites compared to the same sites at Nordre älv sites with the lowest levels. However, bile from
the May sampling. Hjuvik showed significantly higher EROD fish caught in Nordre älv contained approximately 20 times
activities compared to the inner harbor site, Aspholmarna higher PAH metabolites compared to bile from the control fish
(Fig. 3a). In November, fish from Aspholmarna showed sig- used in the laboratory exposure studies (Tables 1 and 2).
nificantly higher levels of EROD that the four other sites in the Results from the sites Hjuvik and Skalkorgarna, where two
River Göta älv estuary and the national reference site, pools obtained at the same sampling time point were analyzed,
Fjällbacka. show similar results. The repeatability in these results demon-
strates robustness of our method. In the Nordre älv and
CYP1A levels Aspholmarna sites, results show higher levels in the fish
captures in the first sampling time point compared to the
Hepatic CYP1A levels were significantly higher in eelpout second (Table 2).
caught at Hjuvik following the oil spill in July when compared
to fish from the same site caught prior to the spill in May.
CYP1A levels did not differ significantly between sites in Correlations
May or in July (Fig. 3b).
Correlations between data obtained from the field and labora-
DNA adduct levels tory exposure studies were calculated. In the field study,
EROD activity correlated positively to DNA adducts (p=
DNA adduct levels did not differ between the sites in May 0.006). Also, CYP1A levels correlated strongly to EROD
(Fig. 3d). In July, the DNA adduct levels were highest at the activity. However, we noted an outlier in the data; samples
Skalkorgarna site; however, the difference was not statistically from Aspholmarna collected in July, 2 weeks after the oil spill,
significant between sites or compared to May. In November, displayed lower EROD activities than would be predicted by
the Skalkorgarna site showed a significantly higher DNA the correlations to CYP1A (Fig. 3a, b). In addition, when
adduct levels compared to Fjällbacka; the Aspholmarna site EROD results from the Aspholmarna site in July were re-
also displayed significantly higher DNA adduct levels com- moved from the analysis, EROD activity also correlated to the
pared to Fjällbacka, and levels in Aspholmarna were signifi- sum of PAHs in bile (p<0.001). Also, a comparison of total
cantly higher than both of the other sites. PAHs measured by FACs and the sum of the PAHs measured
A comparison of the autoradiogram fingerprints from the by chemical analysis revealed a strong correlation (p=0.008).
laboratory exposure control fish (0 sample from the reference Following the exposure to oil in the laboratory, results
site Grundsund) and the fish-exposed i.p. for bunker oil at the showed strong correlations between EROD and CYP1A pro-
laboratory show several similarities (Fig. 4). tein levels and between EROD and total PAHs measured by
FACs in bile. There was also a significant correlation between
FACs the two methods of measuring PAHs (FACS and HPLC mea-
surements) in the fish from the field study (p=0.044).
Result from FACs measurements showed that eelpout cap- The pattern of alkylated PAH metabolites in the field
tured at the Aspholmarna site in May had significantly higher samples and the samples from the laboratory study were
FACs in the bile compare to the other sites, while the lowest almost identical (see Tables 1 and 2).
Environ Sci Pollut Res

Table 1 Amount of hydroxylated naphthalenes, fluorenes, phenan- analyzed PAH and their different grades of methylation are shown in
threnes, and pyrenes in bile from eelpout exposed to the bunker oil in a the column. Identification of methylated PAH is based on interpretation of
laboratory study. Eelpout were exposed to for 96 h to three oil concen- mass spectra and the amount is approximated by assuming the same
trations (10, 100, and 1,000 μg/L) through water exposure. Levels are response factors as for the unmethylated congener
shown as nanogram PAH per gram of bile. The total amount of the

Exposure OH naphthalenes OH fluorenes OH phenanthrenes OH pyrenes Total

C0 C1a C2a C3a C0 C1a C2a C0 C1a C2a C0 C1a C2a

ng/g f.w ng/g f.w ng/g f.w ng/g f.w ng/g f.w ng/g f.w ng/g f.w ng/g f.w ng/g f.w ng/g f.w ng/g f.w ng/g f.w ng/g f.w

Control <0.5 <0.05 <0.5 <0.5 <1 <0.5 <0.5 2 <5 <5 100 <5 <10 102
Low dose <0.5 <0.05 8 2 4 3 3 15 200 13 300 230 30 809
Middle dose 0.8 20 160 25 67 51 46 480 1,300 290 3,500 3,500 300 9,740
High dose 3.5 200 1,400 230 440 410 320 3,100 7,800 2,100 >23,000 >20,000 990 >60,000

a
Values are based on interpretation of mass spectra

Discussion incomplete combustion of organic matter (pyrogenic sources).

PAHs are genotoxic, and their metabolism results in interme-
PAHs are one of the most important groups of organic con- diate compounds that can bind directly to DNA, RNA, lipids
taminants found in the aquatic environment, and one source of and proteins, or mediate their toxic effects through the pro-
this input into waters is from oils spills which unfortunately duction of reactive oxygen species. The oil spill that occurred
occur commonly (petrogenic sources). Another source is in Göteborg harbor provided an opportunity to investigate the

Fig. 3 Effects of oils spill on female eelpout. EROD activities (a), levels point. Asterisk indicates significant difference in levels at a certain site
of CYP1A protein (b), PAH metabolites (FACs) in bile (c), and levels of during or after the oils spill, compared to before (May), p<0.05. NA not
DNA adducts (d) are shown. Results are shown as mean±standard error. available, NM not measured
Letters (a, b, c) indicate statistical differences between sites at one time
 Environ Sci Pollut Res

was spilled in the harbor was almost identical in most cases.

This is evidence that that the wild fish were actually exposed,
and accumulated substances from the spill, even though an-
ecdotal evidence suggested that the oil was insoluble and
formed agglomerates that could easily be removed from the
water (personal conversation with city representatives). The
amounts of specific PAH metabolites present in bile were
measured using GC/MS, and these results were compared to
measurements conducted using the FACs which indicates the
sum of total PAH metabolites. Synchronous fluorescence
spectrometry of PAH metabolites (FAC) is a sum parameter
that measures pyrene-like compounds. This means that you
also measure fluorescing compounds other than PAH metab-
Fig. 4 Autoradiogram images showing DNA adduct patterns in wild
caught eelpout used in a laboratory exposure study. Example of one olites that are present in the oil. It is a rapid and versatile
control fish and one fish exposed to the oil through i.p. injection. Simi- method to compare the amount of pyrene-like compounds in
larities in pattern suggest that the eelpout had been exposed to low levels fish from different sites and within groups. The results from
of the oil prior to capture
FACs are not directly comparable to the GC/MS method,
which determines the amount of specific isomers of PAH
effects of exposure to extremely high levels of PAHs in metabolites after enzymatic hydrolysis of glucuronides.
eelpout in the natural environment, and to address the useful- However, results from both methods correlated well, indicat-
ness of different methods to assess these effects. ing that the more simple FACs method is suitable to assess
The PAHs originating from petrogenic sources are often exposure to PAHs. If there is need for more specific analysis,
dominated by two- and three-ringed aromatics, while PAH e.g., to distinguish between sources of pollution, the GC/MS
from pyrogenic sources are dominated by four- and five- method is preferable. Results from the Nordre älv and
ringed aromatics (Anderson and Lee 2006). The ratio of the Aspholamarna sites, where fish were captured at two different
sum of methylphenanthrene to phenanthrene (MP/P ratio) are time points, show that the levels of specific PAH metabolites
used to distinguish between petrogenic (MP/P >2) and pyro- in the bile decreased rapidly. Fish were apparently able to
genic PAH (MP/P <0.5) (Boonyatumanond et al 2006). The metabolize and excrete a large portion of the PAHs within a
MP/P ratios for the different field sampling sites ranged from few days.
1.5 to 2.2 indicating that the PAH originated from petrogenic Exposure to the oil spill in the harbor resulted in signifi-
sources. Also, the pattern of alkylated PAH metabolites in the cantly elevated levels of CYP1A proteins and to hepatic
bile of fish caught in the recipient of Göteborg harbor and in EROD activity. Fish captured at all sites in the harbor also
the bile of fish exposed in the laboratory to the bunker oil that had elevated levels of PAH metabolites in their bile in July,

Table 2 Amount of hydroxylated naphthalenes, fluorenes, phenan- analyzed PAH and their different grades of methylation are shown in
threnes, and pyrenes in bile from eelpout captured at different locations the column. Identification of methylated PAH is based on interpretation of
(see Fig. 1 for map) approximately 2 weeks after the oil spill. Levels are mass spectra and the amount is approximated by assuming the same
shown as nanogram PAH per gram of bile. The total amount of the response factors as for the unmethylated congener

Exposure Date OH naphthalenes OH fluorenes OH phenanthrenes OH pyrenes Total

C0 C1a C2a C3a C0 C1a C2a C0 C1a C2a C0 C1a C2a

ng/g f.w ng/g f.w ng/g f.w ng/g f.w ng/g f.w ng/g f.w ng/g f.w ng/g f.w ng/g f.w ng/g f.w ng/g f.w ng/g f.w ng/g f.w ng/g f.w

Blank <0.5 <0.05 <0.5 <0.5 <1 <0.5 <0.5 <5 <5 <5 <5 <5 <10 <10
Nordre älv1 4/7 2.6 25 66 7 24 11 10 114 180 41 620 230 <10 1,330
Nordre älv2 10/7 2.0 16 29 5 12 8 5 19 79 12 310 52 <10 549
Hjuvik1 3/7 1.8 17 100 7 40 14 <0.5 140 200 20 960 600 <10 2,100
Hjuvik2 3/7 1.9 28 86 6 29 12 12 120 280 14 850 330 <10 1,769
Skalkorgarna1 4/7 2.0 18 62 5 30 17 15 100 470 28 1,100 340 <10 2,187
Skalkorgarna2 4/7 2.0 19 62 5 25 11 10 110 200 26 730 300 <10 1,500
Aspholmarna1 30/6 13 130 790 110 240 230 100 2,400 4,100 1,100 9,600 4,900 260 23,973
Aspholmarna2 4/7 4.8 190 540 24 63 13 23 250 330 82 1,400 860 99 3,879

a
Values are based on interpretation of mass spectra
Environ Sci Pollut Res

shortly after the spill. PAHs are known to bind to the AH These effects have been seen previously in fish following an
receptor which plays an important role in the induction of oil spill (Lee and Anderson 2005).
CYP1A proteins (Schlenk et al. 2008). In November, EROD A comparison of our measurements of DNA adducts in the
activities and PAH metabolites were at levels similar to those field samples and laboratory-treated fish revealed an interest-
measured in samples taken prior to the spill, indicating that ing result. Even though the increase in DNA adducts in liver
PAH load in the body had decreased. Several studies have tissue is obvious after i.p. oil exposure, it could be observed
addressed the seasonal variations of EROD activity in fish that the control fish, collected at a clean reference site
(Ronisz et al. 1999; Rotchell et al. 1999), indicating that (Grundsund, 65 km north of Göteborg harbor) and not ex-
highest levels are normally recorded in the late winter months posed to laboratory conditions, had an adduct level around
of February and March. In the current study, levels were 2.6 nmol/mol normal nucleotides, a level that suggests previ-
lowest during the spring and fall samplings. The fact that the ous exposure to DNA adduct forming xenobiotics (Aas et al.
oil spill occurred in the end of June, and samples were col- 2003). Furthermore, when analyzing the DNA adduct finger-
lected in July, when EROD levels would normally be at their print pattern in the autoradiogram from the control fish, sam-
lowest, is another indication of the extreme induction of this pled 3 months after the spill at a location 65 km to the north of
enzyme activity in the exposed fish. the spill, an interesting observation appear. Comparison of the
DNA adducts were also measured in selected samples, fingerprints suggest strong similarities in DNA adduct-
and results showed that levels tended to increase following forming substances between the feral control group of fish
the oil spill, but the increase was not significant in July due (control) and the oil-i.p.-injected-exposed fish, thereby sug-
to high variation. However, DNA adducts were present in gesting similar exposure and a long-distance transport of
the harbor at significantly higher levels several months after substances in the oil spill. Previous studies has observed that
the spill indicating more long-term chronic effects, a result corresponding types of DNA adduct patterns can be found
of interactions between PAH metabolites and DNA (Aas over widespread areas, indicating that pollutants are able to
et al. 2000). This relationship between PAH exposure, acute spread and cause similar DNA damage over long distances
induction of EROD, followed by DNA adduct formation (Ericson et al. 1998; Ericson et al. 1999).
after chronic exposure, has been demonstrated in a labora- In addition to further verifying effects of PAH-rich oil on
tory exposure study using flounder exposed via food both expression and activity of the phase I detoxification
(Reynolds et al. 2003). system, we show that, in situations where exposure to PAHs
Our results are in agreement with other studies which have is very high such as those found in an oil spill, CYPA1
indicated that biomarkers such as EROD activity or PAH in catalytical activity (EROD activity) can be inhibited. In vivo
bile are useful to investigate short-term or acute effects of and in vitro studies in Fundulus heteroclitus (Willett et al.
exposures (French et al. 1996; Aas et al. 2000). DNA adducts 2001) have demonstrated this, but propose that the inhibition
may be more useful as biomarkers in chronic effect situations, of EROD activity may be due to a downregulation of CYPA1.
and have been shown to occur in feral fish (Wirgin and Our results do not indicate a downregulation in CYP1A
Waldman 1998) and to increase linearly with exposure to protein, as we see no decrease in levels measured using
increasing concentrations of PAHs and with time (French ELISA, so we propose that the effect is a result of inactivation
et al. 1996). Our laboratory exposure also indicated that of the catalytic activity of the CYP1A protein. A decrease in
EROD activity and CYP1A levels were induced following EROD activity, despite high levels of CYP1A protein, has
an acute 96-h water exposure study, and that PAH metabolites been demonstrated in winter flounder following exposure to
can be found in bile at this time as well. There was a signif- high concentrations of a polychlorinated biphenyl, and these
icant increase in DNA adducts 21 days following a single i.p. effects were also found at polluted field sites (Monosson and
injection of the crude oil. Stegeman 1991). In either case, these results call for caution
The laboratory study was conducted in order to confirm when using EROD as single biomarker of effect in similar
results seen in the field and to provide controlled samples to be situations as results may be misleading. Analysis of additional
used in assessment of DNA adducts. Exposure to the bunker biomarkers, such as PAH levels in bile, may be advantageous
oil via water resulted in uptake of PAHs, which is evident in in order to correctly interpret EROD results.
the PAH metabolites measured in the bile. Fish also had
increased levels of CYP1A protein and EROD activity in liver
Acknowledgments This study was funded by the EU-BEEP project,
tissue. These results were all dose dependent, and levels of Västra Götaland county administration, and the Swedish Environmental
PAHs, CYP1A protein, and EROD activity all correlate sig- Protection Agency.
nificantly with one another (p<0.001, Pearson’s correlation).
Open AccessThis article is distributed under the terms of the Creative
Eelpout were also exposed to the bunker oil via an i.p. injec-
Commons Attribution License which permits any use, distribution, and
tion, which resulted in a significant increase in DNA adducts reproduction in any medium, provided the original author(s) and the
in hepatic tissue, verifying our results from the field samples. source are credited.
 Environ Sci Pollut Res

References French BL, Reichert WL, Hom T, Nishimoto M, Sanborn HR, Stein JE
(1996) Accumulation and dose-response of hepatic DNA adducts in
English sole (Pleuronectes vetulus) exposed to a gradient of con-
Aas E, Baussant T, Balk L, Liewenborg B, Andersen OK (2000) PAH taminated sediments. Aquat Toxicol 36(1–2):1–16
metabolites in bile, cytochrome P4501A and DNA adducts as envi- Frenzilli G, Scarcelli V, Barga ID, Nigro M, Forlin L, Bolognesi C, Sturve
ronmental risk parameters for chronic oil exposure: a laboratory J (2004) DNA damage in eelpout (Zoarces viviparus) from
experiment with Atlantic cod. Aquat Toxicol 51(2):241–258 Goteborg harbour. Mutat Res/Fund Mol Mech Mutagen 552(1–2):
Aas E, Beyer J, Jonsson G, Reichert WL, Andersen OK (2001) Evidence 187–195
of uptake, biotransformation and DNA binding of polyaromatic Gercken J, Förlin L, Andersson J (2006) Developmental disorders in
hydrocarbons in Atlantic cod and corkwing wrasse caught in the larvae of eelpout (Zoarces viviparus) from German and Swedish
vicinity of an aluminium works. Mar Environ Res 52(3):213–229 Baltic coastal waters. Mar Poll Bull 53(8–9):497–507
Aas E, Liewenborg B, Grøsvik BE, Camus L, Jonsson G, Børseth JF, Goksøyr A, Förlin L (1992) The cytochrome P-450 system in fish,
Balk L (2003) DNA adduct levels in fish from pristine areas are not aquatic toxicology and environmental monitoring. Aquat Toxicol
detectable or low when analysed using the nuclease P1 version of 22(4):287–311
the 32P-postlabelling technique. Biomarkers 8(6):445–460 Haritash AK, Kaushik CP (2009) Biodegradation aspects of polycyclic
Amat A, Burgeot T, Castegnaro M, Pfohl-Leszkowicz A (2006) DNA aromatic hydrocarbons (PAHs): a review. J Hazard Mat 169(1–3):1–15
adducts in fish following an oil spill exposure. Environ Chem Haux C, Forlin L (1988) Biochemical methods for detecting effects of
Letters 4(2):93–99 contaminants on fish. AMBIO 17(6):376–380
Anderson JW, Lee RF (2006) Use of biomarkers in oil spill risk assess- Larsson DGJ, Hällman H, Förlin L (2000) More male fish embryos near a
ment in the marine environment. Hum Ecol Risk Assess: Res Int J pulp mill. Environ Toxicol Chem 19(12):2911–2917
12(6):1192–1222 Lee RF, Anderson JW (2005) Significance of cytochrome P450 system
Baird WM, Hooven LA, Mahadevan B (2005) Carcinogenic polycyclic responses and levels of bile fluorescent aromatic compounds in
aromatic hydrocarbon-DNA adducts and mechanism of action. marine wildlife following oil spills. Mar Poll Bull 50(7):705–723
Environ Mol Mutagen 45(2–3):106–114 Leonard JD, Hellou J (2001) Separation and characterization of gall
Beach AC, Gupta RC (1992) Human biomonitoring and the 32P- bladder bile metabolites from speckled trout, Salvelinus fontinalis,
postlabeling assay. Carcinogenesis 13(7):1053–1074 exposed to individual polycyclic aromatic compounds. Environ
Beyer J, Jonsson G, Porte C, Krahn MM, Ariese F (2010) Analytical Toxicol Chem 20(3):618–623
methods for determining metabolites of polycyclic aromatic hydro- Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951) Protein measure-
carbon (PAH) pollutants in fish bile: a review. Environ Toxicol ment with the Folin phenol reagent. J Biol Chem 193(1):265–275
Pharmacol 30(3):224–244 Lyons BP, Stentiford GD, Green M, Bignell J, Bateman K, Feist SW,
Boonyatumanond R, Wattayakorn G, Togo A, Takada H (2006) Goodsir F, Reynolds WJ, Thain JE (2004) DNA adduct analysis and
Distribution and origins of polycyclic aromatic hydrocarbons histopathological biomarkers in European flounder (Platichthys
(PAHs) in riverine, estuarine, and marine sediments in Thailand. flesus) sampled from UK estuaries. Mutat Res/Fund Mol Mech
Mar Pollut Bull 52(8):942–956 Mutagen 552(1–2):177–186
Carney Almroth B, Sturve J, Berglund Å, Förlin L (2005) Oxidative Malmström C, Konn M, Bogovski S, Lang T, Lönnström LG, Bylund G
damage in eelpout (Zoarces viviparus), measured as protein car- (2009) Screening of hydrophobic DNA adducts in flounder
bonyls and TBARS, as Biomarkers. Aquat Toxicol 73:171–180 (Platichthys flesus) from the Baltic Sea. Chemosphere 77(11):
de Hoop L, Schipper AM, Leuven RSEW, Huijbregts MJ, Olsen GH, 1514–1519
Smit MGD, Hendriks AJ (2011) Sensitivity of polar and temperate Monosson E, Stegeman JJ (1991) Cytochrome P450E (P450IA) induction
marine organisms to oil components. Environ Sci Tech 45(20): and inhibition in winter flounder by 3,3′,4,4′-tetrachlorobiphenyl:
9017–9023 comparison of response in fish from Georges Bank and Narragansett
Dunn BP, Black JJ, Maccubbin A (1987) 32P-postlabeling analysis of Bay. Environ Toxicol Chem 10(6):765–774
aromatic DNA adducts in fish from polluted areas. Cancer Res Reddy MV, Randerath K (1986) Nuclease P1-mediated enhancement of
47(24 Part 1):6543–6548 sensitivity of 32P-postlabeling test for structurally diverse DNA
Ericson G, Balk L (2000) DNA adduct formation in northern pike (Esox adducts. Carcinogenesis 7(9):1543–1551
lucius) exposed to a mixture of benzo[a]pyrene, benzo[k]fluoranthene Reichert WL and French B (1994) The 32P-Postlabeling protocols for
and 7H-dibenzo[c, g]carbazole: time-course and dose-response stud- assaying levels of hydrophobic DNA adducts in fish. US
ies. Mutat Res/Fund Mol Mech Mutagen 454(1–2):11–20 Department of Commerce, National Oceanic and Atmospheric
Ericson G, Lindesjöö E, Balk L (1998) DNA adducts and histopatholog- Administration, National Marine Fisheries Service, Northwest
ical lesions in perch (Perca fluviatilis) and northern pike (Esox Fisheries Science Center
lucius) along a polycyclic aromatic hydrocarbon gradient on the Reynolds WJ, Feist SW, Jones GJ, Lyons BP, Sheahan DA, Stentiford
Swedish coastline of the Baltic Sea. Can J Fish Aquat Sci 55(4): GD (2003) Comparison of biomarker and pathological responses in
815–824 flounder (Platichthys flesus L.) induced by ingested polycyclic
Ericson G, Noaksson E, Balk L (1999) DNA adduct formation and persis- aromatic hydrocarbon (PAH) contamination. Chemosphere 52(7):
tence in liver and extrahepatic tissues of northern pike (Esox lucius) 1135–1145
following oral exposure to benzo[a]pyrene, benzo[k]fluoranthene and Ronisz D, Förlin L (1998) Interaction of CYP1A1 inducers in liver of
7H-dibenzo[c, g]carbazole. Mutat Res 427(2):135–145 rainbow trout and eelpout. Mar Environ Res 46(1–5):93–95
Forlin L, Andersson T, Koivusaari U, Hansson T (1984) Influence of Ronisz D, Larsson DGJ, Forlin L (1999) Seasonal variations in the
biological and environmental factors on hepatic steroid and xenobi- activities of selected hepatic biotransformation and antioxidant en-
otic metabolism in fish: interaction with PCB and β- zymes in eelpout (Zoarces viviparus). Comp Biochem Phys Part C:
naphthoflavone. Mar Environ Res 14(1–4):47–58 Pharm Toxicol Endocrin 124(3):271
Forlin L, Haux C, Karlsson Norrgren L, Runn P, Larsson A (1986) Ronisz D, Lindesjöö E, Larsson Å, Bignert A, Förlin L (2005) Thirteen
Biotransformation enzyme activities and histopathology in rainbow years of monitoring selected biomarkers in Eelpout (Zoarces
trout (Salmo gairdneri) treated with cadmium. Aquat Toxicol 8(1): viviparus) at reference site in the Fjällbacka Archipelago on the
51–64 Swedish West Coast. Aquat Ecosys Health Manag 8(2):175–184
Environ Sci Pollut Res

Rotchell JM, Bird DJ, Newton LC (1999) Seasonal variation in Vetemaa M, Sandström O, Förlin L (1997) Chemical industry effluent
ethoxyresorufin O-deethylase (EROD) activity in European eels impacts on reproduction and biochemistry in a North Sea population
Anguilla anguilla and flounders Pleuronectes flesus from the of viviparous blenny (Zoarces viviparus). J Aquat Ecosys Stress
Severn Estuary and Bristol Channel. Mar Ecol Prog Ser 190:263–270 Recov 6(1):33–41
Schiedek D, Broeg K, Baršienė J, Lehtonen KK, Gercken J, Pfeifer S, Whitehead P, Bauchot M, Hureau J, Nielsen J and Tortonese E (1984)
Vuontisjärvi H, Vuorinen PJ, Dedonyte V, Koehler A, Balk L, Fishes of the north-eastern Atlantic and the Mediterranean
Schneider R (2006) Biomarker responses as indication of contami- (Poissons de l'Atlantique du nord-est et de la Méditerranée).
nant effects in blue mussel (Mytilus edulis) and female eelpout Paris, UNESCO
(Zoarces viviparus) from the southwestern Baltic Sea. Mar Poll Willett KL, Wassenberg D, Lienesch L, Reichert W, Di Giulio RT
Bull 53(8–9):387–405 (2001) In vivo and in vitro inhibition of CYP1A-dependent
Schlenk D, Celander M, Gallagher E, George S, James M, Kullman S, activity in Fundulus heteroclitus by the polynuclear aromatic
Hurk PVD and Willett K (2008) Biotransformation in fishes. The hydrocarbon fluoranthene. Toxicol App Pharm 177(3):264–
toxicology of fishes, CRC Press: 153-234 271
Sturve J, Berglund Å, Balk L, Broeg K, Böhmert B, Massey S, Savva D, Wirgin I, Waldman JR (1998) Altered gene expression and genetic
Parkkonen J, Stephensen E, Koehler A, Förlin L (2005) Effects of damage in North American fish populations. Mutat Res/Fund Mol
dredging in Göteborg Harbor, Sweden, assessed by biomarkers in Mech Mutagen 399(2):193–219
eelpout (Zoarces viviparus). Environ Toxicol Chem 24(8):1951–1961 Yanik PJ, O’Donnell TH, Macko SA, Qian Y, Kennicutt MC
van der Oost R, Beyer J, Vermeulen NPE (2003) Fish bioaccumulation (2003) The isotopic compositions of selected crude oil
and biomarkers in environmental risk assessment: a review. Environ PAHs during biodegradation. Organic Geochem 34(2):291–
Toxicol Pharmacol 13(2):57–149 304