Environmental Toxicology and Chemistry, Vol. 26, No. 5, pp.

944953, 2007
2007 SETAC
Printed in the USA
0730-7268/07 $12.00 .00

GILL METAL BINDING AND STRESS GENE TRANSCRIPTION IN BROWN TROUT

(SALMO TRUTTA) EXPOSED TO METAL ENVIRONMENTS: THE EFFECT OF
PRE-EXPOSURE IN NATURAL POPULATIONS
BJRN HENRIK HANSEN,* YVIND A. GARMO, PAL A. OLSVIK, and ROLF A. ANDERSEN
Norwegian University of Science and Technology, Department of Biology,
Department of Chemistry, Hgskoleringen 5, N-7491 Trondheim, Norway
National Institute of Nutrition and Seafood Research, Nordnes, N-5817 Bergen, Norway
( Received 25 July 2006; Accepted 3 November 2006)
AbstractBrown trout (Salmo trutta) from two native populations from the Rros area in Central Norway, acclimated in miningaffected habitats to different levels of Cd/Zn and Cu, together with trout from a nearby unaffected river (reference) were transferred
to a nearby lake with higher levels of Cu, Cd, and Zn than those in their respective native rivers. This experiment was conducted
to gain information about the underlying resistance mechanisms developed in fish exposed to metal environments. The focus was
on gill metal accumulation and transcription of the metal-responsive stress genes metallothionein-A (MT-A), Cu/Zn-superoxide
dismutase (SOD), catalase (CAT), glutathione peroxidase (GPx), glutathione reductase (GR), and heat shock protein 70 (HSP-70).
The only shared response shown between the three groups after transfer were Cu accumulation and MT-A induction. The Cuacclimated trout produced mucus to reduce the uptake of Cu into the gills. The MT-A levels were highest in the Cd/Zn-acclimated
trout both before and after transfer. Before transfer, antioxidant transcription (SOD and GPx) was higher in gills of Cu-acclimated
compared to the Cd/Zn-acclimated trout, but increased transcription of antioxidant stress genes was observed after transfer in both
metal-acclimated groups. The metal-acclimated trout groups also showed an increase in the transcription of HSP-70. Compared to
the reference population not previously exposed to metals, stress gene transcription increased faster in the metal-acclimated populations. The exception was induction of CAT, which appeared to be depressed after transfer in Cd/Zn-acclimated trout. The data
indicate that acclimation to chronic metal exposure involves different strategies to cope with different metals and that these strategies
involve both physiological mechanisms (mucus production) as well as metal-related stress gene transcription.
KeywordsStress gene transcription

Brown trout

Oxidative stress

Gills

Metals

Induction of MTs by metals in salmonids is mediated by

multiple copies of metal-responsive elements (MREs) present
in the 5-regulatory regions of the MT genes [11]. Metal-responsive elements are activated by a Zn-sensitive, metal-responsive transcription factor (MTF-1) [12], but induction of
MT gene transcription also has been shown in fish for a range
of other metals, including Cd, Cu, Ag, and Hg in teleosts [13].
The induction mechanisms of antioxidant enzymes by metals,
however, remain unclear. Recently, MREs have been found
upstream genes encoding proteins and enzymes dealing with
oxidative stress, metal transport, GSH turnover, and general
metabolism [7].
Tolerance to metals has been extensively studied in salmonids, which have been found to be much more sensitive
than other fish families. Studies on gill Cu-binding properties
in rainbow trout (Oncorhynchus mykiss) and yellow perch
(Perca flavescens) revealed that a higher tolerance in yellow
perch was caused by differences in gill binding of Cu [14].
Brook trout (Salvelinus fontinalis) has been found to be more
tolerant to Cu than rainbow trout, and this reduced sensitivity
was explained by a lower affinity for Cu on the gills [15].
Differences in metal tolerance between species also have been
attributed to mutations in the TATA box necessary for MT
transcription [16].
Normally, gill epithelia are covered with an extracellular
polyanionic matrix (mucus) containing a variety of glycoproteins, with a high degree of sialic acids, mucopolysaccharides,
low-molecular-weight compounds, inorganic ions (Na, K,
Ca2, and Cl), and water [17]. This mucus, secreted by mu-

INTRODUCTION

The gills are the major route of uptake for waterborne metals in freshwater fish. Metal exposure through the water gives
rise to a wide range of morphological changes in gills as well
as osmoregulatory deficiencies. Copper and Zn are essential
elements, but high concentrations may cause toxic effects, especially in gills. Metals like Cd and Cu can cause oxidative
stress through several mechanisms: The Fenton reaction [1],
depletion of cellular glutathione (GSH) [2,3], alterations of
mitochondrial electron-transfer chain [4], or inhibition of antioxidant enzymes [5,6]. Zinc, which is an essential cofactor
in more than 200 enzymes, can give cytotoxic effects in the
presence of hydrogen peroxide (H2O2) [7]. Generation of reactive oxygen species (ROS), which occurs naturally during
aerobic metabolism in mitochondria, normally is prevented
from causing harmful effects by antioxidant enzymes, such as
Cu/Zn-superoxide dismutase (SOD), catalase (CAT), and glutathione peroxidase (GPx) [8]. Other cellular proteins involved
in cellular protection against oxidative stress are glutathione
reductase (GR) [8], heat shock protein 70 (HSP-70) [9], and
metallothionein (MT) [10]. The equilibrium between ROS production and removal can be altered by the presence of ROSmediating metals, such as Cd and Cu. If the balance between
ROS and antioxidants favors ROS, then oxidation of macromolecules (DNA, proteins, and lipids) may occur.
* To whom correspondence may be addressed
(bjorn.h.hansen@sintef.no). The current address of B.H. Hansen is
SINTEFMaterials and Chemistry, N-7465 Trondheim, Norway.
944

Environ. Toxicol. Chem. 26, 2007

Gill metals and stress gene transcription in brown trout

cocytes and dispersed throughout the epithelium of the skin

and gills, may function as an ion-exchange system, with different affinities for different metals [18]. The mucus layer is
thought to be protective by binding potentially toxic metals.
The mucus layer may then be sloughed off, hence removing
these metals before reaching uptake binding sites. The equilibrium between mucus production and release is influenced
by metals, and acute metal exposure can result in a massive
increase of mucus synthesis and secretion, which may be harmful [19].
Brown trout (Salmo trutta) populations in the area around
the mining city of Rros in Central Norway have been exposed
to metal-contaminated water for many generations, and they
are considered to be acclimated to these metals in their natural
environments [20,21]. The river Stribekken is convenient for
reference purposes, because it is not influenced by mine tailings and runoffs, whereas the Naustebekken and Rugla rivers
represent considerably Cd/Zn- and Cu-polluted environments,
respectively [21]. We have recently shown [20] that the levels
of Cu in gills of Naustebekken trout were higher than those
found in Rugla trout, despite the fact that levels of Cu in the
Rugla River were twofold higher than in the Naustebekken
River. In the present study, we test the hypothesis that reduced
gill uptake of Cu as a function of increasing mucus secretion
could be a plausible explanation for these differences and,
subsequently, an important mechanism for surviving chronic
exposure to Cu. The Cd/Zn-acclimated Naustebekken trout had
higher levels of gill MT than the Cu-acclimated Rugla trout;
in fact, the MT levels in Rugla trout gills were comparable to
those in the reference Stribekken trout. The protein MT therefore seems to play a vital role in the survival of brown trout
chronically exposed to Cd/Zn, but not for trout chronically
exposed to Cu [20]. When fish were transferred from the uncontaminated Stribekken River to the Cu-contaminated Rugla
River [22] and to the Cd/Zn-contaminated Naustebekken River
(unpublished data), we found increased transcription of antioxidant genes, indicating that trout living in these rivers are
subjected to considerable metal-mediated oxidative stress. To
study responses in these differently metal-acclimated trout
groups, trout were transferred to Lake Orvsjen because of its
high levels of Cd, Cu, and Zn. In addition to the gill metal
uptake studies, transcription of MT-A, SOD, CAT, GPx, GR,
and HSP-70 were studied. After transfer, the trout were expected to respond according to their fundamental acclimation
background from the metal levels in their respective native
habitats, making more visible the defense mechanisms that
have developed and revealing how these trout handle fluctuations in metal levels within their respective habitats.
MATERIALS AND METHODS

Experimental setup and sampling

Brown trout were captured during low water levels (July
2004) in the rivers Stribekken, Naustebekken, and Rugla using
basket traps and electric fishing. The fish were caged in their
natural rivers for a period of 4 to 7 d before they were transferred to, caged, and exposed in Lake Orvsjen. The fish were
divided in groups of six individuals from each river, and exposure times in Lake Orvsjen water were 0 (controls), 6, and
12 h. Because all fish were immature, fish of different sizes
and sexes were used throughout.
The fish were killed by a blow to the head, weighed, and
measured before the second gill arches with filament on each
side were removed. One of these arches was put directly in

945

an acid-washed tube (G1), and the other was soaked briefly

in distilled water and blotted on filter paper before being placed
in a separate acid-washed tube (G2), a sampling method adopted from Wilkinson and Campbell [23]. The fraction of metals
assumed to be associated with mucus or loosely bound to the
gill surface was calculated by the following equation:
(100%([Me]G1 [Me]G2))/[Me]G1. The remaining gill arches
with filaments (80100 mg) were homogenized in 1 ml of
Trizol reagent (Invitrogen, Life Technologies, Carlsbad, CA,
USA) using autoclaved pistils and tubes (Kontes, Vineland,
NJ, USA). All samples were frozen in liquid nitrogen as quickly as possible. The samples were transferred to a 80C freezer
for storage on the same day.

Water chemistry
Water samples were collected from the Stribekken, Naustebekken, and Rugla rivers as well as from Lake Orvsjen
before and at the times of trout transfer. These samples were
preserved by addition of Suprapure HNO3 (14.5 M; Merck,
Darmstadt, Germany) to an acid concentration of 0.1 M and
stored at 4C before measurements of the total metal concentrations in the water. Diffusive gradients in thin films (DGTs)
were used to assess labile metal concentrations in the rivers
and in Lake Orvsjen. The procedure for use of DGT has been
described previously [20]. Conductivity, pH, and temperature
were measured by a portable multiradiometer (pH/cond 340i/
SET; WTW, Darmstadt, Germany). Alkalinity was measured
with a digital titrator (model 16900; Hach, Loveland, CO,
USA).

Metal analysis
The gill arches were lyophilized and weighed before digestion in 800 l of HNO3 (14.5 M) and 400 l of H2O2 (30%,
pro analysis; Merck). The samples were digested by heating
them to a temperature of greater than 100C for an hour. The
samples were subsequently cooled and diluted to 14 ml in the
storp, Swesame polypropylene tubes (AB Cerbo-Hertila, A
den).
Metal measurements in water, DGT, and gill samples were
done by inductively coupled plasmamass spectroscopy (ICPMS; Element2; Thermo Finnegan, Bremen, Germany). Certified reference material (bovine liver 1577; National Institute
of Standards and Technology, Gaithersburg, MD, USA) was
used to assess the accuracy of the analysis. For every tenth
sample, a matrix-modified standard sample was measured to
correct for drift.

Stress gene transcription

Primers and probes for the gene transcription analysis were
selected using Primer Express 2.0 software (Applied Biosystems, Foster City, CA, USA), and they were obtained from
sequences from the National Center for Biotechnology Information (NCBI; Bethesda, MD, USA) GenBank. The sequences
of the primers and probes, the product size, and the NCBI
GenBank accession numbers are given in Table 1. All assays
were originally developed for Atlantic salmon (Salmo salar)
[24] but proved to work well on the closely related brown
trout.
Total RNA was isolated from gill samples homogenized in
Trizol reagent according to the manufacturers description (Invitrogen). Subsequently, samples were treated with DNase I
to remove potentially contaminating DNA (DNAfree kit; Ambion, Austin, TX, USA). First-strand cDNA synthesis was run

946

Environ. Toxicol. Chem. 26, 2007

B.H. Hansen et al.

Table 1. Nucleotide sequences of polymerase chain reaction (PCR) primers and TaqMan probes (Applied Biosystems, Foster City, CA, USA)
used for real-timePCR quantificationa
Gene
18S
EF1AA
MT-A
SOD
CAT
HSP-70
GR
GPx
a

Acc. no.
AJ427629
AF321836
X97274
BG936553
BG935638
BG933934
BG934480
BG934453

Forward primer
CCCCGTAATTGGAATGAGTACACTTT
CCCCTCCAGGACGTTTACAAA
CCTTGTGAATGCTCCAAAACTG
CCACGTCCATGCCTTTGG
GAGGGCAACTGGGACCTTACT
CCCCTGTCCCTGGGTATTG
CCAGTGATGGCTTTTTTGAACTT
GATTCGTTCCAAACTTCCTGCTA

Reverse primer

TaqMan probe

ACGCTATTGGAGCTGGAATTACC
CACACGGCCCACAGGTACA
CAGTCGCAGCAACTTGCTTTC
TCAGCTGCTGCAGTCACGTT
GGACGAAGGACGGGAACAG
CACCAGGCTGGTTGTCTGAGT
CCGGCCCCCACTATGAC
GCTCCCAGAACAGCCTGTTG

CACCAGACTTGCCCTCC
ATCGGTGGTATTGGAAC
ATCTTGCAACTGCGGTGG
ACAACACCAACGGCT
CTAATTAAATGGCAAGCTATC
Not used
Not used
TGAATGGAGACACAGAAC

The stress genes analyzed were 18S ribosomal rRNA (18S), elongation factor 1A- (EF1AA), metallothionein-A (MT-A), superoxide dismutase
(SOD), catalase (CAT), heat shock protein 70 (HSP-70), glutathione reductase (GR), and glutathione peroxidase (GPx). Acc. no. GenBank
accession number.

in triplicate on 96-well reaction plates with the GeneAmp PCR

9700 machine (Applied Biosystems) with the TaqMan reverse
transcription reagent (Applied Biosystems) containing multiscribe reverse transcriptase (RT; 1.67 U/l), MgCl2 (5.5 mM),
10 TaqMan RT-buffer (1), deoxyribonucleotide triphosphates (500 M of each), and RNase inhibitor (0.4 U/l).
Oligo-dT primers (2.5 M) were used for all genes except 18S
rRNA, for which random hexamer primers (2.5 M) were
used. Input RNA amount was 125 ng in each reaction. Reverse
transcription was performed at 48C for 60 min, followed by
a RT inactivation step at 95C for 5 min.
The cDNA (5 l) from each RT reaction was transferred
to new 96-well plates, and real-time polymerase chain reaction
(PCR) was run with the ABI Prism 7000 sequence detection
system (Applied Biosystems). The two-step semiquantitative
RT-PCR protocol developed and described by Olsvik et al.
[24] was performed to measure transcriptional levels of SOD,
CAT, GPx, MT-A, 18S rRNA, and elongation factor 1A-
(EF1AA). For GR and HSP-70, a SYBR Green assay was performed as follows: In addition to cDNA (5 l), each well
contained 2 QuantiTect SYBR Green PCR master mix (1;
Qiagen, Crawley, UK) and primers (0.4 M of each) diluted
with RNase-free water to a total volume of 25 l. Polymerase
chain reaction was run after a 15-min activation and denaturation step at 95C, followed by 40 cycles of 15 s at 95C and
60 s at 55C.
Of the two housekeeping genes tested on this material, 18S
rRNA and EF1AA, the Microsoft Excelbased (Microsoft, Redmond, WA, USA) geNorm software [25] determined 18S
rRNA to be the one least affected by the exposure regime.
Mean normalized expressions for MT-A, SOD, CAT, HSP-70,
GPx, and GR were calculated by the Microsoft Excelbased
Q-gene software (Microsoft) using PCR efficiency and housekeeping gene normalization [26].

Statistical analysis
One-way analysis of variance was used to compare both
metal accumulation and gene transcription between fish from
control populations and between controls and exposed fish.
Pearson correlation analysis was used to study relationships
between metal accumulation and stress gene transcription and
between the different stress genes. For correlation analysis,
control fish and the two exposed groups (6 and 12 h) were
analyzed together for each population. All statistical analyses
were performed using SPSS 12.0 software (SPSS, Chicago,
IL, USA).

RESULTS

Water chemistry
Chemical parameters determined in the waters from the
three rivers and Lake Orvsjen are displayed in Table 2. The
Rugla water had a higher alkalinity and hardness than the water
from the other three localities. However, all the water qualities
can be characterized as relatively soft and neutral with respect
to pH. Organic carbon was not analyzed, but a previous investigation showed that total organic carbon normally is low
(3 mg/L) in these rivers [27].
The rivers Stribekken, Naustebekken, and Rugla represent
three different metal-exposure regimes: Stribekken had low
concentrations of metals; Naustebekken elevated concentrations of Cd and Zn, and Rugla an elevated concentration of
Cu. Previously, variations of the metal concentrations in the
rivers Naustebekken and Rugla have been reported with different water-flow conditions affecting native trout populations.
The Cu concentration in Rugla can rise appreciably above 40
g/L when the water flow is high, whereas concentrations of
Cd and Zn can double in Naustebekken during runoff episodes
[28]. Our data indicate that the Cu concentrations in the Naustebekken River have increased somewhat since 1997 [28]. The
relatively small variations in the total concentrations of Cd,
Cu, and Zn found in the present investigation suggest that the
fish exposure conditions were relatively constant before the
transfer to Lake Orvsjen. The data in Table 3 indicate that
most of the Cd and Zn are present in the DGT-labile fraction
in the Naustebekken and Rugla rivers and in Lake Orvsjen.
In contrast, much of the Cu, Al, and Fe were in a nonlabile
and nonbioavailable fraction; that is, they occurred as colloidal
or particulate species and/or were strongly bound in complexes
with natural organic matter.

Fish data
Because of low trout population density in the rivers, the
amount of trout used in the experiment was low. Weight,
length, and condition factor for the trout are summarized in
Table 3. The trout from the Rugla River were somewhat smaller
in size compared with those from the other rivers. We found
no differences in size and weight between the groups from the
individual rivers, but in the Naustebekken trout groups, we
observed a significant decrease in condition factor during the
exposure ( p 0.05). The high range in size and weight of
the trout may have effects on both metal accumulation and
gene transcription. However, only immature fish were used,
and we consider that such biological factors have no effects

Environ. Toxicol. Chem. 26, 2007

Gill metals and stress gene transcription in brown trout

947

Table 2. Limnochemical data for total and labile metal concentrations of waters from the Stribekken, Naustebekken, and Rugla rivers and Lake
Orvsjen (Rros, Central Norway) for the summer of 2004a
Element

Stribekken River

pH
Conductivity (S/cm)
Alkalinity (Eq/L)
Temperature (C)

7.4
18.50
100
10.9

Naustebekken River

0.2 (4)
0.71 (2)
2.8 (4)
1.8 (4)

7.1
23.33
120.8
11.6

Rugla River

0.1 (5)
0.58 (3)
14.5 (5)
1.6 (5)

7.3
44
261
8.9

Lake Orvsjen

0.1 (4)
5.66 (2)
41.4 (4)
1.9 (4)

6.4
36.5
86.0
13.7

(1)
(1)
(1)
(1)

CD (ng/L)
Total
Labile

4.0 2.0 (5)

1.4 (1)

232.2 46.7 (5)

182.0 (1)

23.9 9.1 (4)

23.0 (1)

698.0 (1)
825.0 (1)

Cu (g/L)
Total
Labile

0.5 0.1 (5)

0.1 (1)

10.7 2.2 (5)

2.2 (1)

22.9 7.9 (4)

7.3 (1)

113.9 (1)
53.7 (1)

Zn (g/L)
Total
Labile

5.7 5.3 (5)

1.1 (1)

130.7 22.5 (5)

87.2 (1)

7.2 2.2 (4)

7.5 (1)

348.7 (1)
403.9 (1)

Al (g/L)
Total
Labile

29.6 8.5 (5)

4.9 (1)

41.6 6.8 (5)

8.1 (1)

107.7 109.6 (4)

8.7 (1)

176.9 (1)
19.9 (1)

Fe (g/L)
Total
Labile

31.4 13.6 (5)

1.35 (1)

21.0 1.2 (5)

0.80 (1)

174.8 187.8 (4)

7.49 (1)

75.9 (1)
3.2 (1)

2.51 0.05 (5)

NA

5.28 0.90 (4)

NA

3.36 (1)
NA

Ca (mg/L)
Total
Labile
a

1.97 0.01 (5)

NA

Labile metals were measured using diffusive gradients in thin films. Values are presented as the mean standard deviation, with the number
of samples in parentheses. NA not analyzed.

Naustebekken than in control fish from Rugla ( p 0.05),

even though the water concentration of Cu was higher in Rugla
than in Naustebekken (Table 2). The amount of mucus-bound/
loosely bound Cu in fish before transfer to Lake Orvsjen was
insignificant. Second, gill-bound Cu concentrations reached a
higher level in Naustebekken fish compared to fish from the
other two localities when exposed for 6 h in Lake Orvsjen
water. Further increases were not observed between 6 and 12
h. In fact, for Rugla fish, the concentration of gill-bound Cu
appeared to decrease. Third, Rugla fish showed the largest
fraction of mucus-bound/loosely bound Cu when exposed to
Lake Orvsjen water (up to 60%), which is consistent with
the qualitative observation that the Cu-acclimated Rugla fish
produced a greater amount of mucus after transfer.
No significant differences were found between mucusbound/loosely bound and total Cd or Zn concentrations in the
trout groups ( p 0.05), indicating that these metals were not
bound to mucus or loosely to the gill surface. Figure 2 shows

on the responses found and the conclusions being made. The

decrease in condition factor seen only in the Naustebekken
trout after transfer may be interpreted as a response to the
exposure. However, we consider it to be coincidental, because
a response on condition factor should not be expected from
such a short exposure time.

Gill metals
Figure 1 shows the concentrations of gill-bound (intracellular/tightly bound) and total (mucus/loosely bound and intracellular/tightly bound) Cu in fish from Stribekken, Rugla, and
Naustebekken rivers before transfer in controls and after exposure in water from Lake Orvsjen for 6 and 12 h. The same
figure also shows the size of the fraction (%) of Cu that was
bound to the mucus or loosely bound to gill surfaces in the
different trout groups during the exposure regime in Lake
Orvsjen. These results illustrate the following observations:
First, gill-bound Cu levels were higher in control fish from

Table 3. Weight, size, and condition factor of brown trout (Salmo trutta) from the Stribekken, Naustebekken, and Rugla rivers transferred to
Lake Orvsjen (Rros, Central Norway) and exposed for 6 or 12 ha
Exposure
Stribekken trout

Naustebekken trout

Rugla trout

Ctrl
6h
12 h
Ctrl
6h
12 h
Ctrl
6h
12 h

Weight (g)
39.83
40.17
34.50
28.00
23.83
19.00
14.33
14.00
14.67

19.48
26.57
18.32
16.24
15.46
8.29
8.09
7.54
4.72

Size (cm)
16.12
16.12
15.53
14.55
14.33
13.98
11.28
11.67
12.12

2.77
3.45
2.94
3.54
3.21
2.17
1.54
0.85
0.93

Condition factor
0.86
0.86
0.85
0.82
0.71
0.66
0.92
0.84
0.81

0.08
0.06
0.07
0.06
0.06*
0.06*
0.11
0.24
0.10

Controls (Ctrl) are fish sampled before transfer. Values are presented as the mean standard deviation. An asterisk indicates a significant
difference from the control (p 0.05). n 6 for all groups.

948

Environ. Toxicol. Chem. 26, 2007

B.H. Hansen et al.

Fig. 1. Concentrations of Cu in gills of brown trout (Salmo trutta) from the Stribekken, Naustebekken, and Rugla rivers after transfer to Lake
Orvsjen (Rros, Central Norway) and exposed for 6 or 12 h. Controls are fish sampled before transfer. Gray bars represent concentrations of
intracellular Cu or Cu tightly bound to gill surface. White bars represents total Cu concentration in gills, including Cu bound to mucus or loosely
bound to gill surface. The line plots show the percentage of Cu that was loosely bound or bound to mucus, as calculated by the equation given
in Materials and Methods. All values are presented as the mean standard deviation (n 6). An asterisk indicates significantly higher metal
levels compared to the corresponding control group ( p 0.05).

the total concentrations of Cd (Fig. 2A) and Zn (Fig. 2B) in

gills of the three trout groups exposed in Lake Orvsjen. No
significant differences in Zn levels were found between exposed trout and their corresponding control group, but for Cd,
a significant increase was observed in the Stribekken trout after
transfer ( p 0.05). In the Naustebekken trout gills, a significant decrease in gill Cd occurred after 12 h of exposure ( p
0.05). The two figures clearly illustrate the high levels of
both Cd and Zn in the Naustebekken trout compared to the
Stribekken and Rugla trout before transfer.

0.628, p 0.01, n 18) and HSP-70 (r 0.585, p 0.05,

n 18) in the Naustebekken trout and between Cu and GR
(r 0.500, p 0.05, n 18) in the Rugla trout.
A high degree of positive correlation between different
stress genes indicated that a battery of genes was turned on
during the exposure in Lake Orvsjen (Table 4). Correlations
between CAT and other stress genes were not found, except
in the Naustebekken trout, for which CAT was negatively correlated with SOD and with GR ( p 0.05).
DISCUSSION

Stress gene transcription

Significant transcription differences were found between
the fish groups from the three rivers before transfer to Lake
Orvsjen (Fig. 3). The MT-A gene transcription appeared to
be highest in the Naustebekken trout (Fig. 3A). Rugla trout
had the highest gene transcript levels of SOD (Fig. 3C) and
GPx (Fig. 3E). Expression of HSP-70 was higher in the Stribekken trout compared to the groups from the other rivers
before transfer (Fig. 3B). Naustebekken trout had lower levels
of GR transcripts compared to Rugla and Stribekken fish (Fig.
3F).
Transfer to Lake Orvsjen resulted in increased transcription of MT-A in all three trout groups ( p 0.05), and MT-A
was the only gene that significantly increased in the gills of
trout from the reference Stribekken River. For the Cd/Zn-acclimated Naustebekken trout, significant increases in gene transcripts of MT-A (12 h) (Fig. 3A), SOD (12 h) (Fig. 3C), GPx
(12 h) (Fig. 3E), GR (6 and 12 h) (Fig. 3F), and HSP-70 (6
and 12 h) (Fig. 3B) were observed ( p 0.05). For CAT,
however, a significant decrease was observed in the Naustebekken trout after 12 h of exposure ( p 0.05) (Fig. 3C). For
the Cu-acclimated Rugla trout, significant increases in gene
transcripts of MT-A (12 h) (Fig. 3A), GPx (12 h) (Fig. 3E),
GR (6 and 12 h) (Fig. 3F), and HSP-70 (6 and 12 h) (Fig. 3B)
were observed ( p 0.05).
The MT-A gene transcription in Stribekken trout was found
to correlate well with gill concentrations of Cd (r 0.643, p
0.01, n 18), Cu (r 0.745, p 0.001, n 18), and Zn
(r 0.605, p 0.01, n 18), but significant correlations
were not found in Naustebekken or Rugla trout ( p 0.05).
We found significant correlations between Cu with GR (r

Gill metals
Visual observations indicated that the Cu-acclimated Rugla
trout produced greater amounts of mucus on the gills than
trout from the Stribekken and Naustebekken rivers after transfer to the same Cu challenge in Lake Orvsjen. The results
displayed in Figure 2 therefore support the hypothesis that the
secretion of mucus may protect gill tissues from toxic metals
[19,29,30]. Previous studies of Al by Wilkinson and Campbell
[23] and by Playle and Wood [31] showed that gill binding
could account for less than 18% of the Al extracted from the
water passing over the gills, and those authors argued that Al
was bound by mucus and quickly sloughed off. No similar
studies have been done on Cu, Cd, and Zn, but studies on
body mucus have confirmed the ability of mucus to bind these
metals [17,29,30]. However, mucus protection probably is not
the only explanation for the lower basal level and accumulation
rate of gill Cu in Rugla trout compared to Naustebekken trout.
Binding of Cu by the protein MT, which occurs at higher
concentrations in the gill tissue of the Naustebekken trout [20],
probably also is important. In addition, increased elimination
of Cu may be involved in acclimation during prolonged exposure to Cu [32] and, subsequently, contribute to the lower
concentration of Cu in the Rugla trout gills.

Stress gene transcription

Induction of MT-A transcription in gills followed the same
pattern in all three trout groups after transfer to Lake Orvsjen.
For the Stribekken reference trout, MT-A transcripts correlated
significantly with Cu and Zn uptake in the gills ( p 0.01).
Synthesis of MT is regulated mainly at the transcriptional level

Gill metals and stress gene transcription in brown trout

Fig. 2. Total concentrations of Cd (A) and Zn (B) in gills of brown

trout (Salmo trutta) from the Stribekken, Naustebekken, and Rugla
rivers after transfer to Lake Orvsjen (Rros, Central Norway) and
exposed for 6 h (white bars) or 12 h (cross-hatched bars). Controls
(solid bars) are fish sampled before transfer. Values are presented as
the mean standard deviation (n 6). An asterisk indicates significantly different metal levels compared to corresponding control group
( p 0.05).

and is induced by various metal ions, such as Cd, Cu, and Zn

[33]. As described elsewhere [20], basal levels of MT-A transcripts were higher in the Naustebekken trout than in the two
other trout groups, which is shown in Figure 3A. This group
also exhibited the highest levels of MT-A transcripts after
transfer to Lake Orvsjen. In addition, synthesis of MT has
been shown previously to be important in acclimation to the
metal levels present in the Naustebekken River [21,28]. This
protein can bind potentially toxic metal ions and act as a radical
scavenger [10]. We have shown previously that metal binding
by MT is not necessarily increased in gills after exposure to
Cu, even after MT-A transcription had increased, and have
argued that MT functions as a radical scavenger rather than
as a metal-binding protein in gills [22]. The nonacclimated

Environ. Toxicol. Chem. 26, 2007

949

Stribekken trout responded by only inducing MT-A transcription after transfer to Lake Orvsjen, in contrast to the two
metal-acclimated trout groups, which exhibited significant increases in all stress genes except CAT. It may be concluded
that a faster induction of several stress genes during shortterm high metal levels is a mechanism by which metal acclimation is achieved.
Exposures to Cd, Cu, and Zn have been shown to result in
an increase in SOD gene transcripts as well as enzyme activity
[34,35], which is in accordance with our findings on gene
transcription. The SOD transcription increased significantly in
the gills of the Naustebekken trout after 12 h of metal exposure
in Lake Orvsjen ( p 0.05). The Rugla trout, which had a
higher basal level of SOD transcripts, also seemed to induce
this gene, but not significantly so ( p 0.05). The two GSHhandling enzymes, GPx and GR, also were induced after transfer of the metal-acclimated trout groups ( p 0.05) (Fig. 3E
and F), indicating a potential oxidation of GSH. Copper can
alter levels of reduced GSH by increasing activity of glutathione S-transferase [36] and GPx [3,34] or inhibit GSH synthetase [36]. This metal also forms stable complexes with GSH,
hence decreasing GSH levels in cytosol [37]. The fact that
GPx and GR are induced in metal-acclimated trout when transferred to Lake Orvsjen may indicate a faster turnover of GSH
in the metal-acclimated trout compared to the reference Stribekken trout.
Positive correlations between the different stress gene responses, except for CAT, might indicate that these genes are
regulated through similar induction mechanisms. After transfer
to Lake Orvsjen, significant induction and correlation between different stress genes were most prominent in the metalacclimated trout groups, which may have two explanations:
Either the oxidative stress in the metal-acclimated trout is higher than in the reference trout, or the metal-acclimated trout
have a faster stress response. Previous findings [38,39] have
showed that antioxidant enzymes like GPx, CAT, and Cu/ZnSOD act in a cooperative or synergistic way to ensure cell
protection against oxidative stress and that this optimal protection is achieved only when an appropriate balance between
the activities of these enzymes is maintained. Their activities
probably are not solely under transcriptional control, but their
transcription might be influenced by the same transcription
factor.
The metal-responsive transcription factor has been found
in salmonids, in which it is the key regulation mechanism for
MT transcription [40]. In addition to MT, MTF-1 has been
shown to mediate the synthesis of other genes that are important in protection against oxidative stress, such as genes
involved in GSH biosynthesis [41], SOD, and CAT [7]. In the
presence of free Zn, MTF-1 will bind to MREs upstream of
the target gene, hence initiating transcription [42]. We did not
find an increase in gill Zn concentrations in the trout, except
for Rugla trout 6 h after transfer, so if stress gene transcription
was to be initiated through MTF-1, then either Zn had to be
replaced by Cu on Zn-binding proteins [43] or MTF-1 had to
be activated by other metals. It has been argued previously
that metals other than Zn can activate the MT-promoter MRE
in salmonids [44], but to our knowledge, this argument has
not yet been made for Cu. High expression of MTF-1 in trout
gills may explain both the elevated response in metal-acclimated trout and the correlation between transcriptions of the
different stress genes.
Heat shock proteins are not induced through the MTF-1

950

Environ. Toxicol. Chem. 26, 2007

B.H. Hansen et al.

Fig. 3. Gene transcriptional levels of metallothionein-A (MT-A; A), heat shock protein 70 (HSP-70; B), superoxide dismutase (SOD; C), catalase
(CAT; D), glutathione peroxidase (GPx; E), and glutathione reductase (GR; F) in gills of brown trout (Salmo trutta) from the Stribekken,
Naustebekken, and Rugla rivers when exposed in Lake Orvsjen (Rros, Central Norway) water for 6 h (white bars) or 12 h (cross-hatched
bars). Controls (ctrl; white bars) are fish sampled before transfer. Values are presented as the mean standard deviation (n 6). Lowercase
letters (a and b) show significant differences between control groups from the different rivers, where a b ( p 0.05). An asterisk indicates
significantly different gene transcription compared to corresponding control group ( p 0.05).

Environ. Toxicol. Chem. 26, 2007

Gill metals and stress gene transcription in brown trout

951

Table 4. Correlation coefficients (r) from correlation analysis (Pearson correlation) between gene transcription levels of metallothionein-A (MT-A),
superoxide dismutase (SOD), catalase (CAT), glutathione peroxidase (GPx), glutathione reductase (GR), and heat shock protein 70 (HSP 70) in the
trout (Salmo trutta) from the Stribekken (S), Naustebekken (N), and Rugla (R) rivers exposed in Lake Orvsjen (Rros, Central Norway)a

MT-A
HSP-70
SOD
CAT
GPx

Trout group

HSP-70

SOD

CAT

GPx

GR

S
N
R
S
N
R
S
N
R
S
N
R
S
N
R

0.518*
0.753***
0.772***

0.399
0.681**
0.372
0.861***
0.681**
0.501*

0.026
0.410
0.186
0.094
0.453
0.433
0.101
0.558*
0.160

0.625**
0.664**
0.812***
0.796***
0.713***
0.657**
0.656**
0.441
0.133
0.029
0.381
0.205

0.722***
0.874***
0.798***
0.875***
0.923***
0.814***
0.831***
0.846***
0.698***
0.001
0.497*
0.154
0.860***
0.728***
0.491*

The sample was consistent throughout (n 18). *p 0.05, **p 0.01, ***p 0.001.

pathway but, rather, by the heat shock factor 1, which binds

to the heat shock promoter element. Heat shock proteins are
regarded as indicators responding to a wide range of stressors
[45], including metals [46]. For the metal-acclimated trout, a
significant increase in HSP-70 transcription was already observed after 6 h of exposure in Lake Orvsjen (Fig. 3B). Ryan
and Hightower [47] showed that HSP-70 protein concentrations in fish were elevated after a sublethal exposure to Cd
and Cu. This is in agreement with our results from Naustebekken trout, in which we found a positive correlation ( p
0.05) between HSP-70 and Cu uptake in gills. This relationship, however, was not found in the Rugla trout. In gills of
rainbow trout (O. mykiss), a mixture of Cd, Cu, Pb, and Zn
caused an increase in HSP-70 during a 21-d exposure regime
[46]. The elevation of HSP-70 transcription in the gills of the
metal-acclimated trout indicated that protein misfolding had
occurred, possibly mediated by ROS caused by the presence
of metals. Heat shock protein 70 is induced when proteins
form nonnative disulfide bonds after oxidative stress [9], so a
correlation between HSP-70 and enzymes dealing with GSH
turnover, which we found in all three trout groups, could be
expected.
After transfer to Lake Orvsjen, the Naustebekken trout
showed a significant decrease in CAT but an increase in SOD
transcription ( p 0.05). A similar negative relationship also
has been found regarding activities of these two enzymes in
Nile tilapia (Oreochromis niloticus) from a site polluted with
polychlorinated biphenyl ethers [48] and in killifish (Fundulus
heteroclitus) experimentally exposed to Cd both in vivo and
in vitro [5]. Pruell and Engelhardt [5] argued that SOD activity
increased with production of O2 by the pollutants, and according to Kono and Fridovich [49], this radical has the ability
to inhibit CAT activity. A negative relationship between CAT
activity and Cd exposure has been described [50,51] and suggested as a direct, metal-mediated structural alteration of the
enzyme. A depression of CAT synthesis also is possible. The
significant negative correlation between SOD and CAT transcriptions in gills of Naustebekken trout further strengthens
this type of relationship between these parameters. In addition,
exposure to Cu has been found not to affect CAT activity [51].
Lake Orvsjen consists of a mixture of these metals, but the
Cu levels are much higher than the Cd levels. The bioavailability of Cd in Lake Orvsjen, on the other hand, was greater

than that of Cu, as shown by the DGT method (Table 2). Metal
mixtures therefore might give different responses, at least on
level of gene transcription.
Naustebekken trout were, as mentioned, acclimated to Cd
exposure in their natural environment. It is well known that
Cd affects the electron-transport chain in mitochondria, enhancing production of O2 [4]. This radical is readily broken
down to H2O2 by SOD, and two enzymes deal with this product: CAT, and GPx [1,8]. Because H2O2 easily crosses biological membranes, both GPx (cytosol) and CAT (peroxisomes)
should be affected. After transfer to Lake Orvsjen, CAT transcription was found to be significantly decreased in Naustebekken trout gills, a response not seen for the other two groups.
Only Naustebekken trout showed a significant positive correlation between SOD and GPx. Adjustments of CAT transcription could be a response indicating a potential pathway
for acclimation to chronic Cd exposure. Naustebekken trout
simultaneously depressed their CAT transcription and increased their GPx transcription. The reason for this might be
that Cd-mediated inhibition of CAT activity results in a higher
demand for GPx to catalyze the oxidation of H2O2 at the expense of GSH. We found an increased transcription of GPx in
the Naustebekken trout but not in the Stribekken trout. This
increase in GPx should be accompanied by an increase in GR.
Indeed, a significant increase in GR was found in the Naustebekken trout gills exposed for only 6 h in Lake Orvsjen
water. Because CAT is known to be inhibited by Cd, a response
to chronic Cd exposure might be to generate more CAT molecules de novo. Another possible response might be to save
energy by reducing the transcription of a gene encoding a
protein that will not function during Cd exposure. The reduction in CAT transcripts in Naustebekken trout might well be
such a mechanism.
From our data, it was impossible to predict if differences
in the response to acute metal exposure were an effect of
acclimation or genetic adaptation. However, some genetic differences between the trout populations of the Rugla and Naustebekken rivers have been found [52]. The Naustebekken and
Stribekken rivers are not physically separated, and trout may
migrate between these rivers. Therefore, genetic differences
between these two populations were considered to be unlikely.
The Stribekken trout sampled, however, were caught upstream,
above a waterfall, and had never been exposed to Naustebek-

952

Environ. Toxicol. Chem. 26, 2007

ken water. Because the responses of stress gene transcription

were clearer in the Naustebekken trout gills than in the other
trout groups, we consider these responses to be a result of
acclimation rather than of genetic adaptation. However, genetic
adaptation cannot be ruled out, because it can happen after
relatively few generations when chronically exposed to metals
[53].
CONCLUSIONS

To conclude, trout seem to develop different strategies toward survival in metal-rich environments. Mucus secretion
appears to reduce Cu-uptake in chronically Cu-exposed fish,
whereas Cd/Zn-exposed trout rely on intracellular binding of
the metals to MT to prevent toxic effects. In addition, induction
of antioxidant enzymes seems to play an important role in both
metal-acclimated trout groups. A lack of significant response
on stress gene transcription, except for MT-A, in the Stribekken
trout during the exposure and significantly increased transcription on all stress genes, except CAT, in the metal-acclimated
trout indicated that acclimation to chronically metal-rich environments involved a shorter response time to a sudden increase in metal exposure. A similar response pattern in the
Naustebekken and Rugla trout further indicated no discrimination between different metals in this respect, at least not for
Cd, Cu, and Zn, but basal levels of antioxidant enzyme transcripts generally were higher in the chronically Cu-exposed
Rugla trout. In addition, a synchronic induction of stress genes
was observed in the trout when transferred to Lake Orvsjen.
This indicated a similar mechanistic pathway in which the
studied stress genes are induced. It can be argued that transcription of stress genes does not represent good evidence in
mechanistic studies, because very often, a correlation between
gene transcription and functional protein activity or concentration is clearly lacking [20,22,54]. Therefore, protein/enzyme
assays should be applied in future studies to further investigate
the mechanistic basis for acclimation to metal exposure in
natural populations.
AcknowledgementThe authors wish to thank Syverin Lierhagen,
Sonja Ylving, Sindre A. Pedersen, and two anonymous reviewers for
valuable comments on the manuscript. This work was financed by the
Norwegian Research Council (project 147474/720).
REFERENCES
1. Stohs SJ, Bagchi D. 1995. Oxidative mechanisms in the toxicity
of metal ions. Free Radic Biol Med 18:321336.
2. Rau MA, Whitaker J, Freedman JH, Di Giulio RT. 2004. Differential susceptibility of fish and rat liver cells to oxidative stress
and cytotoxicity upon exposure to pro-oxidants. Comp Biochem
Physiol C 137:335342.
3. Freedman JH, Ciriolo MR, Peisach J. 1989. The role of glutathione in copper metabolism and toxicity. J Biol Chem 264:5598
5605.
4. Wang Y, Fang J, Leonard SS, Murali Krishna Rao K. 2004. Cadmium inhibits the electron transfer chain and induces reactive
oxygen species. Free Radic Biol Med 36:14341443.
5. Pruell RJ, Engelhardt FR. 1980. Liver cadmium uptake, catalase
inhibition and cadmium thionein production in the killifish (Fundulus heteroclitus) induced by experimental cadmium exposure.
Mar Environ Res 3:101111.
6. Shukla GS, Hussain T, Chandra SV. 1987. Possible role of regional superoxide dismutase activity and lipid peroxide levels in
cadmium neurotoxicity: In vivo and in vitro studies in growing
rats. Life Sci 41:22152221.
7. Chung MJ, Walker PA, Brown RW, Hogstrand C. 2005. ZINCmediated gene expression offers protection against H2O2-induced
cytotoxicity. Toxicol Appl Pharmacol 205:225236.

B.H. Hansen et al.

8. Evans P, Halliwell B. 2001. Micronutrients: Oxidant/antioxidant
status. Br J Nutr 85 (Suppl. 2):6774.
9. McDuffee AT, Senisterra G, Huntley S, Lepock JR, Sekhar KR,
Meredith MJ, Borrelli MJ, Morrow JD, Freeman ML. 1997. Proteins containing non-native disulfide bonds generated by oxidative
stress can act as signals for the induction of the heat shock response. J Cell Physiol 171:143151.
10. Sato M, Bremner I. 1993. Oxygen free radicals and metallothionein. Free Radic Biol Med 14:325337.
11. Kling P, Olsson PE. 1995. Regulation of the rainbow trout metallothionein-A gene. Mar Environ Res 39:117120.
12. Palmiter RD. 1994. Regulation of metallothionein genes by heavy
metals appears to be mediated by a zinc-sensitive inhibitor that
interacts with a constitutively active transcription factor, mtf-1.
Proc Natl Acad Sci U S A 91:12191223.
13. Cheung APL, Lam THJ, Chan KM. 2004. Regulation of tilapia
metallothionein gene expression by heavy metal ions. Mar Environ Res 58:389394.
14. Taylor LN, Wood CM, McDonald DG. 2003. An evaluation of
sodium loss and gill metal binding properties in rainbow trout
and yellow perch to explain species differences in copper tolerance. Environ Toxicol Chem 22:21592166.
15. MacRae RK, Smith DE, Swoboda-Colberg N, Meyer JS, Bergman
HL. 2003. Copper binding affinity of rainbow trout (Oncorhynchus mykiss) and brook trout (Salvelinus fontinalis) gills: Implications for assessing bioavailable metal. Environ Toxicol Chem
18:11801189.
16. Olsson PE, Kille P. 1997. Functional comparison of the metalregulated transcriptional control regions of metallothionein genes
from cadmium-sensitive and tolerant fish species. Biochim Biophys Acta 1350:325334.
17. Handy RD. 1989. The ionic composition of rainbow trout body
mucus. Comp Biochem Physiol A 93:571575.
18. Newman MC, Jagoe CH. 1994. Ligands and the bioavailability
of metals in aquatic environments. In Hamelink JL, ed, Bioavailability: Physical, Chemical, and Biological Interactions.
Lewis, Boca Raton, FL, USA, pp 3962.
19. Eddy FB, Fraser JE. 1982. Sialic acid and mucus production in
rainbow trout (Salmo gairdneri Richardson) in response to zinc
and seawater. Comp Biochem Physiol C 73:357359.
20. Hansen BH, Rmma S, Garmo A, Olsvik PA, Andersen RA.
2006. Antioxidative stress proteins and their gene expression in
brown trout (Salmo trutta) from three rivers with different heavy
metal levels. Comp Biochem Physiol C 143:263274.
21. Olsvik PA, Gundersen P, Andersen RA, Zachariassen KE. 2001.
Metal accumulation and metallothionein in brown trout, Salmo
trutta, from two Norwegian rivers differently contaminated with
Cd, Cu and Zn. Comp Biochem Physiol C 128:189201.
22. Hansen BH, Rmma S, Sfteland LIR, Olsvik PA, Andersen RA.
2006. Induction and activity of oxidative stress-related proteins
during waterborne Cu exposure in brown trout (Salmo trutta).
Chemosphere 65:17071714.
23. Wilkinson KJ, Campbell PGC. 1993. Aluminum bioconcentration
at the gill surface of juvenile Atlantic salmon in acidic media.
Environ Toxicol Chem 12:20832095.
24. Olsvik PA, Kristensen T, Waagb R, Rosseland BO, Tollefsen
KE, Baeverfjord G, Berntssen MHG. 2005. mRNA expression of
antioxidant enzymes (SOD, CAT and GSH-Px) and lipid peroxidative stress in liver of Atlantic salmon (Salmo salar) exposed
to hyperoxic water during smoltification. Comp Biochem Physiol
C 141:314323.
25. Vandesompele J, Preter KD, Pattyn F, Poppe B, Roy NV, Paepe
ADSF. 2002. Accurate normalization of real-time quantitative RTPCR data by geometric averaging of multiple internal control
genes. Genome Biology 3:Research0034.10034.11.
26. Muller PY, Janovjak H, Miserez AR, Dobbie Z. 2002. Processing
of gene expression data generated by quantitative real-time RTPCR. BioTechniques 32:13721379.
27. Gundersen P, Olsvik PA, Steinnes E. 2001. Variations in heavy
metal concentrations and speciation in two mining-polluted
streams in central Norway. Environ Toxicol Chem 20:978984.
28. Olsvik PA, Gundersen P, Andersen RA, Zachariassen KE. 2000.
Metal accumulation and metallothionein in two populations of
brown trout, Salmo trutta, exposed to different natural water environments during a run-off episode. Aquat Toxicol 50:301316.
29. Varanasi U, Markey D. 1978. Uptake and release of lead and

Gill metals and stress gene transcription in brown trout

30.
31.
32.
33.
34.

35.

36.
37.

38.

39.
40.
41.

cadmium in skin and mucus of coho salmon. Comp Biochem

Physiol C 60:187191.
Miller TG, Mackay WC. 1982. Relationship of secreted mucus
to copper and acid toxicity in rainbow trout. Bull Environ Contam
Toxicol 28:6874.
Playle RC, Wood CM. 1989. Water pH and aluminum chemistry
in the gill microenvironment of rainbow trout during acid and
aluminum exposure. J Comp Physiol B 159:539550.
Grosell MH, Hogstrand C, Wood CM. 1997. Cu uptake and turnover in both Cu-acclimated and non-acclimated rainbow trout
(Oncorhynchus mykiss). Aquat Toxicol 38:257276.
Thiele DJ. 1992. Metal-regulated transcription in eukaryotes. Nucleic Acids Res 20:11831191.
Sanchez W, Palluel O, Meunier L, Coquery M, Porcher JM, AitAissa S. 2005. Copper-induced oxidative stress in three-spined
stickleback: Relationship with hepatic metal levels. Environ Toxicol Pharmacol 19:177183.
Cho YS, Choi BN, Kim KH, Kim SK, Kim DS, Bang IC, Nam
YK. 2006. Differential expression of Cu/Zn superoxide dismutase
mRNA during exposures to heavy metals in rockbream (Oplegnathus fasciatus). Aquaculture 253:667679.
Canesi L, Viarengo A, Leonzio C, Filippelli M, Gallo G. 1999.
Heavy metals and glutathione metabolism in mussel tissues.
Aquat Toxicol 46:6776.
Brouwer M, Hoexum-Brouwer T, Cashon RE. 1993. A putative
glutathione-binding site in Cd/Zn-metallothionein identified by
equilibrium binding and molecular-modeling studies. Biochem J
294:219225.
Michiels C, Raes M, Toussaint O, Remacle J. 1994. Importance
of SE-glutathione peroxidase, catalase, and Cu/Zn-SOD for cell
survival against oxidative stress. Free Radic Biol Med 17:235
248.
Bagnyukova TV, Chahrak OI, Lushchak VI. 2006. Coordinated
response of goldfish antioxidant defenses to environmental stress.
Aquat Toxicol 78:325331.
Dalton TP, Solis WA, Nebert DW, Carvan MJ. 2000. Characterization of the MTF-1 transcription factor from zebra fish and trout
cells. Comp Biochem Physiol B 126:325335.
Gunes C, Heuchel R, Georgiev O, Muller KH, Lichtlen P, Bluthman H, Marino S, Aguzzi A, Schaffner W. 1998. Embryonic
lethality and liver degeneration in mice lacking the metal-responsive transcriptional activator MTF-1. EMBO J 17:2846
2854.

Environ. Toxicol. Chem. 26, 2007

953

42. Saydam N, Adams TK, Steiner F, Schaffner W, Freedman JH.

2002. Regulation of metallothionein transcription by the metalresponsive transcription factor MTF-1. J Biol Chem 277:20438
20445.
43. Shaw CF, Savas MM, Petering DH. 1991. Ligand substitution and
sulfhydryl reactivity of metallothionein. Methods Enzymol 205:
401413.
44. Mayer GD, Leach A, Kling P, Olsson PE, Hogstrand C. 2003.
Activation of the rainbow trout metallothionein-A promoter by
silver and zinc. Comp Biochem Physiol B 134:181188.
45. Iwama GK, Thomas PT, Forsyth RB, Vijayan MM. 1998. Heat
shock protein expression in fish. Rev Fish Biol Fish 8:3556.
46. Williams JH, Farag AM, Stansbury MA, Young PA, Bergman
HL, Petersen NS. 1996. Accumulation of HSP70 in juvenile and
adult rainbow trout gill exposed to metal-contaminated water and/
or diet. Environ Toxicol Chem 15:13241328.
47. Ryan JA, Hightower LE. 1994. Evaluation of heavy-metal ion
toxicity in fish cells using a combined stress protein and cytotoxicity assay. Environ Toxicol Chem 13:12311240.
48. Bainy ACD, Saito E, Carvalho PSM, Junqueira VBC. 1996. Oxidative stress in gill, erythrocytes, liver and kidney of Nile tilapia
(Oreochromis niloticus) from a polluted site. Aquat Toxicol 34:
151162.
49. Kono Y, Fridovich I. 1982. Superoxide radical inhibits catalase.
J Biol Chem 257:57515754.
50. Palace VP, Klaverkamp JF. 1993. Variation of hepatic enzymes
in three species of freshwater fish from pre-Cambrian shield lakes
and the effect of cadmium exposure. Comp Biochem Physiol C
104:147154.
51. Romeo M, Bennani N, Gnassia-Barelli M, Lafaurie M, Girard JP.
2000. Cadmium and copper display different responses towards
oxidative stress in the kidney of the sea bass Dicentrarchus labrax. Aquat Toxicol 48:185194.
52. Olsvik PA, Hindar K, Zachariassen KE, Andersen RA. 2001.
Brown trout (Salmo trutta) metallothioneins as biomarkers for
metal exposure in two Norwegian rivers. Biomarkers 6:274288.
53. Xie LT, Klerks PL. 2003. Responses to selection for cadmium
resistance in the least killifish, Heterandria formosa. Environ
Toxicol Chem 22:313320.
54. Misra RR, Crance KA, Bare RM, Waalkes MP. 1997. Lack of
correlation between the inducibility of metallothionein mRNA
and metallothionein protein in cadmium-exposed rodents. Toxicology 117:99109.