Effects of metal exposure on motor neuron development, neuromasts and the escape response of zebrafish embryos Laura Sonnack, Sebastian Kampe, Elke Muth-K¨ohne, Lothar Erdinger, Nicole Henny, Henner Hollert, Christoph Sch¨afers, Martina Fenske PII: DOI: Reference:

S0892-0362(15)00196-8 doi: 10.1016/j.ntt.2015.05.006 NTT 6537

To appear in:

Neurotoxicology and Teratology

Received date: Revised date: Accepted date:

9 September 2014 11 May 2015 12 May 2015

Please cite this article as: Laura Sonnack, Sebastian Kampe, Elke Muth-K¨ohne, Lothar Erdinger, Nicole Henny, Henner Hollert, Christoph Sch¨afers, Martina Fenske, Effects of metal exposure on motor neuron development, neuromasts and the escape response of zebrafish embryos, Neurotoxicology and Teratology (2015), doi: 10.1016/j.ntt.2015.05.006

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

PT

Effects of metal exposure on motor neuron

SC

RI

development, neuromasts and the escape response of

NU

zebrafish embryos

MA

Laura Sonnack*† ‖‖, Sebastian Kampe‡, Elke Muth-Köhne‡, Lothar Erdinger§, Nicole Henny§, Henner Hollert‖‖, Christoph Schäfers‡, Martina Fenske*† Fraunhofer Institute for Molecular Biology and Applied Ecology IME, Aachen, Germany;



Fraunhofer Institute for Molecular Biology and Applied Ecology IME, Schmallenberg,

TE

D



AC CE P

Germany; §Department of Medical Microbiology and Hygiene, Heidelberg University Hospital, Heidelberg, Germany; ‖‖Department of Ecosystem Analysis, Institute for Environmental Research, RWTH Aachen University, Germany Corresponding Author *Laura Sonnack

Fraunhofer Institute for Molecular Biology and Applied Ecology IME Forckenbeckstrasse 6 52074 Aachen Germany [email protected] Phone +49 241 6085 13342

1

ACCEPTED MANUSCRIPT *Martina Fenske Fraunhofer Institute for Molecular Biology and Applied Ecology IME

PT

Forckenbeckstrasse 6 52074 Aachen

RI

Germany

SC

[email protected]

NU

Phone +49 241 6085 12230

MA

KEYWORDS

AC CE P

TE

D

Motor neuron damage, neuromasts, behavior, metal toxicity, zebrafish embryo

2

ACCEPTED MANUSCRIPT ABSTRACT Low level metal contaminations are a prevalent issue with often unknown consequences for

PT

health and the environment. Effect-based, multifactorial test systems with zebrafish embryos to

RI

assess in particular developmental toxicity are beneficial but rarely used in this context. We

SC

therefore exposed wild-type embryos to the metals copper (CuSO4), cadmium (CdCl2) and cobalt (CoSO4) for 72 hours to determine lethal as well as sublethal morphological effects. Motor

NU

neuron damage was investigated by immunofluorescence staining of primary motor neurons

MA

(PMNs) and secondary motor neurons (SMNs). In vivo stainings using the vital dye DASPEI were used to quantify neuromast development and damage. The consequences of metal toxicity

D

were also assessed functionally, by testing fish behavior following tactile stimulation. The

TE

median effective concentration (EC50) values for morphological effects 72 hours post fertilization (hpf) were 14.6 mg/L for cadmium and 0.018 mg/L for copper, whereas embryos

AC CE P

exposed up to 45.8 mg/L cobalt showed no morphological effects. All three metals caused a concentration-dependent reduction in the numbers of normal PMNs and SMNs, and in the fluorescence intensity of neuromasts. The results for motor neuron damage and behavior were coincident for all three metals. Even the lowest metal concentrations (cadmium 2 mg/L, copper 0.01 mg/L and cobalt 0.8 mg/L) resulted in neuromast damage. The results demonstrate that the neuromast cells were more sensitive to metal exposure than morphological traits or the response to tactile stimulation and motor neuron damage.

3

ACCEPTED MANUSCRIPT 1. INTRODUCTION Metals in the environment arise from natural sources such as erosion, volcanic activity and forest

PT

fires (Nriagu, 1989), but also from anthropogenic sources, mainly the processing and

RI

manufacturing of metals, the disposal of metal-containing electronic devices (e-waste) and

SC

chemical products (e.g. paints), and from agriculture (AMAP, 1998). Metal contamination is a global environmental problem because metals are neither chemically nor biologically degradable

NU

and the use of electronic goods and nanomaterials is increasing. Cobalt derivatives are included

MA

in the REACH (Directive 1907/2006/EEC) “Candidate List of Substances of Very High Concern for Authorisation” due to their carcinogenic properties, and cadmium is a priority pollutant

D

according to the European Water Framework Directive (EWFD). Although several metals are

TE

essential for humans and animals, including chromium, iron, cobalt, copper, nickel, vanadium and zinc, because they support vital cell functions, they can still become toxic at higher

AC CE P

concentrations. For instance copper has shown to be highly toxic to aquatic organisms. Juvenile rainbow trout showed an increase in mortality after ten days of exposure starting at 20 μg/L Cu (as CuSO4) as well as accumulation in the gills (Shaw et al. , 2012). Daphnia magna as another example, showed at a copper concentration of 0.4 µg/L Cu that the reproduction after 21 days was inhibited (Dave, 1984). Effects like mortality, apoptosis and hair cell death occurred in zebrafish larvae after exposure to copper up to 76 hpf in a concentration range of 1 to 500 µM CuSO4 (0.06 to 31.77 mg/L Cu) (Hernandez et al. , 2006, Hernandez et al. , 2011, Luzio et al. , 2013). In particular developing fish (Jezierska et al. , 2009) may be affected by copper bioaccumulation. Cobalt and other metals accumulate in adult zebrafish tissues (Reinardy et al. , 2011) and can induce effects like apoptosis and oxidative stress and affect the hatching of zebrafish (Danio rerio) larvae (Cai et al. , 2012, Dave and Xiu, 1991).

4

ACCEPTED MANUSCRIPT Many metals have shown to be harmful to fish, and the underlying mechanisms initiating metal toxicity are complex and often not yet fully understood. However, the knowledge of the initiating

PT

events causing adverse metal effects is important for a realistic estimation of the hazards associated with exposure of fish to metals at different scenarios and conditions, from acute to

RI

chronic. Moreover, the contamination of the environment with diverse metals at trace

SC

concentrations, increasingly in nano particulate form, and with metal containing compounds is a

NU

growing problem while suitable test methods to effectively monitor and assess metal toxicity in aquatic systems are still scarce. The zebrafish embryo toxicity assay (FET) is a recognized

MA

alternative for the assessment of fish acute toxicity (OECD test guideline 236). It is also a powerful tool to assess sub-acute effects, when morphological and sub-morphological

D

parameters are combined and included in the assessment (Braunbeck and Lammer, 2006, Scholz

TE

et al. , 2008, Strähle U. et al. , 2012). The FET combines the benefits of a whole organism and

AC CE P

vertebrate model system with in-vitro scale properties and animal protection advantages (Busch et al. , 2011, Strähle U., Scholz S., 2012). It has shown in numerous studies to be a highly versatile toxicity model which allows the assessment of effects at different levels, from the molecular to the physiological to the organism and even the behavioral in parallel, at short test durations and larger scale (Ali et al. , 2011, Truong et al. , 2014, Yang et al. , 2009). With our study we aimed at exploring the suitability of additional cellular and behavioral endpoints to the FET to improve the evaluation of metal toxicity. Cell or molecule specific fluorescent labels are advantageous as test read-outs because the detection of the fluorescent signals can be computerized and quantified with the help of image processing programs such as ImageJ (Rasband). This reduces the subjectivity of manually assessed endpoints, which rely on variable perceptions and levels of experience of the person conducting the evaluation on the test. Another

5

ACCEPTED MANUSCRIPT advantage is the visualization and identification of molecular and cellular targets of toxicants which can specifically be chosen to indicate mechanism of action. Different fluorescence

PT

endpoints, such as the staining of apoptotic cells with acridine orange or the TUNEL assay have already been used in zebrafish embryos to measure apoptosis caused by cadmium (Chan and

RI

Cheng, 2003, Yu et al. , 2012), mercury (Yang et al. , 2010), copper (Hernandez, Undurraga,

SC

2011) and cobalt (Cai, Zhu, 2012). Muscle fiber development and axon growth have been used

NU

as markers for the myotoxic and neurotoxic effects of cadmium (Hen Chow and Cheng, 2003). The disruption of primary motor neuron (PMN) and secondary motor neuron (SMN)

MA

development was visualized using specific antibody staining (Sylvain et al. , 2010), and accordingly this approach has been used to assess the damage to SMNs in zebrafish embryos 48

D

h post-fertilization (hpf) following exposure to silver nanoparticles (Muth-Kohne et al. , 2013).

TE

Another promising approach is the in vivo staining of neuromast hair cells (Froehlicher et al. ,

AC CE P

2009, Harris et al. , 2003, Ton and Parng, 2005, Williams and Holder, 2000) to detect damaging effects to these hair cells as a developmental toxicity endpoint. For example, the light-sensitive fluorescent vital dye 2-(4-(dimethylamino)styryl)-N-ethylpyridinium iodide (DASPEI) can be used to stain mitochondria in living hair cells allowing the visualization of neuromasts and the lateral line organ (Jorgensen, 1989). The neuromasts, which contain mitochondrion-rich mechanosensory cells, reside on the outer surface of the skin and are easily accessible by the dye. The neural hair cells share many properties with the hair cells in the mammalian inner ear, including selective susceptibility to ototoxins. A recent study already showed effects of metal nanoparticles on the lateral line and the behavior in zebrafish embryos (McNeil et al. , 2014). A combined analysis of motor neuron/neuromast damage and behavioral endpoints can therefore be instrumental in connecting cellular with organism effects because it has been shown that

6

ACCEPTED MANUSCRIPT neurons, especially Mauthner cells, mediate specific reactions known as the escape response, startle response or C-start in the goldfish (Carassius auratus) and zebrafish (Eaton et al. , 1984,

PT

Faber et al. , 1989, Liu and Fetcho, 1999, Zottoli, 1977). The Mauthner cells receive information from the lateral line organ and excite motor neurons via the spinal cord, causing contractions of

SC

RI

the muscles and thus, the escape reflex.

In this study, we investigated the effects of metal exposure on motor neurons and neuromasts,

NU

morphological development and the escape response of wild-type zebrafish embryos. With this

MA

approach, it was intended to link cellular, morphological and functional aspects of adverse effects of metal toxicity. Specifically, we examined the toxic effects of cobalt, cadmium and

D

copper by using 72 h FETs in combination with these additional endpoints, to evaluate whether

TE

an improvement in the sensitivity to metal exposure can be achieved and to discuss the

AC CE P

ecotoxicological significance of the effects observed. 2. MATERIALS AND METHODS 2.1. Chemicals

Suitable exposure concentrations of copper and cadmium cobalt were selected based on the results of previous FET range-finding experiments. For cobalt, the test concentrations were chosen based on published data (Reinardy, Teyssie, 2011, Sylvain, Brewster, 2010) since no concentration-response could be established for morphological effects within the water soluble concentration range. Copper was applied as copper sulfate (CuSO4, Sigma-Aldrich) at concentrations of 0.011, 0.014, 0.024, 0.068, 0.152 and 0.364 mg Cu/L, and was chosen because of its known toxicity towards aquatic organisms, e.g. LC50 = 13.82 μM in zebrafish at 4 dpf (Hernandez, Undurraga, 2011). Cadmium was applied as cadmium chloride (CdCl2, Sigma-

7

ACCEPTED MANUSCRIPT Aldrich) at concentrations of 2.0, 4.2, 8.9, 18.2 and 34.8 mg Cd/L. Cobalt was applied as cobalt(II) sulfate heptahydrate (CoSO4.7H2O, Sigma-Aldrich) at concentrations of 0.8, 2.6, 6.4,

PT

16.6 and 45.8 mg Co/L. Ethanol (ROTIPURAN® ≥99.8%, C2H6O, Carl Roth) was used as a positive control substance for neurotoxic effects. Concentrations of 1.0, 1.5, 2.0, 2.5 and 3.0%

RI

were chosen for the exposure according to the results of previous studies (Muth-Köhne et al.

SC

2012; Sylvain et al. 2010). The antibiotic neomycin (neomycin trisulfate salt hydrate;

NU

C23H46N6O13.3H2SO4 x H2O, Sigma-Aldrich) was applied as a positive control for ototoxic effects and used at concentrations of 1.0, 3.16, 10, 31.62 and 100 mg/L. We also used the anti-

MA

cancer drug cisplatin (cis-diamineplatinum(II) dichloride; H6Cl2N2Pt, Sigma-Aldrich) at concentrations of 1.0, 2.66, 7.07, 18.8 and 50 mg/L to evaluate hair cell regeneration potential

D

(Mackenzie and Raible, 2012). The exposure concentrations of neomycin and cisplatin were

TE

selected according to the results previously performed FET range-finding experiments (data not

AC CE P

shown). 2.2. Analysis of metals

Metal concentrations in the ISO water solutions used for the tests were quantified by inductively coupled plasma mass spectrometry (ICP-MS) using a Perkin Elmer Elan 6000. All measurements were carried out according to DIN EN ISO 17294-2 (17294-2, 2005-02). The instrument was calibrated using at least eight equidistant standard solutions (Merck, Germany). Generally, metals were quantified using the most abundant isotope without isobar or other interferences. All samples were analyzed without any further preparation using an instrument controlled auto sampler. However, samples were diluted if the measured concentrations were outside of the calibration range. The quality of the measurements was confirmed by the analysis of certified reference materials (surface water level 2; SPS, Norway).

8

ACCEPTED MANUSCRIPT To confirm that the concentrations were constant over time, the test solutions were measured at the test start and after 72 h. The total metal concentrations are shown as the means ± standard

PT

error of the mean (SEM) of two measurements.

RI

Table 1 shows that the metal concentrations were in the same range at the beginning and end of

SC

the test for all three metals, indicating that the solutions were stable over time. The concentrations of copper, cadmium and cobalt were in the same range as the nominal

NU

concentrations for all test solutions. The variance between the nominal and empirical

MA

concentrations exceeded 20% in some cases, so the results refer to the measured copper, cadmium and cobalt concentrations and are depicted as mean empirical values.

AC CE P

TE

D

Table 1: Concentrations of copper (Cu), cadmium (Cd) and cobalt (Co) determined by inductively coupled plasma mass spectrometry (ICP-MS). Concentrations are shown as mean values with the standard error of the mean. Cu (mg/L) Cd (mg/L) Co (mg/L) nominal measured nominal measured nominal measured 0h 72 h 0h 72 h 0h 72 h 0.000 ± 0.000 ± 0.00 ± 0.00 0.0 ± 0.0 ± 0.000 0.000 0.000 0.00 0.00 ± 0.00 0.00 0.00 0.00 0.011 ± 0.011 ± 2.05 ± 2.03 ± 0.78 ± 0.77 ± 0.010 0.000 0.000 2.04 0.01 0.02 0.76 0.02 0.01 0.014 ± 0.014 ± 4.19 ± 4.23 ± 2.43 ± 2.85 ± 0.020 0.002 0.002 3.72 0.04 0.03 2.02 0.03 0.04 0.024 ± 0.024 ± 8.88 ± 8.92 ± 6.52 ± 6.30 ± 0.044 0.001 0.001 6.78 0.14 0.06 5.38 0.03 0.00 0.068 ± 0.068 ± 17.80 ± 18.60 ± 16.20 ± 17.00 ± 0.088 0.001 0.001 12.35 0.70 0.80 14.30 0.40 0.40 0.152 ± 0.152 ± 35.45 ± 34.05 ± 45.75 ± 45.90 ± 0.187 0.000 0.000 22.49 0.45 0.05 38.02 0.45 0.10

2.3. Exposure of zebrafish embryos Zebrafish embryo toxicity test were carried out following the principles of the DIN EN ISO 15088 (15088:2009-06). Wild-type zebrafish from established breeding groups at Fraunhofer

9

ACCEPTED MANUSCRIPT IME (originally obtained from West Aquarium GmbH, Bad Lauterberg, Germany) were used for egg production. The zebrafish were cultured in large groups in 200–300-L tanks under flow-

PT

through conditions at 27 ± 1°C with a 14-h photoperiod. For egg collection, shallow glass spawning trays covered with metal mesh lids were placed into the tanks before illumination and

RI

were left for 1 h. Metal exposure experiments were carried out using 96-well U-bottom

SC

polystyrene microtiter test plates (Greiner Bio-one, Germany), with one plate representing one

NU

replicate within a test. The wells of the plates were saturated overnight with 200 µl of the corresponding test solutions, which were replaced prior to the test start. All substances were

MA

prepared in ISO-standard water (prepared according to OECD guideline 203, Annex 2, and diluted 1:5). The 1:5 diluted ISO-water is a soft water with a water hardness of 4° dH and a

D

carbonate hardness of 1° dH. The pH in all tests remained between 7.5 and 8.5.

TE

Due to a lack of morphological effects in the zebrafish embryos after exposure to cobalt in 1:5

AC CE P

diluted ISO water, a supplementary test was performed to investigate a potential carbonate hardness dependent cause. To this end, embryos were exposed to the highest cobalt concentration of 45.8 mg/L using different dilutions of the ISO-standard water as test medium to simulate carbonate hardness conditions between 1° (17.85 mg/L) and 4°dH (89.25 mg/L). This supplementary test was not part of the original study plan and the method description and the results are therefore only included in the Appendix of this paper. After harvest of the eggs, the embryos were transferred to Petri dishes containing the test solutions as quickly as possible to ensure early exposure ( 1 hpf). The eggs were examined under a microscope, and those which were coagulated, unfertilized or younger than the four-cell stage were removed. Afterwards, the eggs were individually transferred from the Petri dishes to the wells (one embryo/well) of the plates and incubated for 72 h at 26 ± 1°C with a 14-h

10

ACCEPTED MANUSCRIPT photoperiod. Each test consisted of two replicate plates for each treatment (chemical exposure or ISO-water control) as technical replicates to account for possible plate effects. Each plate

PT

contained 24 embryos for controls and exposures, accounting for n=48 embryos per treatment and test. An internal plate control of 12 wells and embryos was also included on each plate. All

RI

tests were repeated three times to account for possible inter-assay variations. The total exposure

SC

time for the morphological effects was 72 h, with assessments after 24, 48 and 72 h. The tests

NU

were terminated after 72 h since the neuromast staining works most efficient at this time point and the method requires termination and fixation of the embryos. The 96 h assessment time point

MA

recommended by the OECD guideline 236 was therefore omitted and the final post-hatch mortality rate determined 24h early at 72 hf. Motor neuron and neuromast staining, and

D

behavioral assays, were carried out after 48 h and 72 h exposure, respectively. The assessment

TE

time points are recorded as hpf based on the incubation time, disregarding the development stage

AC CE P

of the embryos  1 hpf at the start of exposure. Sublethal and lethal morphological effects were evaluated according to Braunbeck and Lammer (2006) (Braunbeck and Lammer, 2006). A list of all morphological endpoints (Table A.1) assessed can be found in the appendices. 2.4. Immunofluorescence staining of motor neurons Embryos exposed to metals or ethanol for 48 h (and corresponding controls) were dechorionated manually and fixed in 4% (w/v) paraformaldehyde (Sigma Aldrich) in phosphate-buffered saline (PBS, Invitrogen) for 4 h at room temperature, then washed three times in PBS containing 0.1% (v/v) Triton X-100 (PBST) and stored at 4°C. Whole-mount immunostaining was carried out according to Westerfield (2000) (Westerfield, 2000) and as modified by Muth-Köhne et al. (2012) (Muth-Kohne et al. , 2012), with the modification that permeabilization and blocking was achieved by incubation with 4% Triton X-100 plus 10% (v/v) normal goat serum in PBST

11

ACCEPTED MANUSCRIPT (PBS/4TN) for 30 min. The embryos were then incubated with a primary antibodies mixture: znp1, a mouse monoclonal antibody (IgG2a) specific for PMNs and zn8, a mouse monoclonal

PT

antibody (IgG1) specific for SMNs (Trevarrow et al. , 1990) obtained from the Developmental Hybridoma Bank, University of Iowa, USA). The secondary antibodies were DyLight 549-

RI

conjugated AffiniPure Goat Anti-Mouse IgG, specific for Fc subclass 1, and DyLight 649-

SC

conjugated AffiniPure Goat Anti-Mouse IgG, specific for Fc subclass 2a (Jackson

NU

ImmunoResearch Europe). Antibody incubation, imaging and analysis were carried out as described by Muth-Köhne et al. (2012, 2013) (Muth-Kohne, Sonnack, 2013, Muth-Kohne,

MA

Wichmann, 2012). We examined individual motor neurons in the area above the yolk sac extension (approximately 10 motor neurons per embryo). Each percentage value represents the

D

mean (± SEM) proportion of normally-developed motor neurons, and motor neurons with

TE

toxicity effects that we classed as mild (i.e. delayed development, axons lacking stereotyped

AC CE P

morphology), moderate (i.e. axons with ectopic branches or innervating neighboring myotomes) or severe (e.g. truncated axons) in embryos derived from three independent test replications, with 5–6 individual embryos in each replicate. 2.5. Vital dye staining of the neuromasts Non-hatched zebrafish embryos at 72 hpf were manually dechorionated and transferred to 1.5mL Eppendorf tubes filled with the vital dye solution (0.1 mg/mL DASPEI in ISO water) and stained in the dark for 20 min. The dye was removed carefully by repeatedly washing with 1:5 diluted ISO water. It was important to ensure the vitality of the larvae, otherwise selective staining of the neuromasts would have been impossible and the whole fish would be stained. The embryos were embedded horizontally on microscope slides using one drop of 3% methylcellulose. The living larvae were analyzed under a DMI6000 microscope using a Leica

12

ACCEPTED MANUSCRIPT AF6000 system. Z-stack image series were acquired from each embryo and the images were processed by generating a full-focus image with the ImageJ plugin “Stack Focuser” and then

PT

analyzed by measuring the fluorescence intensity of the neuromasts using ImageJ (Version 1.46, National Institutes of Health). Five neuromasts were measured per embryo in the region between

SC

replications, with 5–6 individual embryos in each replicate.

RI

the distal end of the yolk sac and the end of the tail. We analyzed three independent test

NU

2.6. Behavior assay

MA

In order to evaluate behavioral responses after tactile stimulation, embryos were exposed to the test solutions in 24-well flat-bottom plates, with each well filled with 2 mL of test solution and

D

1:5 diluted ISO water. We examined 12 embryos (72 hpf) at each test concentration. Each

TE

embryo was transferred to a single well. Unhatched embryos were dechorionated. A 10-μL

AC CE P

pipette tip was used to nudge the embryos and the responses were recorded. To measure the embryo behavior, a scoring system was developed defining four different levels of escape behavior (Buck et al. , 2012): the normal escape response (normal startle response or C-start), reduced escape response (1–2 cm movement), minimal escape response (1–5 mm movement) and no escape response (no movement). 2.7. Statistical analysis For all substances were a full concentration-effect relationship could be determined the EC50 and LC50 values with corresponding 95% confidence intervals (CIs) for sublethal and lethal morphological endpoints were calculated by applying probit analysis using ToxRat Professional v2.10.

13

ACCEPTED MANUSCRIPT Further statistical analysis was carried out using GraphPad Prism software v5.0. All data were tested for normality and homogeneity of variance. A one-way ANOVA with Dunnett’s method

PT

of post hoc testing for multiple comparisons (*p = 0.05; **p = 0.01; ***p = 0.001) was conducted to test for significant differences between the control and exposure groups for a

SC

RI

respective test parameter.

NU

3. RESULTS

MA

3.1. Morphological effects in zebrafish embryos

The sublethal and lethal morphological effects of the three metals were assessed after 24, 48 and

D

72 h, as well as for ethanol (the control substance for motor neuron damage) and

TE

neomycin/cisplatin (control substances for ototoxicity). Figure 1 shows the results at 72 hpf,

determined.

AC CE P

whereas Table 2 lists the EC50/LC50 values for 24, 48 and 72 hpf, where these could be

14

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC CE P

Figure 1: Proportion of 72-hpf embryos displaying morphological effects after exposure to different concentrations of (A) cobalt (Co), cadmium (Cd) and copper (Cu) and (B) control substances (neomycin, cisplatin and ethanol). Bar charts show the concentration-dependent increase in embryos displaying any effect (white bars), and exemplary no hatch embryos (light gray), embryos with edema (darker gray) and coagulated eggs (black). The white bars depict the sum of all recorded effects according to table A1 (appendix), other bars show the predominant morphologic effects coagulation, no-hatch and edema; statistical significance versus control groups tested for “eggs with any effect” (white bars) only (Oneway ANOVA with post hoc Dunnett´s test *= p < 0.05, **= p < 0.01, ***= p < 0.001). The embryos exposed to cobalt (Fig. 1 A, left) showed no morphological effects after 72 h at any of the tested concentrations in 1:5 diluted ISO water. Morphological effects after cobalt exposure occurred when in ISO water of higher carbonate hardness was used as test medium. These results can be found in the Appendices, cobalt and carbonate hardness. However, exposure to cadmium resulted in a concentration-dependent increase in coagulation, up to 90% at the highest concentration of 34.8 mg/L (Fig. 1 A, middle). The calculated mean EC50 Cd values remained

15

ACCEPTED MANUSCRIPT fairly constant (13.95 mg/L at 24 hpf, 14.68 mg/L at 48 hpf and 14.59 mg/L at 72 hpf), as did the LC50Cd values, which were approx. 1 mg higher (Table 2).

72 hpf

n.d.

n.d.

n.d.

13.95 (Cl 7.25 36.32) 0.115 (Cl 0.05 0.53)

14.68 (Cl 7.10 48.18) 0.118 (Cl 0.05 0.42)

14.69 (Cl 6.33 69.16) 0.018 (Cl 0.01 0.04)

Neomycin [mg/L]

n.d.

n.d.

Cisplatin [mg/L]

n.d.

Ethanol [%]

1.34 (Cl 1.29 1.39)

MA

D

Copper [mg/L]

TE

Cadmium [mg/L]

n.d.

AC CE P

Cobalt [mg/L]

1.27 (Cl 1.22 1.31)

24 hpf

48 hpf

72 hpf

n.d.

n.d.

n.d.

15.27 (Cl 6.79 72.36) 0.121 (Cl 0.048 0.550)

15.58 (Cl 6.21 129.06) 0.124 (Cl 0.047 0.519)

15.84 (Cl 5.94 219.55) 0.125 (Cl 0.050 0.484)

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

1.87 (Cl 1.46 2.22)

1.76 (Cl 1.47 2.01)

1.59 (Cl 1.4 1.72)

SC

48 hpf

NU

24 hpf

RI

PT

Table 2: Mean EC50 and LC50 values of cobalt, cadmium, copper, neomycin, cisplatin and ethanol with the corresponding 95% confidence intervals for sublethal and lethal morphological effects calculated using probit analysis (ToxRat Professional v2.10, ToxRat Germany). EC50 LC50

n.d.

3.11 (Cl 2.80 3.45) 1.16 (Cl 1.12 1.20)

Embryos exposed to copper also displayed a concentration-dependent increase in coagulation from 24 hpf (data not shown), but we also observed an additional no-hatch effect at 72 hpf. Exposure to copper at a concentration of 0.014 mg/L already suggested a no-hatch effect (Fig. 1 A, right) although this was not significant due to the large deviation among the three biological replicates. The mean EC50Cu values were 0.118 mg/L at 48 hpf and 0.018 mg/L at 72 hpf, with the much lower EC50Cu value at 72 hpf reflecting the no-hatch effect (Table 2). When the metals were compared, copper was found to be the most toxic with the lowest EC50 value of 0.018 mg/L after 72 hpf. The antibiotic control substance neomycin (Fig. 1 B, left) showed no morphological

16

ACCEPTED MANUSCRIPT effects after 72 h at all tested concentrations, as seen before for cobalt. Cisplatin-treated embryos also showed no significant morphological effects at 24 and 48 hpf, regardless of the

PT

concentration (data not shown). However, the 72 hpf embryos demonstrated a significant nohatch effect at a concentration of 2.66 mg/L cisplatin (Fig. 1 B, middle). At 24 and 48 hpf, an

RI

EC50Cis could not be determined because there were no morphological effects, but after 72 hpf

SC

the calculated mean EC50Cis was 3.11 mg/L (Table 2). We were unable to determine the LCx

NU

because only sublethal morphological effects were observed. Zebrafish embryos exposed to ethanol showed a concentration-dependent increase in sublethal and lethal effects such as edema

MA

and coagulation from 24 hpf. Furthermore, many embryos did not hatch until 72 hpf, as was observed also for the copper and cisplatin exposure. Even at the lowest concentration of 1%

D

ethanol was sufficient to induce a significant increase of morphological effects (Fig. 1 B, right).

TE

The calculated mean EC50Ethanol values reduced over time, with 1.34% at 24 hpf and 1.16% at

AC CE P

72 hpf, and the LC50Ethanol value at 72 hpf was 1.59% (Table 2). Because there was 100% coagulation in embryos treated with the highest concentrations of copper (0.364 mg/L), cadmium (34 mg/L) and ethanol (2.5% and 3%), it was not possible to measure motor neuron/neuromast defects or behavior in these treatments.

3.2.Effects on motor neuron development Zebrafish embryos exposed to all three metals were characterized by lower numbers of normally developing PMNs (Fig. 2, A) and SMNs (Fig. 2, B) at high concentrations.

17

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Figure 2: Representative inverted fluorescent images of stained (A) primary motor neurons (PMN) and (B) secondary motor neurons (SMN) in control embryos and embryos treated with cobalt (Co), cadmium (Cd) and copper (Cu). Arrows point to abnormal PMNs and SMNs which are branched, truncated or innervate into neighboring axons. Bar charts show the concentration-dependent reduction in the proportion of normally-developed motor neurons (white bars) and the increase in the proportion of motor neurons with minor (light gray), moderate (darker gray) and severe (black) defects; statistical significance versus control groups tested for “normally-developed motor neurons” (white

18

ACCEPTED MANUSCRIPT bars): (One-way ANOVA with post hoc Dunnett´s test, 3 independent experiments with 5-6 individual embryos; 8-9 motor neurons per embryo *= p < 0.05, **= p < 0.01, ***= p < 0.001)

PT

For cobalt-treated embryos, the reduction was significant for PMNs (Fig. 2 A, left) and SMNs (Fig. 2 B, left) at the highest concentration of 45.8 mg/l. For cadmium-treated embryos, the

RI

effects on both parameters were significant already at 8.9 mg/L (Fig. 2, middle). PMNs were not

SC

significantly affected in copper-treated embryos, but an adverse effect on SMNs was already

NU

significant at a concentration of 0.024 mg/L (Fig. 2 A + B, right). In the other treatments, significant effects on SMNs and PMNs were detected at the same exposure concentrations. The

MA

motor neuron effects of all three metals included moderate to severe defects, from excessively branched and truncated axons to axons innervating neighboring myotomes at the two highest

D

concentrations (see images of Fig 2 A and B, orange arrows). Furthermore, weakly developed or

TE

missing SMNs were frequently observed towards the distal ends of the tail region. This effect

AC CE P

was considered mild and suggests a delayed development of the motor neurons. Ethanol-treated embryos (positive controls) showed a significant reduction in the number of normal motor neurons at the lowest concentration of 1% (Fig. A.1 A, Appendices), and certain severe effects were also observed, such as truncation of the axons. 3.3.Neuromast damage

Neuromast damage was analyzed to measure the disruption of the hair cells of the lateral line. Figure 3 shows fatal effects after 72 h. A concentration-dependent reduction in the fluorescence intensity was clearly observed for all three metals (Fig 3), and significant effects were observed already at the lowest treatment concentration of copper and cadmium respectively the second lowest of cobalt (2.6 mg/L). The highest reduction in fluorescence intensity was seen in the cadmium-treated embryos, where the intensity dropped to less than 30% across all

19

ACCEPTED MANUSCRIPT concentrations (Fig. 3 B, middle), although measurements were not possible at the highest

AC CE P

TE

D

MA

NU

SC

RI

PT

concentration (34.8 mg/L) due to the extent of coagulation (>85%).

Figure 3: (A) Representative inverted fluorescent images of stained neuromasts in control embryos and embryos treated with cobalt (Co), cadmium (Cd) and copper (Cu). The green arrows show five normally fluorescent neuromasts and the orange arrows affected neuromasts expressing weaker or no fluorescence. (B) Relative fluorescence intensity (%) of stained neuromasts in 72-hpf embryos after exposure to cobalt (Co), cadmium (Cd) and copper (Cu); statistical significance versus control groups: (One-way ANOVA with post hoc Dunnett´s test; n = 75 – 90 of 3 independent experiments with 5-6 individual embryos and 5 neuromasts measured per embryo *= p < 0.05, **= p < 0.01, ***= p < 0.001) In copper-treated embryos, the fluorescence intensity was significantly reduced at the lowest concentration to ~70% compared to the control group (Fig. 3 B, right). Between the lowest concentration (0.011 mg/L) and the next higher concentration (0.014 mg/L), there was no effect variation. The positive control substances neomycin and cisplatin induced a substantial and

20

ACCEPTED MANUSCRIPT significant reduction in the fluorescence intensity of neuromasts, starting at 3.16 mg/L neomycin with 20% fluorescence intensity and 2.66 mg/L cisplatin with a fluorescence intensity of

PT

approximately 35% compared to the control (Fig. A.1 B, Appendices).

SC

RI

3.4.Behavioral effects

In addition to the motor neuron and neuromast damage, we also considered the escape behavior

NU

of the embryos after tactile stimulation. The 72-hpf embryos exposed to cobalt displayed a slightly reduced escape response, which was significant at the highest concentration of 45.8

MA

mg/L (Fig. 4, left). This effect threshold was consistent with the motor neuron damage (Fig. 2), but lower than for the damage to neuromasts (Fig.3). In cadmium-treated 72-hpf embryos,

TE

D

exposure to concentrations of 18.2 mg/L affected the escape response which was indicated already at 8.9 mg/L. (Fig. 4, middle). First effects of cadmium on motor neurons were seen at the

AC CE P

same concentration whereas neuromast damage occurred only at 2.0 mg/L.

Figure 4: Tactile stimulation response assay at 72 hpf after exposure to cobalt (Co), cadmium (Cd) and copper (Cu). Bar charts show the concentration-dependent reduction in the normal escape response (white bars) and an increase in a reduced escape (light gray), nearly no escape (darker gray) and no escape (black) after touch; statistical significance versus the control groups was tested for “normally escape response” (white bars): (Oneway ANOVA with post hoc Dunnett´s test; n = 36 per treatment of 3 independent experiments with 12 individual embryos *= p < 0.05, **= p < 0.01, ***= p < 0.001).

21

ACCEPTED MANUSCRIPT The copper-treated embryos showed a significant reduction in the normal escape response at 0.024 mg/L (Fig. 4, right). The lowest observed effect concentration (LOEC) of copper for the

PT

behavioral endpoint did not differ from the morphological effects (Fig. 1A). Copper induced motor neuron (SMN) and neuromast damage instead became significant already at the lower

RI

concentration of 0.011 mg/L. Neomycin and cisplatin were used as positive control substances

SC

for the neuromast damage test, and these substances also induced a significant concentration-

MA

mg/L cisplatin (Fig. A.1 C, Appendices).

NU

dependent reduction in the tactile stimulus response, starting at 10 mg/L neomycin and 2.66

4. DISCUSSION

D

Since many mechanisms of metal toxicity in fish remain to be unraveled, it is difficult to fully

TE

evaluate the adverse outcome of metal exposure based on the read outs of conventional fish tests

AC CE P

and existing knowledge. We used the zebrafish embryo toxicity test (zFET) to integrate mechanistic and functional aspects of subacute metal toxicity with teratogenicity. Hence, the scope of the existing test was extended by the phenotypic endpoints motor neuron and neuromast damage and behavioral impairment. Whether this approach could indicate adverse outcomes of metal toxicity and increase the explanatory power of the zFET was explored for copper, cadmium and cobalt. Cobalt did not induce any detectable morphological effects in 72-hpf zebrafish embryos even at the highest concentration of 45.8 mg/L (Fig. 1 A, left). In comparison, cobalt at concentrations greater than 10.8 mg/L in water (100 mg/L CaCO3) was previously reported to inhibit hatching in 72-hpf zebrafish embryos (Dave and Xiu, 1991) and even 0.1 mg/L cobalt was sufficient to inhibit hatching when Hank’s solution was used as the medium (Cai, Zhu, 2012). The absence of

22

ACCEPTED MANUSCRIPT this effect in our investigation may therefore be medium-dependent because we used cobalt dissolved in 1:5 diluted ISO water (

Effects of metal exposure on motor neuron development, neuromasts and the escape response of zebrafish embryos.

Low level metal contaminations are a prevalent issue with often unknown consequences for health and the environment. Effect-based, multifactorial test...
1MB Sizes 0 Downloads 8 Views