Introduction
A wide variety of organic compounds is frequently detected in streams that receive agricultural, domestic and (or) industrial wastewater effluent. These contaminants include antibiotics, other prescription drugs, non-prescription drugs, animal and plant steroids, reproductive hormones, personal care products, detergent metabolites and other extensively used chemicals. If the drugs, their metabolites and transformation products are not eliminated during sewage treatment, they may enter the aquatic environment and eventually reach drinking water (Kümmerer, Reference Kümmerer2009). In this sense, medical substances (Heberer, Reference Heberer2002), some active pharmaceutical ingredients (Kallenborn et al., Reference Kallenborn, Fick, Lindberg, Moe, Nielsen, Tysklind, Vasskoog and Kümmerer2008) and some persistent organic pollutants (POPs) (Ericson et al., Reference Ericson, Nadal, Van Bavel, Lindström and Domingo2008; Wilhelm et al., Reference Wilhelm, Bergmann and Dieter2010; Ullah et al., Reference Ullah, Alsberg and Berger2011; Eschauzier et al., Reference Eschauzier, Beerendonk, Scholte-Veenendaal and de Voogt2012), have been detected in drinking water. Unfortunately, the complexity of the mixtures of these substances, their interactions and, in general, their variable and random presence at even relatively low levels in drinking water, make their routine chemical detection and control difficult or even impossible. In consequence, bioindicators are used as an alternative to monitor their presence and zebrafish is currently being used to assess certain contaminants in water quality studies (Ansari & Sharma, Reference Ansari and Sharma2009; Molinari et al., Reference Molinari, Troisi, Fallico, Paparella and Straface2009).
In recent years there has been a growing body of evidence that exposure to chemicals in the environment poses a serious threat to human and animal reproduction (Gregoraszczuk & Ptak, Reference Gregoraszczuk and Ptak2013). Moreover, it has been widely demonstrated that different chemical compounds, such as endocrine-disrupting chemicals and reproductive toxicants (diethylstilbestrol (DES), bisphenol A (BPA), POPs, etc.), are able to modify the epigenetic marks and induce persistent and in some cases transmissible changes of epigenetic states (Baccarelli & Bollati, Reference Baccarelli and Bollati2009), but without changes in DNA sequence (Wolffe & Guschin, Reference Wolffe and Guschin2000), which is the main epigenetic feature. As our long-term aim is to detect the possible presence of epigenetic factors in drinking water with effects on reproduction, using the zebrafish as bioindicator, an initial aspect to be considered is the possible barrier effect of the chorion to the passage of these substances into the embryo until hatching.
As in other fish species, the zebrafish (Danio rerio) embryo is surrounded by an acellular envelope, the chorion, between 1.5 and 3.5 μm in thickness (Bonsignorio et al., Reference Bonsignorio, Perego, Del Giacco and Cotelli1996; Rawson et al., Reference Rawson, Zhang, Kalicharan and Jongebloed2000). The chorion contains pores 0.17 μm2 in size (Cheng et al., Reference Cheng, Flahaut and Cheng2007). These pores were reported to be abundant at a rate of approximately 7.2 × 105 pores per chorion (Hart & Donovan, Reference Hart and Donovan1983). The pores might be responsible for size-dependent restrictions on the uptake of chemicals that have been reported for some large compounds exceeding 3 kDa (atomic mass unit), such as fluorescein dextrans (Creton, Reference Creton2004). Nevertheless, despite ongoing speculation it is not clear if these pores are permeable or what is their exact function and, consequently, whether the chorion is or is not an effective barrier (Henn, Reference Henn2011).
In the case of zebrafish, hatching occurs between 48–72 h post fertilisation (hpf), when organogenesis is almost complete (Kimmel et al., Reference Kimmel, Ballard, Kimmel, Ullmann and Schilling1995). Before that time, two relevant reproductive events take place: the first is genital ridge formation, followed by the migration of primordial germ cells (PGCs) to these genital ridges (Yamaha et al., Reference Yamaha, Saito, Goto–Kazeto and Arai2007). Moreover, it must be taken into account that the germ-cell-specific marker was detected in the cleavage planes in 2-cell and 4-cell-stage embryos (Yoon et al., Reference Yoon, Kawakami and Hopkins1997). Obviously, the presence and permeation to the chorion of chemicals could influence their effects on these two reproductive events.
The immediate solution to a possible permeation restriction would be to remove the chorion completely by pronase treatment and/or mechanically. Although removal of the embryo chorion is efficient at epiboly or later stages (Truong et al., Reference Truong, Harper and Tanguay2011) even with automated dechorionation systems (Mandrall et al., Reference Mandrall, Truong, Jephson, Sarker, Moore, Lang, Simonich and Tanguay2012), unfortunately at the mid blastula transition (MBT) stage (3 hpf) the efficiency of complete dechorionation with pronase treatment is low and, moreover, time consuming (Simão et al., Reference Simão, Cardona-Costa, Pérez Camps and García-Ximénez2010a,Reference Simão, Cardona-Costa, Pérez Camps and García-Ximénezb). Taking into account that a reproducible technique to avoid the barrier effect of the chorion with a high survival rate is required, in the present work the chorion will be partially degraded instead of eliminated. Moreover, early dechorionated embryos require a specific culture medium to reach the fry stage (Westerfield, Reference Westerfield2000), so it is essential, in our case, to assess the survival of chorion-degraded embryos when they are cultured in different drinking waters.
Material and methods
Care and maintenance of zebrafish colony
Adult zebrafish were kept in 20 l tanks in a 3:2 ratio (females: males) and fed on granular food supplemented with recently defrosted hen egg yolk and shrimp meat (Simão et al., Reference Simão, Cardona-Costa, Pérez Camps and García-Ximénez2010a). The light cycle was regulated at 14 h light/10 h dark (Brand et al., Reference Brand, Granato, Nüsslein-Volhard, Nüsslein-Volhard and Dahm2002; Matthews et al., Reference Matthews, Trevarrow and Matthews2002).
Fertilised embryos at the MBT stage (the phase in which the MZT maternal zygotic transition (MZT) has been completed (Westerfield, Reference Westerfield2000)) were selected under a stereomicroscope and left in a Petri dish with dechlorinated and decalcified tap water (Westerfield, Reference Westerfield2007). Embryos were used at the MBT stage is because it is the onset of the zygotic phase (Dahm, Reference Dahm, Nüsslein-Volhard and Dahm2002). Damaged embryos were discarded and only intact embryos were used in the experiment. No bleach treatment was applied, but sterilised media and materials (pipettes) in aseptic conditions were used. All chemicals and culture media were purchased from Sigma-Aldrich (Madrid, Spain).
Experimental design
Two consecutive experiments were carried out. The first experiment attempted to establish pronase treatment duration required for maximum partial degradation of the chorion but with a low mortality rate. Pronase solution was 1.5 mg/ml in H10, i.e. Hanks buffered salt solution (HBSS) diluted 10% (v/v) in ultrapure water (Millipore) (Westerfield, Reference Westerfield2007; Simão et al., Reference Simão, Cardona-Costa, Pérez Camps and García-Ximénez2010b). Each vial used contained 1.5 ml of this solution. Throughout the experiments we used three different commercial batches of pronase.
In each assay, batches of embryos at the MBT stage were treated with pronase solution, previously warmed to 28.5ºC, for 5, 10 or 15 min respectively, whereas control group embryos were not treated with pronase. After pronase treatment, embryos were washed twice, for 5 min each wash, in 50 ml of H10 (35 mOsm) at 28.5ºC to remove any pronase residue. Next, the chorion of the embryos was recorded with a camera connected to a stereomicroscope (Nikon SMZ 800, Japan) at ×15 magnification. Pictures of the chorion were captured with the VLC media player program to assess the presence of holes in accordance with the treatment duration. Embryos were kept in H10 at 28.5ºC for 5 days, because by this time the fry were already able to feed. Survival rates were evaluated at 2 h after pronase treatment (pronase lethal effects), and at 5 days post fertilisation (dpf). At least four replicates were performed in all experimental groups.
Once the pronase treatment duration was established, the second experiment tested the effect of the different drinking waters on embryo survival rates when they were cultured in these waters for 5 days after the pronase treatment. This second experiment is justified because it is not exactly clear if all types of drinking water will satisfy the culture requirements of dechorionated embryos.
The selected waters were intended to cover the range of water types to be studied in future assays. Except for H10 and mineral bottled water, sampling of tap water from the distribution network was performed on at least three different dates throughout the experiment.
Thus, pronase-treated and non-treated embryos at the MBT stage were cultured until 5 dpf at 28.5ºC in six different waters depending on their source: H10 stored either in a Pyrex glass (A) or plastic bottle (B), bottled spring water (C) and, finally, three waters (D, E, F) from different tap water distribution networks. Type D was tap water from a city located in a region with intensive agricultural activity from the hydrological basin of the river Xúquer; type E was tap water from a city also located in a region with intensive farming activity, but from the Túria river hydrological basin. Types D and E came from groundwater prospecting. Finally, type F was from the tap water distribution network of a medium-sized city supplied from both the Túria and Xúquer rivers. Survival rates were evaluated at 2 h after pronase treatment and 5 dpf. At least eight replicates were performed in all experimental groups. All water samples whatever the water types were left in open containers prior their use, letting the water stand for several days to eliminate the chlorinated volatile substances used in purification water process, before being heated to 28.5ºC.
In both experiments, results were analysed using the chi-squared test (Statgraphics Plus 5.1). The Yates correction for continuity was used when a single degree of freedom was involved.
Results
Results from the first experiment (see Table 1) showed that at 2 h after pronase treatment, time of exposure to this enzyme (5, 10 or 15 min) progressively penalised the survival rates. The worst result was reached with the group exposed for 15 min. From 2 h after pronase treatment to 5 dpf, ruling out the direct immediate effect of pronase treatment evaluated at 2 h, there were no significant differences between the treated groups, although all of them differed significantly from the non-treated (control) group.
Table 1 Survival rates at 2 h after pronase treatment and until 5 dpf of mid blastula transition (MBT) embryos from zebrafish (Danio rerio) treated with pronase for 5, 10 and 15 min. The control group was not treated with pronase
In each column, rows with different superscripts are statistically different (P < 0.05).
With regard to images captured, after 10 min of pronase treatment holes were observed in some cases, however the most evident presence of holes, in all cases, was at 15 min.
Based upon results from the first experiment, pronase treatment for 15 min was selected to ensure partial degradation of the chorion but without an excessive reduction in embryo survival.
In the second experiment, survival rates at 5 dpf were compared between embryos treated or not with pronase when cultured in each water type. The initial number of embryos was considered at 2 h after pronase treatment in the treated groups (see Table 2).
Table 2 Survival rates of embryos from zebrafish (Danio rerio) treated and not treated with pronase for 15 min and cultured in different waters according to their source until 5 dpf
In each row, columns with different superscripts are statistically different (P < 0.05).
According to the results obtained in the second experiment, in which survival rates were compared between the six types of water, but independently in treated (chorion-degraded) and untreated groups, in the treated groups comparison the least statistically significant (P < 0.05) results were obtained in waters A, B and E. This differential effect of type of water on chorion-degraded embryos can distort the experimental results.
In contrast, in non-treated groups there were no differences between waters (P = 0.8257). This latter result can be explained because all types were drinking water (A and B types were H10) in which no acute toxicity was expected.
Overall, in all types of waters except the F type, degradation of the chorion with pronase reduced the embryo survival rates at 5 dpf, reaching levels of significance in embryos cultured in type A, type B and type E waters.
The absence of differences between water type A (stored in Pyrex glass) and B (stored, like the other waters, in plastic bottles suitable for commercial waters), would indicate that the container does not affect embryo survival rates in any case.
With regard to abnormality rates until 5 days post fertilisation (dpf), it was noted that these rates were low in all groups (1–2%), either treated or non-treated, except in type D water, where the abnormality rates in the pronase-treated group reached 4%.
Discussion
Similar observations to those obtained in the first experiment on the immediate detrimental effect of pronase on embryos have been already made by Henn & Braunbeck (2011) and others (Simão et al., Reference Simão, Cardona-Costa, Pérez Camps and García-Ximénez2010a) in toxicity and chimaerism studies.
In any case, results show that the effects caused by the pronase treatment, without complete dechorionation, affect embryo survival at 5 dpf, whatever the treatment duration.
The survival results of dechorionated embryos at 5 dpf were relatively low in H10 medium in both experiments. This medium was initially proposed by Westerfield (Reference Westerfield2000), although other authors later incorporated modifications to the same (Truong et al., Reference Truong, Harper and Tanguay2011).
On the other hand, results obtained in the second experiment indicated that the type of water affects chorion-degraded embryo survival at 5 dpf. In toxicity substance assessment studies, dechorionated embryos are used (Truong et al., Reference Truong, Harper and Tanguay2011) and a very efficient automated dechorionation system has even been developed (Mandrall et al., Reference Mandrall, Truong, Jephson, Sarker, Moore, Lang, Simonich and Tanguay2012) for denudation with pronase. However, in this type of study, the culture medium is always the same and only the concentration of the substance to be assessed is changed (Henn & Braunbeck, 2011; Truong et al., Reference Truong, Harper and Tanguay2011). Furthermore, embryo dechorionation is performed in epiboly (6 hpf) (Truong et al., Reference Truong, Harper and Tanguay2011) or more advanced stages. These results may, therefore, be the two biggest differences between our case and toxicology studies: the different culture media (drinking water) in which the embryos are cultured and the dechorionation of the embryos at an earlier stage (3 hpf).
The dilemma is therefore whether to keep the chorion intact and assume the loss of information on the initial reproductive effects or, alternatively, to permeabilize the chorion and in doing so introduce a distortion factor in the assessment of subsequent damage to embryos and, perhaps, in reproduction, depending on the water in which the embryos with the degraded chorion are cultured.
A possible solution to avoid such water distortion would be to remove the chorion at 24 h post fertilisation (hpf), with high survival rates at this time (90%) (Henn & Braunbeck, Reference Henn and Braunbeck2011). However, it has been stated that, before the 20 somites (19 hpf) stage, PGCs move to the genital ridge, whereas those after that stage mostly do not (Yamaha et al., Reference Yamaha, Saito, Goto–Kazeto and Arai2007). So, removing the chorion at 24 hpf will not prevent the barrier effect at earlier stages on PGCs and on differentiation of the genital ridge.
Conversely, leaving the chorion intact will prevent any water distortion on embryo survival rates (and, perhaps, on embryo organogenesis and cell differentiation) depending on the tap water source, as was detected in the second experiment. It must be remarked that in water quality studies embryos are used with the chorion intact because the hatching success of an embryo is an endpoint assessment (Hallare et al., Reference Hallare, Pagulayan, Lacdan, Köhler and Triebskorn2005; Wu et al., Reference Wu, Jiang, Zhang, Chen and Zhang2013).
It seems that the chorion is not a major barrier for simple chemicals, however there are exceptions such as cationic polymers, heavy metals and large molecules (e.g. polymers) (Henn, Reference Henn2011). In any case, the chorion retards the free exchange of substances (Harvey et al., Reference Harvey, Kelly and Ashwood-Smith1983). Alternatively, it is suspected that the chorion pores potentially restrict the uptake of compounds depending on their size (Creton, Reference Creton2004). Nevertheless, a reasonable molecular size cut-off value for fish embryo testing cannot be set (Henn, Reference Henn2011). Whatever the case, natural hatching in zebrafish occurs within 48–72 hpf, so the chorion barrier effect, if any, will only be present before that time.
It must be emphasised that in this stage, and in addition to the possible barrier effect of the chorion on the permeability, fish embryos also have low membrane permeability, with the presence of cell layers that act as osmotic barriers making the permeation of substances (e.g. cryoprotectants) extremely difficult (Hagedorn et al., Reference Hagedorn, Kleinhans, Wildt and Rall1997; Cardona-Costa & García-Ximénez, Reference Cardona-Costa and García-Ximénez2007). Obviously, this barrier effect of embryos does not depend on the presence or otherwise of the chorion.
With regard to the following reproductive milestones, these occur much later after hatching, and therefore will not be affected by the previous presence or absence of the chorion, at least directly. Thus, in zebrafish, the first sign of sex differentiation is initiated at 10 to 12 days post fertilisation (dpf) (Tong et al., Reference Tong, Hsu and Chung2010) and is completed at 35 dpf in females and 45 dpf in males (Weiting & Wei, Reference Weiting and Wei2013). It should be highlighted that the initial number of PGCs in the gonad may be related to subsequent sex differentiation (Lo et al., Reference Lo, Hui, Yu, Wu and Cheng2011).
In conclusion, the results obtained in this work advise against the partial or total denudation of early embryos when they are used to detect substances in different drinking waters and assume the possible limitation to the passage of any likely epigenetic factor that could affect the genital ridge or PGC migration.
Acknowledgements
The authors would like to thank Mr Javier Rubio Rubio for his valuable technical support and Mr Neil Macowan for improving the English of this manuscript.
