Using Whole mount In Situ Hybridization to Examine Thyroid Hormone Deiodinase Expression in Embryonic and Larval Zebrafish: a Tool for examining OH-BDE toxicity to early life stages

Abstract

Polybrominated diphenyl ethers (PBDEs) and their oxidative metabolites (hydroxylated PBDEs; OH-BDEs) are known endocrine disrupting contaminants that have been shown to disrupt thyroid hormone regulation both in mammals and in fish. The purpose of this study was to determine the precise organ and tissue locations that express genes critical to thyroid hormone regulation in developing zebrafish (Danio rerio), and to determine the effects of an OH-BDE on their expression. While RT-PCR can provide quantitative data on gene expression, it lacks spatial sensitivity to examine localized gene expression; and, isolation of organs from zebrafish embryos is technically difficult, if not impossible. For this reason, the present study used whole mount in situ hybridization to simultaneously localize and quantify gene expression in vivo. While PBDEs and OH-BDEs have been shown to inhibit the activity and expression of deiodionases, a family of enzymes that regulate thyroid hormone concentrations intracellularly, it is unclear whether or not they can affect regional expression of the different isoforms during early development. In this study we investigated deiodinase 1 (Dio1), deiodinase 2 (Dio2), and deiodinase 3 (Dio3) mRNA expression at the following life stages (2, 8, and 1k-cells; 50%-epiboly, 6 and 18-somites, 22, 24, 48, 72 hpf and/or 10 dpf) in zebrafish and found life stage specific expression of these genes that were highly localized. To demonstrate the use of this technique for investigating potential endocrine disrupting effects, zebrafish embryos were exposed to 1, 10 and 100 nM 6-OH-BDE-47. Significant increases in mean intensity of Dio1 and Dio3 expression in the periventricular zone of brain and pronephric duct, respectively (quantified by measuring intensity of coloration using ImageJ analysis software) were observed, suggesting localized response at the HPT axis with the possibility of impacting neurodevelopment. Our results demonstrate effects of OH-BDEs on thyroid regulating gene expression and provide more insight into potential sites of injury during early life stages.

1. Introduction

In humans, the thyroid hormones thyroxine (T4) and 3, 5, 3′-triiodothyronine (T3) play an important role in regulating development and growth in target tissues (Chan et al., 2003; Johnson and Lema, 2011). This is especially important in regards to the development of the central nervous system (CNS), which begins in utero (Calvo et al., 2002; Fisher, 1997; Kilby et al., 2000). Mild maternal hypothyroidism in early pregnancy has been associated with adverse neuropsychological outcomes, demonstrating that changes in maternal thyroid status affect fetal CNS development (Pop et al., 2003).

In most vertebrates (particularly mammals, fish and amphibians), T4 is considered a pro-hormone, and must first be metabolized to T3 in peripheral tissues by the selenocysteine containing enzymes deiodinase 1 (Dio1) or deiodinase 2 (Dio2) in order to bind to thyroid hormone nuclear receptors (Kohrle, 2000). Dio1 catalyzes both outer ring deiodination (ORD) and inner ring deiodination (IRD), while Dio2 only catalyzes ORD. In contrast, deiodinase 3 (Dio3) only catalyzes IRD and metabolizes T4 into the biologically inactive reverse triiodothyronine (rT3) (Darras and Van Herck, 2012). From the mammalian literature, Dio1 is primarily distributed in kidney, liver and thyroid (Brtko et al., 2002; Richard et al., 1998; Schoenmakers et al., 1992; Visser et al., 1988); Dio2 is distributed in brain, anterior pituitary, placenta, thyroid, skeletal muscle and heart (Bernal, 2002; Bianco et al., 2002; Croteau et al., 1996; Hosoi et al., 1999; Murakami et al., 2001; Salvatore et al., 1996a; Salvatore et al., 1996b); and Dio3 is found in the brain, uterus, and placenta (Darras et al., 1999; Galton et al., 1999; Richard et al., 1998; Salvatore et al., 1995). To date, studies investigating the presence and expression of Dio1, Dio2 and Dio3 have been mostly conducted in juvenile and adult rats.

In contrast to mammals, teleost Dio1 has been detected only in kidney and liver (Johnson and Lema, 2011(; Klaren et al., 2012), Dio2 has been detected in brain, liver, kidney and gonads but not in heart or skeletal muscle (Sambroni et al., 2001); and Dio3 expression was found in gill, brain, eyes and skin (Orozco and Valverde, 2005; Picard-Aitken et al., 2007; Sanders et al., 1999). However, it is worth noting that these studies were restricted to adults. This is due, in part, to the difficulty of characterizing gene expression in embryonic or larval fish that are small in size and undergo rapid differentiation and development (Teraoka et al., 2009).

Zebrafish embryos/larvae has been proposed as a good alternative model for human disease and developmental biology research, particularly for their use in identifying target genes responsive to endocrine disrupting chemicals in the environmental (Hoshijima and Hirose, 2007; Segner, 2009). For example, Walpita et al (2007) reported Dio mRNA expression in early life stages of zebrafish at 8, 12, 23, 30, 36, 48 and 75 hours post fertilization (hpf) using quantitative Real time polymerase chain reactions (qRT-PCR). From 8–48 hpf, Dio2 expression remained at low levels but increased dramatically at 75 hpf. Moreover, Walpita et al knocked down Dio2 expression and found that gene knockdown increased Dio1 mRNA expression, while knocking down Dio1 did not lead to significant change in Dio2 nor Dio3 expression (Walpita et al., 2010; Walpita et al., 2009).

Polybrominated diphenyl ethers (PBDEs) are a class of flame retardant chemicals that are ubiquitous contaminants and have known impacts on thyroid hormone regulation (Birnbaum and Staskal, 2004; Dingemans et al., 2011). They have been added to myriad products, including building materials, furniture,, textiles, plastics, electronic equipment and motor vehicles to decrease risks from fire (Schreiber et al., 2010). However, because of the large volume use of PBDEs, and their widespread contamination and accumulation in human tissues, there has been an increased concern about their potential to elicit human health effects. Maternal transfer (either in utero or through lactation) of PBDEs is one of the major routes of exposure to developing infants, which may be more vulnerable to their effects (de Wit, 2002). Studies suggest that early development may be particularly sensitive to PBDE exposures and toxicity. For example, rodents exposed to PBDEs in utero were found to develop problems with learning, memory function and behavior later in life(Johansson et al., 2008; Viberg et al., 2002, 2003, 2004, 2005, 2007; Viberg et al., 2006). Deficits in neurodevelopment in children have also been associated with PBDE exposure at early life stages (Eskenazi et al., 2013; Herbstman et al., 2010). PBDEs are structurally similar to thyroid hormones, and become even more similar in structure when they are hydroxylated through oxidative metabolism to form OH-BDEs (Qiu et al., 2007; Stapleton et al., 2009). Several studies have demonstrated that PBDE exposure is associated with changes in thyroid hormone concentrations, both in animal exposure studies and with the human population(Chevrier et al., 2010; Fernie et al., 2005; Stapleton et al., 2011; Tomy et al., 2004; Turyk et al., 2008; Zhou et al., 2001). In addition, in vitro studies have demonstrated that the OH-BDEs inhibit thyroid deiodinase activity and likely have greater potency than PBDEs in disrupting thyroid hormone regulation (Butt et al., 2011; Meerts et al., 2001).

Recently, Dio1, Dio2 and Dio3 transcripts were used as biomarkers of endocrine disruption in rats and in other organisms including fish (Schmutzler et al., 2007). Several studies indicated Dio1, Dio2 and Dio3 gene transcript levels were significantly affected by exposure to environmentally realistic concentrations of PBDEs (Egloff et al., 2011; Lema et al., 2008; Schmutzler et al., 2007). For example, after zebrafish embryos were exposed to 0.625 ppm 2,2,4,4′-tetrabromodiphenyl ether (BDE-47) from 24 to 28 hpf, Dio1 and Dio2 mRNA were up-regulated (Usenko et al., 2012). Exposure to the commercial deca-BDE flame retardant (BDE-209) for 14 or 21 days in zebrafish (Li et al., 2011) and Chinese rare minnow (Chen et al., 2012b) significantly up-regulated Dio1 and/or Dio2 mRNA expression. In addition, both BDE-99 and BDE-209 exposure resulted in up-regulation of Dio1 mRNA expression in human hepatocytes (Stapleton et al., 2009). In vivo studies in fish have also demonstrated that exposure to BDE-209 can lead to significant inhibition of Dio activity (Noyes et al., 2011). Furthermore, OH-BDEs have been shown to significantly inhibit activity of Di1 enzyme in vitro in human liver tissues (Butt et al., 2011).

All the above studies measured mRNA expression using qRT-PCR. In the studies with zebrafish and the Chinese rare minnow, results were derived from pools of whole animals and do not provide any information as to the organ/tissues in which these changes occurred, which would be crucial for understanding mechanisms of toxicity. We are then left to wonder whether changes restricted to specific sites would have been determined when diluted by non-responsive organ/tissues of the whole animal. While one could possibly isolate organs/tissues of interest, such tedious procedures are technically difficult in small developing organisms, such as zebrafish embryos. Whole mount in situ hybridization (WISH) has the potential to provide a better platform to localize and quantify gene expression in vivo (Dong et al., 2002). Given these considerations, the aims of this study were to develop a WISH method to examine localized gene expression of Dio1, Dio2 and Dio3 in early zebrafish embryos, and to use this method to investigate the effects of 6-OH-BDE-47, a common PBDE metabolite found in human tissues, and with known potential for Dio inhibition, on mRNA expression of the three deiodinase isoforms.

2. Materials and Methods

2.1. Fish care

Adult zebrafish (Danio rerio) were maintained in a recirculating AHAB system (Aquatic Habitats, Apopka, FL, USA) at 28 °C under a 14:10 light:dark cycle. Adult fish were fed brine shrimp and Zeigler’s Adult Zebrafish Complete Diet (Aquatic Habitats). Embryos were collected after natural spawning of adult zebrafish and were maintained in 30% Danieau solution in petri dishes in an incubator under the same conditions as the adults. From 5 day post-fertilization (dpf), larvae were transferred to 1 L beaker at a density of ~150 larvae/L. Between 6–16 dpf, larvae were fed with dry food LARVAL AP 100 (larval food supplement, Zeigler Bros Inc. Gardeners. PA. USA) 2 times daily. After 16 dpf, larvae were fed with brine shrimp nauplii daily. All larvae were maintained at same conditions as the adults. All care and reproductive techniques were non-invasive and approved by the Duke University Institutional Animal Care & Use Committee (A053-10-03).

2.2. Dosing

OH-BDE standards were purchased neat (>99.5% purity) from Accustandard (New Haven, CT). Stock solutions of 1mM concentration were prepared by dissolving an appropriate amount of the neat standard in dimethylsulfoxide (DMSO; Sigma, St. Louis, MO, USA). Exposure solutions were prepared via serial dilution from stock solution with fish culture system water with concentration of DMSO <0.4%. Zebrafish embryos were exposed to three concentrations of 6-OH-BDE-47 (1, 10 and 100 nM) from 4 hours post fertilization (hpf) until 22 hpf. Around 24 hpf pigmentation of the zebrafish embryo begins to occur. Therefore, exposures ceased at 22 hpf to prevent any confounding of the WISH method from pigmentation in the embryo.

2.3. Whole mount in situ hybridization

Upon reaching desired life stages, zebrafish embryos were fixed in 4% (w/v) paraformaldehyde in phosphate saline solution (pH 7.4) overnight, dechorionated by watchmaker’s forceps, and then stored at −20°C. Whole mount in situ hybridization was carried out according to (Dong et al., 2002). Embryos were hybridized with an antisense probe of 987 base pairs for zebrafish Deiodinase 1 (Dio1, BC_076008.1), an antisense probe of 774 base pairs for zebrafish Deiodinase 2 (Dio2, NM_212789.3), or an antisense probe of 809 base pairs for zebrafish Deiodinase 3 (Dio3, NM_001177935.2), respectively. Dio1, Dio2 and Dio3 probe were cloned with the following primers: Dio1 forward primer, 5′-TGCACTCGCAAATCTCTCACGATG-3′; Dio1 reverse primer, 5′-GCTACTGATGCTACCATTAC-3′; Dio2 forward primer, 5′-ATGGGCTTGCTTAGTGTG GACCTC-3′; Dio2 reverse primer, 5′-TCACTTTCCGTAGCACTTCTCCAG-3′; Dio3 forward primer, 5′-TAGACGTGCAGCACCGCGGA -3′; Dio2 reverse primer, 5′-CGCTCCAGCCAGTCTCTGAG-3′. Following hybridization overnight at 64°C, embryos were washed with 2× SSC (300 mM NaCl, 30 mM sodium citrate, pH 7.0) and 0.2× SSC twice for 30 min, respectively. Next the embryos were blocked with 2% blocking reagent (Roche, Mannheim, Germany), then incubated overnight with 3000× diluted anti-DIG antibody conjugated with alkaline phosphatase (Roche) at 4°C. Lastly, the color reaction was carried out by incubation with BM-purple substrate (Roche).

2.4. Expression localization and quantification

Dio1, Dio2 and Dio3 mRNA expression were recorded with a Nikon Eclipse E600 light microscope, a Nikon DXM 1200 digital camera, and EclipseNet imaging software (Nikon, Melville, NY). To best identify sites of Dio expression indicated by WISH, locations of these sites were compared and referenced with histologic sections of zebrafish of appropriate ages using the online Zfish atlas (http://zfatlas.psu.edu/progress.php).

For quantifying the intensity of Dio1, Dio2 and Dio3 expression, the following procedure was used. First the area, or tissue, of greatest staining intensity was outlined by ImageJ (National Institutes of Health, Bethesda, MD) Analysis Software and the sum of intensity (reaction product plus background) was determined in this region of interest (ROI). Next, the outline was moved to an adjacent region/tissue of the embryo that showed no reaction product and the background intensity was determined. This was then subtracted from the ROI. Using the sum intensity and dividing by the ROI pixel area, the mean intensity was calculated. A minimum of 5 images were examined per developmental stage per treatment.

2.5. Statistical analyses

Expression data are expressed as mean ± SD. GraphPad Prism 4 software was used for statistical analysis. Differences between means were analyzed using one-way ANOVA followed by Dunn’s test. A p value of <0.05 was considered statistically significant.

3. Result

3.1. Localization of Dio1, Dio2 and Dio3 mRNA by WISH

In this experiment, we used WISH to determine the embryonic tissue/organ distribution of Dio1, Dio2 and Dio3 mRNA in various early life stages of zebrafish embryos. Consistent results were found in at least two repeated experiments using 5 embryos or larvae in each developmental stage in each experiment.

A summary of tissue specific mRNA localization and intensity of the three deiodinases is show in Table 1. Dio1 and Dio2 mRNA expression was detected starting at the two cell stage (0.75 hour, Fig. 1A), and in all subsequent stages [four - (1 h, Fig. 1B), 8- (1.25 h, Fig. 1C), and 1k-cell (3 h, Fig. 1D), 50% epiboly (5.3 h, Fig. 1E), 6- (12 h Fig. 1F and G), and 18 somite (18 h, Fig. 1H), 24- (Fig. 1I), 72 -, and 96 hours post fertilization (hpf) and 10 days post fertilization (dpf) (Fig. 1J, K and L)]. Dio1 and Dio2 expression were co-localized from the two cell stage through the 18 somite stage (Table 1). Between two cell - through 1 k-cell stages, Dio1 and Dio2 mRNA was expressed ubiquitously in cells with variable intensity, and the highest expression was found to be at the 50% epiboly stage. From the 6--through the 18 somite stage, Dio1 and Dio2 mRNA expression was more concentrated at the rostral and caudal regions. Expression was also found in the periventricular zone of brain at 22 - and 24 hpf (Fig. 1I, and Fig. 4A and B), approximate to the cerebral cortex. After 24 hpf, Dio1 and Dio2 mRNA expression differed as to localization. There was only weak Dio1 mRNA expression in the mid-brain region at 48 and 72 hpf, and no other regions displayed any significant expression (Fig. 1J). At 96 hpf, Dio1 was strongly expressed in liver, kidney, and proximal intestine (Fig. 1K). From 96 hpf until 10 day post fertilization (10 dpf), Dio1 mRNA expression was found in the proximal intestine and in the internal tissue (adrenal gland equivalent tissue in zebrafish) (Fig. 1L). In contrast, Dio2 mRNA expression was found in adenohypophysis at 48 hpf (Fig 2. A, B and C), in the swim bladder at 96 hpf, and in the proximal intestine at 10 dpf (Fig. 2E and F). Dio3 mRNA expression was first found in pronephric duct from the 6-somite stage (Fig. 3A) and with increasing area and intensity through 22 - (Fig. 3B) and 24 hpf (Fig. 3C). There was also weak expression in the mid-brain region at 24 hpf (Fig. 3C).

Table 1.

A summary of tissue mRNA localization of three deiodinase.

Stage/tissues/Gene	Dio1	Dio2	Dio3
2cell	+	+	−
4cell	+	+	−
8cell	+	+	−
1k cell	+++	+++	−
50% epiboly	+++	+++	−
6 somite	+++ head	+++ head	++ pronephric duct
18 somite	++ head	++ head	++ pronephric duct
24 hpf	++ periventricular zone of brain	++ periventricular zone of brain	+++ pronephric duct
48 hpf	−	++ Pituitary, spinal cord
72 hpf	+ head	++ head
96 hpf	+++ Liver and swimming bladder	+ swimming bladder
10 dpf	+++ Kidney and liver	+ liver

−, no signal; +, positive singal; moderate signal; +++, strong signal.

Fig. 1.

Expression of Deiodinase 1 in different developmental stages of zebrafish embryo and larva as shown by dark coloration using in situ hybridization. A, 2 cell; B, 4 cell; C, 8 cell; D,1 k cell; E, 6-somite; F and G, 6-somite; H, 18-somite; I, 24 hpf; J, 72 hpf; K, 96 hpf; L, 10 dpf. Red arrows indicate the location of expression.

Fig. 4.

Representative images 6-OH-BDE 47 induced Dio1 and Dio2 mRNA expression as detected by whole mount in situ hybridization. A: DMSO control for Dio1; B, DMSO control for Dio2; C 6-OH-BDE 47 exposed for Dio1; D: 6-OH-BDE 47 for Dio2; E: sense control. Red arrows indicated locations of mRNA expression.

Fig. 2.

Expression of Deiodinase 2 in zebrafish larva at different time points as shown by dark coloration using in situ hybridization. A, B and C, 48 hpf; D, 72 hpf; E, 96 hpf; F, 10 dpf.

Fig. 3.

Expression of Deiodinase 3 in different developmental stages of zebrafish embryo as shown by dark coloration using in situ hybridization. A, 6-Sommite; B, 22 hpf; C 24 hpf; D, a section of B. Red arrows indicate location of expression.

3.2. Effects of 6-OH-BDE 47 on Dio1, Dio2 and Dio3 mRNA expression visualized by WISH

At 22 hpf, Dio1 and Dio2 mRNA expression was found in periventricular zone (Fig. 4), and Dio3 mRNA expression in the pronephric duct respectively (Fig. 3 and 6). Therefore, we selected 22 hpf to conduct the exposure assessments with 6-OH-BDE 47 and examined impacts on Dio1, Dio2 and Dio3 mRNA expression by WISH. We found that exposure to 1-, 10- and 100 nM 6-OH BDE47 increased Dio1 mRNA expression by 21.7% (p-value =0.603), 41% (p <0.05) and 44.2% (p <0.05) in the periventricular zone at 22 hpf relative to control embryos, respectively (Fig. 4A, C and Fig. 5A). Similarly, Dio3 mRNA was significantly increased by 21 %, 33.8% and 31.8% (all p <0.05) by 1 -, 10 -and 100 nM 6-OH BDE 47 concentrations, respectively (Fig. 6 and Fig. 5C). On the contrary, Dio2 mRNA expression was slightly, but not significantly, increased by exposure at 10 - or 100 nM of 6-OH BDE 47 (by 23 % and 24.1% respectively) (p-value =0.297), when compared to the control group (Fig. 4B, C and Fig. 5B).

Fig. 6.

Representative images of 6-OH-BDE-47 induced Dio3 mRNA expression detected using whole mount in situ hybridization. A, Control; B 1nM 6-OH-BDE-47; C, 10 nM 6-OH-BDE-47 and D 100 nM 6-OH-BDE-47; E: sense control. Red arrows indicate locations of Dio3 mRNA expression.

Fig. 5.

Gene expression levels as indicated by intensity of coloration in in situ hybridization between control and 6-OH-BDE-47 exposed groups. Embryos were exposed from 4 hpf to 22 hpf to 1, 10 and 100 nM 6-OH-BDE 47. A. Dio1 B. Dio2 C. Dio3. Different letters indicate significant differences (p<0.05). n=10–15.

4. Discussion

4.1. Differential expression and role of Dio1, Dio2 and Dio3 during early zebrafish development

The results of this study demonstrate that in situ hybridization can provide an indispensable tool for examining regional gene expression in early developmental life stages such as embryos, cells and tissues (Schad et al., 2003; Wang et al., 2012). WISH analysis yielded sufficient sensitivity and resolution to determine tissue-specific gene expression of deiodinases at different developmental stages within the intact organism. We demonstrated that Dio1, Dio2 and Dio3 are expressed in a spatially and temporally specific manners during zebrafish early development. Dio1 and Dio2 expression appeared to decrease steadily as zebrafish embryos developed from 2 - through 8 cell stages. Such findings may indicate that the initial Dio1 and Dio2 were from a maternal source and were distributed amongst daughter cells as the embryos underwent subsequent divisions. Such, a process would result in a steady decrease in expression. From 1k cell stage onward, Dio1 and Dio2 expression levels increased and were highly expressed in the 50% epiboly and 6-somite stages, when neuromere development and primary organogenesis take place in the zebrafish (Kimmel et al., 1995).

Notably, at 22 hpf, Dio1 and Dio2 mRNA expression were concentrated at the periventricular zone of the brain (Fig. 4). In contrast, Dio3 mRNA was found in the pronephric duct from 6-somite stage to 24 hpf. This suggests that embryos have specialized tissues/organs concentrating at de-activating the pro-hormone T4. To our knowledge, this is the first study showing locations where Dio1 Dio2 and Dio3 mRNA expression changes in different embryonic stages (from 2 cell and/or 6-somite stage) of zebrafish. There were no other reports of Dio1 and Dio2 expression in early embryonic stage for fish, but some data were available in amphibians, birds and mammals to support the presence of deiodinase expression in embryos. In most cases, significant Dio1 Dio2 and Dio3 expression was found in the developing brain, which is highly sensitive to thyroid hormones (Horn and Heuer, 2010; Johnson and Lema, 2011; Kaplan and Yaskoski, 1981; Tu et al., 1999; Van Herck et al., 2012). Dio1 and Dio2 are known to play an important role in fish embryonic development (Eales et al., 1993; Kohrle, 2000). Walpita et al (Walpita et al., 2010; Walpita et al., 2009) investigated the biological functions of Dio1 and Dio2 using antisense oligonucleotides and combined knockdown of both deiodinases. Their results demonstrated that Dio2 is essential to otic vesicle and head trunk development, and pigmentation. Remarkably, pigmentation inhibition and development delay due to knockdown of Dio2 was rescued by addition of 50 nM T3 treatment to embryos at 31 hpf (Schmutzler et al., 2007). Given these results, it is not surprising that we found Dio2 mRNA expression in periventricular zone of the brain at 22 and 24 hpf (pharyngula period) in zebrafish embryos. During this period (22–24 hpf) zebrafish embryos show early pigmentation in both the retina and skin (Kimmel et al., 1995). Our observations suggest that zebrafish embryos express Dio2 to regulate T3 in these tissues at 22–24 hpf.

Although T3 concentrations were reported to be very low throughout the morphogenesis period in zebrafish (Walpita et al., 2007), the Dio1 or Dio2 mRNA levels detected in this study suggest there may be a significant capacity for T3 synthesis via deiodination of T4 in these early life stages. Dio1 and Dio2 expression in the periventricular zone at 22 hpf also suggests a role for thyroid hormone signaling in very early neurogenesis. Thus, these results support the role of deiodinases in early embryonic organogenesis of the central nervous system and neuron development. Similar expression patterns were also found in early Xenopus embryos (Morvan Dubois et al., 2006), where deiodinase mRNA (Dio1, Dio2 and Dio3) expression was detected in the head region and axial structures of early stages (NF30 and NF35). This expression pattern has been reported in rats as well (Meerts et al., 2004). This conservation of deiodinase expression pattern between different taxa supports zebrafish as a suitable alternative model for studying thyroid biology and toxicology.

Earlier reports on Dio1 and Dio2 mRNA expression in early stages of zebrafish using qRT PCR indicated that Dio1 mRNA gradually increased during development while Dio2 levels were stable until hatching (Walpita et al., 2007). Our results here are a little different. After hatching and organ differentiation, we only found very weak Dio1 and Dio2 mRNA expression in the brain area at 72 hpf. Similar to a study by Thisse (Thisse et al., 2003), we detected strong Dio1 mRNA expression in the liver, kidney and intestinal bulb at 96 hpf (Fig. 1K) and weak Dio2 mRNA expression in intestinal bulb at 96 hpf (Fig. 2E). This is in agreement with previous data of T3 concentrations peaking at the “metamorphic climax”, i.e., that time when progressive filling of melanophores in the lateral stripe, and swim bladder occur. Also, inflation of the swim bladder occurs at this time (Kimmel et al., 1995). A high level of Dio1 mRNA expression in liver may, in addition, suggest that T4 is being converted to T3 and circulated to support pigmentation. Up through 10 dpf, strong Dio1 expression was detected in the inter-renal gland (adrenal equivalent) and intestine. Interestingly, no Dio2 mRNA was found in inter-renal gland, and was only detected weakly in the swim bladder. In summary both Dio1 and Dio2 appeared to play important roles for maintaining larval development and differentiation, and Dio1 may have more influence than Dio2 over this duration.

Dio3 expression is known to affect cell proliferation (Huang, 2009), and Dio3 has been identified as a thyroid hormone target gene involved in metamorphosis in Xenopus (Becker et al., 1997; Kawahara et al., 1999; Marsh-Armstrong et al., 2004). In our study, zebrafish Dio3 mRNA expression was only detected in the pronephric duct from the 6-somite cell stage to 24 hpf. Nakajima et al (Nakajima et al., 2012) reported that Dio3 expression was the highest at metamorphic climax (stage 60) as measured by qRT-PCR in amphibians. Given that the biological function of Dio3 is likely conserved between amphibians and fish, we hypothesize that it has the same influence on cell proliferation in zebrafish embryos. Our results then indicate that Dio3 was likely involved in kidney development in zebrafish, and this result was similar to Xenopus tropicalis, where Dio3 was detected in pro-nephros at stage 28 early tailbud embryos (Tindall et al., 2007). A similar finding of high expression level of Dio3 transcripts in bovine kidney has also been reported (Connor et al., 2005).

4.2. WISH for detection of Dio1, Dio2 and Dio3 mRNA expression after exposure 6-OH BDE 47

In this study we provided evidence demonstrating that WISH analysis provides information on Dio tissue-specific gene expression at different developmental stages in intact organisms. Given the potential thyroid disrupting effects of some PBDE flame retardants, and their metabolites, we therefore sought to examine the impacts of one specific metabolite, 6-OH BDE 47, on Dio1, Dio2 and Dio3 gene expression in zebrafish. According to our results, 6-OH-BDE-47 (10 and 100 nM) significantly upregulated Dio1 mRNA expression, but no significant induction of Dio2 mRNA at 22 hpf was observed (Fig. 5). In fish, T4 is the predominant form of thyroid hormone synthesized but T4 is not biologically active. Dio1 and Dio2 covert T4 into the biologically active T3 (Orozco and Valverde, 2005). T3 is crucial for growth and differentiation in early stages of zebrafish. Dio3, on the other hand, converts T4 into biologically inactive rT3. Chen et al reported that BDE208 induced T3 concentration in zebrafish lavae and also induced Dio1 and Dio2 mRNA expression in same period. Increased Dio1 can lead to increased conversion of T4 into T3, thus increasing T3 concentrations (Chen et al., 2012a). The decrease of Dio3 may further decrease the inactivation of T3 into T2, thus add to the effect of increasing T3 concentrations.

PBDE congeners (BDE 47, 100 and 209) and OH-BDE metabolites (6-OH-BE 47, 4′-OH-BDE 49) have been detected in high concentrations in human serum, human breast milk and in animals (Birnbaum and Staskal, 2004; Koh et al., 2010; Schecter et al., 2010; Stapleton et al., 2011). 6-OH-BDE 47 is a known metabolite of BDE 47, which is a common PBDE in human tissues. It is worth nothing that 6-OH-BDE 47 is also synthesized naturally in marine sponges and red algae, and thus is a natural product found in the marine environment, and likely, seafood consumed (Hamers et al., 2008; Malmvarn et al., 2005). Therefore not all 6-OH-BDE-47 in humans may be attributable to flame retardant exposure. 6-OH-BDE-47 has also been identified as a ubiquitous compound of toxicological concern in human serum (Zota et al., 2011). Comparisons of several PBDEs and their metabolites revealed that 6-OH-BDE-47 is more toxic to zebrafish embryos than 5-OH-BDE-47, 2,2′,4,4′-tetrabromodiphenyl ether (BDE-47), DE-71 (a commercial PBDE mixture), 2,2′,4,4′,5-pentabromodiphenyl ether (BDE-99), BDE-100, 2,2′,4,4′,5,5′-hexabromodiphneyl ether (BDE-153) and 2,2′,4,4′,5,5′6,6′-deca-bromodiphenyl ether (BDE-209). It leads to induced apoptotic cell death, developmental arrest, and disrupted gene regulations in zebrafish (Nyholm et al., 2009; Usenko et al., 2012). Dio1 and Dio2 mRNA expression in zebrafish adult liver was increased by 18 and 7 fold respectively at 168 hpf by qRT-PCR after exposure to 1.243 μM (0.625 ppm) 6-OH-BDE-47 from 24 hpf to 48 hpf (Usenko et al., 2012). Amongst other PBDEs, BDE-209 and DE-71 have also shown to significantly induce Dio1 and Dio2 mRNA expression at 14 dpf in zebrafish larva (Chen et al., 2012b; Gao et al., 2010). Unsurprisingly BDE-99 and BDE-209 were reported to significantly up-regulate Dio1 mRNA expression in human hepatocytes, but Dio2 mRNA expression was not measured in the study (Stapleton et al., 2009). Conversely, Szabo et al reported both the Dio1 enzyme activity and mRNA expression were decreased in liver of rat male pups at postnatal day 4 (PND4) and PND 21 after maternal exposure to DE-71 from gestational day 6 to PND21. However, DE-71 is mixture of PBDEs including BDE-47, BDE-99, BDE-100, BDE-153 and BDE-154, and low concentrations of brominated dioxin and dibenzofuran contaminants (Hanari et al., 2006; Szabo et al., 2009). Therefore, exposure to DE-71 may result in multiple mechanisms and lead to complex results.

6-OH-BDE-47 also significantly induced Dio3 in a dose-dependent manner, but in the pronephric duct at 6-somite, 22 hpf and 24 hpf zebrafish embryos (Fig. 3). This compares with the amplitude of Dio3 induction by T3 in goldfish (Nelson and Habibi, 2008). In rodents, Dio3 is also expressed in placental tissues and is important in maintaining low thyroid hormone concentrations in the developing fetus (Huang et al., 2003). Up-regulation of Dio3 may lead to lower than normal T3 concentrations in the developing fetus (Chan et al., 2005).

In conclusion, we have used WISH to monitor Dio1, Dio2, and Dio3 tissue specific mRNA expression in very early embryonic stages and to show how such patterns differ temporally. Dio1 and Dio2 expression were present from 2 cell stage onwards but the mRNA may be from maternal source from the 2 to the 1k cell stage. Dio3 was not detected until the 6-somite cell stage. After exposure to a ubiquitous contaminant, 6-OH-BDE 47, Dio1 was increased substantially while Dio2 and Dio3 were only weakly increased. There was also no change in the locations/tissues where the respective Dio1, Dio2 and Dio3 were expressed. Our results suggests that WISH may provide a more detailed account of toxic mechanism(s) and the resultant injury caused by exposure to PBDE and OH-BDEs in early life stages.

Highlights.

• Whole mount In situ hybridization (WISH) identifies regional and developmental deiodinase mRNA expression in zebrafish

• Deiodinase 1 and 2 were localized in periventricular zone of the brain in zebrafish embryos

• Deiodinase 3 was localized in the pronephric duct of zebrafish embryos

• 6-OH-BDE-47 exposure increases Deiodinase 1 and 3 mRNA expression in developing zebrafish