Methyl Mercury Suppresses the Formation of the Tail Primordium in Developing Zebrafish Embryos

Toxicological Sciences, Jun 2010

The objective of this study was to characterize the mechanisms of action of the model environmental toxicant methyl mercury (MeHg) in the zebrafish embryo. Zebrafish embryos were exposed to MeHg, and the effective concentration and window of exposure were determined in wild-type and fluorescent reporter transgenic zebrafish embryos. Genes were systematically assessed for altered expression in response to MeHg by in situ hybridization. MeHg impairs development of the fin fold and the tail fin primordium. Alterations in transgene expression were noted at 6 μg/l MeHg, making this shh:gfp line the most sensitive biosensor of MeHg exposure. The matrix metalloproteases mmp9 and mmp13 and eight other genes are induced in the embryonic tail in response to MeHg. Our data suggest that MeHg impairs tail development at least partially by activation of the tissue remodeling proteases Mmp9 and Mmp13.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

Methyl Mercury Suppresses the Formation of the Tail Primordium in Developing Zebrafish Embryos

Lixin Yang 1 Nga Yu Ho 1 Ferenc M uller 0 1 Uwe Strahle 1 0 Present address: Department of Medical and Molecular Genetics, School of Clinical and Experimental Medicine, College of Medical and Dental Sciences, University of Birmingham, Institute of Biomedical Research , B15 2TT Edgbaston, Birmingham , UK 1 Institute of Toxicology and Genetics Forschungszentrum Karlsruhe Karlsruhe Institute of Technology 76021 Karlsruhe , Germany The objective of this study was to characterize the mechanisms of action of the model environmental toxicant methyl mercury (MeHg) in the zebrafish embryo. Zebrafish embryos were exposed to MeHg, and the effective concentration and window of exposure were determined in wild-type and fluorescent reporter transgenic zebrafish embryos. Genes were systematically assessed for altered expression in response to MeHg by in situ hybridization. MeHg impairs development of the fin fold and the tail fin primordium. Alterations in transgene expression were noted at 6 mg/l MeHg, making this shh:gfp line the most sensitive biosensor of MeHg exposure. The matrix metalloproteases mmp9 and mmp13 and eight other genes are induced in the embryonic tail in response to MeHg. Our data suggest that MeHg impairs tail development at least partially by activation of the tissue remodeling proteases Mmp9 and Mmp13. - The teratogenic and neurotoxic effects of methyl mercury (MeHg) became sadly evident after large-scale poisonings in Japan and Iraq in the last century (Ekino et al., 2007). In the environment, inorganic mercury is rapidly transformed into methyl mercury, which then accumulates in the food chain. Populations with a high seafood consumption are at particular risk (Myers et al., 2007). As recognized in follow-up studies of the poisonings in Japan and Iraq, methyl mercury can act as a neurotoxicant in humans. Studies of animal models suggested that low level of MeHg affects specifically the motor and sensory systems (for review, see Ekino et al., 2007). Large-scale, longterm cohort studies of the effects on low-level intake of MeHg via the food chain were conducted in New Zealand, on the Seychelles and the Faroe islands (Davidson et al., 2006; Grandjean and Landrigan, 2006). The U.S. National Academy of Science upon careful review of these and other studiesconcluded that prenatal exposure to low levels of MeHg causes behavioral deficits (Grandjean and Landrigan, 2006, and references therein). Long-term studies of human exposure represent one approach to assess the health risk emanating from low levels of MeHg. Other strategies relied on studies of animal models to understand the mechanisms how MeHg can affect development. Zebrafish embryos may not only be ecotoxicologically relevant models but may also aid in the elucidation of the molecular mechanism underlying the effects of low-level MeHg exposure in humans (Yang et al., 2009 and references therein). Zebrafish embryos develop completely outside the mother and were previously shown to be highly sensitive to MeHg exposure (Samson and Shenker, 2000; Yang et al., 2007), thereby providing a vertebrate model whose development is not complicated by the physiology of the mother as in mammals (Kimmel et al., 1995). Exposure of zebrafish embryos at levels between 20 and 30 lg/l MeHg led to characteristic flexures of the axis, impaired development of the fin fold, and caused embryonic or early larval death presumably due to cardiac malfunction (Samson et al., 2001). Doses between 10 and 20 lg/l were previously reported to show a delayed mortality syndrome and behavioral deficits such as impaired swimming activity and prey-capture performance (Samson et al., 2001). Morphological and behavioral end points are frequently not very informative with respect to underlying mechanisms of the toxic effect. In a toxicogenomic study using zebrafish embryos, we have identified 79 genes that were significantly upregulated in zebrafish embryos exposed to MeHg (Yang et al., 2007). Toxicants such as cadmium (CdCl2), lead (PbCl2), or arsenic (As2O3) showed related but distinct patterns of gene activation. In this and previous studies (Samson and Shenker, 2000; Yang et al., 2007), it was noted that MeHg effects the development of the median fin fold. These effects appeared reminiscent of the phenotypes of mutant zebrafish with an impairment of the sonic hedgehog (Shh) signaling pathway (Hadzhiev et al., 2007). During larval stages, the tail develops from a primordium located in the ventral fin fold, a region that The Author 2010. Published by Oxford University Press on behalf of the Society of Toxicology. All rights reserved. For permissions, please email: coincides with a gap in the melanophore streak located immediately anterior to the posterior tip of the notochord. This region is fortuitously marked by expression of a transgene harboring upstream and downstream sequences of the shh gene of zebrafish (Ertzer et al., 2007; Shkumatava et al., 2004). The transgene does not reflect, however, the reactivity of the endogenous shh gene but presumably represents rather the response of an unknown locus, into which the transgene has integrated in the Tg(-2.4shh:gfpABC line #15) (Hadzhiev et al., 2007). In mutants impairing the activity of Shh signal transduction, both the gap in the melanophore streak as well as transgene expression are abolished (Hadzhiev et al., 2007). The effect of MeHg on the integrity of the fin fold and the potential involvement of the Shh pathway prompted us to investigate the fin primordium in MeHg-treated embryos in detail. MATERIALS AND METHODS Fish maintenance. The Tg(-2.4shh:gfp:ABC#15) transgenic line contains the 2.4-kb shh promoter upstream of and the elements ar-A, ar-B, and ar-C of the shh introns 1 and 2 downstream of the gfp coding sequence as described (Ertzer et al., 2007; Hadzhiev et al., 2007; Muller et al., 1999, 2000). The Tg(-2.4shh:gfp:ABC#15) line and wild-type zebrafish strains AB, ABO, and Tubingen were maintained on a 14 h/10 h light-dark cycle at 28.5 C in recirculization systems (Schwarz Ltd, Germany, and Muller and Pfleger Ltd, Germany) and fed commercial food and in-house hatched brine shrimp as described (Westerfield, 1993). Zebrafish embryos were obtained from spawnings of wild-type or identified Tg(-2.4shh:gfp:ABC#15) fish in spawning cages (Westerfield, 1993). Chemical treatments. MeHg (methyl mercury chloride; CH3ClHg), lead (II) chloride (PbCl2), arsenic (III) oxide (As2O3), and cadmium chloride (CdCl2) were purchased from Sigma-Aldrich (St Louis, MO). Stock solution of 100 mg/l MeHg was prepared in water and working solutions (6120 lg/l) were obtained by diluting the stock solution in embryo medium (60 lg/ml; Instant Ocean, Red Sea, Houston, TX, pH 6.73, Ca 0.8 mg/l; K 0.6 mg/l; Mg 2 mg/l; Na 16 mg/l; S 2 mg/l [Westerfield, 1993]) to the desired concentrations. Working solutions of 20 mg/l PbCl2, 79 mg/l As2O3, and 15 mg/l CdCl2 in embryo medium were prepared in the same way from 5, 10, and 4 g/l stock solutions in fish water, respectively. As described previously (Yang et al., 2007), toxicant concentrations were chosen so that more than 80% treated embryos showed malformation but that embryo mortality was minimal. Unless otherwise stated, embryos were exposed to chemicals from 4 to 72 hours postfertilization (hpf). A control group under the same conditions as the treatment groups but with the embryos reared in embryo medium only was included in all experiments. Embryos were anesthetized with 0.02% tricaine (3-aminobenzoic acid ethyl ester) for documentation with a stereomicroscope (Leica MZ 16F) or a compound microscope (Leica DM 5000 B). In situ hybridization and antibody labeling. The following genes were selected with specific expression in the tail primordium by an RNA in situ hybridization screen. The cDNA of the sequestosome 1 (sqtm1) gene (accession number NM_213173) was obtained by screening a cDNA library, derived from a mix of mRNA from 16, 24, and 36 hpf zebrafish embryos, with a radioactive oligonucleotide probe (5#-GGATCAAGATGTCTCAACAAGTTTC-3#). The digoxigenin (DIG)-labeled antisense RNA of sqtm1 was synthesized by digestion with XhoI followed by transcription with T3 polymerase using standard procedures (Oxtoby and Jowett, 1993; Strahle et al., 1994). Genes for NM_213361 (accession number NM_213361), serum/glucocorticoid regulated kinase 1 (sgk1) (accession number BC052134), AW115990 (accession number AW115990), and BM101698 (accession number BM101698) were purchased as I.M.A.G.E. clones from ImaGenes (Germany). Antisense probes of NM_213361, sgk1, AW115990, and BM101698 were generated by PCR using primers 5#-CCGGATCCGGTGGTGCAAATCAAAGAACTGCTCCTCAGTG-3# and 5#-CAGAGATGCATAATACGACTCACTATAGGGAGAGACAAACCACAA CTAGAATG-3# and DIG-labeled RNA was then transcribed by T7 polymerase from the PCR fragment. The antisense RNA probes of mmp9 (NM_213123), mmp13 (NM_201503), l-plastin (NM_131320), myeloid-specific peroxidase (mpx, NM_212779), and apolipoprotein-E (NM_131098) were transcribed with T7 and Sp6 RNA polymerase (Table 1). Further details are available upon request. Embryos were grown in embryo medium or in chemical solution containing 0.003% phenyl-thiourea to prevent the formation of pigment at 28.5 C. Control and treated embryos were fixed in 4% paraformaldehyde overnight at 4 C and then stored in 100% methanol at 20 C. The In situ hybridization and antibody labeling of caspase3 (Research & Diagnostics Systems, Inc., USA) were performed according to the published protocols (Oxtoby and Jowett, 1993; Strahle et al., 1994). For double RNA, labeling probes were labeled with either DIG (Roche) or DNP-11-UTP (PerkinElmer). These two probes were detected by anti-Dig POD (Roche) and anti-DNP-HRP (PerkinElmer) sequentially and the staining was developed following the manuals of the tyramide signal amplification (PerkinElmer) plus Cyanine 3 and Cyanine 5 kits (PerkinElmer). Once developed, the embryos were dissected from the yolk and mounted in 70% glycerol for whole-mount imaging or embedded in Aqua Polymount (Polyscience) followed by analysis with a Leica SP2 confocal microscope. For sectioning, stained embryos were embedded in 3% agarose, and 20-lm sections were cut with a vibratome. For detection of apoptotic cells, 72 hpf embryos were incubated in 0.5 ll/ml of AO (acridine orange hydrochloride, Sigma) in embryo medium for 1 h. After staining, AO was removed by washing the embryos with embryo medium three times, 5 min each. Real-time PCR. Three independent total RNAs were extracted from each of the 6, 30, 60 lg/l MeHg-treated, and control embryos according to the procedures provided by the RNeasy Mini Kit (Qiagen). The reverse transcription of cDNA was performed with 1 lg total RNA following standard procedures (Sambrook et al., 1989). Real-time PCR of each sample including three biological repeats and two technical replications was carried out following the suppliers instructions (Applied Biosystem) by using SYBR Green and the comparative Ct method. The raw data were processed and analyzed with the Excel software. The mRNA levels were calculated using 2 DDCt. Morpholino knockdown. The p53 antisense morpholino oligonucleotide (GACCTCCTCTCCACTAAACTACGAT) synthesized by GeneTools Ltd (Ekker and Larson, 2001) was resuspended in water at 0.5mM in 0.1% phenol red. The morpholino was injected into the yolk of one- to two-cellstage embryos using a gas-driven microinjector (Muller et al., 2000). Mmp inhibitor experiments. Zebrafish embryos were treated with 0.1mM GM6001 (general Mmp inhibitor; Calbiochem), 1% DMSO, 100 lg/l MeHg or 0.1mM Mmp Inhibitor II (Mmp9- and Mmp13-specific inhibitor; Calbiochem), 1% DMSO, and 100 lg/l MeHg in embryo medium. Controls were exposed to 100 lg/l MeHg and 1% DMSO in embryo medium without the Mmp inhibitors. All the treatments were conducted from 4 to 72 hpf. At 72 hpf, MeHg and the Mmp inhibitors were removed by washing the embryos with embryo medium three times, 5 min each. Photos of the tail region of the embryos were captured by a CCD camera connected to a stereomicroscope, and the area of the fin fold (Supplementary fig. 5) was calculated by the software ImageJ (Abramoff et al., 2004). One-way ANOVA followed by Bonferronis or Dunnetts multiple comparison tests have been performed by comparing the various treatment groups. MeHg Impairs Formation of the Fin Fold and Abolishes the Tail Fin Primordium During previous studies (Samson and Shenker, 2000; Yang et al., 2007), it was noted that exposure of embryos to MeHg resulted in apolipoprotein-E (apoe) E74-like factor 3 (elf3) peroxiredoxin 1 (prdx1) Sequestosome 1 (sqtm1) serum/glucocorticoid regulated kinase 1 (sgk1) Accession number NM_131098 XM_684268 NM_213173 BC052134 F: tctggcagtatgtgtctgaactc R: aatagcctccaagacagacaag F: gagcgaatctgaacgtcacc R: gccggaaggagctctctgta F: gccaaggtggttcatgactt R: ccagtcgatgtcctggtttt F:ggcacgagaaacatgaagacc R:cacacacaaatgattttcaagga F:cagcaacactgcttgcactt R:aggaaacccatcatgaccaa F: ctgcccatggttagactggt R: aagacgttcgcaatggtagg F: gtatcgcgagacttgagcac R: gcattaattcacatcacttctgtc ggatcaagatgtctcaacaagtttc F:ccggatccggtggtgcaaatc aaagaactgctcctcagtg R:cagagatgcataatacgactca ctatagggagagacaaaccaca actagaatg F:ccggatccggtggtgcaaatcaaa gaactgctcctcagtg R:cagagatgcataatacgactcacta tagggagagacaaaccacaact agaatg F:ccggatccggtggtgcaaatcaa agaactgctcctcagtg R:cagagatgcataatacgactca ctatagggagagacaaaccaca actagaatg F:ccggatccggtggtgcaaatcaa agaactgctcctcagtg R:cagagatgcataatacgactcac tatagggagagacaaaccaca actagaatg pGEM-T Easy pGEM-T Easy pGEM-T Easy pGEM-T Easy pGEM-T Easy pBlueScript II SK() pME18S-FL3 Restriction enzyme a malformed tail fin fold (Figs. 1A and 1B). Instead of forming a clear, thin and fan-shaped membrane as in the controls (Fig. 1A), embryos treated with 60 lg/l MeHg from 4 to 72 hpf had a smaller and less well-structured fold surrounding the tail (Fig. 1B). These defects in the fin fold were strikingly correlated with an altered pattern of pigment distribution. At 72 hpf (Fig. 1C), the melanophores form two rows above and below the tail at the boundary to the fin folds. The ventral row is characterized by a gap in the row of melanophores slightly anterior to the tip of the tail (Fig. 1A). This gap was previously shown to correlate with a region which functions as the primordium of the tail fin and which is dependent on Shh signaling (Hadzhiev et al., 2007). This gap was missing in 72 hpf embryos exposed to 60 lg/l MeHg from 4 to 72 hpf (Figs. 1AD) as in embryos with defects in the Shh signaling pathway (Hadzhiev et al., 2007). MeHg (60 lg/l) caused, in addition, pericardial edema, and the head and eyes were smaller as previously reported (Samson and Shenker, 2000, 2001; Yang et al., 2007). Expression of the green fluorescent protein (GFP) from the transgene Tg(-2.4shh:gfpABC) marks the embryonic primordium of the tail fin that will grow out much later during larval stages (Hadzhiev et al., 2007). We tested whether treatment with MeHg impairs expression of this early marker of the tail fin primordium. Transgenic embryos were exposed to 60 lg/l MeHg from 4 to 72 hpf. A high proportion of embryos (86%, n 5, 20 embryos each) showed a complete loss of expression of GFP in the tail fin primordium (Figs. 1E and 1F). This correlated with loss or reduction of the gap in pigmentation (Figs. 1G and H). Expression of the transgene in the floor plate or in the pectoral fins was not affected by the MeHg treatment, indicating that MeHg specifically impaired GFP expression in the tail fin primordium. The Effect on the Tail Fin Primordium Is Specific for MeHg We next wished to test whether other chemicals such as CdCl2, As2O3, or PbCl2 with related toxicogenomic response profiles (Yang et al., 2007) could also induce this malformation. We exposed embryos to water (Fig. 2A, control), to 60 lg/l methyl mercury (Fig. 2B), to 20 mg/l PbCl2 (Fig. 2C), to 79 mg/l As2O3 (Fig. 2D), or to 15 mg/l CdCl2 (Fig. 2E) from 4 to 72 hpf. At the same or lower concentrations, all four compounds induced a profound toxicogenomic response (Yang et al., 2007). Embryos were examined at 72 hpf for defects in fin fold formation and whether the gap in the ventral streak of pigment cells (indicative of the tail fin primordium) is present. Neither PbCl2 nor CdCl2 nor As2O3 treatment led to defects in the fin folds or abolished the gap in pigmentation (Figs. 2AE). These data suggest that, within the substances tested, the defect in the tail primordium is specific to treatment with MeHg. Next, we asked whether MeHg treatment would induce apoptosis. Both acridine orange (Darzynkiewicz et al., 1992) and immunohistochemical staining of caspase3 (Cohen, 1997) were utilized to detect apoptotic cells. The number of cells undergoing programmed cell death was slightly increased in MeHg embryos in comparison to controls (Supplementary fig. 1AD$). We also tested whether we could prevent the defects in the tail fin primordium by blocking apoptosis. To this end, we injected a morpholino that blocks translation of p53 (Robu et al., 2007). The concentration of injected Mo-p53 used blocks effectively the unspecific apoptotic effects of morpholinos in the brain (data not shown, Robu et al., 2007). This Mo-p53 concentration could, however, not abolish the defect in the fin fold (Supplementary fig. 1E) underscoring the notion that cell death is not the primary or only cause of the defects in the tail of MeHg-treated embryos. Concentrations as Low as 6 lg/l MeHg Cause Defects in the Tail Fin Primordium Next, we determined the effective dose of MeHg. Transgenic embryos were exposed and the size of the domain expressing GFP in the tail fin primordium was scored. Two classes of effects were observed (Fig. 3AC): a reduction of the GFP expression domain (type I, Fig. 3B) and a complete loss of the GFP expression in the fin fold (type II, Fig. 3C). Reduced GFP expression was noted in 10% embryos exposed to 6 lg/l MeHg from 4 to 72 hpf (Fig. 3D). More than 80% of embryos showed a reduction of the GFP domain (type I), when 20 lg/l MeHg was administered (Fig. 3D). Bathing embryos in 30 lg/l caused defects in all exposed embryos and an increase to slightly more than 50% embryos with type II defects (Fig. 3D). More than 80% of embryos exposed to 60 lg/l MeHg showed the complete loss of transgene expression in the tail fin primordium (Fig. 3D). A 6-h Exposure Window Any Time between 12 and 48 h Is Sufficient to Induce the Tail Defects We next asked when MeHg exposure during development will affect the formation of the melanophore gap. Embryos were exposed to 60 lg/l MeHg for 6- or 12-h periods in the first 2 days of development. Treatment from 4 to 12 h of development during blastula, gastrula, and early segmentation stages did not significantly effect the formation of the melanophore gap in comparison to untreated controls (Fig. 4). In contrast, all intervals of treatment in the next 1.5 days lead to loss of the gap in melanophore distribution (Fig. 4). Moreover, a 6-h period of exposure at any time during this phase is sufficient to induce the defect. Exposure of embryos after 48 hpf did not elicit any significant effect anymore, thus limiting the sensitive phase to 48 hpf. This suggests that the most sensitive phase coincides with a period, when tail fin development has been initiated by emergence of the tail fin primordium, but prior to the apparent outgrowth of the tail fin. MeHg Treatment Causes Ectopic Expression of Matrix Metalloproteases mmp9 and mmp13 in the Embryonic Tail In previous studies assessing the toxicogenomic responses to MeHg (Yang et al., 2007), we noted that the mRNAs of the metalloproteases mmp9 and mmp13 were strongly upregulated in response to MeHg treatment in whole-embryo extracts. This prompted us to assess whether the expression of mmp9 (NM_213123) and mmp13 (NM_201503) mRNA could be correlated with the observed defects in the tail region. We exposed embryos to 60 lg/l MeHg from 4 to 72 hpf and assessed the pattern of expression at 72 hpf by whole-mount in situ hybridization with antisense probes complementary to mmp9 and mmp13 mRNA. Both mRNAs were ectopically expressed in the fin folds and in individual cells scattered over the tail region in MeHg-treated embryos (Figs. 5A, 5B, 5D, and 5E). This upregulation of mmp9 and mmp13 transcripts was quantitatively confirmed by real-time PCR with reversetranscribed cDNA prepared from total RNA of embryos exposed to 6, 30, and 60 lg/l MeHg. The results demonstrate that exposure to 30 and 60 lg/l MeHg increase the levels of mmp9 and mmp13 transcripts dramatically (Supplementary fig. 2). However, only background levels of mmp9 and mmp13 mRNAs were detected upon exposure to 6 lg/l MeHg (Supplementary fig. 2) as for solvent controls. Double in situ hybridization, with both probes simultaneously showed that increased expression could already be noted at 24 hpf in embryos treated with MeHg from 4 hpf onward (Supplementary figs. 3AF). We found cells that expressed only mmp9 but not mmp13 mRNA (Supplementary fig. 3). Sectioning of embryos hybridized to mmp9 and mmp13 antisense RNA showed a strong enrichment of the expressing cells in the epidermis (data not shown). We wondered whether the scattered mmp9- and mmp13-expressing cells are immune cells involved in tissue clearance. MeHg-treated embryos were hybridized to the l-plastin antisense probe, which is specific for macrophages and monocytic cells (Bennett et al., 2001). Neutrophils were detected with the mpx antisense probe (Bennett et al., 2001). Neither gene was significantly upregulated by MeHg or was expressed in a pattern resembling that of the mmp9- and mmp13-expressing cells (Supplementary fig. 4). Thus, the scattered mmp9 and mmp13 mRNA-positive cells are distinct from neutrophils, monocytic cells, and macrophages. We next tested whether PbCl2, As2O3, and CdCl2 would also induce an increase in expression of mmp9 and mmp13 mRNA in the tail. We exposed embryos to either water or to 20 mg/l PbCl2 or to 79 mg/l As2O3 or to 15 mg/l CdCl2 from 4 to 72 hpf. In situ hybridization of PbCl2-treated embryos revealed only a few scattered cells expressing mmp13 mRNA in the tail ectopically, while mmp9 mRNA was not induced (Figs. 5C and 5F). The other two chemicals did not elicit ectopic expression of mmp9 and mmp13 mRNA at all (data not shown). Thus, the induction of mmp9 and mmp13 expression in the tail is highly specific to treatment with MeHg. MeHg may cause defects by the ectopic activation of Mmp9 and Mmp13 enzymatic activities in the fin fold. To test this notion, we applied the Mmp inhibitors GM6001 or MMP-9/MMP-13 Inhibitor II (Yoshinari et al., 2009) to MeHg-treated embryos (100 lg/l). To quantitatively assess the effect of inhibitors, we calculated the area of the fin fold by drawing a vertical line through the tip of the notochord (Supplementary fig. 5) and calculating the area of the fin fold posterior to this line using the ImageJ software (Abramoff et al., 2004). MeHg treatment from 4 to 72 hpf reduced the area of the posterior fin fold by two-fold (Fig. 5G). Cotreatment of embryos with 0.1mM GM6001 or MMP-9/MMP-13 Inhibitor II led to a small but significant increase in the area of the fin fold with p < 0.05 (one-way ANOVA test) (Fig. 5G). Multiple Genes Are Upregulated in the Tail Fin in Response to MeHg In order to identify further genes that could be linked to the defects in tail fin morphogenesis, we screened a set of 43 genes previously noted to be regulated in response to MeHg exposure in microarray experiments (Yang et al., 2007). We focused the in situ hybridization analysis on 72 hpf embryos that were exposed to 60 lg/l MeHg from 4 to 72 hpf. In addition to the mmp9 (32 out of 33 embryos showed ectopic expression, 32/33) and mmp13 (40/46) genes, we identified eight further genes that are ectopically expressed in the tail region of embryos treated with MeHg. The expression of three genes (NM_213361, 24/36; AW115990, 17/44; BM101698, 30/44) with unknown function was strongly activated by MeHg treatment in a pattern that resembled that of mmp9 and mmp13 activation (Figs. 6AF). The mRNA of sgk1, (NM_199212, 32/40) was prominently expressed in response to MeHg in individual cells in the fin fold in a pattern reminiscent of mmp9 and mmp13 mRNA expression (Figs. 6G and H). In addition, we scored ectopic expression of the peroxiredoxin 1 (NM_001013471, 53/56), E74-like factor 3 (XM_684268, 8/23), and the sqtm1 (NM_213173, 34/34) gene. The latter three genes were expressed only in very few cells of the tail in response to MeHg (Figs. 6IN). Strong expression in response to MeHg was noted for the apolipoprotein-E mRNA (NM_131098, 49/51) (Figs. 6O and 6P). We show here that MeHg impairs development of the fin fold and specification of the tail fin primordium in the zebrafish embryo. The effective window is restricted to the period between 12 and 48 hpf and the effect appears to be specific to MeHg at least within the limited range of substances tested. Moreover, the tail fin primoridum is most sensitive to MeHg exposure; adverse effects of concentrations as low as 6 lg/l MeHg are detectable by a transgenic line expressing GFP in the caudal tail fin primordium. MeHg exposure leads to induction of a number of genes in the fin fold and tail including the metalloproteases mmp9 and mmp13, suggesting that tissue remodeling plays a role in MeHg toxicity. Interference with the Development of the Tail Fin Primordium The tail fin develops as an asymmetric structure that grows out from a restricted zone of the ventral fin fold of the embryonic tail. The location of this zone correlates with a gap in the melanophore streak that runs along the flank of the embryo at 72 hpf (Hadzhiev et al., 2007). MeHg has the same effect as loss-of-function mutations in the Shh signaling pathway, that is, loss of the pigment gap and loss of Tg(-2.4shh:gfpABC #15) transgene expression in the caudal fin primordium. However, MeHg treatment did not interfere with Shh-dependent processes at known other sites of Shh action (Barresi et al., 2000; Schauerte et al., 1998) in the embryo such as motor neuron specification or somite development (Yang, Ho, and Strahle, unpublished data). Moreover, Shh signaling appears to be continuously required for tail fin development also at stages older than 48 hpf (Hadzhiev et al., 2007). In contrast, exposure to MeHg does not impair transgene expression in the caudal fin primordium after 48 hpf. These observations make it rather unlikely that MeHg acts directly on Shh signaling. It has been proposed that neural crest cells migrate to the caudal fin primordium (Wood and Thorogood, 1984; van Eeden et al., 1996). The time frame of MeHg effect (1248 hpf) would fit with an interference by MeHg in this process. However, we did not find impairment of neural crest differentiation and migration as inferred from the patterns of crestin and sox10 expression in 24-h old embryos (Yang, Ho, Strahle, unpublished data). The effect on the caudal fin primordium does not appear to be a reflection of overall tissue degradation in the exposed embryos. Although MeHg, As2O3, PbCl2, and CdCl2 induce related toxicogenomic responses (Yang et al., 2007), only MeHg induced loss of the pigment gap and robust ectopic mmp9 and mmp13 mRNA expression in the embryonic tail. The effects on transgene expression in the tail fin primordium were observed at very low concentrations (6 lg/l), at which tissue degeneration in other embryonic regions is not evident. We do not know the absolute MeHg concentration in the embryo. Previous studies did not detect accumulation of MeHg in the fin fold. These studies reported MeHg accumulation in the lens and the brain (Korbas et al., 2008; Santra et al., 2009). Locomotor deficits of embryos treated with concentrations of 15 lg/l MeHg were attributed to neurotoxic effects (Samson et al., 2001). Our results suggest that these low concentrations can also impair tail fin development, thereby contributing to the behavioral deficits observed previously (Samson et al., 2001). In previous studies, concentrations as low as 27 lg/l and 266 lg/l MeHg triggered luciferase reporter activity in zebrafish embryos and in a zebrafish-derived cell line, respectively (Carvan et al., 2000; Kusik et al., 2008). Reduction of the shh:gfp transgene expression in the caudal fin primordium is thus a highly sensitive readout of MeHg exposure. MeHg Induces Expression of Metalloproteases mmp9 and mmp13 in the Fin Fold of Exposed Embryos Besides impaired specification of the tail fin primordium, administration of MeHg induced defects in the fin fold such as reduced size and increased apoptosis. Most strikingly, MeHg exposure triggered the expression of a set of 10 genes in the tail region, among which the metalloprotease genes mmp9 and mmp13 together with the lipid transporter apolipoprotein-E and three uncharacterized genes (NM_ 213361, AW115990, BM101698) were abundantly induced. The mmp9 and mmp13 expression overlap in some cells. We found, however, also cells that expressed mmp9 mRNA but not mmp13 mRNA, suggesting that the mmp9- and mmp13expressing cell populations are not identical or that the induction of the RNAs follows a stochastic process leading to expression in some but not all cells. We did not detect upregulation of markers such as l-plastin and mpx that are expressed by macrophages/monocytic cells and neutrophils, respectively (Bennett et al., 2001), suggesting that the mmp9 and mmp13 expression are not the reflection of an inflammatory response mediated by these cells invading the MeHg-damaged tissue. Mmp-expressing cells maybe dendritic, antigen-presenting cells, especially as expressing cells appear to be strongly enriched in the epidermis of MeHg-treated embryos (Yang, Ho, Strahle, unpublished data). However, very little is so far known about dendritic or Langerhans cells in the zebrafish embryo. The mmp9, mmp13, and sgk1 genes as well as the unknown genes NM_213361, AW115990, and BM101698 are expressed in cells scattered throughout the fin fold. These genes could have a role in repair of the MeHg-inflicted tissue damage (Bai et al., 2005) or could be directly involved in the toxicological mechanism. It is tempting to speculate that misregulated mmp9 and mmp13 expression leads to tissue damage and thus to the observed effects seen in the fin folds. CdCl2, As2O3, and PbCl2 did not elicit a reduction of the tail fin primordium and induced ectopic expression of mmp9 and mmp13 mRNA only in a few cells or not at all. Upregulation of mmp9 and mmp13 mRNA expression was noted already at 24 h in MeHg-treated embryos, suggesting that an increased transcription of the mmp9 and mmp13 genes is an early consequence of MeHg exposure. In addition, blockage of Mmp9 and Mmp13 enzymatic activity with specific small molecule inhibitors reduced the observed defect in fin fold development. This argues in favor of Mmp9 and Mmp13 being at least part of the toxic mechanisms leading to impaired fin fold development. We noted a moderate increase in apoptotic cells in the fin fold. In addition, the literature reports a wide variety of molecular mechanisms, how MeHg can disturb cell homeostasis (Cambier et al., 2009; Castoldi et al., 2001; Hoffmeyer et al., 2006; QU et al., 2003; Senger et al., 2006; Vicente et al., 2004). The effect on mmp9 and mmp13 is one novel further aspect in the pleiotropic toxicological response to MeHg that may have, however, a very strong bearing on the developmental toxicity of MeHg. Whether misregulated mmp9 and mmp13 could also cause the defects in the tail fin primordium remains to be addressed. We could not detect an elevation of mmp9 and mmp13 mRNA by RT-PCR relative to controls in embryos treated with 6 lg/l MeHg. This concentration caused impairment of the tail fin primordium as inferred from a reduction of the transgene expression domain in 10% of treated embryos. It is questionable whether a 10% increase in mmp9/mmp13 mRNA levels above those of controls would have been detected by in situ hybridization or the RT-PCR assay. This underscores the sensitivity of the transgene-readout relative to these assays. The transcriptional profiling of MeHg-treated embryos suggested that exposed embryos suffer from oxidative stress as they upregulated expression of oxidative stress genes (Kusik et al., 2008; Yang et al., 2007). With the exception of the peroxiredoxin 1 gene that was upregulated in the lateral line neuromasts located in the gap of the melanophore, expression of oxidative stress response genes was not significantly increased in the fin fold. Altogether, we have identified 10 genes whose expression is specifically upregulated in the tail of MeHg-treated embryos and that haveto our best knowledgenot been linked to MeHg toxicity in the past. The gfp transgenic line, whose expression is abolished in the caudal fin primordium, represents a highly sensitive readout of MeHg-inflicted tail defects. Our experiments underscore the usefulness of the zebrafish as model to assess developmental toxicity of MeHg and expand its capabilities to obtain mechanistic insights by the identification of a number of novel genes that can serve as tools for further studies. These genes could serve as biosensors to monitor the effect of toxicants in living zebrafish embryos. Especially, the fluorescent readout of toxic impact in transgenic responder lines will not only allow the study of live animals but is probably the only way how, together with automation of embryo handling and image acquisition (Gehrig et al., 2009; Lam et al., 2009; Yang et al., 2009), to achieve the throughput that is necessary for the systematic assessment of the toxic effect of the large numbers of chemicals demanded by regulators (Hartung and Rovida, 2009). SUPPLEMENTARY DATA Supplementary data are available online at http://toxsci Forschungszentrum Karlsruhe and HGF Additional Funding Toxicogenomics; the European commission IP ZF-MODELS (contract N LSHC-CT-2003-503466); BMBF Gen-DarT2 (PTJ-BIO/0315190B to F.-Z.). ACKNOWLEDGMENTS We thank N. Borel and her fish house team for fish care, S. Rastegar for organizing the laboratory, and M. Rastegar for help with microscopy.

This is a preview of a remote PDF:

Lixin Yang, Nga Yu Ho, Ferenc Müller, Uwe Strähle. Methyl Mercury Suppresses the Formation of the Tail Primordium in Developing Zebrafish Embryos, Toxicological Sciences, 2010, 379-390, DOI: 10.1093/toxsci/kfq053