Gene expression changes in developing zebrafish as potential markers for rapid developmental neurotoxicity screening

Neurotoxicology and Teratology 32 (2010) 91–98

Contents lists available at ScienceDirect

Neurotoxicology and Teratology j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / n e u t e r a

Gene expression changes in developing zebrafish as potential markers for rapid developmental neurotoxicity screening Chun-Yang Fan a,b,1, John Cowden a,2, Steven O. Simmons a, Stephanie Padilla a, Ram Ramabhadran a,⁎ a b

Cellular and Molecular Toxicology Branch, Neurotoxicology Division, United States Environmental Protection Agency, Research Triangle Park, NC 27711, USA Curriculum in Toxicology, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, USA

a r t i c l e

i n f o

Article history: Received 12 November 2008 Received in revised form 25 March 2009 Accepted 23 April 2009 Available online 19 May 2009 Keywords: Zebrafish Developmental neurotoxicity Nervous system Quantitative real time PCR Gene expression profile

a b s t r a c t Hazard information essential to guide developmental neurotoxicity risk assessments is limited for many chemicals. As developmental neurotoxicity testing using rodents is laborious and expensive, alternative species such as zebrafish are being adapted for rapid toxicity screening. Assessing the developmental neurotoxicity potential of chemicals in a rapid throughput mode will be aided by the identification and characterization of transcriptional biomarkers that can be measured accurately and rapidly. To this end, the developmental expression profiles of ten nervous system genes were characterized in 1 to 6 days post fertilization zebrafish embryos/larvae using quantitative real time PCR (qRT-PCR). Transcripts of synapsinII a (syn2a) and myelin basic protein (mbp) increased throughout development, while transcripts of gap43, elavl3, nkx2.2a, neurogenin1 (ngn1), α1-tubulin, and glial fibrillary acidic protein (gfap) initially increased, but subsequently declined. Transcripts for nestin and sonic hedgehog a (shha) decreased during development. We tested the responses of these potential biomarkers to developmental neurotoxicant exposure, and found that the expression profiles of a subset of genes were altered both during and after exposure to sublethal doses of ethanol, a known developmental neurotoxicant. Collectively, these data indicate that transcript levels of the candidate genes change during development in patterns which are consistent with literature reports, and that the expression of the transcripts is perturbed by treatment with a developmental neurotoxicant (ethanol). These results suggest that the expression profiles of these genes may be useful biomarkers for rapid evaluation of the developmental neurotoxicity potential of chemicals. Published by Elsevier Inc.

1. Introduction Defining the developmental neurotoxicity potential of environmental chemicals is important because chemical exposure may adversely affect brain function in childhood, with persistent effects into adulthood [19]. There are about 85,000 registered chemicals in commercial use in the United States, with more than 2000 new chemicals being registered each year. It is estimated that about 3–5% of industrial chemicals (2500–4200) cause neurotoxic effects; however most have never been tested for neurotoxicity, and even fewer have been tested for developmental neurotoxicity [12,22]. Traditionally, rats or mice are used for developmental neurotoxicity testing, but

⁎ Corresponding author. Neurotoxicology Division (Mail Drop B105-03), United States Environmental Protection Agency, 109 T.W. Alexander Drive, Research Triangle Park, NC 27711, USA. Tel.: +1 919 541 3558; fax: +1 919 541 3335. E-mail address: [email protected] (R. Ramabhadran). 1 Current address: Syngenta Biotechnology, Inc., Research Triangle Park, NC 27709, USA. 2 Current address: Integrated Risk Information System, National Center for Environmental Assessment, the United States Environmental Protection Agency, Research Triangle Park, NC 27711, USA. 0892-0362/$ – see front matter. Published by Elsevier Inc. doi:10.1016/j.ntt.2009.04.065

these tests are costly, time consuming and labor intensive [13]. Given the large number of chemicals with insufficient developmental neurotoxicity data, it is critical to develop novel methods for costeffective screening of chemicals for developmental neurotoxicity testing [15,30,34,54], and combining in vitro cell-based assays with in vivo assays in alternative animal models such as zebrafish [14,23,47]. Zebrafish (Danio rerio) is a vertebrate animal model that has many advantages for toxicity screening. These include economical husbandry, high fecundity, small size, rapid development, and transparency of embryos that permits non-invasive observations during development [23,26,60,61]. Because many aspects of embryonic development and organ system biology are conserved between zebrafish and mammals, zebrafish have been widely used for the study of genetics, developmental biology, and human disease processes [4,33]. Recently, zebrafish have also been used in drug discovery and environmental toxicity testing [45]. Several endpoints have been used in toxicity testing using zebrafish, including lethality, teratogenicity, behavioral changes and alterations in gene expression [2,41,45,50]. Among these endpoints, measuring gene expression changes using quantitative real time polymerase chain reaction (qRT-PCR) has the advantages of sensitivity, reproducibility and rapidity, all of which make this technique ideal

92

C.-Y. Fan et al. / Neurotoxicology and Teratology 32 (2010) 91–98

for large-scale screening [57]. To apply this approach in developmental neurotoxicity testing in zebrafish, however, it is essential to identify genes which can be monitored as biomarkers of zebrafish development in general, as well as more specific biomarkers of nervous system development. Although the zebrafish genome is fully sequenced, an appropriate set of transcriptional markers of development, the PCR primer pairs, methods of analysis, and the sensitivity of qRT-PCR have not been established in this system to enable rapid screening. To this end, it is necessary to identify candidate genes for nervous system development, develop and validate PCR primers, and also optimize the procedures and parameters for measurement of transcript levels in embryos/larvae. Before they can be applied to large scale screening, it is also important to determine if the candidate genes can serve as biomarkers for developmental neurotoxicology by demonstrating their responsiveness to prototypical developmental neurotoxicants. In this report, we assessed the temporal expression profiles of ten candidate genes in developing zebrafish embryos/larvae from day 1 to day 6 post fertilization (i.e., from approximately the 26+ somite stage until 2 days after hatching — at a time when the fish begin swimming and feeding). Seven genes known to be expressed in neuronal stem cells and/or in developing neurons were selected for this study: nestin [39], elavl3 (encoding HuC) [28,44], sonic hedgehog a (shha) [17], nkx2.2a [46], growth associated protein 43 (gap43) [55], α1-tubulin [21], and neurogenin1 (ngn1) [8,29]. Synapsin IIa (syn2a) was chosen as a marker of synapse formation [25]. Two genes expressed in other cell types comprising the nervous system were also included: glial fibrillary acidic protein (gfap) expressed in astrocytes [42] and myelin basic protein (mbp), expressed by oligodendrocytes as a component of the myelin sheath that surrounds many neurons [10]. Two genes that are ubiquitously-expressed in all cell types were included as normalization controls: ribosomal protein L13A (rbl13a) and elongation factor 1 alpha (ef1α) [53]. Six of these twelve genes in the set are expressed exclusively in the nervous system: mbp [10], elavl3 [28], ngn1 [29], gap43 [55], α1-tubulin [21] and gfap [42]. Three other genes are expressed in the nervous system, and to varying extents in other tissues: nestin [39], nkx2.2a [46], shha [17]. The expression pattern of syn2a has not been described in zebrafish, but it has been shown to be specifically expressed in the nervous system of other species [25]. We examined the developmental expression profiles of these genes using qRT-PCR, a quantitative and sensitive method to measure RNA levels [49]. First we determined the developmental expression profiles of the ten test genes in normally developing (unexposed) zebrafish embryos/larvae. Subsequently, as a proof-of-concept for using this approach for rapid developmental neurotoxicity testing, we assessed the expression of these genes in embryos/larvae exposed to a prototypical neurotoxicant, ethanol [18,20,43]. We conclude that developmental gene expression profile changes in the candidate genes examined have potential as biomarkers for the rapidly evaluating chemicals for developmental neurotoxic effects.

ing Inc., San Diego, CA). Fifteen fish were held in each 9 liter aquarium at a ratio of one male to two females. The photoperiod was set at 14 h light and 10 h of dark with lights coming on at 8:30 AM. Zebrafish were fed twice daily with flake food (Aquatic Eco-Systems Inc., Apopka, FL) and once daily with newly-hatched brine shrimp (Brine Shrimp Direct, Ogden, UT). Matings were set up 1 h prior to onset of the light cycle. Zebrafish pairs were transferred into mating tanks (Aquatic Habitats, Apopka, FL) with a mesh at the bottom of the inner chamber, and embryos were collected 1 h after the beginning of the light period. Fertilized embryos were sorted under an Olympus SZH10 stereo microscope (Olympus America, Center Valley, PA), rinsed in 10% Hanks' buffer (13.7 mM NaCl, 0.54 mM KCl, 25 µM Na2HPO4, 44 µM KH2PO4, 130 µM CaCl2, 100 µM MgSO4, 420 µM NaHCO3) to remove debris, and were subsequently disinfected in 0.01% of bleach (sodium hypochlorite) for 5 min to remove contaminating bacteria, fungi and parasites. Subsequently, the embryos were rinsed several times in 10% Hanks' buffer to remove residual bleach. All procedures used were approved in advance by the IACUC (Institute Animal Care and Use Committee) of the US EPA National Health and Environmental Effects Research Laboratory.

2. Materials and methods

2.5. qRT-PCR

2.1. Materials

Total RNA was extracted (n = 3) from the 10 pooled zebrafish embryos (not dechorionated) and larvae using RNeasy Mini Kit. RNA quality and concentration were determined respectively using an Agilent 2100 bioanalyzer (Agilent Technologies, Santa Clara, CA) and a NanoDrop 1000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA). Complementary DNAs (cDNAs) were synthesized using ‘SuperScript III First-Strand Synthesis SuperMix’ for qRT-PCR according to the manufacturer-recommended protocol. For qRT-PCR analysis, synthesized cDNAs were diluted to a working concentration of 1 ng/μl (calculated based on input RNA concentration). The 20 μl qRT-PCR reaction mix consisted of 10 μl of SYBR GreenER qPCR SuperMix Universal (2×), 0.5 μM of each PCR primer, and 1 ng cDNA template. qRT-PCR was performed using an ABI 7500 Real-Time PCR system

RNeasy Mini Kit was purchased from Qiagen (Valencia, CA). DNase was purchased from Ambion (Austin, TX). SuperScript III First-Strand Synthesis SuperMix for qRT-PCR, SYBR GreenER qPCR SuperMix Universal, and PCR primers were purchased from Invitrogen (Carlsbad, CA). Wild type zebrafish were purchased from Aquatic Research Organisms (Hampton, NH). 2.2. Zebrafish maintenance Zebrafish were maintained in an AAALAC-accredited animal facility at 26 °C in 9-liter aquaria in a recirculating system (Aquaneer-

2.3. Chemical exposure and sample collection Zebrafish embryos/larvae were raised in 3 ml 10% Hanks' buffer [58] in 6-well plates, with 20 embryos/larvae in each well and daily buffer replacement. Embryos/larvae were collected daily between day 1 and day 6 post fertilization (all references to ‘day’ in the rest of the sections are ‘day post fertilization’). For ethanol exposure experiments, a 1% (174 mM) ethanol solution was prepared by diluting 95% ethanol (Sigma–Aldrich, St. Louis, MO) in 10% Hanks' buffer. Groups of 20 embryos/larvae were exposed to ethanol in 20 mL borosilicate glass scintillation vials (Thermo Fisher Scientific, Waltham, MA) for 5 days (from 0 until day 5) with daily solution changes, and were then raised in 10% Hanks' buffer. At day 3 or day 6, 10 embryos/larvae (n = 3) were collected, pooled and euthanized in 250 mg/L of tricaine methanesulfonate (MS-222, Argent Chemical Laboratories, Redmond, WA), snap-frozen in liquid nitrogen, and stored at − 80 °C until RNA extraction. 2.4. Primer design Zebrafish-specific primers for qRT-PCR were designed using primer design tools provided by Integrated DNA Technologies Inc. (http://www.idtdna.com/Scitools/Applications/Primerquest/Default. aspx). Primer lengths varied between 18 and 25 base pairs, G/C contents were about 50% and melting/annealing temperatures (Tm's) were approximately 60 °C. The amplicons resulting from the qRT-PCR using these primers migrated between 80 and 150 base pairs in length when analyzed by agarose gel electrophoresis. Details of the primers are listed in Table 1.

C.-Y. Fan et al. / Neurotoxicology and Teratology 32 (2010) 91–98

2.6. Statistical analysis

Table 1 Primer pairs used in qRT-PCR analysis. Gene name rbl13a

Primer sequence

Forward: TCT GGA GGA CTG TAA GAG GTA TGC Reverse: AGA CGC ACA ATC TTG AGA GCA G gap43 Forward: TGC TGC ATC AGA AGA ACT AA Reverse: CCT CCG GTT TGA TTC CAT C elavl3 Forward: AGA CAA GAT CAC AGG CCA GAG CTT Reverse: TGG TCT GCA GTT TGA GAC CGT TGA gfap Forward: GGA TGC AGC CAA TCG TAA T Reverse: TTC CAG GTC ACA GGT CAG myelin basic Forward: AAT CAG CAG GTT CTT CGG AGG protein AGA Reverse: AAG AAA TGC ACG ACA GGG TTG ACG Synapsin IIa Forward: GTG ACC ATG CCA GCA TTT C Reverse: TGG TTC TCC ACT TTC ACC TT α1-tubulin Forward: AAT CAC CAA TGC TTG CTT CGA GCC Reverse: TTC ACG TCT TTG GGT ACC ACG TCA sonic Forward: GCA AGA TAA CGC GCA ATT CGG hedgehog a AGA Reverse: TGC ATC TCT GTG TCA TGA GCC TGT neurogenin1 Forward: TGC ACA ACC TTA ACG ACG CAT TGG Reverse: TGC CCA GAT GTA GTT GTG AGC GAA nkx2.2a Forward: CGG CAA ACC TTG CCA TAC GCT AAA Reverse: GCG CGT TAT ATT GCA TGT GCT GGA nestin Forward: ATG CTG GAG AAA CAT GCC ATG CAG Reverse: AGG GTG TTT ACT TGG GCC TGA AGA ef1α Forward: TTG AGA AGA AAA TCG GTG GTG CTG Reverse: GGA ACG GTG TGA TTG AGG GAA ATT C

93

GenBank ID

Amplicon size (bp)

NM_212784

148

NM_131341

82

NM_131449

107

NM_131373

97

AY860977

NM_001002597

102

All data are shown as mean± standard error of the mean (SEM). All statistical analyses were conducted using StatView 5.0.1 for Windows (SAS Institute, Inc., Cary, NC). Gene expression time course data obtained from untreated samples (Figs. 1–3), were analyzed initially using a global analysis of variance (ANOVA) with time and gene as the independent variables and gene expression level as the dependent variable (all genes analyzed in each sample were included together as a repeated measure in this initial analysis). Significant interaction of time by gene was observed (p b 0.0001; F-value= 12.48). Subsequent stepdown ANOVA analyses for the time course of expression of each gene were followed by Fisher's Protected Least Significant Difference post-hoc test to test for differences between each time point for each gene. For the analyses of the changes in gene expression induced by ethanol exposure (Fig. 4), all data for both the day 3 and day 6 time points were first analyzed using a global, 3-way ANOVA using ethanol treatment, day and

80

NM_194388

117

DRU30711

117

NM_131041

116

Dr.75083

113

XM_001919887

80

NM_131263

91

(Applied Biosystems, Foster City, CA). The PCR reaction cycle conditions used were: 50 °C for 1 min, 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min, and the fluorescent signals were measured at the annealing/extension step. Following completion of each qRT-PCR reaction, melting curve analyses were performed to validate the specificity of the PCR amplicons. Data were processed by ABI 7500 system SDS software version 1.3.1 (Applied Biosystems, Foster City, CA), and analyzed using Microsoft Excel 2003. For each gene, the relative level of transcript was determined and this relative level was further normalized to the RNA level of an internal control gene, ribosomal protein L13A (rpl13a). Transcript levels of each gene measured at different time points were normalized to their respective levels at day 1 or to the level at day 3 for mbp because its expression was undetectable at earlier times. The fold change relative to day 1 (or day 3 for mbp) RNA levels were calculated using the equation: fold change = 2− ΔΔCt [35], where ΔCt= Ct(target gene) − Ct(rpl13a) and ΔΔCt(day) = ΔCt(day) − ΔCt(day 1). In ethanol exposure experiments, the RNA transcript level of each gene was normalized to rpl13a levels on the day the RNA was collected. In this study, the RNA levels of the internal control gene, rpl13a, were not affected by 1% ethanol exposure. The fold changes for the ethanol exposure data were calculated as the ratio of the rpl13a-normalized RNA level from the 1% ethanol treated sample to the rpl13a-normalized RNA level of the untreated controls.

Fig. 1. Developmental expression profiles of gene class that show increasing and sustained expression during the first six days of zebrafish development. (A) ef1α; (B) syn2a; (C) mbp. Total RNA was purified from embryos/larvae pools collected daily from day 1 to day 6, and the RNA levels were determined by qRT-PCR as described in Materials and Methods with transcript of ribosomal protein gene rpl13a as the internal control. The RNA levels in each sample were normalized to that of internal control. Foldchanges in transcript levels shown at different days are relative to the level at day 1 for syn2a and ef1α, and relative to the level at day 3 for mbp (because no mbp expression was detectable before day 3). Values shown are the mean from three independent groups of embryos +/− SEM (n = 3). ⁎ Denotes statistically significant changes as compared to day 1 for ef1α and syn2a, and compared to day 3 for mbp (p b 0.05). At day 1, the level of expression (% relative to rpl13a) was 250 for ef1α and 0.05 for syn2a. The level of expression of mbp (% relative to rpl13a on day 3) was 0.3. Note: the ranges of the y axes differ among panels A, B, and C.

94

C.-Y. Fan et al. / Neurotoxicology and Teratology 32 (2010) 91–98

Fig. 2. Developmental expression profiles of gene class that show increased expression followed by decrease during the first six days of zebrafish development. (A) elavl3, (B) ngn1, (C) nkx2.2a, (D) α1-tubulin, (E) gfap, (F) gap43. Total RNA was purified from embryos/larvae pools collected from day 1 through day 6, and the RNA levels were determined by qRT-PCR as described in Materials and Methods with transcript of ribosomal protein gene rpl13a as the internal control. The transcript levels in each sample were normalized to that of the internal control. Fold-changes in transcript levels shown at different days are relative to the levels at day 1. Values shown are the mean from three independent groups of embryos +/− SEM (n = 3). ⁎ Denotes statistically significant changes as compared to the day 1 levels (p b 0.05). At day 1, the level of expression for each gene (% relative to rpl13a) was 1.0 for elavl3; 0.1 for ngn1; 0.1 for nkx2.2a; 6.0 for α1-tubulin; 2.0 for gfap; and 1.0 for gap43. Note: the ranges of the y axes differ among the 6 panels.

gene as the independent variables and level of gene expression as the dependent variable. This initial, global ANOVA revealed a significant (p b 0.0001; F-value = 17.19) treatment× gene× time point interaction. Subsequent step-down ANOVAs were used to assess the differences in gene expression in the ethanol-treated group versus the control group at day 3 and day 6. In all cases p b 0.05 was considered statistically significant.

protein synthesis translational elongation factor that is constitutively expressed in all cell types, showed no change relative to the value at day 1 (Fig. 1A). All of the other ten genes assessed showed distinctive patterns of expression relative to their levels at day 1 or day 3 for mbp (Figs. 1–3). These patterns of expression fall into three major categories as described below. 3.1. Gene class that shows increasing and sustained expression

3. Results The expression profiles of a number of genes were examined in normally developing zebrafish embryos/larvae during the first 6 days of development. (All references to ‘day in the text are ‘day post fertilization’). The RNA level of the internal control gene, ef1α, a

Expression of two genes, synapsin IIa (syn2a) and myelin basic protein (mbp), showed sharp increase at early times and sustained expression thereafter throughout the study period. As shown in the qRT-PCR assay (Fig. 1B), transcript levels of syn2a were dramatically up-regulated between day 2 and day 6. At day 3, syn2a transcript

C.-Y. Fan et al. / Neurotoxicology and Teratology 32 (2010) 91–98

95

other hand, was stable from day 1 to day 4 with an approximately a 25% decrease on day 5 and 50% decrease at day 6. 3.4. Ethanol exposure alters gene expression

Fig. 3. Developmental profiles of gene class that show decline in expression throughout development during the first six days of zebrafish development. (A) nestin, (B) shha. Total RNA was purified from embryos/larvae pools collected from day 1 through day 6, and the transcript levels were determined by qRT-PCR as described in Materials and Methods with transcript of ribosomal protein gene rpl13a as the internal control. The RNA levels in each sample were normalized to that of the internal control. Fold-changes in transcript levels shown at different days are relative to the levels at day 1. Values shown are the mean from three independent groups of embryos +/− SEM (n = 3). ⁎ Denotes statistically significant changes as compared to the day 1 levels (p b 0.05). At day 1, the level of expression for each gene (% relative to rpl13a) was 0.3 for nestin and 0.2 for shha. Note: the ranges of the y axes differ between panels A and B.

levels were elevated nearly 12-fold compared to day 1. After day 3, syn2a transcript levels remained at a plateau for up to 6 days. As shown in Fig. 1C, mbp transcript was not detected at day 1 or day 2. By day 3, however, mbp transcript was detectable and continued to increase nearly 10-fold between day 3 and day 6. 3.2. Gene class that shows increased expression followed by decline The expression pattern of this second class of genes was marked by high level of RNA early in embryonic development followed by decreasing levels over the following 6 days of development (Fig. 2). The expression of elavl3 transcript (Fig. 2A) peaked at day 2 and thereafter returned to day 1 levels. The transcript levels for ngn1 (Fig. 2B) and nkx2.2a (Fig. 2C) also peaked at day 2, but fell below day 1 levels during the following 4 days (i.e., day 3 through day 6). α1tubulin transcript (Fig. 2D) showed peak expression at day 2 and day 3, with expression returning to day 1 levels between day 4 and day 5, and then dropping below day 1 level on day 6. A slightly different pattern was observed for transcripts of gfap (Fig. 2E) and gap43 (Fig. 2F): the peak expression between day 2 and day 3 was followed by declining expression, but the level of expression always remained above the day 1 expression level. 3.3. Gene class that shows decline in expression levels throughout development The RNA levels for two genes, nestin and sonic hedgehog a (shha), were highest early in development with a decrease thereafter (Fig. 3A and B). nestin transcript level was highest on day 1 and day 2 followed by dramatic decreases between day 3 to day 6. shha RNA level, on the

Because ethanol is well-recognized developmental neurotoxicant known to cause fetal alcohol syndrome in humans [20] and developmental defects in zebrafish [11], we used ethanol to perturb embryonic/ larval zebrafish development. Zebrafish were dosed beginning 6 h postfertilization with a sublethal dose of ethanol (1%, 174 mM [48]). This concentration of ethanol is below levels that cause high degree of lethality but one that produces teratological [32,48] and behavioral effects [16,36]. Control and exposed embryos/larvae were collected at day 3 and day 6 and assessed for RNA levels. At day 3, five genes (gfap, nestin, ngn1, nkx2.2a, α1-tubulin) showed increased RNA levels, and one (mbp) showed decreased levels (Fig. 4A) relative to untreated controls. The increases in transcript level compared to control embryos/larvae were as follows: 1.5-fold for gfap, 1.7-fold for nestin, 2.7-fold for ngn1, 1.4-fold for nkx2.2a, and 1.3-fold for α1-tubulin. Conversely, mbp transcript level declined to 0.3-fold of control following developmental ethanol exposure. At day 6 (24 h after the cessation of ethanol exposure), the levels of two of the genes, gfap and nestin, remained elevated by 1.8fold and 1.5-fold respectively relative to control (Fig. 4B). At day 6, there was also a small, but statistically significant, 15% increase in the expression of the ef1α gene, which has been used as a control to normalize gene in other studies [53]. In this study, the transcript levels of rpl13 were used to normalize gene expression levels as its levels were unaffected by 1% ethanol exposure. 4. Discussion Rodents are traditionally used for developmental neurotoxicity testing; however these tests are expensive, time consuming and labor intensive [13]. Because a large number of chemicals have no available developmental neurotoxicity data, it is important to implement alternate methods for cost-effective screening and prioritization of chemicals for further detailed testing in other established systems [15]. Alternative species-based animal models such as zebrafish have the potential for developmental neurotoxicity screening. In order to apply the rapid and sensitive method of qRT-PCR as a screening tool in zebrafish, we characterized temporal transcription profiles of a set of genes that are associated primarily with the zebrafish nervous system. The expression profiles showed that some genes were expressed very early in development, while others peaked during mid-development, and yet others were only marginally expressed early in development, but showed dramatically increased expression during mid to later development. In contrast to our qRT-PCR approach, most reports in the literature have used in situ hybridization to localize RNA and/or immunostaining to localize proteins. This study therefore is the first detailed report of expression analysis using qRT-PCR a technique that allows the rapid determination of gene expression profiles required in rapid screens. Our data show that in normal (unexposed) embryos/larvae the expression profiles of the candidate genes studied here fall into three classes and these are discussed below in the context of literature reports. We further show that the well-recognized developmental neurotoxicant, ethanol, perturbs the expression profiles of a subset of neuronal gene expression profiles suggesting that these genes might provide biomarkers for the rapid assessment of neurotoxic potential of chemicals. The transcripts for two genes, nestin and shha, were maximally expressed early in development of the zebrafish embryo. This pattern of expression is consistent with the role of these genes in development as suggested by the previously characterized expression pattern in developing zebrafish. Nestin is a type IV intermediate filament protein widely expressed in the developing mammalian nervous system, in neuronal progenitor cells and neuronal stem cells, and is considered a marker for stem cells committed to the neural fate [31]. It has been

96

C.-Y. Fan et al. / Neurotoxicology and Teratology 32 (2010) 91–98

Fig. 4. Ethanol-induced changes in gene transcript levels in developing zebrafish embryos/larvae. (A) day 3 and (B) day 6 following ethanol exposure as described in the text. The fold-changes in gene expression were normalized to an internal control gene rpl13a. The values of fold-change for ethanol treated samples at day 3 or day 6 reflect the ratio to the transcript levels relative to untreated controls at the corresponding time points. The values shown are the mean from five independent groups of embryos +/−SEM (n = 5). ⁎ Denotes changes in transcript level relative to normalized control embryos that are statistically significant at p b 0.05. Note that the transcript level of ef1α was slightly increased at day 6 following 5 days of ethanol exposure.

shown using in situ hybridization to be expressed very early during zebrafish embryonic development where nestin transcripts were first detected at 11 h post fertilization and were expressed throughout the developing nervous system by 24 h post fertilization. From day 2 through day 4, nestin signal was reduced and restricted to proliferative zones of in the central nervous system [39]. These reported temporal patterns agree with our qRT-PCR results where nestin transcription was maintained at stable levels from day 1 to day 2, and subsequently decreased through day 6 post-fertilization (Fig. 3A). Sonic hedgehog (shha) is a morphogen that patterns many organ systems (including the nervous system), and is considered an important signaling molecule in zebrafish embryogenesis [24,40]. In an shha promoter-GFP transgenic reporter zebrafish, GFP signal was first detected in the embryonic shield at 6 h post fertilization. At the 8-somite stage (about 11 h), the transgene was expressed in the notochord, the floor plate, and in the ventral forebrain. By 48 h, GFP was expressed widely in the nervous system. At 3 weeks post fertilization, GFP continued to be detectable in the brain and the ventral midline of the spinal cord [17] of the larvae. Although we did not assess gene expression any earlier than 24 h (i.e., day 1), we also found very early expression of shha, and no significant decrease until day 5 (Fig. 3B). The gene expression profile of shha reported in this study is in general agreement with the GFP transgene expression in the transgenic zebrafish line. Transcripts for six other genes — elavl3, ngn1, nkx2.2a, α1-tubulin, gfap, and gap43 — showed an expression patterns that differed from those of nestin and shha, with peak expression levels occurring after day 1, but before day 6 (Fig. 2A–F). HuC (encoded by elavl3) and ngn1

proteins have been reported to be early neuronal markers in zebrafish and mammals [28,29,38]. Zebrafish elavl3 is an ortholog of the Drosophila elav gene which encodes the RNA-binding protein, HuC [28]. During zebrafish development, elavl3 is expressed in the neuronal precursors in the neural plate, with highest level observed at 32 h post fertilization [28,29,44]. This result concurs with our observation using of peak RNA expression level at day 2 (Fig. 2A). Ngn1 is a helix–loop–helix transcription factor expressed in the proliferating progenitor cells during initiation of neuronal differentiation and plays a crucial role in directing neurogenesis [7,8]. In an in situ hybridization study in zebrafish, ngn1 transcripts were found to be expressed early in the neural plate at 3-somite stage (about 10 h post fertilization) and by 24 h ngn1 expression was found in many regions throughout the brain of the embryo [7]. A study using ngn1 promoter-GFP transgenic zebrafish showed that by 48 h post fertilization GFP was expressed strongly in the neurons along the neural tube of embryos [8]. No literature reports have examined the expression profile of ngn1 after 48 h. In the present study, the peak of ngn1 transcription observed at 48 h (Fig. 2B) which is in agreement with GFP transgene expression. This result suggests that ngn1 may play a role in early zebrafish nervous system development, similar to the reported roles for ngn1 homologues in other species [9]. nkx2.2a is a member of a family of homeobox genes which may be involved in the dorsoventral patterning of the mammalian and zebrafish nervous system as well as of other structures [3,51]. In situ hybridization first detected nkx2.2a expression at 10 h post fertilization, and the gene continued to be expressed for at least the first 48 h [3]. In an nkx2.2a promoter-GFP transgenic zebrafish model, GFP signal was detected at

C.-Y. Fan et al. / Neurotoxicology and Teratology 32 (2010) 91–98

12-somite-stage (about 14 h post fertilization). At 24 h post fertilization, GFP was found in the brain and the ventral spinal cord. The transgene was expressed robustly up to day 4, the final time point reported [46]. These results agree with our results indicating nkx2.2a to be an earlyexpressed gene with a peak of expression at day 2 followed by a decline thereafter (Fig. 2C). α1-tubulin is an intermediate filament protein forming the microtubule cytoskeleton in developing or regenerating axons and dendrites. It was reported α1-tubulin transcripts were first detectable at 12 h post fertilization in the zebrafish embryo, reaching a peak level at day 2 and declining between day 3 and day 7 [21]. The gene expression profile of α1-tubulin obtained in this study (Fig. 2D) closely matches the reported pattern. Glial fibrillary acidic protein (GFAP) is an intermediate filament protein (class III) that is highly expressed in astrocytes and radial glial cells of the central nervous system and is considered a marker of astroglia in the brain [42]. In a gfap promoter-GFP transgenic zebrafish line, GFP signal was first detected at 12 h post fertilization with the signal increasing until day 4 [5]. The gfap transcription profile observed in the present study (Fig. 2E) agrees with the reported time course in gfap-GFP transgenic zebrafish. Growth associated protein 43 (gap43) is expressed at high levels in zebrafish neurons during development and axonal regeneration [56]. In gap43 promoter-GFP transgenic zebrafish, GFP was detected at 17 h post fertilization, with the GFP signal increasing until day 3 [55]. The gap43 gene expression profile observed in this study, showing elevated levels of gap43 transcripts between day 2 and day 5 is concordant with the observations in the transgenic strain. In general, both synaptogenesis and myelination are known to be processes that occur during later stages of nervous system development. Consistent with this observation, the transcript levels of the two putative biomarkers for these processes demonstrate peak expression levels at later stages of development. Transcripts for myelin basic protein (mbp: a biomarker of myelination) and synapsin IIa (a biomarker of synapse formation) showed the most dramatic increase around days 3–4, and expression remained high at day 6. Mbp is required for myelination of axons in the developing central nervous system of the zebrafish [10,59]. It has been reported that the mbp transcripts were first detected at day 2 in a small number of cells in the ventral hindbrain, and by day 4 more cells in both the hind- and mid-brain showed mbp expression with a considerably increased expression from day 4 to day 7 [10,59]. These results are consistent with data obtained in this study which show that mbp transcription is elevated significantly between day 3 and day 6. Synapsin IIa (syn2a) is a neuronal phosphoprotein that binds small synaptic vesicles to induce further synaptogenesis in mammals, playing an important role in both synaptogenesis and neurotransmitter release [25]. No reports on transcription of syn2a during zebrafish development are available, but our data are consistent (compare Fig. 1B with Fig. 1C) with synapse formation preceding myelination; syn2a transcription was up-regulated from day 2 to day 6, while mbp expression emerged later and did not reach a peak until day 4 to day 6. Ethanol is recognized as a chemical that adversely affects mammalian and zebrafish development [6,11,32,48]. Using a sublethal dose of ethanol capable of inducing teratological and behavior changes, we showed that the expression profiles of a subset of genes in this study were altered [48]. gfap and nestin expression levels remained elevated relative to control levels at 24 h after the exposure to ethanol had ceased (i.e., day 6). Interestingly, genes whose expression was altered at day 6 were a subset of the genes that were affected at day 3 (compare Fig. 4A with Fig. 4B). It is significant that the observed effects of ethanol did not stem from a global suppression or activation of transcription as expression of all genes was not uniformly altered, nor were all genes affected in response to ethanol exposure. Although mechanisms of ethanol action were not the focus of this study, a limited comparison of our data can be made with literature observations on two genes, shha and gfap, performed using microarrays and qRT-PCR. These observations are made, however, with caveats regarding differences in experimental factors such as developmental

97

stage at exposure, the dose, the time of measurement, and most importantly, the specific strain of zebrafish used. Consistent with our observation, Kily et al. [27] reported that gfap remained up-regulated following 174 mM (1%) ethanol exposure in a microarray study of 4month old adult zebrafish exposed 20 min per day for 4 weeks followed by 21 days with no ethanol exposure. Contrary to our findings, however, two studies in zebrafish have reported a reduction of shha transcript levels by ethanol exposure at doses similar to the one used in this study [32,37]. This discrepancy may be attributed to differences in the experimental factors, and/or to differences in the specific stains used in the studies. Three groups have reported strain-dependent differences in responses to ethanol-induced lethality and behavior deficits [16,36,37]. Given this inter-strain variability within zebrafish, comparisons with other classes within the vertebrate subphylum such as chicken [1] may not be informative. Microarray studies have been performed in zebrafish to investigate gene expression changes on a genome scale [27]. Although microarrays permit the evaluation of large gene sets, the microarray results require confirmation using qRT-PCR which is a more sensitive and reproducible technique [52]. Thus the qRT-PCR approach used in this study is ideal for characterizing small sets of candidate genes for use in toxicity testing in a relative high throughput manner. In summary, our results show that expression levels of zebrafish genes examined were influenced by developmental stage of the embryos/ larvae. The observed profiles are consistent with the published data obtained using other methods, indicating that qRT-PCR is an efficient and rapid method to characterize gene expression changes in comparison to methods such as immunostaining or in situ hybridization. The ethanol exposure data serves as a proof-of-concept indicating that transcript levels of some genes in this study are responsive to sublethal doses of a developmental neurotoxicant. Because some of the affected genes are located exclusively in the nervous system (α1-tubulin, ngn1, mbp and gfap), we believe that changes in gene expression profiles has potential use as an endpoint for efficient screening for developmental neurotoxicity using zebrafish. These transcriptional biomarkers are currently being used to compare chemicals with known mammalian developmental neurotoxicity potential with those devoid of developmental neurotoxic potential to assess further the method and to relate the observed gene expression changes to a mode of action where possible. Conflict of interest statement Nothing to declare. Acknowledgments We thank Beth Padnos and Dr. David Kurtz for their care and oversight of the zebrafish breeding colony. We also would like to thank Dr. Joyce Royland for her helpful comments, and Drs. Kevin M Crofton and Susan Hester for their critical review of the manuscript. This work was supported by a training grant: EPA CR 833237. Disclaimer The research described in this document has been funded wholly by the U.S. Environmental Protection Agency. It has been reviewed by the National Health and Environmental Effects Research Laboratory and approved for publication. Approval does not signify that the contents necessarily reflect the views of the Agency nor does mention of trade names or commercial products constitute endorsement or recommendation for use. References [1] S.C. Ahlgren, V. Thakur, M. Bronner-Fraser, Sonic hedgehog rescues cranial neural crest from cell death induced by ethanol exposure, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 10476–10481.

98

C.-Y. Fan et al. / Neurotoxicology and Teratology 32 (2010) 91–98

[2] T.P. Barros, W.K. Alderton, H.M. Reynolds, A.G. Roach, S. Berghmans, Zebrafish: an emerging technology for in vivo pharmacological assessment to identify potential safety liabilities in early drug discovery, Br. J. Pharmacol. 154 (2008) 1400–1413. [3] K.A. Barth, S.W. Wilson, Expression of zebrafish nk2.2 is influenced by sonic hedgehog/vertebrate hedgehog-1 and demarcates a zone of neuronal differentiation in the embryonic forebrain, Development 121 (1995) 1755–1768. [4] J. Berger, P. Currie, The role of zebrafish in chemical genetics, Curr. Med. Chem. 14 (2007) 2413–2420. [5] R.L. Bernardos, P.A. Raymond, GFAP transgenic zebrafish, Gene Expr. Patterns 6 (2006) 1007–1013. [6] J. Bilotta, J.A. Barnett, L. Hancock, S. Saszik, Ethanol exposure alters zebrafish development: a novel model of fetal alcohol syndrome, Neurotoxicol. Teratol. 26 (2004) 737–743. [7] P. Blader, N. Fischer, G. Gradwohl, F. Guillemot, U. Strahle, The activity of neurogenin1 is controlled by local cues in the zebrafish embryo, Development 124 (1997) 4557–4569. [8] P. Blader, C. Plessy, U. Strahle, Multiple regulatory elements with spatially and temporally distinct activities control neurogenin1 expression in primary neurons of the zebrafish embryo, Mech. Dev. 120 (2003) 211–218. [9] P. Blader, C.S. Lam, S. Rastegar, R. Scardigli, J.C. Nicod, N. Simplicio, C. Plessy, N. Fischer, C. Schuurmans, F. Guillemot, et al., Conserved and acquired features of neurogenin1 regulation, Development 131 (2004) 5627–5637. [10] C. Brosamle, M.E. Halpern, Characterization of myelination in the developing zebrafish, Glia 39 (2002) 47–57. [11] M.J. Carvan , E. Loucks, D.N. Weber, F.E. Williams, Ethanol effects on the developing zebrafish: neurobehavior and skeletal morphogenesis, Neurotoxicol. Teratol. 26 (2004) 757–768. [12] L. Claudio, An analysis of the U.S. Environmental Protection Agency neurotoxicity testing guidelines, Regul. Toxicol. Pharmacol. 16 (1992) 202–212. [13] L. Claudio, W.C. Kwa, A.L. Russell, D. Wallinga, Testing methods for developmental neurotoxicity of environmental chemicals, Toxicol. Appl. Pharmacol. 164 (2000) 1–14. [14] S. Coecke, A.M. Goldberg, S. Allen, L. Buzanska, G. Calamandrei, K. Crofton, L. Hareng, T. Hartung, H. Knaut, P. Honegger, et al., Workgroup report: incorporating in vitro alternative methods for developmental neurotoxicity into international hazard and risk assessment strategies, Environ. Health Perspect. 115 (2007) 924–931. [15] D.J. Dix, K.A. Houck, M.T. Martin, A.M. Richard, R.W. Setzer, R.J. Kavlock, The ToxCast program for prioritizing toxicity testing of environmental chemicals, Toxicol. Sci. 95 (2007) 5–12. [16] C.A. Dlugos, R.A. Rabin, Ethanol effects on three strains of zebrafish: model system for genetic investigations, Pharmacol. Biochem. Behav. 74 (2003) 471–480. [17] R. Ertzer, F. Muller, Y. Hadzhiev, S. Rathnam, N. Fischer, S. Rastegar, U. Strähle, Cooperation of sonic hedgehog enhancers in midline expression, Dev. Biol. 301 (2007) 578–589. [18] C.R. Goodlett, K.H. Horn, F.C. Zhou, Alcohol teratogenesis: mechanisms of damage and strategies for intervention, Exp. Biol. Med. (Maywood) 230 (2005) 394–406. [19] P. Grandjean, P.J. Landrigan, Developmental neurotoxicity of industrial chemicals, Lancet 368 (2006) 2167–2178. [20] M. Guizzetti, M. Catlin, L.G. Costa, The effects of ethanol on glial cell proliferation: relevance to the fetal alcohol syndrome, Front. Biosci. 2 (1997) e93–e98. [21] A. Gulati-Leekha, D. Goldman, A reporter-assisted mutagenesis screen using alpha 1-tubulin-GFP transgenic zebrafish uncovers missteps during neuronal development and axonogenesis, Dev. Biol. 296 (2006) 29–47. [22] T. Gura, Toxicity testing moves from the legislature to the petri dish-and back, Cell 134 (2008) 557–559. [23] A.J. Hill, H. Teraoka, W. Heideman, R.E. Peterson, Zebrafish as a model vertebrate for investigating chemical toxicity, Toxicol. Sci. 86 (2005) 6–19. [24] P.W. Ingham, A.P. McMahon, Hedgehog signaling in animal development: paradigms and principles, Genes Dev. 15 (2001) 3059–3087. [25] H.T. Kao, B. Porton, A.J. Czernik, J. Feng, G. Yiu, M. Haring, F. Benfenati, P. Greengard, A third member of the synapsin gene family, Proc. Natl. Acad. Sci. U. S. A. 95 (1998) 4667–4672. [26] G. Kari, U. Rodeck, A.P. Dicker, Zebrafish: an emerging model system for human disease and drug discovery, Clin. Pharmacol. Ther. 82 (2007) 70–80. [27] L.J. Kily, Y.C. Cowe, O. Hussain, S. Patel, S. McElwaine, F.E. Cotter, C.H. Brennan, Gene expression changes in a zebrafish model of drug dependency suggest conservation of neuro-adaptation pathways, J. Exp. Biol. 211 (2008) 1623–1634. [28] C.H. Kim, E. Ueshima, O. Muraoka, H. Tanaka, S.Y. Yeo, T.L. Huh, N. Miki, Zebrafish elav/HuC homologue as a very early neuronal marker, Neurosci. Lett. 216 (1996) 109–112. [29] C.H. Kim, Y.K. Bae, Y. Yamanaka, S. Yamashita, T. Shimizu, R. Fujii, H.C. Park, S.Y. Yeo, T.L. Huh, M. Hibi, et al., Overexpression of neurogenin induces ectopic expression of HuC in zebrafish, Neurosci. Lett. 239 (1997) 113–116. [30] P. Lein, P. Locke, A. Goldberg, Meeting report: alternatives for developmental neurotoxicity testing, Environ. Health Perspect. 115 (2007) 764–768. [31] U. Lendahl, L.B. Zimmerman, R.D. McKay, CNS stem cells express a new class of intermediate filament protein, Cell 60 (1990) 585–595.

[32] Y.X. Li, H.T. Yang, M. Zdanowicz, J.K. Sicklick, Y. Qi, T.J. Camp, A.M. Diehl, Fetal alcohol exposure impairs Hedgehog cholesterol modification and signaling, Lab. Invest. 87 (2007) 231–240. [33] G.J. Lieschke, P.D. Currie, Animal models of human disease: zebrafish swim into view, Nat. Rev. Genet. 8 (2007) 353–367. [34] E. Linney, L. Upchurch, S. Donerly, Zebrafish as a neurotoxicological model, Neurotoxicol. Teratol. 26 (2004) 709–718. [35] K.J. Livak, T.D. Schmittgen, Analysis of relative gene expression data using realtime quantitative PCR and the 2(-Delta Delta C(T)) method, Methods 25 (2001) 402–408. [36] B. Lockwood, S. Bjerke, K. Kobayashi, S. Guo, Acute effects of alcohol on larval zebrafish: a genetic system for large-scale screening, Pharmacol. Biochem. Behav. 77 (2004) 647–654. [37] E. Loucks, M.J. Carvan, Strain-dependent effects of developmental ethanol exposure in zebrafish, Neurotoxicol. Teratol. 26 (2004) 745–755. [38] Q. Ma, C. Kintner, D.J. Anderson, Identification of neurogenin, a vertebrate neuronal determination gene, Cell 87 (1996) 43–52. [39] J. Mahler, W. Driever, Expression of the zebrafish intermediate neurofilament Nestin in the developing nervous system and in neural proliferation zones at postembryonic stages, BMC Dev. Biol. 7 (2007) 89. [40] F. Muller, B. Chang, S. Albert, N. Fischer, L. Tora, U. Strähle, Intronic enhancers control expression of zebrafish sonic hedgehog in floor plate and notochord, Development 126 (1999) 2103–2116. [41] J. Muncke, R.I. Eggen, Vitellogenin 1 RNA as an early molecular biomarker for endocrine disruption in developing zebrafish (Danio rerio), Environ. Toxicol. Chem. 25 (2006) 2734–2741. [42] A.L. Nielsen, A.L. Jorgensen, Structural and functional characterization of the zebrafish gene for glial fibrillary acidic protein, GFAP, Gene 310 (2003) 123–132. [43] J.W. Olney, D.F. Wozniak, N.B. Farber, V. Jevtovic-Todorovic, P. Bittigau, C. Ikonomidou, The enigma of fetal alcohol neurotoxicity, Ann. Med. 34 (2002) 109–119. [44] H.C. Park, C.H. Kim, Y.K. Bae, S.Y. Yeo, S.H. Kim, S.K. Hong, J. Shin, K.W. Yoo, M. Hibi, T. Hirano, et al., Analysis of upstream elements in the HuC promoter leads to the establishment of transgenic zebrafish with fluorescent neurons, Dev. Biol. 227 (2000) 279–293. [45] C. Parng, In vivo zebrafish assays for toxicity testing, Curr. Opin. Drug Discov. Dev. 8 (2005) 100–106. [46] S. Pauls, E. Zecchin, N. Tiso, M. Bortolussi, F. Argenton, Function and regulation of zebrafish nkx2.2a during development of pancreatic islet and ducts, Dev. Biol. 304 (2007) 875–890. [47] R.T. Peterson, R. Nass, W.A. Boyd, J.H. Freedman, K. Dong, T. Narahashi, Use of nonmammalian alternative models for neurotoxicological study, Neurotoxicology 29 (2008) 546–555. [48] M.J. Reimers, A.R. Flockton, R.L. Tanguay, Ethanol- and acetaldehyde-mediated developmental toxicity in zebrafish, Neurotoxicol. Teratol. 26 (2004) 769–781. [49] T.D. Schmittgen, K.J. Livak, Analyzing real-time PCR data by the comparative C(T) method, Nat. Protoc. 3 (2008) 1101–1108. [50] S. Scholz, S. Fischer, U. Gundel, E. Kuster, T. Luckenbach, D. Voelker, The zebrafish embryo model in environmental risk assessment—applications beyond acute toxicity testing, Environ. Sci. Pollut. Res. Int. 15 (2008) 394–404. [51] K. Shimamura, D.J. Hartigan, S. Martinez, L. Puelles, J.L. Rubenstein, Longitudinal organization of the anterior neural plate and neural tube, Development 121 (1995) 3923–3933. [52] S.L. Spurgeon, R.C. Jones, R. Ramakrishnan, High throughput gene expression measurement with real time PCR in a microfluidic dynamic array, PLoS ONE 3 (2008) e1662. [53] R. Tang, A. Dodd, D. Lai, W.C. McNabb, D.R. Love, Validation of zebrafish (Danio rerio) reference genes for quantitative real-time RT-PCR normalization, Acta. Biochim. Biophys. Sin. (Shanghai) 39 (2007) 384–390. [54] R.L. Tanguay, M.J. Reimers, Analysis of ethanol developmental toxicity in zebrafish, Methods Mol. Biol. 447 (2008) 63–74. [55] A.J. Udvadia, 3.6 kb genomic sequence from Takifugu capable of promoting axon growth-associated gene expression in developing and regenerating zebrafish neurons, Gene Expr. Patterns 8 (2008) 382–388. [56] A.J. Udvadia, R.W. Koster, J.H. Skene, GAP-43 promoter elements in transgenic zebrafish reveal a difference in signals for axon growth during CNS development and regeneration, Development 128 (2001) 1175–1182. [57] D. Voelker, C. Vess, M. Tillmann, R. Nagel, G.W. Otto, R. Geisler, K. Schirmer, S. Scholz, Differential gene expression as a toxicant-sensitive endpoint in zebrafish embryos and larvae, Aquat. Toxicol. 81 (2007) 355–364. [58] M. Westerfield, The zebrafish book, A Guide for the Laboratory Use of Zebrafish (Danio rerio), 4th Edition, University of Oregon Press, Eugene, 2000. [59] M. Yoshida, W.B. Macklin, Oligodendrocyte development and myelination in GFPtransgenic zebrafish, J. Neurosci. Res. 81 (2005) 1–8. [60] L.I. Zon, Zebrafish: a new model for human disease, Genome Res. 9 (1999) 99–100. [61] L.I. Zon, R.T. Peterson, In vivo drug discovery in the zebrafish, Nat. Rev. Drug Discov. 4 (2005) 35–44.