General and Comparative Endocrinology 172 (2011) 120129

Contents lists available at ScienceDirect

General and Comparative Endocrinology

journal homepage: www.elsevier.com/locate/ygcen

The knockdown of the maternal estrogen receptor 2a (esr2a) mRNA affects

embryo transcript contents and larval development in zebrash
Andrea Celeghin a,1, Francesca Benato a,1, Surachai Pikulkaew b, Md. Golam Rabbane a, Lorenzo Colombo a,
Luisa Dalla Valle a,
a
Comparative Endocrinology Laboratory, Department of Biology, University of Padova, Via U. Bassi 58/B, 35131 Padova, Italy
b
Faculty of Veterinary Medicine, Chiang Mai University, Chiang Mai 50100, Thailand

a r t i c l e i n f o a b s t r a c t

Article history: In zebrash, ovulated oocytes are loaded with maternal estrogen receptor 2a (esr2a) mRNA which is
Available online 1 January 2011 spread as granular and lamentous structures throughout the central ooplasm and is promptly relocated
inside the blastodisc area at the 1-cell stage (0.2 h post-fertilization, hpf), as shown by in situ hybridiza-
Keywords: tion. This transcript is available for translation until its sharp decline from 4 to 8 hpf, being replaced by
Estrogen receptor low levels of zygotic esr2a mRNA mainly localized in the head region and around the yolk sac from 24 hpf
esr2a until hatching at 48 hpf. To test the functional role of the maternal esr2a mRNA, 1- or 2-cell embryos were
Zebrash
injected with 10.3 ng each of morpholino (MO) to knockdown translation (MO2-esr2a) of both maternal
Development
Morpholino
and zygotic esr2a transcripts, with a missplicing MO (MO3-esr2a) to effectively block post-transcription-
ally the zygotic transcript alone, and with a non-specic MO-control. Treatment with MO2-esr2a
increased apoptosis in embryos, especially in the brain, and caused severe malformations in 63% of 1
5 dpf larvae, as compared to 1011% in those treated with MO3-esr2a and MO-control. Defects included
body growth delay with curved shape, persistent yolk sac with reduced sub-intestinal veins and swollen
yolk extension, abnormal brain and splanchnocranium development, smaller eyes and otic vesicles, peri-
cardial oedema, uninated swim bladder and rudimentary caudal n with aberrant circular swimming.
Affected larvae could survive for only 1214 days. The MO2-esr2a phenotype was rescued with co-injec-
tion of 30 pg/embryo of mutated zebrash esr2a mRNA encoding the full length of Esr2a, but containing
eight silent mutations in the region recognised by MO2-esr2a. A lower dosage (15 pg) failed to recover
mortality and abnormality. Raising the dosage to 60 and 90 pg increased abnormality, but not mortality,
whereas with 120 pg both mortality and abnormality worsened, indicating a strict quantitative require-
ment of Esr2a. Co-injection of an anti-p53 MO failed to rescue the MO2-esr2a phenotype, eliminating the
possibility of off-target effects. Pangenomic microarray analysis revealed that 240 and 219 signicantly
expressed transcripts were up- and down-regulated, respectively, by maternal Esr2a protein deciency in
8-hpf MO2-esr2a embryos. Also at 48 hpf, 162 and 120 presumably zygotic transcripts were up- and
down-regulated, respectively, but only 18 were in common with each of the 8-hpf sets. In total, the tran-
scripts from 705 genes were affected by Esr2a knockdown. These ndings suggest the involvement of
maternal esr2a mRNA, presumably transactivated by maternal 17b-estradiol stored in the oocyte from
enveloping granulosa cells, in the epigenetic programming of zebrash development.
2010 Elsevier Inc. All rights reserved.

1. Introduction
Estrogen signaling participates in the control of a wide range of
processes related to vertebrate reproductive and non-reproductive
functions throughout embryonic development, postnatal growth
and adult life. Multiple and interacting transduction mechanisms
mediate estrogen induction of both genomic and non-genomic
responses via nuclear and membrane receptors, respectively. In
Corresponding author. Fax: +39 0498276199.
E-mail address: luisa.dallavalle@unipd.it (L. Dalla Valle).
1
Both authors contributed equally to this work.
mammals, there are two nuclear estrogen receptors, Esr1 and
Esr2 (formerly Era and Erb), which belong to a superfamily of tran-
scription factors. They affect the transcription of target genes
either by binding to estrogen responsive elements (ERE) on their
promoters as estrogen-activated complexes or by associating in li-
ganded or unliganded states with other transcription factors. They
are structurally homologous, but functionally different as to tissue
specicity, splicing processing, estrogen and agonist afnity, ligand
requirement for transactivation, transcriptional modulator selec-
tivity and cell proliferation control [38]. In modied forms, these
receptors may also be diverted to the cytoplasmic face of the
plasma membrane and trigger the formation of various second

0016-6480/$ - see front matter 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.ygcen.2010.12.020
 A. Celeghin et al. / General and Comparative Endocrinology 172 (2011) 120129
messengers for quick responses to estrogens, a role assigned also to
a 7-transmembrane G protein-coupled receptor, GPR30, that
equally responds to estrogens with rapid cellular signaling [31].
During evolution, the two nuclear estrogen receptors have been
remarkably conserved, particularly in their DNA and ligand-bind-
ing domains, but two Esr2 subtypes, named Esr2a and Esr2b (for-
merly Erb2 and Erb1) have been identied in several euteleosts,
including the zebrash [23]. This duplication of the esr2 gene is
probably a remnant of a round of tetraploidization/rediploidization
events that occurred during the early evolution of the actinoptery-
gian lineage with following subfunction partitioning between the
two paralogs, while an esr1 duplicate was lost [39].
In the zebrash, a basal species in the taxon radiation and an
ideal model for developmental studies, the three receptors cover
differential roles during early embryogenesis and later organogen-
esis, as suggested by the dissimilar proles of their respective total
mRNAs of maternal or zygotic origins. Among maternal transcripts
for steroid hormone receptors loaded in unfertilized eggs, esr2a
mRNA is abundant and second only to the glucocorticoid receptor,
as measured by qRT-PCR [29]. Conversely, the maternal transcripts
of Esr1 and Esr2b are absent or negligible, respectively. Within
6.5 h post-fertilization (hpf), maternal esr2a mRNA is rapidly de-
graded, becoming barely detectable at 8 hpf and being replaced
by the zygotic transcript increasing at low levels from 12 till 72
hpf. At the same time, there is a sharp increase of esr2b mRNA
starting at 12 hpf and becoming the prevalent receptor messenger
at 48 hpf. Esr1 mRNA begins to increase at 24 hpf and surpasses the
other receptors at 72 hpf. Hence the three receptors transcripts
peak in succession during development, as inferred from several
studies [2,18,24,29,37].
Presumably owing to the brief existence of the maternal esr2a
messenger and the scarce opportunity for zygotic transcription
during the rst 3 hpf till mid-blastula transition (MBT) - because
cell divide synchronously every 1415 min, devoting 6 min to
mitosis and the rest to DNA replication [16] it has been consid-
ered of lesser signicance as compared to the zygotic receptor
transcripts [2]. These have been localized in different embryonic
and larval tissues, like the forebrain, neuromasts, epidermis, hatch-
ing gland, pectoral n buds, liver [10,24,37] and to be required for
the development of the central and peripheral nervous systems,
cardiovascular functioning, normal sensory-motor behaviour and
body growth [13,25].
However, since growing evidence indicates that maternal ste-
roid hormones may participate in the maternal programming of
developmental processes in both sh [1] and mammals [4], that
sh oocytes may be loaded with estrogens secreted by contouring
granulosa cells during vitellogenesis [14], and that initial regula-
tions may exert profound effects on subsequent ontogenesis, these
facts prompted us to investigate in greater detail the possible
developmental involvement of the maternal esr2a mRNA. We have
found that Esr2a knockdown by morpholino (MO) treatment
causes marked alterations in the contents of a great number of
transcripts in early embryos and results in lethal malformations
in hatched larvae.
2. Materials and methods
2.1. Fish maintenance
Zebrash of wild-type (WT) strain, purchased from local pet
shops, and the Tg(i1a:EGFP)y1 line were maintained as described
in Westereld [40] and Lawrence [19]. Ovulated oocytes were ob-
tained by stripping, while embryos were yielded by natural spawn-
ing and cultured in zebrash Fish Water at 28.5 C with a
photoperiod of 14 h light/10 h dark. The developmental stages
121
were determined according to the hours post-fertilization (hpf)
and morphological features, as described by Kimmel et al. [17].
All larvae were fed from the beginning with the inert diet Novo
Tom (JBL, Neuhofen, Germany) thrice a day.
2.2. Morpholinos
One antisense MO oligonucleotide against the ATG translation
initiation site of esr2a mRNA, MO2-esr2a, was designed according
to the following sequence: 50 -TGT CTC GTT CCG CAT AGT CCG
ACA T-30 . A non-specic oligonucleotide, MO-control, was used as
a negative control: 50 -CCT CTT ACC TCA GTT ACA ATT TAT A-30 .
An antisense splice variant MO, MO3-esr2a, was designed on the
sequence at the exon 3-intron 3-junction: 50 -GTA CTT TCA GAG
AGT CTT ACC TTG T-30 (the exon complementary sequence is
underlined). Finally, we used an MO against the ATG translation
initiation site of p53 mRNA, MO-p53: 50 -GCG CCA TTG CTT TGC
AAG AAT TG-30 (MO4-tp53) [32]. All the MOs were synthesized
by Gene Tools (Philomath, OR, USA) and were reconstituted at
1 mM in nuclease-free water. MO working solutions were prepared
diluting stock solutions in 1 Danieaus buffer (58 mM NaCl,
0.7 mM KCl, 0.4 mM MgSO4, 0.6 mM Ca(NO3)2, 5 mM HEPES, pH
7.6) supplemented with phenol red. Before injection, MOs were
denatured at 65 C to avoid aggregate formation.
Different MO concentrations were tested in the range of 5.2
20.3 ng/embryo. Injections were performed under a dissecting
microscope using a microinjector attached to a micromanipulator
(Leica Microsystems, Milan, Italy) and controlled by a pedal to in-
ject the desired amount of reagent. MOs and mRNAs (see below)
were injected into the yolk mass of 1-cell embryos. These were
then incubated in 1 Fish Water (50: 25 g Instant Ocean,
39.25 g CaSO4, and 5 g NaHCO3 for 1 l) at 28.5 C. MO- and/or
mRNA-injected embryos were raised to the desired stages for
observations or collected for further analysis.
The effectiveness of MO3-esr2a for exon skipping was con-
rmed by the presence of a signicant amount of a shorter RNA
product detected by RT-PCR with two specic primers covering
the splice site (z-esr2a-F-BamHI and esr2a-R, see Table 1).
2.3. In vitro transcription
The full-length open reading frame (ORF) of zebrash esr2a
(accession number: NM_180966) was amplied from 2-dpf zebra-
sh embryo total cDNA with z-esr2a-F-BamHI and z-esr2a-R-XbaI
oligonucleotides (see Table 1) and cloned into pGEM-T Easy vector
(Promega, Milan, Italy). Both strands were sequenced.
For rescue experiments, since the injected synthetic mRNA
should not contain the MO target sequence, a modied esr2a cDNA
was amplied with a forward primer containing eight silent muta-
tions (z-esr2a-mis-F) and the same reverse primer as above, using
the pGEM-T-esr2a as the template. The resulting PCR product was
cloned into the pGEM-T vector and sub-cloned into the pCS2+
expression vector. The expression construct, after complete
sequencing to check for nucleotide changes that could alter the
amino acidic sequence or interrupt the translation reading frame,
was linearized with XbaI digestion, electrophoretically separated,
and puried by gel extraction with Zymoclean Gel DNA Recovery
Kit (Societ Italiana Chimici, Roma, Italy). The linearized plasmid
(1 lg) was used as a template for the synthesis of capped synthetic
mRNA with the SP6 MEGAscript transcription kit (Ambion, Milan,
Italy). The synthetic mRNA was precipitated by centrifugation in
LiCl and its integrity conrmed by formaldehydeagarose electro-
phoresis. Stock of synthetic mRNA solutions were quantied by
absorbance and diluted to desired concentrations in 1 Danieaus
buffer for microinjection.
122 A. Celeghin et al. / General and Comparative Endocrinology 172 (2011) 120129

Table 1
Primers used in PCR analyses.

Primer Orientation Sequence Position Accession number

0 0
z-esr2a-F-BamHI Sense 5 -CGGGATCCCGGAACTCATCCGCCTTCACC-3 +276 ? +294 AF516874
z-esr2a-R-XbaI Antisense 50 -GCTCTAGAGCGTGTTTAGGGTCCGTGCTGT-30 +1960 ? +1941
z-esr2a-mis-F Sense 50 -ACCATGAGCGAATACCCAGAGGGTGATAGTCC-30 +292 ? +323
esr2a-F Sense 50 -CAGACCTCTGTCTCAGCA-30 +1421 ? +1438
esr2a-F1 Sense 50 -CAACAGAGTCGACTTCAACAG-30 +607 ? +627
esr2a-R Antisense 50 -CAGCAGACACAGCAGCTT-30 +1641 ? +1624
ef1a-F Sense 50 -GACAAGAGAACCATCGAG-30 +177 ? +194 NM_131263
ef1a-R Antisense 5-CCTCAAACTCACCGACAC-3 +447 ? +430
ccng1-F Sense 50 -GATTGAGGATCAGCACGAG-30 +804 ? +822 NM_199481
ccng1-R Antisense 50 -CAGTTATGGGCACTCAACAC-30 +1099 ? +1080
anx2a-F Sense 50 -GCACAGATGTGAAGTGCTG-30 +646 ? +664 NM_181761
anx2a-R Antisense 50 -CAGTCGTCTCCATTGCAC-30 +1059 ? +1042
hmox1-F Sense 50 -CCACACACCGATATGCAC -30 +582 ? +599 NM_00127516
hmox1-R Antisense 50 -CAACGTGATGCCCACTCC-30 +1030 ? +1013

The restriction sites of BamHI and XbaI are in bold italic letters. The mutated bases on z-esr2a-mis-F primer are double underlined.

The pGEM plasmid containing the full-length z-esr2a-ORF was
used to synthesize a digoxigenin (DIG)-labelled riboprobe for
whole-mount in situ hybridization (WMISH). Another riboprobe
was made from a plasmid containing the last 536 coding nucleo-
tides plus the 30 -UTR region of z-esr2a. In this case, the insert
was amplied with the primer z-esr2a-for by 30 -RACE analysis
using the kit FirstChoice RLM-RACE kit (Ambion), following the
manufacturers instructions. The PCR fragment was then cloned
into pGEM-T Easy vector. Templates for the synthesis of these ribo-
probes were linearized by restriction enzyme digestion (ApaI and
BamH1, respectively), electrophoretically separated, puried by
gel extraction, and resuspended in sterile water. DIG-labeled ribo-
probes were synthesized by in vitro transcription with SP6 and T7
RNA polymerases, respectively (Roche, Milan, Italy), following the
manufacturers instructions.
2.4. Whole-mount in situ hybridization and staining of cartilages and
blood vessels
WMISH was performed, on normal and treated embryos, essen-
tially as reported by Thisse and Thisse [36]. Embryos were xed
overnight in 4% paraformaldehyde (PFA) in phosphate-buffered sal-
ine (PBS) at the desired stage of development. For embryos older
than 24 hpf, pigmentation was suppressed by raising embryos in
0.03% PTU (1-phenyl-2-thiourea; Sigma, Italy) in Fish Water. After-
wards, the embryos were washed in PBT (PBS plus 0.1% Tween 20;
Sigma) and dechorionated with forceps. They were treated with
methanol and stored at 20 C. Methanol-stored embryos were
rehydrated in methanol/PBS series, permeabilized by proteinase K
(10 lg/ml), prehybridized, and then hybridized overnight at 65 C
with 1 ng/ll of the appropriate riboprobe in the Hybridization
Mix (HM: 50% formamide, 5  SSC, 0.1% Tween 20, 50 lg/ml hepa-
rin and 500 lg/ml tRNA). After HM/SSC and SSC/PBT washing series,
embryos were preincubated in blocking solution (2% sheep serum
and 2 mg/ml BSA in PBT) and then incubated overnight at 4 C with
alkaline phosphatase-conjugated anti-DIG antibodies (Roche) di-
luted 1:3000 in blocking solution. After PBT washing, embryos were
pre-soaked in staining buffer and then incubated in NBT/BCIP (5-ni-
tro-blue tetrazolium chloride/bromo-4-chloro-30 -indolyphosphate
p-toluidine salt) for blue staining. WMISH-stained embryos were
mounted in 80% glycerol in PBT or cleared and mounted in 2:1 ben-
zyl benzoate/benzyl alcohol, observed under a Leica DMR micro-
scope and photographed with a Leica DC500 digital camera.
For cartilage staining, larvae were xed overnight in 4% PBS-
buffered PFA, rinsed in distilled water and stained overnight in a
0.1% Alcian blue solution. Larvae were then cleared by washing
sequentially in 3% hydrogen peroxide and 1% KOH, and whole
mounted.
Blood vessels were visualized by endogenous alkaline phospha-
tase activity at 3 dpf following the protocol of Serbedzija and co-
workers [33]. We further characterized morphant vasculature by
utilizing the Tg(i1a:EGFP)y1 zebrash transgenic line that labels
vascular endothelial cells with GFP [20].
For histology analysis zebrash larvae at 6 and 8 dpf were xed
in 4% paraformaldehyde for 2 h and processed into an alcoholxy-
lene series followed by parafn embedding. For the histological
examination, 6-lm-thick sections were stained with hematoxy-
lineosin (HE).
2.5. Apoptotic assays
Apoptotic cells in the embryos were detected by the TUNEL as-
say. The TUNEL assay (TdT-mediated uorescein-dUTP nick end
labeling) was performed using alkali stable digoxigenin-dUTP
and TdT (terminal deoxynucleotidyl transferase) (Roche). Embryos
were xed in 4% PFA (overnight, 4 C), treated with methanol and
stored at 20 C. Methanol-stored embryos were rehydrated in
methanol/PBS series and permeabilized by proteinase K (10 lg/
ml). Afterwards, the embryos were washed in PBT (5  5 min, RT)
and in ethanol/acetic acid (2:1) (20 min RT). After incubation with
TUNEL buffer (30 min, RT), embryos were incubated in 100 ll TUN-
EL reaction mixture (overnight, RT). The reaction was stopped by
washing samples with PBT/EDTA 1 mM (2  1 h, RT).
Samples were then processed as for the whole mount in situ
hybridization.
2.6. Microarray hybridization
Agilent Whole Zebrash Genome Oligo Two-Color Microarrays
(Agilent Technologies, Santa Clara, CA, USA) were performed to
analyze the gene expression of embryos treated with MO2-esr2a
after 8 and 48 hpf against MO-control embryos. RNA was isolated
using TRIzol reagent (Invitrogen) and its quality assessed by micro-
chip analysis on an Agilent 2100 Bioanalyzer. Microarray hybrid-
izations were performed according to the Agilent 60-mer oligo
microarray processing protocol using the Agilent Gene Expression
Hybridization Kit on Agilent Whole Zebrash Microarray 4  44 K
arrays. Briey, an aliquot of the corresponding Cy3- and Cy5-la-
beled fragmented cRNAs were combined and hybridized overnight
(17 h, 65 C) using Agilents recommended hybridization chamber
and oven. Finally, the microarrays were washed once with 6 SSPE
buffer (3.0 M NaCl, 0.2 M NaH2PO4, 0.02 M EDTA, pH 7.4) contain-
ing 0.005% N-lauroylsarcosine for 1 min at RT, followed by a second
wash with pre-heated (37 C) 0.06 SSPE buffer containing 0.005%
N-lauroylsarcosine for 1 min. The last washing step was performed
with acetonitrile for 30 s. Hybridization, scanning and data
 A. Celeghin et al. / General and Comparative Endocrinology 172 (2011) 120129
analyzing protocols were performed according to the instructions
in the Agilents manuals. After scanning the microarray images,
the quantied signal and background intensities for each feature
were substantially normalized by the rank consistency ltering
Lowess method. The data have been deposited in NCBIs Gene
Expression Omnibus [6] and are accessible through GEO Series
accession number GSE24934.
2.7. RNA extraction, reverse transcription and quantitative polymerase
chain reaction (qPCR)
Total RNA was extracted from pools of 2060 embryos at the
desired stages of development using TRIzol reagent. The number
of embryos was chosen according to the embryo size at each time
point.
One microgram total RNA obtained from three different pools of
embryos at each developmental stage was reverse transcribed to
rst-strand cDNA using ThermoScript RT-PCR system
(Invitrogen).
qPCR was performed using the 7500 Real-Time PCR System (Ap-
plied Biosystems) and the DyNAmo HS Syber Green qPCR Kit (Finn-
zymes, Euroclone, Milan, Italy) according to the manufacturers
protocol. The cycling parameters were 95 C for 10 min, followed
by 45 cycles at 95 C for 30 s and 57 C for 60 s. Threshold cycles
(Ct) and dissociation curves were generated automatically by Ap-
plied Biosystems software. Sample Ct values were normalized with
Ct values from zebrash elongation factor-1a (ef1a), which was
invariant in WT and morphant embryos at the same developmental
stage. All analyses were performed in triplicate. The Relative
Expression Software Tool 2009 (REST 2009) [28] was used to esti-
mate relative fold changes in the genes of interest, using the ratio
of the Ct values and the PCR amplication efciencies of both the
genes of interest and the housekeeping gene. REST 2009 uses ran-
domization and bootstrapping methods to test the statistical signif-
icance of the gene expression ratios and calculate 95% condence
intervals for relative fold changes [27]. Primer sequences are
reported in Table 1.
3. Results
3.1. Developmental expression of zebrash esr2a mRNA
To determine the spatio-temporal expression of maternal and
zygotic esr2a mRNAs during zebrash embryogenesis, WMISH as-
Fig. 1. Spatio-temporal expression of esr2a mRNA during zebrash development as evidenced by whole-mount in situ hybridization performed at the indicated stages. All
embryos are lateral views with the animal pole up (6 and 10 hpf) and head up pointing to the left (24 and 48 hpf).
123
says were performed on ovulated oocytes, WT and morphant em-
bryos from 0.2 to 72 hpf, using two DIG-labeled antisense esr2a
mRNA probes, one spanning the entire coding region and the other
covering the C-terminus and the 30 -UTR of the transcript. Both
probes revealed identical expression patterns. In oocytes, maternal
esr2a mRNA was spread as granular and lamentous structures
throughout the central ooplasm, as shown in Fig. 1. At the 1-cell
stage (0.2 hpf), almost all transcript was already relocated in the
blastodisc area with few residual streaming lines beneath. At the
2- and 8-cell stages, the transcript was equally distributed between
the blastomeres. At 6 hpf (shield stage), barely detectable esr2a
mRNA expression was observed in the blastodisc. At 10 hpf, the
signal was more concentrated in the head region, as it was at 24
hpf particularly in the eye, otic vesicle and hindbrain. At 48 hpf,
a marked signal was observed just anteriorly and ventrally to the
yolk sac. Expression at 72 hpf was barely detectable (not shown).
No signicant difference was evidenced between WT and mor-
phant samples. The results of spatio-temporal esr2a mRNAs
expression are in agreement with those obtained by qRT-PCR [29].
3.2. Esr2a MO injections
To assess the developmental roles of maternal vs. zygotic Esr2a
protein during zebrash development, we selectively blocked the
translation of esr2a mRNA by injecting MO2-esr2a, antisense to
the 50 -UTR and ATG site, into 1-cell embryos. Moreover, MO3-
esr2a, an antisense splice variant MO designed on the sequence
at the exon 3-intron 3 junction, was injected to impair the appro-
priate splicing of zygotic transcripts. MO-control, a non-specic
oligonucleotide, was used as negative control.
In order to optimise MO injections, 5.2, 10.3 and 20.3 ng of
MO2-esr2a and MO3-esr2a, as well as MO-control, were injected
per embryo in triplicates. As expected, the highest concentration
(20.3 ng) caused slightly higher lethality and more malformed lar-
vae in the MO-control. Conversely, the lowest dosage (5.2 ng) was
scarcely effective in the MO2-esr2a group (Table 2). Accordingly,
10.3 ng per embryo of MOs seemed the best concentration to
examine the effects of Esr2a protein deciency. As shown in Table 2,
almost two thirds of the embryos treated with MO2-esr2a were af-
fected by graded abnormalities, while the effects of MO3-esr2a at
the same dosage were minimal and equal to those of MO-control.
To verify the anti-splicing effectiveness of MO3-esr2a, total
RNAs extracted from 4, 6 and 8 hpf WT and MO3-esr2a-injected
124 A. Celeghin et al. / General and Comparative Endocrinology 172 (2011) 120129

Table 2
Effects of different dosages of morpholinos (MO2-esr2a, MO3- esr2a, MO-control) on the percentages of dead embryos and of normal and abnormal phenotypes calculated from
the number of surviving larvae at 5 days. Data were pooled from 3 experiments.

n. Dead (%) (means SD) No. surviving Normal (%) (means SD) Abnormal (%) (means SD)
MO-control (ng per embryo)
5.2 251 12 3 221 96 2 42
10.3 291 12 1 256 89 1 11 1
20.3 258 16 5 221 75 2 25 2
MO2-esr2a (ng per embryo)
5.2 268 14 3 231 84 1 16 1
10.3 246 16 5 207 37 4 63 4
20.3 346 23 3 266 16 6 84 6
MO3-esr2a (ng per embryo)
5.2 279 10 5 251 93 3 73
10.3 270 10 5 243 90 4 10 4
20.3 384 15 5 322 87 7 13 7

consistent with the loss of the targeted exon 3, as conrmed by

sequencing. Furthermore this analysis, that is only qualitative as
the PCR was performed at 40 cycles, demonstrated that zygotic
esr2a expression takes place already at 6 hpf, but at very low level,
so that most of the transcript is still of maternal origin. Thus, the
up- and down-regulated genes revealed by the microarray analysis
at 8 hpf (see below) are mainly due to the knockdown of the
maternal Esr2a protein.

3.3. Analysis of morphant phenotypes

During early development, MO2-esr2a-injected embryos could

Fig. 2. Panel A: MO2-esr2a causes deletion of exons 3 (arrow) of the esr2a gene
with introduction of a premature stop codon. The arrow-head indicates the residual
wild-type transcript detected at 6 and 8 hpf. Panel B: sequence of the misspliced
esr2a transcript showing the loss of exon 3 and the introduction of a premature stop
codon.
embryos were analysed by RT-PCR with two esr2a-specic primers
covering the splice site. As shown in Fig. 2, MO3-esr2a almost com-
pletely abolished the properly processed esr2a mRNA, as indicated
by the prevalence of the predicted misspliced transcript detected
as a lower band together with a minimal fraction of normally
spliced esr2a transcript at 8 hpf. The size of this product was
not be morphologically distinguished from the controls (non-in-
jected and MO-control embryos) or MO3-esr2a embryos. Later
on, MO2-esr2a embryos often exhibited a slight growth delay that
was more evident at 1 dpf. Morphant larvae at 3 and 5 dpf were af-
fected by body growth delay and curved shape, abnormal brain and
splanchnocranium development, pericardial oedema, persistent
voluminous yolk sac and reduced or oedematous yolk extension,
uninated swim bladder and rudimentary caudal n with aberrant
circular swimming after mechanical stimulus. Brain, eyes and otic
vesicles were reduced (Fig. 3). Affected larvae could survive for 12
14 days.
The effects of the MO2-esr2a are comparable to those obtained
by Froehlicher and coworkers [10] with a different MO-esr2a

Fig. 3. Phenotypes of embryos or larvae at 1, 3 and 5 dpf after treatment with MO2-esr2a, MO3-esr2a, MO2-esr2a together with MO-p53 as compared to control groups (WT
or treated with MO-control) or after rescuing with mutated z-esr2a mRNA. Animals are presented as lateral view, anterior to the left. Bar = 500 lm.
 A. Celeghin et al. / General and Comparative Endocrinology 172 (2011) 120129 125

Table 3
Results of MO2-nr3c1 rescue by mutated zebrash esr2a mRNA.

MO2-esr2a (10.3 ng) + various concentrations n. Dead (%) (means SD) No. surviving Normal (%) (mean SD) Abnormal (%) (mean SD)
of mutated z-esr2a mRNA
+15 pg mRNA per embryo 261 35 7 171 40 7 60 7
+30 pg mRNA per embryo 288 23 5 222 65 8 35 8
+60 pg mRNA per embryo 320 20 2 258 51 3 49 3
+90 pg mRNA per embryo 313 20 5 238 56 2 44 2
+120 pg mRNA per embryo 205 45 2 112 41 1 59 1

Data were pooled from 3 experiments. MO2-esr2a dose was 10.3 ng per embryo. The percentages of normal and abnormal embryos were calculated from surviving embryos
at 5 dpf.

Fig. 4. Panel A: TUNEL analysis to detect apoptotic nuclei in uninjected control embryos (WT) and MO2-esr2a embryos at 24 hpf. Panel B: histology of zebrash WT and MO2-
esr2a-injected embryos at 6 and 8 dpf (transverse sections). Note hypertrophy of ventral epithelium in morphant embryos. Panel C: alkaline phosphatase staining showing
well-organized sub-intestinal veins (SIV, indicated by arrow) in WT larvae at 3 dpf, while reduced or absent SIV are evident in MO2-esr2a-injected larvae. The same results
were obtained microinjecting MO2-esr2a in Tg(i1a:EGFP)y1 zebrash transgenic line (Panel D).
126 A. Celeghin et al. / General and Comparative Endocrinology 172 (2011) 120129
against the ATG translation initiation site, although hatching rates
and larval survival were much higher in our experiment. This pro-
vides further evidence that the observed developmental defects
were specic for Esr2a knockdown.
3.4. Rescuing of MO2-esr2a knockdown with synthetic mutated esr2a
mRNA
To rescue the knockdown phenotype, the MO2-esr2a was co-in-
jected with in vitro-transcribed mutated zebrash esr2a mRNA
encoding the full length of Esr2a, but containing eight silent muta-
tions in the region recognised by MO2-esr2a.
Embryos injected with the MO2-esr2a (10.3 ng/embryo) and
mutated zebrash esr2a mRNA at different concentrations (15,
30, 60, 90, 120 pg/embryo) showed varying degrees of rescue of
the MO2-esr2a phenotype. As shown in Table 3, with the lowest
dose of 15 pg/embryo, there was 35% mortality, while 60% of the
surviving larvae presented the MO2-esr2a phenotype. The injec-
tion of 30 pg/embryo gave the best rescue, with 23% mortality
and 65% of surviving larvae at 5 dpf developing a normal pheno-
type. The swim bladder was inated and no persistent yolk sac
was found (see Fig. 3). Raising the dosage to 60 and 90 pg/embryo
increased abnormality, but not mortality, whereas with 120 pg/
embryo both mortality and abnormality worsened.
To further prove the specicity of the observed MO2-esr2a
knockdown phenotype, we performed an additional experiments
in which we co-injected MO2-esr2a and MO-p53. The rationale for
these studies is based on a report that MOs can nonspecically acti-
vate the tumor suppressor p53-induced apoptosis, causing off-tar-
get phenotypic effects that are not caused by the specic MO used
[32]. Actually, the p53 transcript was shown to be up-regulated in
the microarray analysis of MO2-esr2a embryos at 8 and 48 hpf
(see below). However, the co-injection of MO-p53 failed to eliminate
the MO2-esr2a phenotype (Fig. 3), thus conrming that it was spe-
cically caused by the translation knockdown of esr2a mRNA.
3.5. Apoptosis analysis
To examine whether Esr2a protein deciency causes apoptosis,
whole-mount TUNEL-staining was used to detect apoptotic cells in
control and morphant embryos analyzed at 10, 24 and 48 hpf.
Staining of WT and MO-control embryos revealed minimal evi-
dence of apoptosis (thus excluding off-target effects), whereas
the number of TUNEL-positive cells in MO2-er2a-embryos was
clearly increased, in particular in the brain region, suggesting that
the developmental defects of MO2-esr2a morphants may be par-
tially caused by an increase of apoptosis, possibly mediated by
an increase in p53 protein (Fig. 4, panel A).
3.6. Sub-intestinal vessels staining
At 3 dpf, the network of sub-intestinal veins (SIV) at the yolk
stalk, as evidenced by endogenous alkaline phosphatase activity,
was reduced and mainly composed of a single series of loops in-
stead of 23 series as in WT. The pattern of intersegmental vessels
in the trunk and tail was also less dened (Fig. 4, panel C). These
results were conrmed by using a transgenic zebrash line
expressing gfp under the guidance of the i1 promoter ensuring
endothelial-specic expression of the reporter and its continuous
monitoring during embryonic vascular development (Fig. 4, panel
D).
3.7. Cartilage staining
Alcian blue staining revealed that the pattern of head cartilage
formation was altered in MO2-esr2a larvae with respect to WT
controls at 6 dpf. The ceratohyals (ch) are not oriented obliquely,
but are almost perpendicular to the basibranchial (bb). The palato-
quadrate cartilages (pq) are bent, giving rise together with the
Meckels cartilages (m) to a more stubby snout. Moreover, the cer-
atobranchial cartilages seem shorter (Fig. 5). The same spectrum of
craniofacial abnormalities was not produced using MO3-esr2a.
3.8. Microarray analysis
To identify specic transcripts affected by Esr2a knockdown,
transcriptome analysis was performed using Agilent zebrash
microarrays. Only the transcripts that were at least 2-fold up- or
down-regulated in MO2-esr2a morphants as compared to MO-con-
trol embryos and with green or red processed signals higher than
1000 were selected. After this rst selection, we found 240 up-
and 219 down-regulated transcripts at 8 hpf and 162 up- and
120 down-regulated transcripts at 48 hpf. We found only 18 com-
mon up-regulated and down-regulated transcripts between 8 and
48 hpf. This nding is in agreement with the report by Froehlicher
and coworkers [10] on 72 hpf-larvae after MO-esr2a treatment.
From the raw data of their Affymetrix microarrays submitted to
GEO website, a comparison was made with the probes identied
as in common with those in our Agilent microarrays. At 8 hpf,
out of 40 transcripts, 21 were discordant, while 5 up- and 14
down-regulated transcripts were concordant with those reported
by them at 72 hpf. At 48 hpf, out of 54 transcripts, 3 were discor-
dant, while 40 up- and 11-down-regulated transcripts were con-
cordant with those at 72 hpf. This provides further evidence that
the observed transcript changes were specic for Esr2a knock-
down. Moreover, a distinct functional role of maternal esr2a mRNA
is suggested by the fact that the set of transcripts altered by MO2-
esr2a at 8 hpf, when zygotic transcription of esr2a is minimal [29],
diverges from those at both 48 and 72 hpf, when only zygotic esr2a
mRNA is present.
To assign the affected transcripts into functional classes based
on gene ontology, public databases were consulted using the NCBI
website (http://www.ncbi.nlm.nih.gov/sites/entrez). Transcripts
were classied into 17 groups, as shown in Table 4. Of the 741
Fig. 5. Alcian blue staining of zebrash WT and MO2-esr2a-injected larvae at 6 dpf
showing alterations of the cartilaginous structures in morphant larvae. bb,
basibranchial; cb1-5, ceratobranchial 1-5; ch, ceratohyal; ep, ethmoid plate; hs:
hyosympletic; ih, interhyal; m, Meckels cartilage; not, notochord; pch, parachor-
dal; pq, palatoquadrate; tr, trabecula.
 A. Celeghin et al. / General and Comparative Endocrinology 172 (2011) 120129 127

Table 4
Differentially expressed genes at 8 and 48 hpf in MO2-esr2a injected embryos, divided in functional classes.

8 hpf 48 hpf In common

Gene ontology Up Down Up Down Up Down
Embryo development 8 11 9 10 1
Metabolic process 30 30 27 24 2 2
Signaling 3 22 4 7 3
Transcription 12 7 19 8 3 2
RNA processing 1 1 1
Translation 9 2 11 2 3
Biological process 16 24 12 6 2 1
Chromatin organization 2 2 4
Cell structure and cytoskeletal organization 11 6 2 3
Intra- and intercellular transport 10 9 5 5 2
Negative regulation of cell proliferation 5 1 3 2 1
Positive regulation of cell proliferation 1 1 1 1
Cellular adhesion 2 2 1 2
Cell cycle and meiosis 1 3 4 2
Immune response 3 2
DNA repair 4
Unknown 122 92 60 48 6 7
Total 240 219 162 120 18 18

lated at 48 hpf by microarray analysis, and at both 8 and 48 hpf by

qRT-PCR, with a greater increase at the latter time (Fig. 6).
The third transcript was that of annexin a2a (anxa2a), a gene ex-
pressed during gastrulation in the notochord and later in the peri-
derm [7], but of unknown function. In the microarray analysis, this
messenger was up-regulated at both 8 and 48 hpf, whereas qRT-
PCR showed that the increase was signicant only at 48 hpf (Fig. 6).

4. Discussion

The present report demonstrates that the maternal esr2a mRNA

Fig. 6. Fold changes in gene expression in MO2-esr2a-injected embryos compared
to WT (set at 1). Values represent the mean S.D. (n = 3). indicates that the
difference in the expression levels are signicantly different (P < 0.05); (P < 0.01);
controls in the MO experimental design. Aspecic cytotoxic effects
(P < 0.001).
transcripts from 705 genes differentially expressed, 45.7% have un-
known function, 15.7% are involved in metabolic process, 8.2% in
biological process, 5.4% in embryo development, 5.1% in intra-
and intercellular signaling, 6.5% in transcription, and 3.4% in trans-
lation. Transcripts encoding estrogen receptors were not affected.
3.9. Validating gene expression data
To independently conrm the microarray analysis, three regu-
lated transcripts were measured by relative qRT-PCR at 8 and 48
hpf. The relative quantication measures the fold-difference in
the morphant amount of a target transcript compared with a con-
trol (WT), whose expression level is set at 1.
In particular, we analyzed the content of heme oxygenase (decy-
cling) 1 (hmox1) mRNA, which encodes a protein that, in mammals,
catabolizes heme to biliverdin, carbon monoxide, and free iron and
is involved in iron homeostasis [30]. In zebrash, this gene is ex-
pressed in the extraembryonic yolk syncytial layer, lens, and a
small population of blood cells, as shown by Thisse et al. [35]. A
statistically signicant down-regulation of this transcript was con-
rmed by qRT-PCR at both developmental times (Fig. 6).
The transcript of cyclin G1 (ccng1), that encodes a protein with
intrinsic growth inhibitory activity [41], was found to be up-regu-
present in zebrash eggs is required for normal embryogenesis and
larval development, because the MO-induced deciency of this
estrogen receptor is characterized by the early (at 8 hpf) and late
(48 hpf) up- or down-regulated transcript contents of hundreds
of genes, resulting in severe malformations and compromised lar-
val viability. This conclusion is supported by the inclusion of four
due to MO injection per se are excluded on the basis of the very low
number of dead embryos and abnormal 5-dpf larvae after treat-
ment with MO-control at the same dosage. The possibility of off-
target apoptosis induced by the activation of the tumour suppres-
sor p53 protein is refuted by the failure of MO-p53 to improve
embryonic and larval phenotypes in MO2-esr2a-co-injected sh.
Specicity of the detrimental effects of Esr2a knockdown is en-
dorsed by rescuing to normality sh co-injected with an adequate
amount of mutated zebrash esr2a mRNA. Finally, the observed
developmental alterations by MO2-esr2a can be ascribed to the
suppressed translational processing of maternal rather than zygo-
tic esr2a mRNA, because of the normal phenotype of sh treated
with MO3-esr2a despite its high effectiveness in missplicing the
zygotic esr2a transcript.
This conclusion is also validated by the microarray analysis
showing that a greater number of transcripts were up- or down-
regulated at 8 hpf than at 48 hpf and that the two sets were essen-
tially unrelated, given the small number of transcripts in common.
There are several reasons why the transcript changes at 8 hpf
should derive from the knockdown of the maternal receptor. First,
the amount of maternal esr2a mRNA at the start of cleavage is
some orders of magnitude greater than those from 8 to 48 hpf
[18,24,29]. It is likely available for translation blockage during
the rst cleavage cycles, because it is rapidly internalized into
the rst blastomeres, where it can be targeted by a still scarcely
diluted MO2-esr2a. Second, the zygotic esr2a mRNA is barely
128 A. Celeghin et al. / General and Comparative Endocrinology 172 (2011) 120129
detectable at 6 hpf, as reported by Lassiter et al. Lassiter et al. [18]
and shown by the small amount of misspliced mRNA in MO3-esr2a
embryos, indicating that the normally spliced band was still likely
due to the residual maternal mRNA. Third, more maternal estrogen
would be available for Esr2a transactivation at the beginning of
embryonic development than at 8 hpf, when zygotic transcription
of esr2a mRNA is going on at a low level, as discussed below.
While the recognition of the relevance of the translational
knockdown of the maternal receptor to account for transcript
changes at 8 hpf is straightforward, more problematic is the inter-
pretation of these changes. Specically, it remains to be established
whether they arose from altered degradation rates of maternal
transcripts or modied transcription of zygotic genes. It is known
that, in the zebrash oocyte, about half of the entire genome is ac-
tively transcribed during the primary growth phase (stage IB) [26]
and that hundreds of these transcripts are preserved from degrada-
tion by translational repression that is removed at ovulation [8].
While a fraction of them is subsequently translated and degraded
before the end of the blastula stage (around 5 hpf), several persist
throughout the gastrulation period (till 10 hpf) [22]. On the other
hand, zygotic transcription starts at the mid-blastula transition
(3 hpf) with the progressive lengthening of the interphase time
in cellular cycles and becomes prominent from 5 hpf onwards
[16,34]. Hence, during gastrulation, both maternal and zygotic
transcripts, even derived from the same gene, may coexist in inver-
sely related proportions during time. At 8 hpf, it may be assumed
that most of the up- or down-regulated transcripts were of zygotic
origin and that the changes induced by Esr2a deciency were med-
iated through transcription-dependent pathways, though indepen-
dent ones for degradation cannot be excluded.
At 48 hpf, it is really remarkable that the altered zygotic tran-
scripts were mostly derived from genes different from those at 8
hpf, while roughly belonging to the same functional classes of gene
ontology. It is known that different transcript sets peak at gastru-
lation, segmentation and pharyngula stages in zebrash [22], fol-
lowed by hatching at around 48 hpf. Thus, Esr2a, either
persisting from high-level maternal transcripts or newly translated
from lower-level zygotic transcripts, might be required also for the
transcriptional control of these later gene sets. This is evident from
the strong up-regulation of the ccng1 transcript measured by qRT-
PCR and microarray at 48 hpf, but not at 8 hpf. Alternatively, the
altered transcripts at 48 hpf might be vertically linked to those
at 8 hpf through genetic cascades in a network of intergenic inte-
gration. In any case, what is certain is that the effects of the MO-in-
duced deciency of Esr2a cannot be compensated by the increasing
amounts of zygotic esr1 and esr2b mRNAs.
Our data not only indicate that Esr2a is required for normal
development, but also that its quantity must be adjusted within
a restricted range. At least this is suggested by the response ob-
tained in the experiment with different doses of rescue mRNA.
The fact that diminished mortality and abnormality were observed
by increasing the dosage from 15 up to 30 pg/embryo can be ex-
plained by the need to provide an adequate amount for rescuing.
However, abnormality increased with 30 and 60 pg/embryo and
both mortality and abnormality were further worsened by elevat-
ing the dosage to 120 pg/embryo, an outcome that suggests that
such quantity was functionally excessive rather than generally
toxic. There is evidence that the degradation of injected mRNA
may not parallel that of the corresponding maternal transcript
due to differences between articial A-tails and natural repolyade-
nylation at 30 end [9]. Moreover, messenger degradation rates may
differ in embryonic cells and overloading even with an isospecic
mRNA may unbalance this ne tuning. This would be hardly sur-
prising considering all the molecular mechanisms the oocyte rst
and then the embryo must rely upon to timely inactivate, reacti-
vate and nally degrade the maternal esr2a transcript.
The fact that 745 transcripts from 720 genes were affected by
Esr2a loss-of-function is evidently connected to the enhanced rates
of embryonic and larval apoptosis and anatomical malformations,
but the reconstruction of the chains of genetic events leading to
the observed abnormalities is beyond the scope of the present
work. Three defects reported here, namely lower hatching rate,
curved tail and circular swimming, were also described after esr2a
MO treatment in a study by Froehlicher et al. Froehlicher et al. [10],
which was centered on the disrupted development of the neuro-
masts along the lateral line. Such an effect was attributed to the
knockdown of the zygotic Esr2a that is co-expressed, at lower lev-
els, with Esr1 in both the sensory hair cells and supporting cells of
this mechanoreceptive system, starting from 60 to 72 hpf and cul-
minating at 6 dpf [37]. Esr1 was found to regulate primordium
migration along the posterior lateral line [12], while Esr2a pro-
motes the differentiation of supporting cells into hair cells. The lat-
ter were absent in morphants neuromasts, as a consequence of an
aberrant activation of the Notch signaling pathway [11]. Signi-
cantly, her12 and notchless homolog 1 were signicantly up-regu-
lated at 8 and 48 hpf, respectively, and aberrant Notch activation
might have occurred also in our 5-dpf morphant larvae. As to cir-
cular swimming, we prefer to impute it to curved body shape
and reduced tail n than to defective neuromasts, because these
serve to avoid obstacles and predators or to spot prey by detecting
pressure changes in the surrounding water rather than for straight
swimming. The lack of swim bladder ination can be interpreted as
a secondary consequence of the swimming impairment, prevent-
ing gulping of air bubbles at the water surface to force dilation of
the bladder wall.
It is tempting to relate the high in situ hybridization signal of
esr2a mRNA around the yolk sac in 48-hpf embryos to both the ob-
served hypertrophy of the covering ventral epithelium and the
poor development of the sub-intestinal veins in 3-dpf morphant
larvae, with consequent delayed utilization of yolk reserves and
stunting. This may partly account for the decient growth of brain,
sensory organs, such as eyes and otic vesicles, as well the derange-
ment of splanchocranial and branchial cartilages, which are mostly
derived from migrating neural crest cells, but the actual mecha-
nisms involved remain to be claried.
A point to be considered is that the term maternal to connote
the transcriptomic load of unfertilized eggs refers to a burst of
transcriptional activity in early oocytes that occurs during the dip-
lotene stage of meiosis I in loops of lampbrush chromosomes [26].
These are formed by two partially uncoiled bivalents, each consist-
ing of two recombined sister chromatids. Hence, the ploidy status
is both recombined and tetraploid and different from that of
maternal somatic cells as to gene dosages and positions. The
long-lived mRNAs that sustain the initial steps of embryogenesis
represent a transcriptomic inheritance that is neither maternal
nor zygotic and is best described as prezygotic, because it precedes
zygotic transcription.
Truly maternal is instead the so-called maternal hormonal
inheritance given by the complex of lipophilic hormones, such as
steroids, secosteroids, thyroid hormones and retinoids, taken up
from the maternal circulation or follicular envelope and retained
in the yolk mass [21]. The case of 17b-estradiol is a privileged
one, because it is synthesized by granulosa cells, which are just
in contact with the oocyte. This estrogen is abundant during vitel-
logenesis, has a high permeability through cell membranes and, in
labeled form, is rapidly incorporated into zebrash oocytes accord-
ing to a rst-order kinetics (data not shown), similar to what was
reported in rainbow trout [3].
The content and fate of maternal 17b-estradiol in ovulated oo-
cytes and early embryos of zebrash have yet to be investigated,
but this estrogen was reported in eggs of salmon (Oncorhynchus
keta) [5], tilapia (Oreochromis niloticus) [14] and medaka, where it
 A. Celeghin et al. / General and Comparative Endocrinology 172 (2011) 120129
occurs at higher levels than testosterone of thecal source [15]. All
three zebrash Esr are transactivated by 17b-estradiol with high
afnity and greater specicity than their mammalian counterpart,
because catecholestrogen and tamoxifen are ineffective [2]. There-
fore, it is conceivable that Esr2a, translated from maternal esr2a
transcript, can be transactivated by maternal 17b-estradiol and
that the alterations of transcript contents in 8-hpf embryos and
the aberrant morphology of 24-hpf embryos were caused by the
shutdown of the prezygotic estrogen transduction pathway in-
duced by MO2-esr2a treatment. To what extent the zygotic Esr2a
from 24 hpf to hatching is also dependent upon maternal estrogen
or from the aromatization of maternal androgen following tran-
scription of cyp19a1b gene from 25 hpf onwards [24] is still an
open question.
In conclusion, powerful regulatory adjustments linked to
maternal life experience in terms of estrogen response can be di-
rectly transferred into descendants before their actual sensing of
surrounding environmental cues. In this way, the oocyte loading
with estrogen and esr2a transcript may anticipate an epigenetic
plasticity in zebrash development, as a maternal contribution to
the adaptive value of the offspring.
Acknowledgments
The mnx1 plamid was kindly provided by Prof. Mayer Dirk
(Institute for Molecular biology, Leopold Franzens University, Inns-
bruck, Austria). The Tg(i1a:EGFP)y1 line was kindly provided by
Prof. Francesco Argenton and Dr. Natascia Tiso (University of Pado-
va, Italy). We are grateful to Dr. Luigi Pivotti for sh husbandry and
technical assistance. Grant support: Progetti di Ricerca di Ateneo
(CPDA078853), University of Padova, Padova, Italy.
References
[1] B. Auperin, M. Geslin, Plasma cortisol response to stress in juvenile rainbow
trout is inuenced by their life history during early development and by egg
cortisol content, Gen. Comp. Endocrinol. 158 (2008) 234239.
[2] P.L. Bardet, B. Horard, M. Robinson-Rechavi, V. Laudet, J.M. Vanacker,
Characterization of oestrogen receptors in zebrash (Danio rerio), J. Mol.
Endocrinol. 28 (2002) 153163.
[3] P. Belvedere, S. Vianello, L. Dalla Valle, A. Ramina, A. Manzalini, B. Salvato, L.
Colombo, Long-term effects on body growth induced by oocyte exposure to
exogenous cortisol and estradiol-17b in rainbow trout, in: E.W. Roubos, S.E.
Wendelaar Bonga, H. Vaudry, A. De Loof (Eds.), Recent Developments in
Comparative Endocrinology and Neurobiology, Shaker Publishing, Maastricht,
The Netherlands, 1999, pp. 258260.
[4] M. Darnaudry, S. Maccari, Epigenetic programming of the stress response in
male and female rats by prenatal restraint stress, Brain Res. Rev. 57 (2007)
571585.
[5] E.G. de Jesus, T. Hirano, Changes in whole body concentrations of cortisol,
thyroid hormones, and sex steroids during early development of the chum
salmon, Oncorhynchus keta, Gen. Comp. Endocrinol. 85 (1992) 5561.
[6] R. Edgar, M. Domrachev, A.E. Lash, Gene expression omnibus: NCBI gene
expression and hybridization array data repository, Nucleic Acids Res. 30
(2002) 207210.
[7] S.A. Farber, R.A. De Rose, E.S. Olson, M.E. Halpern, The zebrash annexin gene
family, Genome Res. 13 (2003) 10821096.
[8] B.M. Farley, S.P. Ryder, Regulation of maternal mRNAs in early development,
Crit. Rev. Biochem. Mol. Biol. 43 (2008) 135162.
[9] M. Fink, G. Flekna, A. Ludwig, T. Heimbucher, T. Czerny, Improved translation
efciency of injected mRNA during early embryonic development, Dev. Dyn.
235 (2006) 33703378.
[10] M. Froehlicher, A. Liedtke, K. Groh, H. Lpez-Schier, S.C.F. Neuhauss, H. Segner,
R.I.L. Eggen, Estrogen receptor subtype b2 is involved in neuromast
development in zebrash (Danio rerio) larvae, Dev. Biol. 330 (2009) 3243.
[11] M. Froehlicher, A. Liedtke, K.S. Groh, S.C.F. Neuhauss, H. Segner, R.I.L. Eggen,
Zebrash (Danio rerio) neuromast: promising biological endpoint linking
developmental and toxicological studies, Aquat. Toxicol. 95 (2009) 307319.
[12] L. Gamba, N. Cubedo, A. Ghysen, G. Lutfalla, C. Dambly-Chaudire, Estrogen
receptor ESR1 controls cell migration by repressing chemokine receptor
129
CXCR4 in the zebrash posterior lateral line system, Proc. Natl. Acad. Sci. U S A.
107 (2010) 63586363.
[13] A. Hamad, M. Kluk, J. Fox, M. Park, J.E. Turner, The effects of aromatase
inhibitors and selective estrogen receptor modulators on eye development in
the zebrash (Danio rerio), Curr. Eye Res. 32 (2007) 819827.
[14] G.A. Hines, L.R. Boots, T. Wibbels, S.A. Watts, Steroid levels and steroid
metabolism in relation to early gonadal development in the tilapia
Oreochromis niloticus (Teleostei: cyprinoidei), Gen. Comp. Endocrinol. 114
(1999) 235248.
[15] T. Iwamatsu, H. Kobayashi, S. Hamaguchi, R. Sagegami, T. Shuo, Estradiol-17b
content in developing eggs and induced sex reversal of the medaka (Oryzias
latipes), J. Exp. Zool. A 303 (2005) 161167.
[16] D.A. Kane, C.B. Kimmel, The zebrash midblastula transition, Development
119 (1993) 447456.
[17] C.B. Kimmel, W.W. Ballard, S.R. Kimmel, B. Ullmann, T.F. Schilling, Stages of
embryonic development of the zebrash, Dev. Dyn. 203 (1995) 255310.
[18] C.S. Lassiter, B. Kelley, E. Linney, Genomic structure and embryonic expression
of estrogen receptor beta a (ERaa) in zebrash (Danio rerio), Gene 299 (2002)
141151.
[19] C. Lawrence, The husbandry of zebrash (Danio rerio): a review, Aquaculture
269 (2007) 120.
[20] N.D. Lawson, B.M. Weinstein, In Vivo imaging of embryonic vascular
development using transgenic zebrash, Dev. Biol. 248 (2002) 307318.
[21] E. Lubzens, G. Young, J. Bobe, J. Cerd, Oogenesis in teleosts: how sh eggs are
formed, Gen. Comp. Endocrinol. 165 (2010) 367389.
[22] S. Mathavan, S.G.P. Lee, A. Mak, L.D. Miller, K.R.K. Murthy, K.R. Govindarajan, Y.
Tong, Y.L. Wu, S.H. Lam, H. Yang, Y. Ruan, V. Korzh, Z. Gong, E.T. Liu, T. Lufkin,
Transcriptome analysis of zebrash embryogenesis using microarrays, PLoS
Genet. 1 (2005) e29.
[23] A. Menuet, E. Pellegrini, I. Anglade, O. Blaise, V. Laudet, O. Kah, F. Pakdel,
Molecular characterization of three estrogen receptor forms in zebrash:
binding characteristics, transactivation properties, and tissue distribution,
Biol. Reprod. 66 (2002) 18811892.
[24] K. Mouriec, J.J. Lareyre, S.K. Tong, Y. Le Page, C. Vaillant, E. Pellegrini, F. Pakdel,
B.C. Chung, O. Kah, I. Anglade, Early regulation of brain aromatase (cyp19a1b)
by estrogen receptors during zebrash development, Dev. Dyn. 238 (2009)
26412651.
[25] B.P. Nelson, R.P. Henriet, A.W. Holt, K.C. Bopp, A.P. Houser, O.E. Allgood Jr., J.E.
Turner, The role of estrogen in the developmental appearance of sensory-
motor behaviors in the zebrash (Danio rerio): the characterization of the
listless model, Brain Res. 122 (2008) 118128.
[26] F. Pelegri, Maternal factors in zebrash development, Dev. Dyn. 228 (2003)
535554.
[27] M.W. Pfaf, Rest 2009 Software User Guide, Qiagen, Hilden, Germany,
2009.
[28] M.W. Pfaf, G.W. Horgan, L. Dempe, Relative expression software tool (REST)
for group-wise comparison and statistical analysis of relative expression
results in real-time PCR, Nucleic Acids Res. 30 (2002) e36.
[29] S. Pikulkaew, A. De Nadai, P. Belvedere, L. Colombo, L. Dalla Valle, Expression
analysis of steroid hormone receptor mRNAs during zebrash embryogenesis,
Gen. Comp. Endocrinol. 165 (2010) 215220.
[30] K.D. Poss, S. Tonegawa, Heme oxygenase 1 is required for mammalian iron
reutilization, Proc. Natl. Acad. Sci. USA 94 (1997) 1091910924.
[31] E.R. Prossnitz, J.B. Arterburn, L.A. Sklar, GPR30: A G protein-coupled receptor
for estrogen, Mol. Cell. Endocrinol. 265266 (2007) 138142.
[32] M.E. Robu, J.D. Larson, A. Nasevicius, S. Beiraghi, C. Brenner, S.A. Farber,
S.C. Ekker, p53 activation by knockdown technologies, PLoS Genet. 3 (2007)
e78.
[33] G.N. Serbedzija, E. Flynn, C.E. Willett, Zebrash angiogenesis: a new model for
drug screening, Angiogenesis 3 (1999) 353359.
[34] W. Tadros, H.D. Lipshitz, The maternal-to-zygotic transition: a play in two acts,
Development 136 (2009) 30333042.
[35] B. Thisse, C. Thisse, Fast Release Clones: A High Throughput Expression
Analysis. ZFIN Direct Data Submission 2, 2004 (http://zn.org).
[36] C. Thisse, B. Thisse, High-resolution in situ hybridization to whole-mount
zebrash embryos, Nat. Protoc. 3 (2008) 5969.
[37] A. Tingaud-Sequeira, A. Andr, J. Forgue, C. Barthe, P.J. Babin, Expression
patterns of three estrogen receptor genes during zebrash (Danio rerio)
development: evidence for high expression in neuromasts, Gene Expr. Patterns
4 (2004) 561568.
[38] O.I. Vivar, X. Zhao, E.F. Saunier, C. Grifn, O.S. Mayba, M. Tagliaferri, I. Cohen,
T.P. Speed, D.C. Leitman, Estrogen receptor b binds to and regulates three
distinct classes of target genes, J. Biol. Chem. 285 (2010) 2205922066.
[39] J.N. Volff, Genome evolution and biodiversity in teleost sh, Heredity 94
(2005) 280294.
[40] M. Westereld, The Zebrash Book. A Guide for the Laboratory Use of Zebrash
(Danio rerio), third ed., University of Oregon Press, Eugene, OR, 1995.
[41] L. Zhao, T. Samuels, S. Winckler, C. Korgaonkar, V. Tompkins, M.C. Horne, D.E.
Quelle, Cyclin G1 has growth inhibitory activity linked to the ARF-Mdm2-p53
and pRb tumor suppressor pathways, Mol. Cancer Res. 1 (2003) 195206.