The oncoprotein SET/I2PP2A (protein phosphatase 2A inhibitor 2) participates in various cellular mechanisms such as transcription, cell cycle regulation and cell migration. SET is also an inhibitor of the serine/threonine phosphatase PP2A, which is involved in the regulation of cell homeostasis. In zebrafish, there are two paralogous set genes that encode Seta (269 amino acids) and Setb (275 amino acids) proteins which share 94% identity. We show here that seta and setb are similarly expressed in the eye, the otic vesicle, the brain and the lateral line system, as indicated by in situ hybridization labeling. Whole-mount immunofluorescence analysis revealed the expression of Seta/b proteins in the eye retina, the olfactory pit and the lateral line neuromasts. Loss-of-function studies using antisense morpholino oligonucleotides targeting both seta and setb genes (MOab) resulted in increased apoptosis, reduced cell proliferation and morphological defects. The morphant phenotypes were partially rescued when MOab was co-injected with human SET mRNA. Knockdown of setb with a transcription-blocking morpholino oligonucleotide (MOb) resulted in phenotypic defects comparable with those induced by setb gRNA (guide RNA)/Cas9 [CRISPR (clustered regularly interspaced short palindromic repeats)-associated 9] injections. In vivo labeling of hair cells showed a significantly decreased number of neuromasts in MOab-, MOb- and gRNA/Cas9-injected embryos. Microarray analysis of MOab morphant transcriptome revealed differential expression in gene networks controlling transcription in the sensory organs, including the eye retina, the ear and the lateral line. Collectively, our results suggest that seta and setb are required during embryogenesis and play roles in the zebrafish sensory system development.
The oncoprotein SET, also known as I2PP2A (protein phosphatase 2A inhibitor 2) or TAF1-β (template activating factor 1β), was initially identified as the SET–CAN fusion gene associated with myeloid leukemogenesis [1,2]. The protein is expressed in rapidly dividing cells of a wide variety of tissues such as liver, lung and brain , and its intracellular localization is mainly regulated by phosphorylation [1,4–7]. SET shuttles between the nucleus and the cytosol and is also found in the endoplasmic reticulum and the plasma membrane [4,8]. In the nucleus, SET acts both as a transcription cofactor and histone chaperone as well as an epigenetic regulator. As a component of the INHAT (inhibitor of histone acetyltransferases) complex, it binds to histones and inhibits histone acetylation , thus suppressing the activity of several transcription factors and nuclear receptors [10–12]. In addition, SET interacts with the transcription co-activator CBP [CREB (cAMP-response-element-binding protein)-binding protein] and the H1 chaperone prothymosin α, and stimulates CBP-mediated transcription [13–15]. Moreover, SET is involved in the replication of viral chromatin templates [16,17] and the epigenetic regulation of histone modifications . In addition, SET plays important roles in DNA repair pathways [19,20].
A key cytoplasmic/membrane role of SET is to inhibit PP2A (protein phosphatase 2A) . PP2A has been characterized as a tumor-suppressor gene that regulates multiple pro-growth/pro-survival signaling pathways associated with cancer progression, such as Akt and β-catenin . The interaction of SET with PP2A raises the interesting possibility that most of the oncogenic activities of SET may be exerted through inhibition of PP2A activity .
SET is also involved in cell cycle regulation. It synergizes with the cell cycle inhibitor p21Cip1 to inhibit cyclin B–CDK1 (cyclin-dependent kinase 1) activity . Furthermore, SET modulates the function of p53 in cell cycle arrest and apoptosis by inhibiting p53 acetylation . Increased cellular levels of SET lead to neuronal apoptosis, a finding consistent with the involvement of SET in Alzheimer's disease  and with the pathogenesis of primary microcephaly .
SET targeting on the plasma membrane has been shown to affect cell migration through the binding of SET with Rac1 . Upon reduction of SET levels by COG112, an apoE (apolipoprotein E)-mimetic peptide that blocks SET, its interaction with Rac1 is inhibited leading to decreased cell migration and invasion of cancer cells . In addition, FTY720 (Fingolimod), a sphingosine analog that blocks SET-PP2A association, mediates lung tumor suppression [28,29]. Therefore SET targeting by small molecules may manipulate multiple pathways involved in cancer progression [28,29].
Despite the plethora of information, the molecular basis of the cellular function of oncoprotein SET is still not well understood. In the present study, we exploited the zebrafish model in an attempt to bridge the knowledge gap between cellular and animal studies and gain further insights into the physiological roles of this multifunctional protein. Zebrafish provides an ideal vertebrate model for gene studies in early embryogenesis. The accessibility and optical properties of zebrafish embryos combined with the use of high-resolution microscopy allow analysis of biological processes and uncovering of the underlying mechanisms. Using a multidisciplinary approach by combining biochemical, molecular and gene expression analyses, we investigated the role of the zebrafish orthologs of human SET in early development. Our results show that the seta and setb paralogs encode proteins that share high homology with the human protein and are expressed during embryogenesis in the sensory organs of zebrafish larvae. Knockdown of set genes with specific MOs (morpholino oligonucleotides) and injections with gRNA (guide RNA)/Cas9 ([CRISPR (clustered regularly interspaced short palindromic repeats)-associated 9] induced similar morphological defects and resulted in reduced number of neuromasts in the injected embryos. Gene expression profiling of set morphants revealed differential expression of genes with key roles in the sensory system and the lateral line system development.
Zebrafish (Danio rerio) were maintained at 28.5°C under 14 h light/10 h dark cycle conditions in the zebrafish facility of the Institute of Molecular Biology and Biotechnology, Ιοannina, and staged according to Kimmel et al. . Zebrafish were maintained in accordance with the European Directive 2010/63 for the protection of animals used for scientific purposes and the Recommended Guidelines for Zebrafish Husbandry Conditions. The experimental protocols described were carried out with zebrafish embryos and larvae up to 7 dpf (days post-fertilization) and were approved by the Animal Bioethics Committee of the Institute of Molecular Biology and Biotechnology (ITE-FORTH).
The anti-Set antibody was raised against the N-terminal peptide MSASAAKVSRKEQNSNHDGADET of Setb and was affinity-purified over a peptide column (Davids Biotechnologie). Anti-goat I2PP2A antibody (sc-5655) and rabbit polyclonal anti-phospho-histone H3 (Ser10) antibody (pH3) (sc-8685-R) were obtained from Santa Cruz Biotechnology. Anti-cleaved caspase 3 (Asp175) (5A1E) rabbit polyclonal antibody was purchased from Cell Signaling Technology (9664L) and monoclonal anti-acetylated tubulin (T7451) was obtained from Sigma.
Total RNA was extracted from 7 dpf larvae using the RNeasy Mini Kit (Qiagen) according to the manufacturer's instructions. cDNAs were synthesized using the PrimeScript RT reagent Kit with gDNA Eraser (TaKaRa Bio) as described in the supplier's manual. To amplify the coding sequence of setb from the cDNA the following primers were used: setb-forward 5′-GCGGGGAATTCATGTCGGCCTCGGCC-3′ and setb-reverse 5′-GCACGCGTCGACTTAGTCATCCTCTCCATC-3′. The PCR protocol consisted of 2 min at 94°C, then 25 cycles of 98°C for 20 s, 65°C for 30 s and 72°C for 30 s, followed by 5 min at 72°C using KAPA HiFi PCR Kit (Kapa Biosystems). The PCR product was cloned into EcoRI/SalI sites of pEGFPC2 vector and setb sequence was verified by sequencing.
Morpholino oligonucleotides and microinjections
Translation-blocking and transcription-blocking MOs were obtained from GeneTools and suspended in sterile water at a concentration of 1 mmol/l. The MO sequences used were as follows: MOab 5′-ACTTTCGCCGCCGAGGCCGACATTA-3′ (translation blocking against 216–240 of mRNA NM_201468.1), transcription blocking MOb 5′-TGAGTTATATTCCCTCTCACCTTGT-3′ (transcription blocking setb E3I3 splice junction) and control MO 5′-CCTCTTACCTCAGTTACAATTTATA-3′. For the MO microinjection experiments, 3.5–7.5 ng of MOs were injected into one–two-cell-stage embryos in a volume of 2.3 nl, using the microprocessor-controlled injector Nanoliter 2000 (WPI). For mRNA microinjection, human SET cDNA was in vitro transcribed using the mMessage mMachine SP6 kit (Ambion). SET mRNA was titrated and 70 pg was co-injected with 7.5 ng of set MOab. For microinjection of gRNA/Cas9 mRNA in zebrafish embryos, the CRISPR target sequence for setb (5′-AAATAATGTCGGCCTCGG-3′) was identified by http://crispr.mit.edu and cloned into pT7-gRNA plasmid vector (Addgene) according to Jao et al. . gRNA and Cas9 mRNA (pT3TS-nCas9n; Addgene) were synthesized using the T7 and T3 mMessage mMachine kit (Ambion) respectively. Amounts of 230 pg of gRNA plus 690 pg of Cas9 mRNA were injected into zebrafish embryos at the one-cell stage, and the targeting efficiency of gRNA was verified by sequencing following a PCR amplification of the genomic region flanking the target region (forward primer 5′-CGCACTTCTAAACTGAGCTCGA-3′, reverse primer 5′-AGGAAACATCATGGCCTCCG-3′).
Whole-mount in situ hybridization
For in vitro synthesis of seta and setb riboprobes, 485 bp and 442 bp fragments were amplified by PCR and cloned into pGEMT Easy Vector (Promega). seta primers: forward, 5′-TCCCCAGTCCTGTTTCCA-3′ and reverse, 5′-GTCTATCAGCTGATTAAACC-3′. setb primers: forward, 5′-GAAGTCATCCTCCAACCAC-3′ and reverse, 5′-AACAGTGTCAACTCTTTGGG-3′.
For the sense riboprobe production, the plasmids were linearized with SalI, and with SphI for the antisense riboprobes. The linear DNAs were then used as templates for the synthesis of seta and setb sense and antisense RNA probes using a DIG RNA labeling kit with T7 and SP6 RNA polymerase respectively (Roche Applied Sciences). For detection of seta and setb expression patterns, embryos were collected at 24 hpf (hours post-fertilization), 48 hpf and 5 dpf. Embryos that developed past 24 hpf were pretreated with 0.003% PTU (phenylthiourea) to prevent pigmentation before fixation in 4% paraformaldehyde overnight at 4°C. Before hybridization, embryos were treated with 10 μg/ml proteinase K for 8, 20 and 45 min for the developmental stages of 24 hpf, 48 hpf and 5 dpf respectively. For the neuromast staining, 5 dpf larvae were permeabilized with 3 μg/ml proteinase K for 25 min. All subsequent steps of whole-mount in situ hybridization procedure were performed as described previously . The stained embryos were observed under a Nikon SMZ800 stereomicroscope and an Olympus BX41 microscope.
To detect Seta/b in zebrafish, embryos were fixed in 4% PFA (paraformaldehyde) at 4°C, overnight. After fixing, samples were washed twice in PBS and incubated in PBS/0.1% Triton X-100 for 10 min at room temperature, then blocked in blocking buffer (10% FBS and 0.1% Triton X-100) at room temperature for 30 min and incubated with rabbit anti-Set polyclonal antibody at 1:40 dilution overnight in blocking buffer at 4°C. After washing, embryos were incubated with anti-rabbit IgG–FITC in blocking buffer overnight, washed with PBS six times for 10 min and mounted onto coverslips. For the phospho-histone H3 staining, dechorionated embryos were fixed in 4% PFA for 2 h at room temperature. After fixation, embryos were washed with PBS, transferred to 100% methanol and stored at −20°C overnight. Methanol was removed and embryos were permeabilized by incubation in 100% acetone for 7 min at −20°C. Acetone was removed and embryos were rehydrated in 50% methanol/50% PBS for 1 h at −20°C and washed once with distilled water and twice with PBST (PBS/0.1% Tween 20). After rehydration, embryos were incubated in blocking solution (10% FBS, 1% DMSO and 0.1% Tween 20) for 1 h at room temperature. Subsequently, embryos were incubated with rabbit anti-pH3 (Ser10) polyclonal antibody at 1:200 dilution in blocking solution at 4°C overnight. Embryos were washed four times for 15 min with PBS-DT (PBS, 1% DMSO and 0.1% Tween 20) and incubated with Alexa Fluor® 488-conjugated anti-rabbit antibody at 1:200 dilution in blocking solution for 3 h at room temperature under dark conditions. After washing (four times for 15 min) embryos were stained with TOPRO-3 in PBS for 30 min at room temperature, protected from light. Finally, embryos were washed with PBS and mounted onto coverslips. Active caspase was detected using the anti-rabbit cleaved caspase 3 polyclonal antibody at 1:100 dilution in blocking solution, following the same protocol as for the pH3 staining. For the acetylated tubulin staining, larvae were fixed in methanol, then incubated in acetone for 7 min at −20°C and rehydrated in graded series of methanol/PBS at 95, 75, 50 and 25% for 10 min each at room temperature. Next, larvae were washed four times for 5 min with PBS/0.5% Triton X-100 and incubated with proteinase K (10 μg/ml) at room temperature for 10 min. After washing, larvae were post-fixed in 4% PFA at room temperature for 30 min, washed again in PBS/0.5% Triton X-100 and incubated in blocking solution (10% FBS, PBS and 0.5% Triton X-100) at room temperature for 1 h and with anti-mouse acetylated tubulin diluted 1:1000 in blocking solution overnight at 4°C. Subsequently, larvae were washed with PBS/0.5% Triton X-100, incubated with Alexa Fluor® 594-conjugated anti-mouse antibody at 1:800 dilution in blocking solution for 4 h at room temperature in the dark, washed and mounted. For the double staining of Seta/b-acetylated tubulin, embryos were incubated with the two antibodies sequentially. Staining of actin cytoskeleton was performed by incubating embryos with rhodamine–phalloidin (Molecular Probes, R415) at 1:20 dilution in 10% FBS, overnight at 4°C, under dark conditions. Double labeling of acetylated tubulin and cleaved caspase 3 was performed sequentially after fixation of 48 hpf embryos with methanol, as described above. Imaging was performed using a Leica TCS SP5 II scanning confocal microscope. Captured images were subsequently processed in LAS AF and Adobe Photoshop.
Western blot analysis
Zebrafish PAC2 fibroblast cells were cultured in Leibovitz's L-15 medium (Gibco) supplemented with 15% FBS at 28°C and 0.5% CO2. HEK (human embryonic kidney)-293 cells were cultured in DMEM (Dulbecco's modified Eagle's medium) supplemented with 10% FBS, antibiotics, 1% l-glutamine and 0.3% glucose. For transient transfection, cells were grown in six-well plates to 50–70% confluence and were transfected using Fugene 6 (Roche) according to the manufacturer's instructions. Cells were harvested 48 h later. Total protein extracts obtained from zebrafish embryos, PAC2 and HEK-293 cells were separated on an SDS/12% polyacrylamide gel and transferred onto nitrocellulose membranes for 1 h. Detection of Seta/b proteins was performed by incubation of the membranes with anti-goat I2PP2A antibody (1:1000 dilution, Santa Cruz Biotechnology; sc-5655) and anti-Set (1:1000). Reactions were visualized using ECL (Pierce Thermo Scientific).
Acridine Orange staining
Uninjected, MOab- and control MO-injected embryos were collected at the stage of 48 hpf and soaked in 5 μg/ml Acridine Orange (Sigma) diluted in E3 embryo medium (5 mM NaCl, 0.17 mM KCl, 0.33 mM CaCl2 and 0.33 mM MgSO4) at room temperature for 30 min under dark conditions. The embryos were washed three times with E3 medium, anesthetized with 0.016% tricaine in E3 medium and observed by confocal microscopy.
In vivo staining of neuromast hair cells
The vital fluorescent dye SynaptoGreen C4 (FM1-43) (Biotium) was used as a mechanotransduction marker of functional hair cells. Larvae were incubated in 0.3 μΜ FM1-43 diluted in E3 medium at room temperature for 1 min in the dark, followed by three rinses with embryo medium. Larvae were anesthetized with 0.016% tricaine in E3 medium, and neuromasts with labeled hair cells were counted unilaterally under a confocal microscope.
For statistical analysis, data were first examined for homogeneity of variances by Levene's test. Statistical differences were evaluated by one-way ANOVA followed by Tukey's HSD (honest significant difference) post-hoc test for equal variances or by Games–Howell test for the case of unequal variances. Data are presented as mean ± S.E.M. considering P ≤ 0.05 as statistically significant. Statistical analysis was performed using SPSS Statistics version 22 (IBM).
Total RNA extraction
Total RNA was extracted from three pools (50 embryos each) of 48 hpf embryos that were injected with MOab, MOb or control MO and purified using the RNeasy Mini kit (Qiagen), according to the manufacturer's instructions. RNA samples were quantitated using a spectrophotometer (Nanodrop 1000, Thermo Scientific) and checked for integrity with the 2100 Bioanalyzer system (Agilent Technologies).
Affymetrix platform transcriptional microarray analysis was performed on RNA samples obtained from set MOab morphants and control MO-injected embryos (three biological replicates from each group) using the Affymetrix GeneChip Zebrafish Gene 1.0 Arrays. RNA samples were processed and labeled for array hybridization using the Ambion WT Expression kit (Life Technologies, catalog number 4411974). Labeled fragmented cDNA (Affymetrix GeneChip® WT Terminal Labeling and Controls kit, catalog number 901524) was hybridized to Zebrafish Gene 1.0 arrays for 16 h at 45°C (at 60 rpm) (Affymetrix GeneChip® Hybridization, Wash, and Stain kit, catalog number 900720). Arrays were washed and stained using the Affymetrix Fluidics Station 450, and scanned using the Hewlett-Packard GeneArray Scanner 3000 7G. Quality of array data was assessed using Expression Console (version 1.3) software (Affymetrix) before importing and analyzing the data in GeneSpring 13.0 (Agilent Technologies).
Differentially expressed transcripts were determined by comparing normalized average intensities (log2 GC-RMA signal intensity values) of biological triplicates of set MOab morphants and control MO-injected embryos and filtering for statistically significant genes using an unpaired Student's t test and a cut-off value of P < 0.05 and fold change >1.7. Microarray data have been deposited in the Gene Expression Omnibus (GEO) database under accession number GSE70049.
Quantitative real-time PCR
Real time PCR was performed on a Lightcycler capillary system 2.0 (Roche) with SYBR Green fluorescent label. cDNA was prepared with the SuperScript® III First-Strand Synthesis SuperMix kit (Invitrogen). Quantitative real-time PCR (qPCR) reaction mix (10 μl) contained 5 μl of 2× Precision SYBR Green mastermix (PrimerDesign), 0.25 μl of each PCR primer (10 μM stock), 2 μl of diluted cDNA template (12-fold) and 2.5 μl of RNase-free water. The qPCR parameters consisted of initial denaturation at 94°C for 15 min followed by 45 cycles at 94°C for 15 s, 60°C for 20 s and 72°C for 10 s. All reactions were performed in triplicate. Standard curves were constructed for each gene using serial dilutions of stock cDNA. The primers of the target genes were designed based on Danio rerio gene sequences found in NCBI. Gene names, gene ID numbers, forward and reverse primer sequences are listed in Supplementary Table S2. Gene expression levels were quantified relative to the expression of sdha and rpl13 reference genes provided in the geNorm reference kit (PrimerDesign). Normalization and data analysis was carried out by the qBase Plus analysis program (Biogazelle) according to MIQE guidelines. Statistical analysis was performed using the unpaired Student's t test (two-tailed). Data are presented as mean ± S.E.M. considering P ≤ 0.05 as statistically significant.
Expression patterns of seta and setb during early zebrafish development
A search in ZFIN database for zebrafish homologs of mammalian SET identified two set paralogs; seta (1835bp, ZDB-GENE-030131-2221, ENSDARG00000031495) located on chromosome 8 which encodes a protein of 269 amino acids and setb (1862bp, ZDB-GENE-030131-433, ENSDARG00000003920) which is located on chromosome 21 and encodes a protein of 275 amino acids (Figure 1A).
Phylogenetic and expression analysis of Set proteins.
To evaluate the evolutionary relationships between zebrafish Seta/Setb and SET proteins from other species, we searched for sequences originating from various groups of organisms such as mammals, fish, amphibians, rodents and birds. Phylogenetic analysis revealed that SET is highly conserved among different species. Danio rerio Seta and Setb are approximately 96% identical with the human protein (Figure 1A), despite the fact that they are more closely related to their fish and amphibian orthologs than to mammalian SET proteins which cluster as a distinct group (Figure 1B).
To study the physiological roles of seta and setb, we first examined their expression patterns in zebrafish embryos and larvae. Whole-mount in situ hybridization labeling of 24 hpf and 48 hpf embryos and 5 dpf larvae revealed similar expression patterns of seta and setb. At 24 hpf setα and setb were detected in the brain with increased expression in the midbrain–hindbrain boundary (MHB) and the eye retina. In the tail region, of these embryos, seta was detected in the spinal cord, whereas setb was more restricted in the tail bud (Figure 2A,B). At the stage of 48 hpf, both set genes were highly expressed in the eye, the otic vesicle and the MHB (Figure 2C,D). The same expression pattern persisted at the larval stage of 5 dpf at which the transcription of set genes was more focused on the RPE (retinal pigment epithelium) (Figure 2E–I,K) and the RGCL (retinal ganglion cell layer) (Figure 2E,G,I,K). The transcripts were also detected in the MHB and the otic vesicle (Figure 2F,H). Imaging at higher magnification revealed staining of the anterior, lateral and posterior cristae of the inner ear that contain patches of sensory hair cells (Figure 2J,L). At this stage, the lateral line mechanosensory organs of zebrafish, the neuromasts  were clearly visible (Figure 2M–R). Figure 2 shows in more detail the expression of both set genes in the accessory and hair cells of the neuromasts (Figure 2P,R respectively). These results suggest that the expression pattern of set genes is correlated with the development of the sensory system organs in zebrafish.
Whole-mount in situ hybridization labeling for seta and setb.
Expression pattern of Seta/b proteins in zebrafish embryos and larvae
To analyze in more detail the expression of Seta/b in zebrafish embryos and larvae, we performed whole-mount immunofluorescence experiments. To this end, we generated a rabbit polyclonal antibody against a peptide containing the 1–23 amino acid residues of Setb that recognizes both Seta and Setb proteins in Western blots (Supplementary Figure S1A,D). In 48 hpf embryos the expression of Seta/b was mainly detected in the periphery of the eye and the lens, as well as in the olfactory pit (Figure 3A, a, arrows). In addition, staining was detected in the spot-like pattern arranged along the length of embryos that consists of the neuromasts (Figure 3A, b,c, arrows). The number and size of the anti-Set-labeled neuromasts in the anterior and posterior lateral line were increased in 7 dpf larvae (Figure 3A, d–f, arrows). Expression of Seta/b was also detected at the RGCL (Figure 3B, a, arrow) and the olfactory pit (Figure 3B, a, rectangle). A view of the olfactory pit at higher magnification shows the expression of Seta/b in the olfactory cilia (Figure 3B, b,c). These data are in agreement with the patterns of set expression (Figure 2) and point to a function of Seta/b proteins in the zebrafish sensory system during early development.
Immunostaining of Seta/b proteins in embryos and larvae.
Localization of Seta/b in lateral line neuromasts
We next checked in more detail the localization of Seta/b in the lateral line system. The lateral line originates from ectodermal placodes that develop anteriorly and posteriorly to the otic placode generating the anterior lateral line (ALL) and posterior lateral line (PLL) respectively . As the lateral line placode migrates towards the tail tip, regularly spaced proneuromasts are deposited and differentiate into hair cells and accessory cells [34,35]. Hair bundles are composed of a single primary microtubule-based cilium (kinocilium) on top of the hair cell, which is flanked by multiple rows of actin-filled projections (stereocilia) (Figure 4A, schematic representation). To study the localization of Seta/b proteins at the lateral line neuromasts, we used confocal microscopy to visualize actin and acetylated tubulin with Seta/b proteins in hair cells. As illustrated in Figure 4A, B cells labeled by the anti-Set antibody were arranged in a typical rosette pattern. Seta/b were localized at the kinocilia and the stereocilia of hair cells at 48 hpf (Figure 4A, a,b). This staining pattern was more visible in the neuromast hair cells of 7 dpf larvae (Figure 4B, a–c). Co-staining of Seta/b and actin, using the anti-Set antibody and phalloidin, confirmed the localization of Seta/b in the stereocilia (Figure 4C). In a parallel experiment, double staining of 7 dpf larvae with anti-Set and anti-acetylated tubulin antibodies demonstrated the presence of Seta/b in the kinocilia (Figure 4D). The expression of Seta/b in the lateral line system suggests a potential role of these proteins in the development of neuromasts and/or their stereotypical distribution and function.
Localization of Seta/b proteins in the lateral line neuromasts.
Morpholino knockdown of set genes
To investigate the role of seta and setb in embryonic development we proceeded with knockdown experiments by injecting MOs into one–two-cell-stage embryos. Injection of increasing amounts of a translation-blocking MO (MOab) that inhibits the translation of both seta and setb mRNAs resulted in a reduction in Seta/b protein levels in a dose-dependent manner, indicating the efficacy of the MO strategy (Figure 5A). Whole-mount immunofluorescence analysis of 48 hpf control MO- and MOab-injected embryos using the anti-Set antibody also showed a significant decrease in the staining intensity, supporting further the MO blocking efficiency (Figure 5B). We next analyzed the effect of Seta/b knockdown in zebrafish morphogenesis. Embryos injected with 7.5 ng of MOab (MOab morphants) exhibited morphological abnormalities compared with the uninjected and control MO-injected ones. The morphants were characterized by a shortened anterior–posterior axis, small heads, bent trunks and curled tails. MOab injection also had a profound effect on eye development, since MOab morphants had smaller and distorted eyes. Depending on the severity of the morphological defects, the morphants were classified into four groups: ‘normal-like’ with no significant phenotypic differences compared with the uninjected or control MO-injected embryos, apart from their smaller bodies; ‘mild’ with slightly curved tails; ‘moderate’ morphants exhibiting defects in the size and shape of their heads/eyes as well as a curvature in the body axis; these embryos were also unable to swim; and ‘severe’ consisting of highly deformed morphants (Figure 5C). The scoring of these phenotypes upon injection of increasing MOab doses is shown in Figure 5D. The control MO up to 15 ng did not cause any morphological defects (results not shown). The abnormalities of seta/b morphants were partially rescued by co-injection of 70 pg of human SET mRNA with 7.5 ng of MOab, indicating that these defects were specifically due to inhibition of seta/b translation by the MO (Figure 5E).
Morpholino-induced knockdown of seta and setb.
Injection of a transcription-blocking MO (MOb) that blocks only setb pre-mRNA splicing (Figure 5F) reduced Seta/b protein levels less efficiently, as expected (Figure 5G,H). The percentage of protein reduction in the embryos injected with MOb (Figure 5G) suggests that, although seta and setb have overlapping expression patterns (Figure 2), setb is the predominantly expressed gene at this stage. MOb morphants (7.5 ng of MOb) exhibited similar defects to MOab morphants in the head and tail and were classified in the mild and moderate groups (Figure 5I). The variability of MOab and MOb effects is possibly due to the differences between the MO blocking mechanism (translation compared with splicing interference) and sequence-dependent efficiency. In addition, as shown in Supplementary Figure S1D, Seta/b proteins were detected at 4 hpf, suggesting that they are maternally deposited (either as protein or mRNA) which can also account for the lower efficiency of the splice-blocking MOb.
To validate further the effects of the two MOs, we designed a gRNA that targeted the first exon of setb when injected with Cas9 mRNA into one-cell-stage embryos. The level of knockdown achieved by gRNA/Cas9 injection was approximately 60% compared with wild-type or Cas9-injected embryos at 48 hpf (Figure 5J). gRNA/Cas9-injected embryos exhibited mild and moderate phenotypes, compared with uninjected or Cas9-injected embryos (Figure 5K).
To determine whether cell proliferation was affected by Seta/b protein depletion, MOab embryos of moderate phenotypes were fixed and immunostained with the anti-phospho-histone H3 (pH3) antibody, which marks late G2-phase and early mitosis. pH3 labeling was significantly decreased in the morphants mainly in the head and trunk/tail, compared with uninjected or control MO-injected embryos (Figure 6A). In vivo staining of embryos with Acridine Orange, a vital dye that specifically stains apoptotic cells through binding to nucleic acids, detected apoptotic foci mainly in the head and trunk of the morphants (Figure 6B). Furthermore, active caspase 3, a key regulator of the apoptotic pathway, was mainly detected in the head and trunk, where decreased proliferation was also observed (Figure 6C).
Effects of set knockdown on cell proliferation and apoptosis.
Taking into account the localization pattern of Seta/b in the neuromast hair cells, we next asked whether the neuromast number was changed in set morphants. To this end, we used the fluorescent styryl dye FM1-43 that labels in vivo active hair cells in neuromasts (representative images shown in Figure 7A). As illustrated in Figure 7B, at 4 dpf in uninjected and control MO-injected larvae, the total number of neuromasts derived from the ALL and PLL was approximately 18. In contrast, only three neuromasts were detected in MOab morphants, whereas MOb larvae displayed ∼13 neuromasts. These data suggest a combined function of seta and setb genes in the lateral line neuromast development. In agreement with the MO knockdown experiments, injection of gRNA/Cas9 mRNA in one-cell-stage embryos also resulted in a severe decrease in the neuromast number in 7 dpf larvae (Figure 7C). To investigate further whether apoptosis is the underlying mechanism for the decreased number of neuromasts upon inhibition of seta/b translation, we co-stained MOab morphants with the anti-acetylated tubulin and anti-caspase 3 antibodies. As shown in Figure 7D, the apoptotic foci and the acetylated tubulin in the head (Figure 7D, g–i) or the trunk (Figure 7D, j–l) of the morphants display no significant degree of co-localization suggesting that apoptosis is not responsible for the reduced neuromast number. These data taken together propose an important role for set genes in neuromast formation, although the underlying mechanism is not clear yet.
Effects of set knockdown on the neuromast development.
Gene expression profiling of set morphants
The localization pattern of set genes and Seta/b proteins in the sensory organs of zebrafish larvae and the decreased number of neuromasts upon the reduction in Seta/b levels prompted us to determine the impact of seta and setb knockdown on the differential expression of genes involved in the sensory system development. Hence we employed microarray analysis to compare the gene expression profiles of MOab morphants with those of control MO-injected embryos at 48 hpf, a time point that corresponds to the stage that the migrating lateral line primordium reaches the tip of the tail. The total transcriptional activity was slightly decreased in the morphants. By selecting a threshold higher than 1.7-fold change and P ≤ 0.05 in set morphants compared with control MO-injected embryos, 1128 genes displayed significantly altered expression. Of these genes, 427 were up-regulated and 701 were down-regulated (Figure 8A). In more detail, this analysis revealed significant changes in the expression of genes involved in gene transcription (84, P = 5.72 × 10−13) and cell adhesion (31, P = 3.95 × 10−10). Gene ontology characterization indicated an effect of set genes in the sensory organ development (Figure 8B), in agreement with seta and setb expression patterns (Figure 2). To validate the results of the microarray analysis, we examined the expression of selected genes by qPCR using as template RNA obtained from MOab- and MOb-injected embryos, focusing on genes likely to be critical for the sensory organs and lateral line development (Supplementary Table S1 and Figure 8C,D).
Transcriptome analysis of set morphants.
Concerning eye development, we validated the expression of the following genes: crx (cone-rod homeobox protein) which is expressed in proliferating retinal progenitors; optc (opticin) a leucine-rich repeat protein that is present in the vitreous of the eye; six7 (sine oculis) a homeobox transcription factor that governs expression of green opsin genes in zebrafish; cryaa (α-crystallin) that is required for normal lens development and the genes of bHLH (basic helix–loop–helix) transcription factors neurod4 and atoh7. Atoh7 was the most down-regulated gene in embryos injected with MOab (Figure 8C), consistent with the microarray data (Supplementary Table S1). This gene is crucial for the development of RGCs (retinal ganglion cells) in the vertebrate retina [36–39]. Comparison of atoh7 expression levels in MOab and MOb embryos revealed that inhibition of the translation of both seta and setb resulted in 200-fold down-regulation of atoh7 (Figure 8C). In contrast, blocking of setb by MOb did not cause such a severe effect on atoh7 expression (2-fold down-regulation, Figure 8D).
To analyze the effect of set knock down on lateral line-related genes we checked the expression of neuroD that is implicated in neuromast hair cell survival , junba which is involved in the regenerative process of the hair cells  and ovol1 that links Wnt signaling with cell migration . In addition, we determined the expression levels of slc12a3 gene which belongs to the solute-linked carrier (slc) family of proteins that regulate the apicobasal polarity of the hair cell in functional neuromasts. As shown in Figure 8C,D, all of the above genes were differentially regulated in both MOab- and MOb-injected embryos. To analyze further the role of set genes in lateral line development, we checked the expression levels of atoh1a, that encodes for a bHLH transcription factor with a key role in this system [43,44]. Using qPCR, we found that atoh1a was significantly up-regulated in both MOab- and MOb-injected embryos, whereas the expression of atoh1b paralog was less affected (Figure 8C,D, insets).
To obtain a more complete picture on the gene expression profile of set morphants, we then examined the transcript levels of three genes that play important roles in otic development: foxj1a, which is required for the genesis of the motile cilia of the otic vesicle and for otolith tethering ; foxi1, which is expressed in otic precursor cells ; and oc90 (otoconin-90) that is required for otolith mineralization in zebrafish . Comparing the expression pattern of these genes between MOab- and MOb-injected embryos, it was apparent that the effect of MOab which blocks translation of both seta and setb was more prominent than that of MOb alone.
Consistent with the published roles of SET in cell cycle regulation , p21 (cyclin-dependent kinase inhibitor 1A, cdkn1a) was the most up-regulated gene (Supplementary Table S1 and Figure 8B,C). In contrast, akt3 (protein kinase B, PKB), a serine/threonine kinase that regulates cell proliferation and the formation of bHLH transcription factor–co-activator complexes  was down-regulated. Τhe expression pattern of these cell cycle-related genes is in agreement with the data presented in Figure 6 that show reduced proliferation and increased apoptotic index in the morphants.
Finally, among several development-related transcription factors listed in our microarray data, we have confirmed using qPCR analysis the expression pattern of the homeobox genes otpa and otpb (homeodomain transcription factor Orthopedia) that regulate the specification of defined subsets of dopaminergic neurons (Supplementary Table S1 and Figure 8C,D). Collectively, our data reveal a complex gene network in which seta and setb may play roles in zebrafish sensory organ development.
SET/I2PP2A is a multifunctional protein involved in several cellular mechanisms including the cell cycle , cell migration  and gene regulation [10–12]. Zebrafish has two set paralog genes, seta and setb, that might exist after a gene duplication event, raising an interesting question regarding their functional relationship. It is important, but still difficult, to differentiate between seta and setb biological functions. These genes code for proteins with almost identical sequences (94% identity), suggesting that there are no changes in one or both genes that would be beneficial to allow the evolution of new functions. Thus the set gene duplicates may act together to perform the functions of their mammalian ortholog. Our in situ hybridization experiments demonstrated similar expression patterns of zebrafish seta and setb genes, consistent with the expression of setb reported by Rauch et al. . More specifically, the transcripts of the set genes were detected in the MHB, the neuromasts and the RGCs which are the first-born retinal cell type that transmit all visual information from the eye to the brain via their axons. Recently, it has been reported that Set-β is developmentally regulated in cultured RGCs , suggesting an agreement between animal and cell culture studies. In addition, Set acts as a downstream cytosolic effector of NF-protocadherin which regulates RGC axon and dendrite outgrowth .
The zebrafish Seta/b proteins were detected in early embryogenesis at the stage of 3.5–4.0 hpf, suggesting the presence of maternal set mRNA in fertilized eggs (Supplementary Figure S1A). The expression of Seta/b in 48 hpf embryos and 7 dpf larvae were mainly detected in the periphery of the eye and the lens, the cilia of the olfactory pit and the sensory organs of the lateral line system. Concerning the Seta/b localization in the neuromasts, the proteins were found in the kinocilium and the stereocilia of hair cells. These findings are interesting, since in mammals, hair cells exist within the sensory epithelium of the cochlea in the inner ear, which is a highly tuned frequency analyzer driven by the co-ordinated motion of hundreds of hair cells . Therefore the localization pattern of Seta/b suggests specific roles in the cilia development of the zebrafish sensory organs.
Knockdown of seta and setb in zebrafish embryos by MO injections caused distinct morphological defects mainly characterized by growth retardation, curved posture and bent tail. The defects of set morphants could be due to unbalanced cell proliferation and increased cell death, which might result in growth imbalances and ventral curvature of the body axis. Indeed, higher apoptosis levels were detected in the morphants' trunks and tails compared with the controls. In contrast, in the same regions of the morphant embryos, the number of proliferation foci was reduced. These results are in line with previous studies which showed that loss of Set-β function, as a result of granzyme A or asparagine endopeptidase proteolytic cleavage, is implicated in cell death [52,53]. In addition, it has been shown recently that an N-terminal cleavage product of Set protein can induce neuronal death in purified RGCs . In agreement with the suggested role of Seta/b in cell proliferation, our analysis of signaling pathway enrichment in MOab transcriptome indicated that Wnt signaling pathway was down-regulated in the morphant transcriptome (results not shown). This is an interesting finding because the Wnt/β-catenin pathway is required for the control of proliferation in the neuromasts and stimulates hair cell regeneration after damage .
The use of MOs in studying development is often accompanied by off-target effects that cause developmental defects [55,56]. In the present study, we have systematically addressed this issue in several ways by: (i) comparing the effects of set splicing- and translation-blocking MOs performing dose–response and time-course experiments; (ii) analyzing protein expression levels in set morphants with Western blot and immunofluorescence analysis; (iii) performing RNA rescue experiments; and (iv) comparing the MO phenotypes with the morphology of embryos obtained upon set targeting with gRNA/Cas9 injections. Besides the morphological abnormalities, set morphants also presented with a significant loss of neuromasts, as revealed by FM1-43 staining. Consistent with this result, the number of neuromasts of gRNA/Cas9-injected embryos was also significantly reduced.
A systematic evaluation of the gene expression profile of set morphants by microarray analysis revealed interesting insights into the function of set genes in zebrafish. First, gene ontology characterization of the microarray genes indicated an effect of set in the sensory system development, consistent with both transcript and protein expression patterns. Secondly, the atonal genes atoh7 and atoh1a were significantly affected according to qPCR analysis. These genes are central factors for the sensory system development because they are linked with the specification of photo- and mechano-receptor cells . Previous work in Drosophila has proposed that photoreceptive and mechanosensitive cell development is evolutionarily linked with an atonal-dependent proto-sensory organ . In vertebrates, these two crucial functions of atonal factors have been separated by gene duplication. Atoh7 (formerly Ath5) specifies mechanoreceptors, olfactory receptors and photoreceptors and is required for the genesis of retinal ganglion cells [36,39]. The level of atoh7 expression determines the ultimate number of RGCs; mutant atoh7 larvae lack RGCs causing blindness  and atoh7 mutant mice lack optic nerves and have a severe reduction in the ganglion cell population . In humans, a deletion of a remote ATOH7 element that impairs the levels and timing of atoh7 expression results in the NCRNA (non-syndromic congenital retinal non-attachment) disease . On the other hand, atoh1a is specific for the generation of the mechanosensitive cells in the ear and the lateral line and its expression defines the cell potential to form hair cells . A failure to down-regulate atoh1a in differentiating hair cells results in immature hair cells that ultimately die .
The activity of atonal transcription factors is context-dependent and may be regulated by region-specific transcription factors that bind to adjacent sites in atonal-dependent enhancers . It is also well established that neural bHLH proteins function as transcriptional activators by recruiting the transcriptional co-activators CBP/p300 to target genes . Since SET binds to CBP/p300 and regulates gene expression [13,14], it is not surprising that the expression of transcription factors whose function depends on CBP/p300 may be affected by Set depletion. On the basis of the results of the present study, it is tempting to propose that set genes may modulate in vivo the transcriptional activity of bHLH factors involved in the sensory system development in several ways. First, by modifying the epigenetic state of atoh targets through the regulation of histone post-translational modifications that are essential for the differentiation of hair cells and RGCs. Therefore it is possible that atoh transcriptional targets have been epigenetically modified in set morphants and are not suited for transcription. Secondly, by affecting the expression of regulatory factors and other transcription factors that are required for atoh-dependent transcription.
It will be of great interest to investigate the relationship of seta and setb genes with atoh1 and atoh7 spatial and temporal transcription focusing on chromatin structure and epigenetic modification changes upon inhibition of set translation or upon gene targeting by the CRISPR/Cas9 system.
In conclusion, our results propose a mechanism for the function of set genes underlining the development of zebrafish sensory system organs and future studies will aid in the dissection of the role of Set proteins during these processes.
anterior lateral line
CREB (cAMP-response-element-binding protein)-binding protein
clustered regularly interspaced short palindromic repeats
human embryonic kidney
protein phosphatase 2A inhibitor 2
posterior lateral line
protein phosphatase 2A
quantitative real-time PCR
retinal ganglion cell
retinal ganglion cell layer
retinal pigment epithelium
I.S. and E.T. performed experiments, interpreted and analyzed the data. K.S. and Z.K. analyzed and interpreted the data. D.B. provided reagents, advice and edited the paper before submission. T.P. conceived the study, designed the experiments and interpreted the data. T.P. and I.S. wrote the paper.
This work has been co-financed in part by the European Union (European Regional Development Fund, ERDF) and Greek national funds through the Operational Program ‘Thessaly-Mainland Greece and Epirus-2007–2013’ of the National Strategic Reference Framework (NSRF 2007–2013).
We are grateful to Professor Ioannis Leonardos for his contribution to the establishment of the zebrafish facility. We also thank Dr Apostolos Mpatsidis for his advice on the statistical analysis and Villy Karali for her help in the initial stages of this project.
The Authors declare that there are no competing interests associated with the manuscript.