Ftr82 Is Critical for Vascular Patterning during Zebrafish Development

Abstract

Cellular components and signaling pathways are required for the proper growth of blood vessels. Here, we report for the first time that a teleost-specific gene ftr82 (finTRIM family, member 82) plays a critical role in vasculature during zebrafish development. To date, there has been no description of tripartite motif proteins (TRIM) in vascular development, and the role of ftr82 is unknown. In this study, we found that ftr82 mRNA is expressed during the development of vessels, and loss of ftr82 by morpholino (MO) knockdown impairs the growth of intersegmental vessels (ISV) and caudal vein plexus (CVP), suggesting that ftr82 plays a critical role in promoting ISV and CVP growth. We showed the specificity of ftr82 MO by analyzing ftr82 expression products and expressing ftr82 mRNA to rescue ftr82 morphants. We further showed that the knockdown of ftr82 reduced ISV cell numbers, suggesting that the growth impairment of vessels is likely due to a decrease of cell proliferation and migration, but not cell death. In addition, loss of ftr82 affects the expression of vascular markers, which is consistent with the defect of vascular growth. Finally, we showed that ftr82 likely interacts with vascular endothelial growth factor (VEGF) and Notch signaling. Together, we identify teleost-specific ftr82 as a vascular gene that plays an important role for vascular development in zebrafish.

1. Introduction

The establishment of blood vessels is critical for embryonic development, and occurs in two distinct stages: vasculogenesis and angiogenesis. The de novo formation of new vessels through angioblast aggregation, migration, specification, and lumenization is termed vasculogenesis. Sprouting and growth of new vessels from preexisting vascular structures is termed angiogenesis. The regulation of vasculogenesis and angiogenesis is involved in multiple cell types, signaling molecules, and interaction pathways to coordinate the formation of the vascular system [1,2,3,4,5]. The zebrafish Danio rerio is an excellent model for the study of gene functions and signaling pathways in the field of vascular development because of the transparency of embryonic stages and many useful transgenic lines that express fluorescent tags in endothelial cells [6,7]. During zebrafish development, formation of the dorsal aorta and the posterior cardinal vein via the specification of angioblast precursor cells is considered vasculogenesis, and the subsequent sprouting and extension of the intersegmental vessels (ISVs) from the dorsal aorta or the process of caudal vein plexus (CVP) formation from the posterior cardinal vein is termed angiogenesis [8,9].

Specific genes controlling artery/vein specification of vasculogenesis have been identified in many vertebrate species [5,10]. In zebrafish embryos, sonic hedgehog (Shh) and vascular endothelial growth factor (VEGF) activate Notch signals via PLC-γ-PKC-MEK pathways and are required for arterial specification [11,12,13,14]. A distinct genetic program that activates PI3K/AKT signaling in angioblasts promotes venous specification [15,16,17]. However, the precise mechanism of signaling interactions in vascular development has yet to be elucidated. From the angiogenesis aspect, VEGF-Notch signaling also plays a central role in ISV patterning, including the specification, proliferation, and migration of endothelial cells [18,19]. Meanwhile, BMP (bone morphogenetic proteins) signaling regulates angiogenic sprouting from the posterior cardinal vein to form the caudal vein plexus (CVP), which is a distinct mechanism of angiogenesis from ISV growth [20,21,22]. Thus, a complex coordination among these signals and molecules has a critical role in controlling endothelial tip cell fates during angiogenesis. However, it is still not fully known which molecules are required for ISV and CVP formation. Identification of the regulators in these processes will enhance our understanding of the patterning of vascular networks, and could provide novel therapeutic targets.

In mammals, the tripartite motif proteins (TRIM) family are involved in various cellular processes, including cell growth, differentiation, and apoptosis [23,24,25]. The biological function of TRIM proteins in many studies is related to innate immunity against viral infection [24]. ftr82 (finTRIM family, member 82) belongs to a large new subset family of tripartite motif proteins, called finTRIMs, which contain a RING-Bbox-coiled coil motif followed by different C-terminal domains. It was first described in teleost rainbow trout by virus induction [24]. FinTRIMs harbor nearly identical RING/B-box regions and C-termini with a B30.2 domain compared to TRIMs of fish and mammals. The zebrafish genome is very diverse, with 84 finTRIM genes on different chromosomes. A phylogenetic analysis showed the number of fintrim genes varies among fish species. The evolution and features of finTRIMs suggest an important role in the innate immunity of fish. The closest mammalian relatives are trim16 and trim25, which have been shown to function in anti-viral infection, but are not true orthologs. It has been shown that bloodthirsty (bty) is an RBCC/TRIM gene required for erythropoiesis in zebrafish [26]. Other than these, there is limited information about TRIMs in fish studies, and no description of TRIM proteins related to vascular function is known.

In this study, we examine the function of ftr82 in vascular development. We found that ftr82 mRNA is expressed in developing vessels, and the loss of ftr82 by morpholino (MO) knockdown impairs vascular patterning which reduces the expression of vascular markers. We further confirmed that ftr82 MO knockdown works efficiently and is gene-specific. We next showed that the growth defect of endothelial cells in ftr82 morphants is likely due to a decrease of cell proliferation and migration, but not the death of endothelial cells. Finally, we showed that ftr82 likely interacts with VEGF and Notch signaling. Together, we identify teleost-specific ftr82 as a vascular gene that plays an important role for the vascular development in zebrafish.

2. Results

2.1. The Expression of ftr82 Is in Developing Vessels

We are interested in characterizing molecules required for vascular patterning [27,28,29]. Our unpublished transcriptome screening suggests that a teleost-specific ftr82 gene is potentially involved in vascular development. To characterize the potential role of ftr82 in the vascular development of zebrafish, we performed whole-mount in situ hybridization to examine the expression pattern of ftr82 during zebrafish embryogenesis. At the 16 somite stage (S), ftr82 is expressed in the optic placode (op), brain, head (h), and lateral plate mesoderm (lpm) (Figure 1A). A dorsal view of embryos showed that ftr82 is also expressed at the lpm (Figure 1A’). The expression at the lpm corresponds to the location in the developing vessels. At 24 h post-fertilization (hpf), ftr82 is expressed in the eye (e), brain, in the vessels (v), and in the caudal vein plexus (CVP) of the trunk (Figure 1B,B’). Transverse sections of embryos at 24 hpf further confirmed that ftr82 is expressed in vessels, including dorsal aorta (da), posterior cardinal vein (pcv), and CVP (Figure 1C,D). At 30 hpf, ftr82 expression continues in the head, trunk vessels, and CVP (Figure 1E,E’). The data show the expression of ftr82 in developing vessels and suggest the function of ftr82 in vascular development.

2.2. Knockdown of ftr82 Causes Vascular Defects during Zebrafish Development

In mammals, the TRIM protein family are involved in various cellular processes, and in many studies, concerned with the innate immunity against viral infection. There is limited information about TRIM or finTRIM proteins related to vascular function in vertebrates to date. Meanwhile, the function of ftr82 is still unknown in zebrafish. To identify the role of ftr82 in vascular development, we use transgenic Tg (kdrl:eGFP)^la116 embryos that express green fluorescent protein (GFP) in endothelial cells. We knocked down ftr82 expression by injecting 3.4 ng of splice-blocking morpholino that targeted the exon 1-intron 1 boundary (ftr82^e1i1 MO). Loss of ftr82 showed two obvious vascular phenotypes: ISV growth defect and CVP mis-patterning (Figure 2C–H). At 26 and 30 hpf, a high percentage of ISV showed growth at the mid-somite, and did not reach the dorsal longitudinal anastomotic vessel (DLAV) (Figure 2D,F,K,L). There is only 20% of complete ISVs (n = 28 embryos) in ftr82^e1i1 morphants, compared to 80% of complete ISVs in uninjected controls (n = 32 embryos). We also observed defects in CVP formation as a second phenotype. At 26, 30, and 48 hpf, endothelial cells in the uninjected embryos’ CVP region showed angiogenic sprouting, migration, and the formation of honeycomb structures/loops (Figure 2G,M,O). Less or no sprouting was observed and quantitated at 26 hpf (Figure 2H,Q), and disruption of the honeycomb structure was observed in the CVP when compared to uninjected controls at 30 hpf (Figure 2M). At 48 hpf, a swollen plexus or fewer CVP capillary loops could be observed (Figure 2S). Quantification of CVP defects showed a dose-dependent decrease in ftr82 morphants at 48 hpf. Thus, our data indicated that ftr82 may play a critical role in controlling endothelial tip cell behaviors and in regulating ISV and CVP formation during angiogenesis. We further tested the efficiency of ftr82^e1i1 morpholino knockdown by real-time PCR (RT-PCR) analysis. We showed that the injection of 3.4 ng of ftr82^e1i1 morpholino diminished the normal fragment of ftr82 (Figure 2U), indicating the loss of ftr82 expression caused by morpholino inhibition.

2.3. Loss of ftr82 Showed Edema, Circulation Defect, and Absent Parachordal Chain

By using Tg (fli1a:eGFP)^y1 fish as endothelial specific promoters of fli1a-driven GFP allows for the examination of parachordal chain (PAC) formation. The PAC is a string of endothelial cells located at the horizontal myoseptum and forms at 48 hpf and PAC cells eventually migrate ventrally along the trunk vasculature to the space between the dorsal aorta and posterior cardinal vein to form the main vessel of the lymphatic system, the thoracic duct (TD). Loss of ftr82 showed incomplete ISV (yellow hollow arrow in Figure 3B) and the absence of PAC formation (white hollow arrowhead in Figure 3B) when compared to the control. The defect of PAC formation was quantitated in the defective or partial growth in Figure 3G. At 72 hpf, we found that ftr82 morphants showed increasing blood accumulation at CVP, edema at brain or pericardium regions (Figure 3C–F). Embryos showed ~80% edema at 72 hpf (n = 28 in control and n = 31 in ftr82 MO) (Figure 3H). Blood accumulation at CVP is consistent with the defects of CVP and loss of circulation. While we examined the circulation in uninjected controls and morphants, we noted limited blood flow in the trunk of ftr82 morphants at 48–72 hpf. ftr82 morphants showed ~10% of embryos with circulation in the axial vessels and/or ISV circulation (n = 38 in control and n = 45 in ftr82 MO) (Figure 3I). Since edema and the blockage of circulation are common secondary effects of vessel impairment, the data is consistent with the vasculature defects in the ftr82 knockdown embryos.

2.4. Knockdown of ftr82 Can Be Rescued by Overexpression of ftr82

We showed that the loss of ftr82 causes vascular defects by splice-blocking morpholino knockdown (Figure 2). To test that the knockdown of ftr82 is being targeted specifically, we first tested the efficiency of ftr82^e1i1 morpholino knockdown by analyzing ftr82 expression products (Figure 2U). To further test phenotypic specificity of the ftr82 morpholino knockdown, we tested the effects of an additional morpholino targeting the translation initiation site (ftr82^atg MO). Our results showed similar vascular defects in ftr82^atg MO (Figure S1) when compared to ftr82^e1i1 morphants. We showed vascular defects in CVP sprouting at 26 hpf (four-fold decrease; Figure S1E,F,O), ISV growth at 30 hpf (50% decrease; Figure S1G,H,P), and ISV cell number at 30 hpf (1.9-fold decrease; n = 22 in controls and n = 25 in ftr82 morphants; Figure S1K,L,Q). The vascular defect at CVP capillary loop formation is also observed at 48 hpf (Figure S1M,N). To confirm the vascular defects caused by morpholino knockdown specifically, we performed rescue experiments by overexpression of ftr82 in control and ftr82^e1i1 morphant embryos. At 30 hpf, we found that overexpression of ftr82 mRNA in ftr82 MO restores ISV growth to form DLAV by 45% compared to injection of ftr82 morpholino alone (ftr82 MO) (Figure 4C–E), while overexpression of ftr82 in control embryos has no obvious effect on the vascular development (Figure 4B) when compared to uninjected controls (Figure 4A). The data suggest that knockdown of ftr82 indeed impaired vascular development. In addition, we have assessed the gross developmental process which is unaffected in ftr82 morphants, because knockdown of ftr82 did not alter the expression of heart, gut, somite, and neural systems (Figure S2). Meanwhile, loss of ftr82 did not cause developmental delay by measuring the heart rate at 24–25 hpf (Figure S3).

2.5. Knockdown of ftr82 Impairs the Growth of ISV Cells

Knockdown of ftr82 results in vascular growth defects, suggesting the interruption of endothelial cell migration and/or proliferation, or an increase of cell death. To test these possibilities, we first performed a TdT-mediated dUTP-X nick end labeling (TUNEL) assay and acridine orange (AO) staining to detect apoptotic cells. We showed that ftr82 morphants increase apoptotic cells at the epidermis of the dorsal tail region compared to uninjected embryos; however, morpholino-induced cell death has no significant increase in the vessel region of trunk (Figure 5A–D,A’–B’), suggesting that the vascular defect is not due to the endothelial cell death. To test if loss of ftr82 would affect cell proliferation, we analyzed the numbers of endothelial cells per ISV in the Tg (kdrl:mCherry^ci5; fli1a:negfp ^y7) embryos, where GFP was expressed in the nuclear of endothelial cells and the mCherry tag on endothelial ISV cells. Loss of ftr82 showed reduced ISV cells significantly when compared to the uninjected control (Figure 5E–G, n = 18 in ftr82 morphants and n = 16 in control, p < 0.0001). In addition, only a small portion of ISV cells can reach to the top of embryo to form DLAV, suggesting that endothelial cell migration is impaired in ftr82 morphants. To confirm the migration defect, we measured the difference of ISV length from 24 hpf to 28 hpf in control and ftr82 MO. The difference of ISV length in ftr82 MO is significantly shorter than control, suggesting that the migration ability is impaired (Figure 5H–L). We further demonstrated the proliferation defect in ftr82 MO by examining the expression levels of proliferation marker phosphohistone H3 (pHH3) and cell cycle negative regulators p21 and p27. Immunostaining with anti-phosphohistone H3 antibody recognizes mitotic cells, and the amount of mitotic cells in the trunk of ftr82 MO showed less brown spots compared to control (arrows in Figure 5M,N). Western blotting also showed reduced pHH3 signal in ftr82 MO. In addition, the increase of the expression of cell cycle inhibitors p21 and p27 suggested the impairment of cell proliferation. Together, the data suggest that ftr82 is important for endothelial cell growth to pattern ISV and/or CVP, likely via regulation of the proliferation and/or migration of the cells.

2.6. Knockdown of ftr82 Affects the Expression of Vascular Markers

The vascular defects in ISV growth and CVP formation suggest that ftr82 is critical for vascular development, and likely modulates vascular identity. To test this hypothesis, we examined the expression of several typical vascular markers, flk, flt4, mrc1, stabilin, and ephrinb2, at 24 hpf by whole-mount in situ hybridization. We found that the expression of the pan-vascular markers flk and stabilin, venous/ISV specific marker flt4, venous marker mrc1, and arterial marker ephrinb2 were all decreased in ftr82 morphants (dashed arrow in Figure 6B,D,F,H,J) compared to the uninjected controls. To determine the extent of vascular marker expression, we quantified flk, flt4, mrc1, and ephrinb2 transcript levels by quantitative PCR (qPCR) analysis and identified a 40%–80% decrease in expression in ftr82 morphants (Figure 6K). These results suggest that ftr82 regulates several vascular genes to control vessel development.

2.7. Regulation between ftr82 and VEGFR2-Notch Signaling

We showed that knockdown of ftr82 by morpholino injection impaired ISV growth with a decrease of ISV endothelial cells. In addition, loss of ftr82 causes the CVP defects of sprouting and capillary network formation. Notch signaling has been shown to be important for vein identity and the interaction of ISV tip cell with VEGFR2 signaling during angiogenesis. Thus, we tested the regulatory relationships between ftr82 and Notch or VEGFR2. We inactivated Notch and VEGFR2 signaling by DAPT and SU5416 treatment, respectively. We found that ftr82 expression was downregulated in vessels and CVP when Notch or VEGFR2 signals were inhibited by whole-mount in situ hybridization and qPCR analysis (Figure 7A–D). On the other hand, we induced VEGF signal by adding a pro-angiogenic compound GS4012 or VEGF in fish medium at 6 hpf, and found that the increase of ftr82 expression though VEGF-treatment was not statistically significant (Figure 7D’). Although SU5416 or DAPT interrupt the growth of ISV and CVP, the main vascular structure does not disappear (Figure 7E–G), suggesting that the down-regulation of ftr82 in drug treatment is not due to the loss of vasculature. Thus, ftr82 plays an important role in vascular development, likely mediated by the Notch/VEGF pathways.

3. Discussion

In this study, we examined the function of ftr82 in vascular development. We found that ftr82 mRNA is expressed in developing vessels (Figure 1), and the loss of ftr82 by morpholino knockdown impairs the growth of ISV and CVP (Figure 2), reduces circulation, impairs PAC formation (Figure 3), and changes the expression of vascular markers (Figure 6). We further showed that the ftr82 MO knockdown works efficiently and functionally (Figure 2, Figure 4 and S1). Next, we demonstrated that the growth defect of endothelial cells in ftr82 morphants is due to a decrease in cell proliferation and migration, but not cell death (Figure 5). However, without a test of individual endothelial cell migration and analysis of molecular markers, our data suggest that ftr82 more likely modulates the proliferation of endothelial cells rather than their migration. Finally, we showed that ftr82 likely interacts with VEGF and Notch signaling (Figure 7).

TRIM proteins contain a tripartite motif, a RING zinc finger, one or two B-boxes, and a coiled-coil domain. They have been shown as novel players in antiviral defense. Many members of this family contain E3 ubiquitin ligases activity that transfers the ubiquitin to a substrate protein, and this activity is RING domain dependent at N-termini [30]. Phylogenetic analyses suggest that finTRIM proteins appeared and diversified during the teleost evolution, while TRIM25 and TRIM39 (bty) are more ancient genes common to all vertebrates [25]. The sharing of B30.2 domains suggests a putative target-binding region between two major protein families involved in immune recognition of pathogen-specific motifs and expects the role of finTRIMs in antiviral innate immune system [24]. Our future work will be aimed at a further understanding of the biological functions of Ftr82, in addition to its vascular function; for example, to test if Ftr82 has a function in antiviral immunity. Interestingly, from our observation in this study, we found that the loss of ftr82 causes the defect of parachordal chain (PAC) formation in the lymphogenesis system at an early stage (Figure 3). In this aspect, ftr82 contributes to the establishment of the lymphatic network for immune defense in addition to innate response of its potential antiviral defense. Meanwhile, the molecular mechanisms that ftr82 control vascular patterning are unknown. Since Ftr82 shares the coiled-coil domain and B30.2 domains for protein–protein interaction, there might be other protein targets associated with the regulation of vascular patterning. This will be interesting to address in the future.

Alternatively, the RING finger domain of Ftr82 can serve as a putative interaction site with E2 ubiquitin-conjugating enzymes, and acts as E3 ubiquitin ligases. It will be interesting to test if Ftr82 shows E3 ubiquitin ligase activity. If yes, what is its substrate target? There are a couple of examples suggesting that E3 ubiquitin ligase functions in the regulation of vascular growth, such as Mib1 [31], Lnx1 [32], Fbxw7 [33], and Atrogin-1 [34]. Mindbomb homolog1 (Mib1) promotes endocytosis of the Notch ligands delta and Jagged. Loss of mib1 causes overgrowth of vessels due to the impairment of tip-stalk cell identity [31]. The E3 ubiquitin ligase Fbxw7 has been shown responsible for the polyubiquitination and proteasomal degradation of substrates such as Notch and c-Myc. Izumi et al. showed that Fbxw7 controls angiogenesis by degradation of active Notch activity in a mouse retina and a zebrafish embryonic trunk [33]. In addition to the regulation of Notch signaling by E3 ubiquitin ligase, VEGF signaling can be regulated by E3 ubiquitin ligase, such as Ftr82 in this study.

4. Materials and Methods

4.1. Zebrafish Line, Maintenance, and Chemicals Treatment

Zebrafish (Danio rerio) were raised and maintained in a 28.5 °C environment as previously described [28,35,36]. The regulations are approved by the National Sun Yat-sen University Animal Care Committee (approval reference #10231). Wild-type AB or TL fish and transgenic lines Tg (flk:GFP)^la116, Tg (kdrl:mCherry)^ci5, and Tg (fli1a:negfp)^y7 [37,38,39] were obtained from the Taiwan Zebrafish Core Facility at Academia Sinica. To prevent pigment formation, embryos were treated with 0.003% 1-Phenyl-2-Thiourea (PTU, Sigma, St. Louis, MO, USA) at six hours post-fertilization (hpf). DAPT (N-[(3,5-Difluorophenyl)acetyl]-l-alanyl-2-phenylglycine-1,1-dimethylethyl ester, Calbiochem, Nottingham, UK), SU5416 (Calbiochem), GS4012 (Calbiochem) or VEGF (Sigma) chemical treatment is for the modification of signaling pathways. A stock solution of 25 mM DAPT, 10 mM SU5416, 20 mg/mL GS4012, and 10 mM VEGF dissolved in dimethyl sulfoxide (DMSO) were diluted to the working concentration and added to embryos fish medium at 6 hpf. Control embryos were treated with 0.3% of DMSO.

4.2. Embryo Morpholino and mRNA Injection

Morpholinos (MOs) were obtained from Gene-Tools (Philomath, OR, USA). Morpholinos are designed to block translation by targeting the first codon of ftr82 (ftr82^atg MO) or to disrupt splicing of ftr82 (ftr82^e1i1 MO) by targeting the exon1-intron1 boundary and the sequences as follows: ftr82^atg MO sequence: 5′-TGGAGACATTTGCTCAGCCATGTTT-3′; ftr82^e1i1 MO sequence: 5′-GCGCTATGTTTTCCTTACCTGTTTT-3′; the morpholinos were injected with 2 and 3.4 ng, respectively. Unless specified, all images show the phenotype of ftr82 morphants (ftr82 MO) with 3.4 ng injection of splicing morpholino (ftr82^e1i1 MO). For rescue experiments, capped and polyadenylated mRNA of ftr82 was synthesized in vitro using the mMESSAGE mMASCHINE kit (Ambion, Austin, TX, USA). All injections were performed into one-cell zebrafish embryos by using a FemtoJet microinjector (Eppendorf AG, Hamburg, Germany).

4.3. RNA Extraction, cDNA Preparation, and Quantitative Real-Time RT-PCR (qPCR) Analysis

Total RNA from certain stages of embryos were prepared using the RNeasy kit (QIAGEN, Valencia, CA, USA) according to the manufacturer’s instructions. One microgram of total RNA was used to generate cDNA using a mix of oligo-dT primer (Roche Applied Science, Branford, CT, USA) and the reverse transcriptase (RTase, Roche) according to the manufacturer’s instructions. Real-time PCR was performed using the LightCycle 96 real-time PCR detection system and SYBR Green I Master (Roche). Real-time RT-PCR primers are listed in Table S1. Relative cDNA amounts were calculated using the LC96 program and normalized to the expression of β-actin based on the ΔΔC_t method. All reactions were performed as biological triplicates.

4.4. Morpholino Efficiency

The efficiency of morpholinos that cause mis-splicing or reduced band signals was determined using polymerase chain reaction (PCR) with primers franking on exon1 and exon 5 (Figure 2T) RNAs extracted from splicing morpholinos injected and uninjected embryos and reverse transcript to cDNA. Primers ftr82mo_f: 5′-TCCCTGTGACTTCTGCACTG-3′ and ftr82mo_r: 5′-CCTCTGACCTCATCCTCTCG-3′ were used to examine ftr82 knockdown efficiency, while primers GAPDH_f: 5′-TGCTGTAACCGAACTCATTGTC-3′ and GAPDH_r: 5′-CAAGCTTACTGGTATGGCCTTC-3′ were used as a loading control.

4.5. Whole-Mount In Situ Hybridization and Cryosection

Whole-mount in situ hybridization was followed by the high-resolution protocol based on [40,41]. An antisense in situ probe for ftr82 was obtained by PCR using primers ftr82_f: 5′-TGCAGCCGTACTCTGACAAC-3′ and ftr82_rT7: 5′-TAATACGACTCACTATAGTGTCCATCAATCATGCCTTC-3′ and in vitro transcription using T7 Polymerase (Roche) with digoxigenin-labeled UTP. The flk, stabilin, flt4, mrc1, and ephrinb2 probes have been described [11,42]. Briefly, embryos were fixed using 4% paraformaldehyde (PFA) and stored at −20 °C in methanol. After rehydration and permeabilization, the probes were added to embryos in hybridization buffer at 65 °C overnight, then washed and blocking with 1% bovine serum albumin (BSA). After incubation with AP-conjugated anti-Dig antibody, they were washed, and NBT/BCIP substrate (Roche) was added for color development. Embryos were then embedded in 3% methylcellulose for photographing. For cryosection, embryos were fixed with tissue-freezing medium Tek O.C.T (optimum cutting temperature) compound and sectioned at 10 µm using a Leica CM3050S cryostat and photographed on an IX71 inverted microscope (Olympus, Tokyo, Japan).

4.6. Imaging and Photo Processing

Embryos were mounted either in 3% methyl cellulose (Sigma) or 1.5% low melt agarose (Invitrogen, Philadelphia, PA, USA). Images were photographed with an AxioCam high-resolution camera and processed by AxioVision software (Carl Zeiss, Jena, Germany). For the confocal images, embryos were immobilized and embedded in 1.5% low-melting-point agarose with 5% tricaine (Invitrogen), and images were collected on a Zeiss LSM700 or Nikon Eclipse 90i C1 confocal microscope and processed with ImageJ software (NIH, Bethesda, MD, USA). Final figures were made using Adobe Photoshop.

4.7. TUNEL Assay

Embryos were fixed with PFA and stored in methanol. Embryos were rehydrated, digested with 10 µg/mL Proteinase K (Roche), and fixed again in 4% PFA for apoptotic cell detection by using a TdT-mediated dUTP-X nick end labeling (TUNEL) assay kit (Roche). Briefly, fixed embryos were treated with 3% hydrogen peroxide for one h at room temperature and washed to eliminate endogenous peroxidase. Then, 45 µL TUNEL label solution was mixed with 5 µL of TUNEL enzyme and added into the embryos for three hours at 37 °C. After reaction, embryos were washed to remove unincorporated nucleotides and blocked with 5% horse serum, then incubated with peroxidase (POD) conjugated anti-fluorescein antibody (Roche) overnight at 4 °C, washed four times in PBT, and visualized using a 3,3′-diaminobenzidine (DAB) colormetric kit (Roche).

4.8. Acridine Orange Staining

The dechorionated embryos were soaked in E3 fish medium containing 2 µg/mL acridine orange (Sigma) for one hour. After washing six times with fresh E3 medium, the embryos were mounted in 3% methylcellulose with 5% tricaine and photographed.

4.9. Protein Extraction and Western Blotting

Embryo lysates were prepared using RIPA lysis buffer (50 mM Tris-HCl pH 7.4, 1% NP-40, 0.25% sodium deoxycholate, 150 mM NaCl, 1 mM phenylmethylsulfonyl fluoride (PMSF), and protease inhibitors) and phosSTOP (Roche). The lysates were spun down, added with SDS sample buffer, and boiled for 10 min. The proteins were separated by SDS-PAGE and then transferred to a polyvinylidene fluoride (PVDF) membrane (BioRad, Hercules, CA, USA). The membranes were washed in TBS (Tris-buffered saline) with Tween-20 (TBST) and blocked with nonfat milk in TBST for 1 h. The membranes were then incubated with anti-phospho histone H3 (pHH3) (Millipore, Billerica, MA, USA), anti-p21, anti-p27 (Santa Cruz Biotechnology, Santa Cruz, CA, USA) polyclonal antibodies, and anti-β-actin monoclonal antibody (Sigma) overnight at 4 °C. The membranes were washed with TBST and incubated with a peroxidase conjugated secondary antibody (Cell Signaling Technology, Danvers, MA, USA) at room temperature for 2 h. The signal was detected by ECL chemiluminescence (Bio-Rad).

4.10. Anti-Phospho Histone H3 Staining

Embryos were fixed with PFA, permeabilized with pre-chilled acetone for 10 min, washed in phosphate-buffered saline (PBS) with 0.1% Tween-20 (PBSTw), and blocked with 10% normal goat serum (NGS) in PBS with 0.8% Triton X-100 (PBSTx) for 1 h at room temperature. Then, embryos were incubated overnight at 4 °C in the anti-phospho histone H3 polyclonal antibody (Millipore; 1:300 in 1% NGS/PBSTx). After washing four times for 10 min each in PBSTw, the embryos were incubated overnight with the secondary antibodies (anti-rabbit IgG peroxidase conjugated, 1:300 in 1% NGS/PBSTx). Embryos were washed five times and incubated with DAB peroxidase substrate (Roche). The color reaction was stopped by washing with PBSTw, fixed for 30 min in 4% PFA, washed in PBSTw, and cleared in 70% glycerol/PBSTw.

5. Conclusions

In summary, we identify teleost-specific ftr82 as a vascular gene that plays an important role for vascular development in zebrafish. We also showed that ftr82 likely interacts with VEGF and Notch signaling to control vascular patterning. It will be interesting to address how the molecular function of ftr82 acts in the future. Understanding this mechanism can provide the evolutionary difference of vascular development and immunity systems between mammals and fish.