Syndecan-4 modulates the proliferation of neural cells and the formation of CaP axons during zebrafish embryonic neurogenesis

Syndecan-4 (Syn4), a single-pass transmembrane heparin sulphate proteoglycan (HSPG), plays significant role in the formation of focal adhesions and interacts with many growth factors to regulate cell migration and neural induction. Here, we show the new roles of syndecan-4(syn4) in zebrafish embryonic neurogenesis. Syn4 is broadly and dynamically expressed throughout the early stages of embryonic development. Knockdown of syn4 increases the expression of the marker genes of multiple types of neural cells. The increased expression of the marker genes is resulted from excessive proliferation of the neural cells. In addition, disrupting syn4 expression results in truncated and multiple aberrant branching of caudal primary (CaP) axons. Collectively, these data indicate that Syn4 suppresses the cellular proliferation during neurogenesis and is crucial for the formation of CaP axons during zebrafish embryogenesis.


Results
Spatio-temporal expressing patterns of syn4 during zebrafish embryonic development. The zebrafish was used as a model organism to investigate the function of syn4 in early embryonic development. The semi-quantitative reverse transcription-PCR (sqRT-PCR) and whole mount in situ hybridization (WISH) showed that syn4 was expressed from 1-cell stage to 72 hours post fertilization (hpf). Notably, the syn4 expressing level was higher at 12-somite stage. These results demonstrate that syn4 is both maternally and zygotically expressed during zebrafish embryogenesis (Fig. 1A).
WISH analysis revealed that the spatial expression of syn4 was dynamic during embryonic development. The syn4 was ubiquitously distributed before the gastrulation (Fig. 1B,C). Then zygotic syn4 mRNA highly expressed in the shield (Fig. 1D) and enriched in the thin evacuation zone and prechordal plate from the late gastrulation stage onward (Fig. 1E). Since the segmentation stage, syn4 was predominantly expressed in rhombomere, ventral diencephalon, midbrain, hindbrain and neural crest till 24 hpf (Fig. 1F-K). Then syn4 mRNA was restricted to the ventricular zone in the forebrain, posterior midbrain and hindbrain at 48 hpf (Fig. 1L,M). The detailed expressing patterns of syn4 are consistent with previous observations in Xenopus embryos and zebrafish embryonic brains 3,14,16 . The results indicate that syn4 has a highly dynamic expressing pattern and may play a role in the nervous development.
Syn4 regulates the migration of neural crest and muscle patterning. In the view of specific distribution of syn4 in the neural tissues, we tried to investigate the function of syn4 during zebrafish neural development. Syn4 antisense morpholino oligonucleotide (syn4 MO), which was previously reported to block the syn4 translation 17 , was injected into embryos at the one-cell stage. As syn4 is involved in neural crest (NC) migration 3 , we assessed whether the syn4 MO we used were able to induce defective NC migration by analyzing the expression of the NC markers crestin and sox10 and observed a strong inhibition of trunk NC migration in syn4 MO-injected embryos at 18 somite stage (18hpf) (supplemental Fig. S1C-F). The results indicate that syn4 MO is able to induce the defective NC migration as described t 3,14 . Given the role of Syn4 in muscle development 18,19 , we also performed the WISH to detect the expressing pattern of myod, which is a transcription factor that controls skeletal myogenesis 20 . The results showed that the expression of myod appeared abnormal curvature in the morphants (supplemental Fig. S2A-E). Fast and slow muscles in the trunk are stained with phalloidin, which recognizes actin fibers of both muscle types, at 18 somite stage (supplemental Fig. S3A-C") and 26hpf (supplemental Fig. S3D-E"). Injection of syn4 MO severely altered the morphology of muscle such that it appeared abnormally loose and wavy. Consistent with the previous observations 18, 19 , the results indicate that syn4 is involved in muscle development.
To test the efficiency and specificity of syn4 MO, we co-injected the syn4 MO with the pEGFP-N1-syn4 recombinant plasmid that contained the syn4 MO binding 5′ -UTR sequences and evaluated the efficiency by analyzing the fluorescence intensity of syn4-GFP fusing protein. As expected, the green fluorescence intensity was severely reduced in the co-injected embryos (supplemental Fig. S1A,B). Co-injection of syn4 mRNA without 5′ -UTR syn4 MO binding sequences was able to rescue the defective NC migration and muscle patterning (supplemental Fig. S1C-F). Altogether, these data demonstrate that the syn4 MO is able efficiently to block syn4 translation as described previously.
To further confirm the MO specificity, we design a series of truncated mutants based on a previous observations 21 (supplemental Fig. S4A). All of the mutant mRNAs were injected into zebrafish embryos. The results demonstrated the mutant lacking the C2 domain in plasma membranes (syn4-Δ C2) showed the phenotypes highly similar to the ones induced by the syn4 MO (supplemental Fig. S4B-J). The results clearly indicate that the ectopic expression syn4 Δ C2 mRNA is able to disrupt the syn4 functions in zebrafish embryos and the syn4-Δ C2 is disrupts syn4 function by acting as a dominant negative. Altogether, the results indicate the syn4 MO and the mutant syn4-Δ C2 are able to reveal the specific roles of syn4 involved in zebrafish embryonic development. The following results were mostly obtained from injection of syn4 MO in zebrafish embryos. The mutant syn4-Δ C2 was used to confirm the specificity of the phenotypes caused by syn4 MO in the following experiments.
Knockdown of syn4 expression increases neural progenitors. Since syn4 was expressed in neural tissues and required for neural crest migration, we then investigated whether syn4 was involved in the embryonic neurogenesis. The effect of down-regulated syn4 expression on neural stem cells and progenitor cells were examined by WISH with nestin 22 . The results showed that nestin expressing pattern was identical between the syn4 MO-injected and the control embryos at 24 hpf ( Fig. 2A,B). However, the expression level was significantly increased in the syn4 morphants at 48 hpf ( Fig. 2D-E"). Accordingly, sqRT-PCR showed the similar results (Fig. 2G,H). The syn4 mRNA was able to rescue the defective phenotypes ( Fig. 2F-F") caused by the syn4 MO and the dominant negative mutant syn4-Δ C2 mRNA (supplemental Fig. S4B-G"). The results show that syn4 negatively regulates nestin expression during embryonic neural development.
Loss of syn4 results in the increase of neuroglial cells. Since the neural stem cells/progenitor cells were regulated by Syn4, we were curious whether other neural cell types such as neuroglial cells were modulated by Syn4 as well. The expression of glial fibrillary acidic protein (gfap), an astrocyte marker, was not altered in the morphants and controls at 24 hpf (Fig. 3A,B), but up-regulated in the ventricular zone, especially in posterior midbrain and the spinal cord in the syn4 morphants by 48 hpf (Fig. 3D-E'). The syn4 mRNA was able to abrogate the increased gfap expression (Fig. 3F,F'). Next, we tested the expression of slc1a3a, also known as glast that expresses in radial glia-astrocyte lineage 23,24 , in embryonic neural tissues. The results showed that there was little difference between the controls and the syn4 morphants at 24 hpf ( Fig. 3G-I'). The expressing levels of slc1a3a were obviously increased in dorsal spinal cord in the morphants at 48 hpf ( Fig. 3J-K"). We then examined the expression of oligodendrocyte transcription factor 2 (olig2) in zebrafish embryos. Olig2 is a marker of Scientific RepoRts | 6:25300 | DOI: 10.1038/srep25300   oligodendrocytes and motor neuron (MN) precursors 25,26 . Not surprisingly, the olig2 expression was increased in the spinal cord in the syn4 morphants at 48 hpf (Fig. 3P-R"). The sqRT-PCR analysis confirmed the altering patterns of the expressing levels of gfap, slc1a3a, and olig2 in zebrafish embryos (Fig. 3S-W). Altogether, these data indicate that the neuroglial cells are increased when syn4 expression is down-regulated.
Knockdown of syn4 expression promotes the generation of neurons. Since syn4 modulated the number of neural progenitors and neuroglial cells, we further tested whether syn4 was involved in neuron specification. Thus, we examined the expression of elavl3, the earliest marker of pan-neuronal cells 27 , in zebrafish embryos. Knockdown of syn4 didn't alter the expression of elavl3 at 24 hpf (Fig. 4A,B). However, injection of syn4 MO resulted in a dramatic increased expressing levels of elavl3 in spinal cord compared with the controls at 48 hpf ( Fig. 4D-E").
To test the formation of motor neurons in syn4-depleted embryos, we analyzed the expression of islet1 and islet2a, which label primary motoneurons (PMNs), sensory Rohon-Beard neurons, and retina neurons. Rostral primary (RoP) and middle primary (MiP) MNs express islet1 but not islet2a, while CaP and variable primary (VaP) MNs express islet2 28  To gain further insight into the functions of syn4, we also analyzed another member of Syndecan family in zebrafish, syndecan2 (syn2) and found that syn2 mRNA was unable to rescue the phenotypes caused by the syn4 MO (supplemental Fig. S5A-F'). Taken together, the observations demonstrate that Syn4 modulates the embryonic development of neural stem cell/progenitors, neuronal glial cells, and neurons.

The proliferation of neural cells is increased by down-regulation of the syn4 expression.
To reveal how the neural cells were regulated by the syn4 expression, we performed immunohistochemistry staining of whole-mount embryos to detect the proliferating cell nuclear antigen (PCNA), which is a marker of cells with proliferative potential 29 . Statistical results revealed that the numbers of PCNA positive cells were increased in the syn4 morphants at 48 hpf. The phenotypes were fully rescued by the syn4 mRNA ( Fig. 5A-B",E). We also counted phospho-Histone H3 (Ser10), which is another marker to label the proliferative cells, positive cells in zebrafish embryos. The results showed that the number of phospho-Histone H3 (Ser10) positive cells was increased in the morphants at 48 hpf (Fig. 5C-D",F). The increased staining cells were more obvious in zebrafish embryonic brains (Fig. 5E,F, supplemental Fig. S6A-C'). Examination of the embryo sections also showed that the phospho-Histone H3 (Ser10) positive cells were increased in the 48hpf zebrafish embryos that were disrupted the Syn4 function (supplemental Fig. S6D-G). In addition, we also performed Two-Color Whole-Mount Staining with DAB and BCIP/NBT to identify whether neural cells, including neuroglial cells and motor neurons, carried increasing signal of phospho-histone H3 in zebrafish embryos after disruption of syn4 function. The results showed that proliferation of all tested types of neural cells was up-regulated in the morphants at 48 hpf ( Fig. 5G and data not shown). Taken together, our data indicate that Syn4 suppresses the proliferation of neural cells during zebrafish embryonic neurogenesis.
Syn4 is required for the formation of CaP axons. As the expression of several markers of motor neuron was altered in the syn4 MO injected embryos, we considered that syn4 might be associated with motor neuron development. There are two distinct classes of spinal motor neurons: PMNs and secondary motor neurons (SMNs) in zebrafish 30 . The PMNs are classified into CaP, MiP, and RoP MNs 31 . These PMNs can be identified by their stereotypical cell body positions and axon projection patterns. Specially, the CaP MNs is easy to be observed as their locates in the middle of each spinal cord and axonal projection extends ventrally 32 . Therefore, we selected CaP MNs as the target objects in Tg [hb9: GFP] ml2 transgenic zebrafish. The the expression of islet1/2a was not altered before 48hpf (supplemental Fig. S2F-K), however, 56% syn4-MO injected embryos showed abnormal PMNs with multiple aberrant branching axons and truncated axons at 26 hpf of the zebrafish embryos ( Fig. 6A-B' ,E). To assess whether overexpression of syn4 was able to disturb the outgrowth of CaP axons, we injected syn4 mRNA at one-cell stage and quantitated the number of PMNs with defective phenotypes at 26 hpf. The result showed that 53% syn4 mRNA-injected embryos had shortened CaP axons and/or abnormally branched axons (Fig. 6C). The length of CaP axon outgrowth in the morphants averaged 79.9 um, which was significantly reduced, compared to the control embryos's CaP axons that averaged 100.2 um (Fig. 6F,G). In addition, the averaged branch number of per CaP axon in syn4 MO-injected embryos was 4.71, while control embryos carried only 3.57 branches in average (Fig. 6F,H). The syn4 mRNA co-injection was able to rescue the defective CaP axon morphology (Fig. 6D). These results indicate that accurate syn4 expression is crucial.

Discussion
Previous observations have shown that Syn4 is an element of the PCP non-canonical Wnt signaling pathways to regulate the neural tube closure and neural crest migration during Xenopus and zebrafish embryonic development 3,14 . Syn4 also modulates the neural induction through the ERK or PKC-dependent signaling pathways 15 . Here, the new role of syn4 is revealed in our works. We identify that the neural cells are prominently increased after the inhibition of syn4 during the early stage of zebrafish embryonic neurogenesis. Further evidences reveal that syn4 inhibits the cell proliferation during the development of neural system and maintains motorneuron morphology.
The prior data have indicated that Syn4 has a positive effect on cell proliferation by regulating PKCα activation to mediate growth factor-stimulated proliferation in the wounded adult tissue and adult skeletal muscle cells 2,33-35 . Syn4 has negative effects on the cell proliferation in multiple tumor cell lines such as breast carcinoma  (A-F' ,G-L' ,M-R'). Dorsal view, anterior to the left in (D"-F",J"-L",P"-R").  and glioblastoma cells. In the tumor cells, the ability of Syn4 to bind to the fibronectin is competitively blocked by tenascin-C in integrin signaling, resulting in the proliferation of tumor cells 36 . Since Syn4 is indispensable to the cell spreading by interaction with the HepII site in fibronectin, disrupting the interaction between Syn4 and FNIII13 results in cell spreading obstacle 37,38 . The data suggest that there exists a mechanism that fibronectin signaling attenuates cell proliferation by Syn4 36 . In consistent with the results obtained from the tumor cell lines, we find that Syn4 inhibits the proliferation of neural cells. Therefore, further work is required to determine whether fibronectin is involved in the inhibition of neural proliferation during embryonic neural development.
The extension of axons and dendrites to form stereotypic neuronal connections is crucial to neural development. It has been reported that the biosynthesis of heparan sulfate (HS) is required for normal axon elongation and branching in animals, including mouse, C. elegans and zebrafish. HSPGs, the carrier proteins of HS, are involved in properly axon architecture. Syndecan-3 and glypican-2 are shown to be required for axon guidance 5,39 . However, little is known about the role of syn4 in the architecture of axon. We identify that perturbed expression of syn4 results in aberrant elongation and abnormal additional branching of CaP axons. The C2 domain of Syn4 is essential for the function of syn4 in the formation of CaP axons. These evidences is similar to the phenotype that defects in the extension and targeting of axons caused by abnormal FGF signaling, suggesting that Syn4 may regulate the axon architecture via the FGF signaling in the zebrafish 5,40,41 . Future experiments are required to explore the factors that are combined with syn4 in regulating CaP axons formation.
Our results show that muscle patterning is altered and the morphology of muscle severely altered in the syn4 morphants since 18hpf. The axons of motor neuron show aberrant elongation and additional branching in the morphants at 26hpf. In view of the axon of primary motor neuron start to project to the myotomes at 18hpf, while the myotomes have formed before the time point 42,43 . Our results suggest that the early defective muscle patterning induced by knockdown of Syn4 functions may be independent of the proliferation of the neural cells.
The results also suggest that the early defective muscle patterning may contribute to defective axon branching and length in the syn4 morphants. However, further experiments are required to address the question. Syn2 and syn4 have a high degree of structural similarity 44 . However, we identify that overexpression of syn2 is unable to rescue the phenotypes caused by knockdown of Syn4 functions in present works. The results indicate that Syn4 specifically modulate the proliferation in neural cells and axon patterning of motor neurons and are consistent with previous observations that Syn2 and Syn4 are engaged in a number of distinct developmental processes 3,45,46 .

Methods
Zebrafish maintenance and strains. All experiments performed on zebrafish (Danio rerio) were according to standard procedures 47 . Embryos were staged as described and raised at 28.5 °C in the same system as adults (Aquatic Ecosystems) 48 . The Tg[hb9:GFP] ml2 transgenic lines 49 and AB strains were used in our studies.
All zebrafish experiments were performed in accordance with the guidelines of the animal ethical committee of West China Hospital. All experimental protocols were proved by the Animal Ethical Committee, West China Hospital, Sichuan University. Syn2/4 cloning. Syn2/4 cDNA was amplified from zebrafish embryos using the primers shown in the supplemental Table S1 and subcloned into the pGEM-T easy vector to synthetize antisense probes for in situ hybridization. The coding region of syn2/4 was subcloned into the pcDNA3.1+ vector for mRNA synthesis. The syn4 primer1 resultant PCR products (with 5′ -UTR sequence) were ligated into pEGFP-N1 vector.
Construction of syn4 mutants. Five mutants of zebrafish syn4 were designed based on a previous observations 21 and generated by PCR. The primers were shown in the supplemental Table S1. The C-terminal domain deletion mutants included Δ C1, Δ C2, Δ V and Δ C1-2. The N-terminal domain deletion mutant Δ Ec lacks the extracellular domain.

Total RNA isolation and Semi-quantitative Reverse transcription-PCR (sqRT-PCR) and
Statistical analysis. Total RNAs of an approximate 100 embryos at different stages, including 1-cell, K-cell, 75% epiboly, 12-somite, 24 h, 28 h, 48 h, 72 h post-fertilization (hpf) were isolated. Each reaction used 20μ l total RNA with the Prime Script Reverse-transcription PCR kit (TaKaRa DRR014A). The resultant cDNA were used as the templates for PCR. The primer pairs used for sqRT-PCR analysis were shown in the supplementary (supplemental Table S1). Statistical analysis was performed using Student's unpaired t-test. Differences were considered significant for p < 0.05.
In vitro transcription of Antisense RNA probes and mRNAs. Antisense RNA probes for in situ hybridization were synthesized in vitro following manufacturer's instructions of Riboprobe system kit (Promega). Plasmid was linearized and the probes syn4, elavl3, gfap, islet1, islet2a, olig2, slc1a3a and nestin were synthesized. Capped syn4 mutant mRNA and full-length mRNA that lacked the morpholino sequence were synthesized using the T7 Mmessagem MACHINE kit (Ambion). Antisense morpholino and mRNA or plasmid injections. Antisense morpholino oligonucleotides (MO) against syn4: TGAGGTAAACTTTCAACAT CTTCTC were purchased from Gene Tools (Philomath, OR, USA) and used as previously described 17 . The mRNAs and MO were injected into the yolk and the plasmid was injected into the cell at the one cell stage. Unless stated otherwise, a volume of 1 nl was injected into embryos with the concentration of 6 ng/nl of MO, 100-150 ng/ul of mRNA and 100-200 ng/ul of plasmid.

Immunohistochemistry Staining of Whole-mount Embryos and Statistical analysis.
Immunohistochemistry was performed using these sections as previously described 51 , the following primary antibodies were used: mouse anti-PCNA (Calbiochem 1:500) and rabbit anti-phospho-Histone H3 (Ser10) (Millipore 1:500). We counted the positive cells at the brain and the whole body of each fish, and calculate the average to compare the difference among groups. Statistical analysis was performed using Student's unpaired t-test. Differences were considered significant for p < 0.05. Immunofluorescence staining. Immunofluorescence staining was performed as previously described 52 using the following antibodies: rabbit anti-phospho-Histone H3 (Ser10) (Millipore 1:200), phalloidin (Sigma 1:1000). In brief, Zebrafish embryos were fixed in 4% paraformaldehyde at 4 °C overnight, washed three times in PBS for 5 min each wash, then incubated in acetone at −20 °C for 7 min and washed three times in PBS for 5 min each wash. Blocking the embryos with 5% BSA and dilute the antibodies with blocking solution used for immunostaining at 4 °C overnight.
For cross-section samples, embryos were equilibrated in 30% sucrose prepared in PBS and embedded in optimal cutting temperature compound at −20 °C. Sections with a thickness of 8-10 um.