microRNA-1 Regulates NCC Migration and Differentiation by Targeting sec63

Background/Aims: Neural crest cells play a vital role in craniofacial development, microRNA-1 (miR-1) is essential in development and disease of the cardiac and skeletal muscle, the objective of our study is to investigate effects of miR-1 on neural crest cell in the craniofacial development and its molecular mechanism. Methods: We knocked down miR-1 in zebrafish by miR-1 morpholino (MO) microinjection and observed phenotype of neural crest derivatives. We detected neural crest cell migration by time-lapse. Whole-mount in situ hybridization was used to monitor the expressions of genes involved in neural crest cell induction, specification, migration and differentiation. We performed a quantitative proteomics study (iTRAQ) and bioinformatics prediction to identify the targets of miR-1 and validate the relationship between miR-1 and its target gene sec63. Results: We found defects in the tissues derived from neural crest cells: a severely reduced lower jaw and delayed appearance of pigment cells. miR-1 MO injection also disrupted neural crest cell migration. At 24 hours post fertilization (hpf), reduced expression of tfap2a, dlx2, dlx3b, ngn1 and crestin indicated that miR-1 deficiency affected neural crest cell differentiation. iTRAQ and luciferase reporter assay identified SEC63 as a direct target gene of miR-1. The defects of miR-1 deficiency could be reversed, at least in part, by specific suppression of sec63 expression. Conclusion: miR-1 is involved in the regulation of neural crest cell development, and that it acts, at least partially, by targeting sec63 expression.


Introduction
Malformation caused by craniofacial dysplasia is a common birth defect that seriously affects human health. Most hard and soft craniofacial tissues are derived from the neural crest cells. The neural crest cells, a multipotent cell population, are induced at the neural plate border during neurulation [1]. After epithelial-mesenchymal transition, neural crest cells migrate along a specific route. Depending on extrinsic cues, they differentiate into various cell types and form structures and organs such as the nervous system, neuroendocrine cells, smooth muscle cells of the cardiac and pigment cells [2]. In the head, neural crest cells migrate from hindbrain rhombomeric regions to the pharyngeal arches where they form the facial cartilage, bone and connective tissue. Three distinguished major streams can be characterized: the mandibular, hyoid, and branchial streams [3]. Any effect on neural crest cells during development can lead to craniofacial defects [4][5][6]. microRNAs (miRNAs) are small, single-stranded noncoding RNAs of 18-25 nucleotides. They regulate gene expression at the posttranscriptional level by Ivyspring International Publisher base-pairing to partially complementary sites in the 3ʹ-untranslated region (3ʹUTR) of their target genes [7,8]. miRNAs may regulate up to 30% of genes, and genes regulated by miRNAs play an essential role in embryonic development [9]. In zebrafish, miR-1 is encoded by dre-miR-1-1 and dre-miR-1-2. The mature sequences of the resulting miRNAs are identical, i.e., 5ʹ-uggaauguaaagaaguauguau-3ʹ. In humans, miR-1 is encoded by hsa-miR-1-1 and hsa-miR-1-2, while their mature sequences are also the same, i.e., 5ʹ-uggaauguaaagaaguauguau-3ʹ. The sequences of mature miR-1 from different species, including zebrafish and humans, are identical. This degree of conservation suggests that they have similar functions in different vertebrates. miR-1 has been reported to show a tissue-specific expression pattern in zebrafish muscle, which is similar to that found in mammals [10].
The muscle-specific expression pattern of miR-1 is essential in cardiac and skeletal muscle development and disease. During cardiac development, miR-1 has been reported to control the balance between cell proliferation and differentiation [11]. In addition, miR-1 promotes myogenic differentiation and is involved in cell cycle regulation and migration processes. miR-1 also plays an essential role in Chinese sika deer-derived chondrocytes by directly targeting the 3ʹUTR of insulin-like growth factor 1 and by inhibiting the proliferation of these cartilage cells [12]. It has also been reported that miR-1 can inhibit cell proliferation and promote cell differentiation and apoptosis [13,14]. miR-1 and miR-133 form clusters on the same chromosomal loci of mouse chromosome 2 and chromosome 18. miR-1 promotes myogenesis by targeting histone deacetylase 4 [15]. Therefore, miR-1 plays a crucial regulatory role in embryonic development and disease. However, it is likely that many more functions of miR-1 are yet to be identified, and its function during neural crest development and pharyngeal cartilage formation remains unknown.
Based on the known role of miR-1 in embryonic development of vertebrates and the high degree of conservation of miR-1 during evolution, we hypothesized that miR-1 might be involved in embryonic development of neural crest-derived tissues in zebrafish through direct targeting of transcripts of proteins responsible for the differentiation of neural crest-derived tissues.

Animal maintenance
Wild-type zebrafish and the transgenic reporter strain Tg (sox10: EGFP) were raised on a 14 h/10 h light/dark cycle at 28.5 °C in the zebrafish facility of the Model Animal Research Center, Nanjing University. All work was performed with the approval of the Ethics Committee of the Stomatological School of Nanjing Medical University. All procedures were carried out according to the guidelines of the Animal Care Committee of Nanjing Medical University.

Alcian blue staining
Zebrafish embryos at 4 days post fertilization (dpf) were fixed in 4% paraformaldehyde in PBS before staining with Alcian blue (Sigma Chemical Co, St. Louis, MO, USA). Embryos were transferred to 30:70 glycerol/1% potassium hydroxide and then to 60:40 glycerol/1% potassium hydroxide before incubation for 2-3 days until they were sufficiently translucent. The key angle and length were measured as described previously [27,28].

Time-lapse imaging
Tg (sox10: EGFP) embryos were anaesthetized and embedded in 0.8% low melting point agarose (Takara Biotechnology, Dalian, China) [29]. Confocal stack pictures of the pharyngeal arch region were taken at the indicated time points using a META Zeiss 810 confocal microscope.

Isobaric tags for relative and absolute quantitation (iTRAQ)
Each sample for iTRAQ was composed of heads of at least 100 embryos. Embryos were anaesthetized at 24 hpf in 1× PBS on ice and stripped for heads. Total proteins were extracted from the heads of embryos. For iTRAQ labeling, approximately 100 μg of proteins were reduced and alkylated by 10 Mm dithiothreitol and 55 Mm iodoacetamide. And then, each sample was digested and labeled with iTRAQ reagents. The analysis was performed by the State Key Laboratory of Reproductive Medicine at Nanjing Medical University [30].

Western blot
50 embryos were transferred to sterilized centrifuge tubes at 24 hpf. The total proteins were extracted by the ultrasonic fragmentation method. Western blot analysis was conducted according to the previous publication [31]. Primary antibody recognizing Sec63 was purchased from Santa Cruz.

Luciferase reporter assay
The 3ʹUTR of sec63, which contains miR-1 binding site, was amplified by PCR with the following primers: F: 5ʹ-CACAACTCGAGCTGTCCTGCTGTT CCACAAAT-3ʹ and R: 5ʹ-TGAGGATCCTTGAGC CCTATTGCCATCCAC-3ʹ. The 3ʹUTR of pax7 was amplified with the following primers: F: 5ʹ-ATCGCTCGAGGAACCCGAGGTTTGTACG-3ʹ and R: 5ʹ-CACAACACAAGCGGCCGCATTCAAT TAATTCTGTCTTCA-3ʹ. The partial 3ʹUTR fragment was cloned into the pMIR-REPORT Luciferase vector (Promega, Madison, WI, USA). For use as a control, the binding site was mutated and then cloned into the same vector.
HEK293T cells (5 × 10 4 cells per well) were seeded in 24-well plates and cultured in DMEM/F12 containing 10% FBS (Gibco, Grand Island, NY, USA) overnight. After 24 h, dre-miR-1 mimics or negative control and wild-type or mutated reporter vector were transfected into HEK293T cells (duplex as a control). After 48 h, the cell cultures were split and the luciferase activity was determined according to the manufacturer's protocol (Promega, Madison, WI, USA). The luciferase activity of each type of sample was measured at least three times [32].

Statistical analysis
Data were presented as the mean ± standard deviation. Experiments were conducted separately at least three times. Student's t-test (SPSS 19.0) was used to assess differences between groups. P value < 0.05 was considered a significant difference.

miR-1 is expressed in zebrafish embryos
miR-1 is expressed in the pharyngeal region in zebrafish at 24 hpf [10]. To investigate whether miR-1 is expressed in zebrafish neural crest cells, sox10-positive (i.e., EGFP-expressing) cells were extracted and isolated by flow cytometry at 11 hpf, 24 hpf, 48 hpf, and 72 hpf from Tg (sox10: EGFP) embryos. qPCR analysis showed that miR-1 was expressed at 12 and 24hpf, and the highest level of expression was observed at 24 hpf, suggesting that miR-1 may play an important role in cranial neural crest cell migration and differentiation (Fig. S1).

Knockdown of miR-1 affects neural crest derivatives
Then we used an antisense MO oligonucleotide (miR-1 MO) to inactivate mature miR-1 (Fig. 1A). First, we observed morphological defects induced by miR-1 MO. At 3 dpf, there were fewer melanophores in the yolk sac and in the body of miR-1 morphants than in the negative control MO group (Fig. 1B). There were also fewer iridophores in the eyes and yolk sac extensions in the miR-1 morphants (Fig. 1C). These results indicate that normal expression of miR-1 is required for the development of neural crest-derived chromatophores.
We also found that the mandibular length was reduced and the mouth opening was ventrally rotated (Fig. 1D, E) in the miR-1 MO-injected group. Next, we analyzed the craniofacial morphology at 4 dpf by Alcian blue staining. In miR-1 morphants, the first pair of the cartilaginous pharyngeal arches (mandibular arches) was smaller than in negative controls, and the 3rd to 5th pairs were barely detectable. Simultaneously, the ceratohyal cartilage (second pharyngeal arch) was shifted posteriorly (Fig.  F). We measured the positional relationships among joints of the palatoquadrate and hyosymplectic cartilage (Meckel's cartilage) as well as the ceratohyal cartilage by analyzing three lines (Fig. 1F), which have been defined previously [27]. The statistical differences of the ratios (B:A and C:A) between the miR-1 MO-injected group and the negative control MO-injected groups are shown in Fig. 1G and H. These results indicate that knockdown of miR-1 expression affected craniofacial chondrogenesis. Therefore, we hypothesized that miR-1 might modulate the induction, migration or differentiation of cranial neural crest cells.

miR-1 deficiency does not affect neural crest induction and specification, but affects neural crest cell migration and differentiation
In order to determine the stage of neural crest development affected by miR-1 functions, we examined the expression of genes associated with specific stages of neural crest development in miR-1 morphants. A number of genes, including msxb, pax3, dlx3b, snail1b, foxd3, dlx2, ngn1, crestin and tfap2a have been shown to play important roles in the induction, specification, migration, and differentiation of neural crest cells [33,34].  We found that the neural crest specification marker foxd3 displayed normal expression in miR-1 morphants at 11 hpf ( Fig. 2A), approximately at the 3-somite stage (ss) before neural crest cell migration. Expression of the markers msxb, pax3, dlx3b, and snai1b in the miR-1 morphants was also normal (Fig.  2B, C, D, E). The normal expression indicated that miR-1 deficiency did not affect the induction and specification of neural crest cells.
Next, we analyzed whether neural crest cell migration was altered in miR-1 morphants. miR-1 MO was injected into Tg (sox10: EGFP) embryos. Then, we mapped sox10 expression during neural crest cell migration at 6 ss, 10 ss, 16 ss and 20 ss. At 6 ss, 10 ss, 16ss and 20 ss, in miR-1 morphants, neural crest cells displayed ectopic EGFP expression in the dorsal midline, while few neural crest cells appeared in the midline in negative control embryos (Fig. 3A, B, C, D). The expression of neural crest migration marker snai1b and tfap2a was also reduced in neural crest streams in the head (Fig. 3E, F). These data implicated miR-1 regulated neural crest cell migration.
To determine whether neural crest cell differentiation was altered, we examined the expression of associated genes. tfap2a controls the differentiation of cranial neural crest cells into mature chondrocytes. At 24 hpf, tfap2a has been reported to be expressed in a highly restricted pattern in the anterior domains [35,36]. We found that, compared with negative control embryos, tfap2a gene expression was reduced in miR-1 MO embryos, as can be most notably seen in the pharyngeal arches (Fig. 3F). At 24 hpf in miR-1 MO morphant embryos, the expression of dlx2a, a marker for the chondrogenic crest, was clearly reduced in the anterior domain of stream 1 and 2, which would give rise to the mandibular and hyoid arches, and dlx2a expression nearly disappeared in the more posterior domain of stream 3, which would migrate into the mandibular arches, the hyoid arches, and the posterior pharyngeal arches (Fig. 4A). Similarly, expression of dlx3b in the first and second arches was extremely reduced in the miR-1 MO embryos (Fig. 4B). The gene ngn1 is a critical marker for dorsal root ganglia development in zebrafish [37,38]. Loss of miR-1 resulted in notable reduction of ngn1 expression (Fig. 4C). At 24 hpf in negative control embryos, three distinct streams of migrating cranial neural crest cells can be detected by the expression of crestin in hindbrain rhombomeres. In miR-1 MO-injected embryos, the expression of crestin was nearly absent in the anterior streams and was reduced in the posterior neural crest stream (Fig. 4D). These data implicated miR-1 regulated neural crest cell differentiation.

miR-1 directly regulates sec63 expression
We analyzed the differential expression of proteins between miR-1 morphants and negative control embryos by iTRAQ analysis. Compared with negative control embryos, 32 proteins were significantly altered after knockdown of miR-1, including upregulation of 9 proteins and downregulation of 23 proteins (Fig. 5A, Fig. S2). Our results revealed that SEC63 was among the upregulated proteins after miR-1 knockdown, and we chose SEC63 for further analysis because of the reasons below. We found a significant difference in the level of sec63 Mrna and protein between the miR-1 morphants and negative control embryos at 24 hpf (Fig. 5B, C). We established that the gene sec63 was expressed in the very domain where the neural crest cells are found (Fig. 5D). Moreover, using the TargetScan search program (Release 6.2), miR-1 was found to be complementary to a RNA sequence from the sec63 3ʹUTR. We used a luciferase assay to test the functional interaction between miR-1 and the sec63 3ʹ-UTR in vitro (Fig. 5E, F, G). The results showed that the miR-1 precursor significantly repressed the expression of luciferase-sec63-3ʹUTR but not luciferase-sec63 3ʹ-mut-UTR (Fig. 5H). Together, these data indicate that miR-1 directly targeted the 3ʹUTR of sec63, thereby modulating the level of sec63 mRNA and protein.   dlx3b (B). ngn1, a marker of neuron differentiation, was downregulated in miR-1 morphants (arrows) (C). The expression of neural crest marker crestin was reduced in the cranial region and the branchial primordia (arrows) (D). microRNA-1, miR-1; morpholino, MO. Taken together, our data suggest that miR-1 loss of function interfered with the development of normal neural crest derivatives through disruption of neural crest cell migration and differentiation. Embryos were viewed laterally, with the anterior to the left. sec63-3ʹUTR binding site of miR-1 was predicted using TargetScan (E, F). (G) Mutant 3ʹUTR of sec63 in dual luciferase reporter plasmids. Overexpression of miR-1 (miR-1-pre) reduced sec63-3ʹUTR luciferase activity in vitro, but not the luciferase activity of mutated sec63-3ʹUTR. Three independent experiments were performed, and each experiment was carried out in duplicate (H).

Reducing sec63 partly restored the neural crest derivative defects in miR-1 morphants
We hypothesized that upregulation of sec63 was one of the likely causes for the defects of neural crest derivatives in miR-1-deficient embryos. To test this possibility, we knocked down sec63 using a MO oligonucleotide in the miR-1 morphants. Coinjection of miR-1 and sec63 MOs greatly reversed the delayed appearance of pigment cells (Fig. 6A). Alcian blue staining demonstrated that the defects in craniofacial cartilage were also corrected in coinjected zebrafish at 4 dpf (Fig. 6B). The reduced expression of dlx2, ngn1 and crestin was partly corrected at 24 hpf (Fig. 6C). Compared with miR-1 morphants, the number of neural crest cells showing ectopic expression of EGFP during the migration was reduced (Fig. 6D). We concluded that miR-1 regulates neural crest cell migration and derivation in zebrafish by targeting sec63.

Discussion
In this study, we discovered a novel role for miR-1 in neural crest cell s, pharyngeal cartilage, and lower jaw development in zebrafish embryos. Inactivation of miR-1 by antisense MO oligonucleotides led to marked phenotypic differences between miR-1 morphants and negative controls. The majority of the defects in the miR-1 morphants were in cells and tissues derived from the neural crest, including the heart, pigment cells, and craniofacial cartilage, confirming the essential role for miR-1 in neural crest cell development. miR-1 inactivation did not affect neural crest cell induction and specification because there was no major reduction in the number of premigratory neural crest cells at 11 hpf. Neural crest cells displayed ectopic localization in miR-1 morphants at 10 ss and 20 ss, representing early migration and late migration, respectively. At both 10 ss and 16 ss, the movement of cranial neural crest cells was delayed in the miR-1 morphants. Based on our findings, we conclude that the primary role of miR-1 in neural crest cells development is to regulate neural crest cell directional migration and differentiation. In addition, it is possible that miR-1 plays a role in neural crest cell apoptosis and proliferation, but this will require further study.
Several direct targets of miR-1 have been determined in vivo, including Hand2 and the Notch ligand delta in Drosophila. In addition, miR-1 regulates EDN1 in hepatocellular carcinoma progression [11,14,39]. Sera [40] predicted that pax7 might be the direct of miR-1. We carried out luciferase assay to test the functional interaction between miR-1 and the pax7 3ʹ-UTR in vitro (Fig. S3A). The results revealed that the miR-1 precursor significantly repressed the expression of luciferase-pax7-3ʹUTR but not luciferase-pax7 3ʹ-mut-UTR (Fig. S3B). However, coinjection of pax7 and miR-1 MOs could not reverse the delayed appearance of pigment cells and the defects in craniofacial cartilage. In order to gain insight into the molecular mechanism of the regulatory effect of miR-1 in neural crest development, we searched different databases for matched sequences between miR-1 and potential target genes in iTRAQ results. After confirmation of function by luciferase assays, we determined zebrafish sec63 as a direct target of miR-1.
SEC63 is a member of the endoplasmic reticulum (ER) translocon [41]. sec63 encodes an integral membrane protein of the ER that is highly conserved from yeast to humans. SEC63 is required in both post-translational and co-translational pathways [42]. Mutations in sec63 cause polycystic liver disease in humans, and even zebrafish sec63 mutants develop liver pathology. In zebrafish embryos, sec63 is expressed at all stages examined and in all tissues, and it is broadly expressed in liver and pancreas. Disruption of sec63 in zebrafish leads to abnormalities in myelinating glia in both the central and peripheral nervous system [43]. The function of sec63 during neural crest development and craniofacial cartilage formation has not been studied yet.
SEC63 is part of post-translational quality control and processing of polycystin-1 (PC1) and polycystin-2 (PC2) [44]. The polycystin complex has been implicated in MAPK/ERK and cAMP pathways [45]. Function of the PCs has been found in different cellular and tissue, is their role in regulation of cell migration and tissue morphogenesis [46]. When the expression of PCs was changed, it could lead to edema and cardiac defects in embryos [46]. And we also observe the edema and cardiac defects in miR-1 MO groups. Then we supposed that Sec63 was predicted target of miR-1 and responsible for proper neural crest cell migration and differentiation.
To determine whether sec63 expression is regulated by miR-1, we carried out rescue experiments. Embryos co-injected with both miR-1 MO and sec63 MO displayed relatively normal body length, head size, heart, and the appearance of pigment cells was also normal. Significantly, the lower jaw extended more anteriorly compared with miR-1 morphant embryos. And the Meckel's cartilage could extend anteriorly beyond (but close to) the eyes. Whole-mount hybridization analyses demonstrated that the expression was similar to that in negative control embryos at 24 hpf. Migration of neural crest cells in miR-1 MO and sec63 MO morphants was relatively normal. This partial effect of sec63 MO in miR-1 MO morphants confirms that sec63 is a target gene of miR-1. These results confirmed the effect of miR-1 on sec63 that was observed in the iTRAQ experiments. In addition, luciferase reporter assays confirmed that sec63 is a direct target of miR-1.
Our findings provide the first evidence that miR-1 regulates sec63 by directly targeting the 3ʹUTR of sec63. Taken together, our study suggests a molecular mechanism in which miR-1 participates in the control of sec63 expression and regulates neural crest migration and formation of Meckel's cartilage during embryonic development in zebrafish. These findings will hopefully help us to better understand craniofacial development and promote effective therapy and treatment for craniofacial malformations.