Transforming growth factor‐β signal regulates gut bending in the sea urchin embryo
Abstract
During gastrulation, one of the most important morphogenetic events in sea urchin embryogenesis, the gut bends toward the ventral side to form an open mouth. Although the involvement of transforming growth factor‐β (TGF‐β) signals in the cell‐fate specification of the ectoderm and endoderm along the dorsal–ventral axis has been well reported, it remains unclear what controls the morphogenetic behavior of gut bending. Here, using two sea urchin species, Hemicentrotus pulcherrimus and Temnopleurus reevesii, we show that TGF‐β signals are required for gut bending toward the ventral side. To search for the common morphogenetic cue in these two species, we initially confirmed the expression patterns of the dorsal–ventral regulatory TGF‐β members, nodal, lefty, bmp2/4, and chordin, in T. reevesii because these factors are appropriate candidates to investigate the cue that starts gut bending, although genetic information about the body axes is entirely lacking in this species. Based on their expression patterns and a functional analysis of Nodal, the dorsal–ventral axis formation of T. reevesii is likely regulated by these TGF‐β members, as in other sea urchins. When the Alk4/5/7 signal was inhibited by its specific inhibitor, SB431542, before the late gastrula stage of T. reevesii, the gut was extended straight toward the anterior tip region, although the ectodermal dorsal–ventral polarity was normal. By contrast, H. pulcherrimus gut bending was sensitive to SB431542 until the prism stage. These data clearly indicate that gut bending is commonly dependent on a TGF‐β signal in sea urchins, but the timing of the response varies in different species.
1 INTRODUCTION
In bilaterians, gastrulation is required to form functional bodies composed of three germ layers: ectoderm, mesoderm, and endoderm; and Lewis Wolpert notes that gastrulation is the most important event in our lives (Lewis, 2008). In the process of gastrulation in deuterostomes, the invagination forms a blastopore, generally a future anus, and the mouth opens later on the opposite side of the body. The position of the mouth in some deuterostomes is independent of gut formation because the stomodeum can be formed within the ventral ectoderm by itself without any endomesodermal structures in those animals (e.g., Wikramanayake & Klein, 1997). In most vertebrates, the invaginated gut is extended toward the anterior at the ventral side of the body and fuses to the ventral ectoderm to open the mouth around the base of the future head region. Because the gut in most vertebrates is at the relatively ventral side of the body from the beginning of invagination, only the extension toward the anterior is required to make “ventral” structures. However, in some deuterostomes, such as in sea urchins, that have a large blastocoel, the gut is initially extended toward the anterior in the middle axis of the body, and then, the tip of the gut has to bend toward the ventral side to form the “ventral” gut.
Gastrulation of sea urchin embryos is composed of primary and secondary invaginations. The primary invagination is driven by the apical constriction of the vegetal plate cells. The highly dense actin bundle at the apical side of vegetal plate cells produces the force to constrict the cell surface and unbalance the area of the cell surface between the apical and basal sides; the apical surface is smaller than that of the basal side; therefore, the vegetal plate is kinked into the blastocoel (reviewed in Kominami & Takata, 2004). The secondary invagination leads to the elongation of the gut, which is managed with the pulling force by secondary mesenchyme cells and with conversion extension (Kominami & Takata, 2004). Although these two steps of gastrulation are well studied, the mechanism of how the invaginated gut is fused to stomodeum to form the open mouth has not been well described. During the early gastrula, the gut elongates straight toward the anterior end of the body, but the tip of the gut reaches to the stomodeum located at the ventral ectoderm at the end of gastrulation; thus, the gut bends toward the ventral side during the process of gastrulation (Figure 1). This phenomenon is conserved among some sea urchin species, at least in Hemicentrotus pulcherrimus and Temnopleurus reevesii, on which we focus in this paper. Because the gut in sea urchin embryos never bends in the wrong direction, such as to the dorsal side, the ventral ectoderm is expected to attract gut cells to proceed with the extension toward the mouth region. The dorsal–ventral patterning of the endomesoderm is regulated by TGF‐β superfamily members, such as Nodal and BMP2/4, in sea urchin embryos (Duboc et al., 2010), suggesting that the ventral attractant is among these factors. However, to date, no evidence indicates that TGF‐β factors are involved in the morphogenetic regulation of gut bending.
In Japanese developmental biology using sea urchins, H. pulcherrimus has been the prominent model organism for decades. In 2015, however, our group reported that T. reevesii could be another model sea urchin that can be used in developmental biology because of its rapid development, tolerance of high temperature and high transparency (Yaguchi et al., 2015). In fact, T. reevesii is more beneficial for observing neurogenesis than H. pulcherrimus, which has a relatively longer developmental time to reach the neurogenic larval stages. Thus, because of the mutual compensation for their disadvantages and benefits, we use these two sea urchins in laboratory experiments to demonstrate the biological mechanisms that are shared beyond the differences in species. Here, with the advantage of the availability of the embryos of the two sea urchin species, we show that regulation by TGF‐β signals is the common mechanism for gut bending in sea urchin embryos.
2 MATERIALS AND METHODS
2.1 Animals and embryo culture
Adult sea urchins H. pulcherrimus and T. reevesii were collected around the Shimoda Marine Research Center (University of Tsukuba, Shizuoka, Japan) and Marine and Coastal Research Center (Ochanomizu University, Chiba, Japan). The gametes were collected by intrablastocoelic injection of 0.5 mol/L KCl. The embryos of H. pulcherrimus and T. reevesii were incubated at 15 and 22°C, respectively, and cultured in glass beakers or plastic dishes filled with filtered natural seawater (FSW) containing 50 μg/ml of kanamycin (Nacalai, Kyoto, Japan). In FSW, the final concentrations were 5.0 µmol/L and 0.5 mmol/L for SB431542 (Merck, Darmstadt, Germany) and NiCl2, respectively. For the negative control of SB431542, we added the same volume of dimethyl sulfoxide (DMSO; Merck) into FSW. The timing of the application of these reagents is described in the text.
2.2 Cloning of TGF‐β family genes
Full‐length of Nodal (LC369130), Lefty (LC369131), Bmp2/4 (LC369132) and Chordin (LC369133) cDNA were obtained from T. reevesii cDNA by PCR using KOD FX DNA Polymerase (TOYOBO, Tokyo, Japan). Based on T. reevesii transcriptome sequence, the detail of which will be published elsewhere, the following primers were designed and used:
- Tr‐Nodal‐F1: 5′‐TCGATTATCGATATGCATCGTTTAGCCGAC‐3′ (underline shows ClaI site),
- Tr‐Nodal‐R1: 5′‐GTGATTCTCGAGGCGGTATGGTATTCGAAC‐3′ (underline shows XhoI site),
- Tr‐Lefty‐F1: 5′‐TCGATTGGATCCATGGAATTATCACTCGGC‐3′ (underline shows BamHI site),
- Tr‐Lefty‐R1: 5′‐GTGATTCTCGAGCGCCATGCATCCGCAATC‐3′ (underline shows XhoI site),
- Tr‐Bmp2/4‐F1: 5′‐TCGATTGAATTCATGGTTACCACCACACAC‐3′ (underline shows EcoRI site),
- Tr‐Bmp2/4‐R1: 5′‐GTGATTCTCGAGCTACCGGCACCCACAGCC‐3′ (underline shows XhoI site),
- Tr‐Chordin‐F1: 5′‐CATCGATTCGAATTCATGTATCGTGCTGTG‐3′ (underline shows EcoRI site),
- Tr‐Chordin‐R1: 5′‐TTCTAGAGGCTCGAGTTACGAGAGCTTCTC‐3′ (underline shows XhoI site).
Obtained cDNA were inserted into the pCS2+ vector and used for the following experiments.
2.3 Microinjection, whole‐mount in situ hybridization and immunohistochemistry
Detailed methods for microinjection, whole‐mount in situ hybridization, and immunohistochemistry are previously described (Yaguchi, Takeda, Inaba, & Yaguchi, 2016). The microinjection reagent was modified from 24% glycerol‐DW to 24% glycerol‐HK buffer (40 mmol/L HEPES, pH 8.0, 120 mmol/L KCl) for H. pulcherrimus, and glycerol was removed from the injection reagent for T. reevesii. The concentration and the sequence of Nodal morpholinos were as follow:
Anti‐serotonin (Sigma‐Aldrich, St. Louis, MO, USA), anti‐phosphorylated‐Smad1/5/8 (pSmad1/5/8; Cell Signaling Technology, Danvers, MA, USA) and anti‐pSmad2/3 (Abcam, Eugene, OR, USA) antibodies were diluted to 1/2,000, 1/1,000 and 1/100, respectively.
2.4 Quantitative polymerase chain reaction (qPCR)
Quantitative polymerase chain reaction was performed as previously described (Ransick, 2004; Yaguchi, 2010) with some modification. Total RNA from T. reevesii embryos were purified using ISOGEN (Nippon Gene) and cDNA was synthesized with the reverse transcriptase SuperScript III (Thermo Fischer Scientific). GoTaq qPCR Master Mix (Promega) was used for PCR with a Thermal Cycler Dice Real Time system (Takara). The primer sets used for qPCR are as follows:
- Mitochondrial COI RNA‐qF1: 5′‐CCGCATTCTTGCTCCTTCTT‐3′,
- Mitochondrial COI RNA‐qR1: 5′‐TGCTGGGTCGAAGAAAGTTG‐3′,
- Nodal‐qF2: 5′‐GCATCGTTTAGCCGACATCT‐3′,
- Nodal‐qR2: 5′‐GGTCGATGATTTCGACTCGT‐3′,
- Lefty‐qF2: 5′‐GCATTCGTCGTGCATACATC‐3′,
- Lefty‐qR2: 5′‐GAGTTCGGCCATGATGATCT‐3′,
- BMP2/4‐qF1: 5′‐CTTGTATGCTGCGTTGTCGT‐3′,
- BMP2/4‐qR1: 5′‐GATCGCCCTCGTTATGGTTA‐3′,
- Chordin‐qF1: 5′‐ACAACCTTGAGGACGAATGG‐3′, and
- Chordin‐qR1: 5′‐CCAGCAGGACACCTTCAA AT‐3′.
Relative amounts of each mRNA were normalized with mitochondrial COI Ct values.
3 RESULTS AND DISCUSSION
3.1 Temnopleurus reevesii embryos can be a comparative sea urchin model for studying the formation of dorsal–ventral polarity
Based on previous works using four different sea urchin species, Strongylocentrotus purpuratus, Paracentrotus lividus, Lytechinus variegatus, and H. pulcherrimus, the dorsal–ventral axis formation is regulated by the combinational functions of Nodal, Lefty, BMP2/4, and Chordin, which are all expressed at the ventral ectoderm during early embryogenesis of sea urchins (Bradham et al., 2009; Coffman, McCarthy, Dickey‐Sims, & Robertson, 2004; Duboc, Röttinger, Besnardeau, & Lepage, 2004; Flowers, Courteau, Poustka, Weng, & Venuti, 2004; Nam et al., 2007; Yaguchi et al., 2016). This observation suggests that embryos of T. reevesii also use the same system to establish the dorsal–ventral polarity because of the similar shapes of the embryonic/larval body. However, expressions of these four genes have not been reported in this species; therefore, the expression and function of these genes during embryogenesis required initial verification. To investigate the spatial expression patterns of these four TGF‐β members, we performed in situ hybridization analyses using early developmental stage embryos of T. reevesii from 8 to 18 hr with 2 hr intervals.
nodal was expressed at one side of 8 hr embryos, with the expression continuing until at least 18 hr (Figure 2), and based on previous reports using other sea urchin species (Duboc et al., 2004; Saudemont et al., 2010), this expression was likely at the ventral ectoderm. The other three gene expression patterns were very similar to that of nodal, suggesting that these genes were also expressed at the ventral ectoderm (Figure 2a). In fact, in 18 hr embryos, in which the dorsal–ventral axis becomes prominent based on the position of PMC ventrolateral clusters (Duloquin, Lhomond, & Gache, 2007), those genes are clearly expressed at the ventral side. To confirm that the expression of these four genes were localized at the ventral ectoderm, we used NiCl2 to ventralize sea urchin embryos (Hardin, Coffman, Black, & McClay, 1992) and checked the expression of the genes. As expected, all of the genes were expressed radially to cover the entire ectoderm, except for the animal plate (anterior neuroectoderm) (Figure 3I–P). Because the initiation of the expression of these genes remains uncertain (Duboc et al., 2004; Range et al., 2007), we attempted to perform in situ hybridization of TGF‐β members before 8 hr, but the spatial expression patterns were not reproducible in individuals, and therefore, the precise expression patterns could not be determined. qPCR data supported this since each batch showed different temporal patterns with low amounts of mRNAs (Figure 2b). Based on in situ hybridization and qPCR, we concluded that the gene was not maternally expressed, but by 8 hr, the expression had begun.
To investigate whether Nodal was on top of the hierarchy of dorsal–ventral gene regulation of these four genes in T. reevesii, as in other sea urchins (e.g., Duboc et al., 2004), we investigated the gene expression patterns of the four genes in Nodal‐deficient embryos. First, we blocked the Alk4/5/7‐mediated TGF‐β signaling pathway using SB431542, which is a specific inhibitor of TGF/Activin type IB receptor (Inman et al., 2002). In SB431542‐treated blastulae from 1 hr post‐fertilization, nodal was not expressed (Figure 3Q,R), whereas control embryos expressed the gene normally (Figure 3A,B). Expression of lefty, bmp2/4, and chordin was also not detected in the absence of Alk4/5/7 activity (Figure 3C–H,S–X). Blocking the Alk4/5/7 receptor, SB431542 inhibited multiple TGF‐β family members, including Nodal and TGF‐β. Therefore, to focus on the specific function of Nodal, we used a morpholino oligonucleotide against nodal (Nodal‐MO) to block its translation and then determined the expression of the four genes. Several works have reported on the radialized morphology, significant decrease of ventrally expressed TGF‐β family genes, and radially distributed serotonergic neurons in Nodal morphants in other species, such as in H. pulcherrimus (Figure 4b,d,h,j,l) (Yaguchi et al., 2016), and a similar radialized phenotype, lack of gene expressions (Figure 3Y–f), and serotonergic neural patterns (Figure 4a,c,g,i,k) were obtained in T. reevesii by injecting Nodal‐MO; thus, the morpholino used in this experiment was supposed to be sufficiently specific to block translation of nodal, which showed that the Nodal‐deficient effect was conserved beyond species. In Nodal morphants, nodal and the other three TGF‐β genes were not expressed in blastula stages (Figure 3Y–f), which indicated that the expression of all four genes depended on Alk4/5/7 activity and the ligand was likely Nodal. Additionally, because Nodal morphants lost the expression of all of the genes, Nodal was at the top of the hierarchy of the signaling pathway involving these TGF‐β members, as reported in other species (Duboc et al., 2004; Saudemont et al., 2010). Collectively, nodal, lefty, bmp2/4, and chordin were expressed at the ventral ectoderm of T. reevesii embryos, and Nodal was required for their expression as in other sea urchin species (Bradham et al., 2009; Coffman et al., 2004; Duboc et al., 2004; Flowers et al., 2004; Nam et al., 2007; Yaguchi et al., 2016).
Although all of these four TGF‐β members are diffusing, notably, the expression site of their genes were well conserved among sea urchin species (Figure 2), which indicates that the control system of the diffusion of Nodal, the top molecule in the regulatory hierarchy, is precisely organized in sea urchin groups (Duboc, Lapraz, Besnardeau, & Lepage, 2008; Nam et al., 2007; Range et al., 2007). Nodal may diffuse only a few cell diameters in S. purpuratus embryos (Yaguchi, Yaguchi, & Burke, 2007), and Lefty, which can diffuse farther than Nodal, blocks the extra diffusion of Nodal at the future ciliary band region in P. lividus and S. purpuratus (Duboc et al., 2008; Yaguchi, Yaguchi, Angerer et al., 2010; Yaguchi et al., 2010). Based on four TGF‐β RNA expression patterns localized only at the ventral region, it is suggested that Nodal‐Lefty antagonism might function similarly in T. reevesii although we did not show the detailed expression patterns like using multi‐color in situ hybridization. Therefore, the further analyses in future like the spatial expression profiling of four genes or the molecular interaction of both proteins might help us to understand Nodal‐Lefty antagonism in T. reevesii embryos. According to the spatial expression data of nodal, we could not confidently conclude where the initial nodal was expressed. Because the results of previous reports using other species also remain debatable (Duboc et al., 2004; Range et al., 2007), spatial regulation might be unstable only at the beginning of gene expression, which could also be attributed to the faint amount of mRNA at the beginning and lack of a protein detection tool, such as an anti‐Nodal antibody, for sea urchin immunohistochemistry. Because of these points, the specific time when the first protein begins to function is difficult to determine. Alternatively, although anti‐pSmad2/3, an intracellular mediator of the Nodal signal, can be one of the indicators of when and where the Nodal protein functions, unfortunately, the pSmad2/3 antibodies we used in this paper (Figure 6) did not work on T. reevesii because of the difference in amino acid sequence at the C‐terminal region. Thus, the understanding of the initial Nodal pathway functions in this sea urchin species must wait until useful antibody reagents are developed.
The expression and the functional data for TGF‐β family members in the early developmental stages of T. reevesii indicated that the gene regulatory network driving the dorsal–ventral axis formation was similar to that in other indirect developer sea urchin species. Combined with a previous report (Yaguchi et al., 2015), this result indicated that this species could be one of the model sea urchins in the study of developmental biology, particularly in analyzing axis formation during early embryogenesis. Currently, the identification of common features or biological phenomena and analyzing the detailed molecular mechanisms to show reproducible results is an important process. In this process, using different species in the same family is highly useful because when the same phenomena are identified or the same experimental results occur, the indication is that the sameness extends beyond the species differences and is likely a shared feature in the sea urchin family. For the examples above, the gut was straightly invaginated at the very beginning of the gastrula stage but bent toward the mouth in both H. pulcherrimus and T. reevesii by the prism stage (Figure 1). The question in this case was what controls the morphogenetic behaviors of the sea urchin gut, which we could consider based on the results obtained from the experiments using two different species. With this strategy, a stronger conclusion could be reached than that when using a single species.
3.2 Alk4/5/7 signal is required for gut bending to the ventral side in the sea urchin embryo
To investigate the involvement of TGF‐β members in gut bending in sea urchin development, we first blocked translation of the dorsal–ventral master gene, nodal. However, because the morpholino injection interfered with the function of Nodal from the beginning, completely radialized embryos were produced in which the ectoderm lost dorsal–ventral polarity and the gut elongated straight toward the anterior pole in both T. reevesii and H. pulcherrimus, as reported previously in other species (cf. Figure 4c,d with a,b) (Duboc et al., 2004). These results were consistent with those of the Alk4/5/7 inhibitor SB431542 treatment, by which TGF‐β, including Nodal, activities were blocked at the reception level (Figure 4e,f). Because these embryos completely lost the secondary body axis before invagination, we could not determine the effect of TGF‐β signals on gut bending. Therefore, we temporarily treated embryos with SB431542 to block the function of Alk4/5/7 during a certain period. Because the TGF‐β‐dependent specification of dorsal–ventral ectoderm is completed by the gastrula stage (Duboc, Röttinger, Lapraz, Besnardeau, & Lepage, 2005), we applied SB431542 from the early‐gastrula until the early pluteus stage. In SB431542‐treated T. reevesii, in which the dorsal–ventral organization of the ectoderm was morphologically normal, the gut did not precisely bend toward the mouth region but did elongate to the animal pole region, whereas the gut bent to the mouth region normally in control embryos (cf. Figure 5f with a). Because the alkaline phosphatase activity in the mid‐ and hindgut was normal in SB431542‐treated embryos and similar to that in the controls (Figure 5b,g), the specification/differentiation of the gut cells was not severely affected, but gut bending was blocked by Alk4/5/7 inhibition (Figure 5e,j). Similarly, in H. pulcherrimus, the gut elongated straightly toward the animal pole and not toward the mouth region with SB431542 treatment (Figure 5c,d,h,i).
To quantify how much the gut bends in Alk4/5/7‐deficient embryos and to estimate when the TGF‐β signal affects the gut, we measured the angle between the tip of the gut and anus in embryos treated with SB431542 at various stages (Figure 5m). In T. reevesii, the data from two independent batches indicated that the gut did not precisely bend to the mouth when inhibition of the Alk4/5/7 signal began from the early gastrula to mid gastrula stages (Figure 5k). After the late gastrula stage, the gut did not depend on the Alk4/5/7 signal for bending activity. By contrast, in H. pulcherrimus, the responsive period was longer than that in T. reevesii and occurred from the early gastrula to early prism stages (Figure 5l), which might be attributed to the different styles of gastrulation between the species. The gut of T. reevesii elongates only halfway through the body (Yaguchi et al., 2015) and bends to the ventral side at the late gastrula stage, whereas the gut in H. pulcherrimus elongates straight to the animal pole at the late gastrula stage (e.g., Harada, Yasuo, & Satoh, 1995). Although we do not have information on what decides the gut length during gastrulation in sea urchin development, the results of SB431542 treatment suggested that the Alk4/5/7 signal was required for gut morphogenetic behaviors until the timing immediately before bending in both species.
Although it has been previously reported that the dorsal–ventral pattern of the gut cell specification is dependent on TGF‐β signaling and that BMP molecules specify the dorsal characteristics of the gut (Duboc et al., 2010), whether Nodal/Activin/TGF‐β directly bind to the receptor on endodermal cells at SB431542‐sensitive timing for gut bending remains unclear. For clarification, we attempted to detect phosphorylated‐Smad2/3 (pSmad2/3) in gut cells during bending. At the late gastrula stage of H. pulcherrimus, a small amount of pSmad2/3 was in the nuclei at the ventral side of the tip of the gut and was abundant at the dorsal side of the gut (Figure 6a). However, because of the similarity of the amino acid sequences, the anti‐pSmad3 antibody recognizes both phosphorylated C‐termini of pSmad2/3 (…KQCSS*VS*; *expected phosphorylation site) and pSmad1/5/8 (…NPISS*VS*) (Yaguchi et al., 2007), and therefore, a definitive conclusion that the ventral faint signal was pSmad2/3 was difficult. To overcome this problem, we attempted to detect pSmad1/5/8 with an anti‐pSmad1/5 antibody in the same stage embryos because this antibody only recognizes pSmad1/5/8 in H. pulcherrimus (Yaguchi et al., 2011). Comparing the patterns of pSmad2/3 with those of pSmad1/5/8, we did not detect any pSmad1/5/8 on the ventral side of the gut (Figure 6b), indicating that pSmad2/3 was in the nuclei of the ventral side of the bending gut. Importantly, pSmad2/3 in this region clearly disappeared with SB431542 treatment (Figure 6c,d), supporting the idea that the direct Alk4/5/7 signal is required for gut bending toward the mouth region in H. pulcherrimus. At the dorsal side of the gut, whether pSmad2/3 was in the nuclei was unclear. Thus, we cannot discuss the involvement of Alk4/5/7 in the morphogenetic pathway at the dorsal side of the gut. Unfortunately, although the C‐terminal region of Smad2/3 in T. reevesii (…KVCSS*MS*) is similar to that of H. pulcherrimus, the anti‐Smad2/3 antibody did not cross‐react between the species.
At the late gastrula stage, Nodal from the ventral ectoderm may bind to the tip of the gut and induce endodermal nodal, which is important for the later left–right asymmetric expression of nodal on the ectoderm (Bessodes et al., 2012; Molina, De Crozé, Haillot, & Lepage, 2013). Thus, it is highly possible that the derivation of the nuclear pSmad2/3 at the tip of the gut was from the ventral Nodal activity, although we could not completely eliminate the possibility that Univin, TGF‐β and/or Activin was bound to the gut. In either case, the Alk4/5/7‐pSmad2/3‐mediated signaling pathway was required for gut bending. In sea urchin development, TGF‐β signals are certainly involved in morphogenesis. For example, the Nodal and BMP2/4 pathways specify the ventral and dorsal ectoderm, respectively (Saudemont et al., 2010), and the cell‐shape of the latter is much more squamous than that of the former. Although the mechanism by which those TGF‐β molecules regulate the cell‐shape in sea urchin embryos remains unclear, the resultant ectoderm contributes to the formation of a pluteus shape in which the dorsal ectoderm occupies a wider area than that of the ventral ectoderm. Therefore, the Alk4/5/7 signal might mediate the cytoskeletal conformation to bend the gut to the mouth region. In fact, the actin filamentation of endodermal cells in vertebrates depends on a Nodal signal during gastrulation (Woo, Housley, Weiner, & Stainier, 2012). Because Nodal‐Alk4/5/7‐Smad2/3 pathway induces foxA and foxD in the ventral endoderm (Duboc et al., 2010), these transcription factors might be suitable candidates for analyzing the mechanisms by which the TGF‐β‐cytoskeleton pathway contributes to induce the morphological changes in endodermal cells of sea urchins. Direct regulation of the cytoskeleton by the Nodal pathway is also expected, and a detailed analysis of the relationship between the signaling pathway and the cytoskeletal regulation in the cells of the tip in the gut requires further study. By contrast, gut bending does not involve the BMP2/4‐Smad1/5/8 pathway because the endoderm formation is normal in the absence of BMP2/4 activity (Yaguchi, Yaguchi, Angerer et al., 2010; Yaguchi et al., 2010).
4 CONCLUSIONS
In this paper, we reached two conclusions that contribute to sea urchin developmental biology: (1) the embryos of T. reevesii are useful for studying axis specification/formation and (2) the Alk4/5/7 signal is required for gut bending toward the mouth. Our scientific claims are strengthened because these conclusions were based on the reproducible results of two different species, demonstrating that using multiple species in the lab can be a powerful tool. In Japanese sea urchin developmental biology, to date, H. pulcherrimus has been the most frequently used model organism. However, based on a previous study (Yaguchi et al., 2015) and this paper, we show that T. reevesii can be used as a routinely available model sea urchin in Japanese labs. Therefore, the similar behavior of gut bending was the focus in this additional species, and the Alk4/5/7‐Smad2/3 pathway was involved in the gut‐bending step, which adds new information regarding morphogenesis of the sea urchin embryo. Because the invaginating gut undergoes conversion‐extension, which is a process in which almost no cells divide (Kominami & Takata, 2004), morphological changes of individual cells are important for the gut to bend. Thus, in future experiments, identification of the molecular linkage between TGF‐β signals and the regulation of the cytoskeleton in the endoderm of the sea urchin embryo is essential.
ACKNOWLEDGMENTS
We thank Drs. J. Yaguchi, R. Burke and M. Kioyomoto for the essential reagents and technical advice. We thank Mrs. T. Sato, D. Shibata, M. Ooue, T. Kodaka, J. Takano, and M. Yamaguchi for collecting and keeping the adult sea urchins. This work is supported by Takeda Science Foundation and Kishimoto Foundation (Senri Life Science Foundation).
