Background

Light plays crucial roles in biological processes such as photosynthesis and vision. Because visual systems involving eyes are very important for animal behaviors, a number of previous studies have investigated how the integrated circuits that mediate light stimulus develop and function [1,2,3,4,5]. In addition, recent studies have suggested that non-visual systems dependent on light also play essential roles in the life activities of animal, such as circadian rhythms [6, 7]. Many of these light-dependent systems rely on photoreceptor Opsin members, which belong to the group of sensory G-protein-coupled receptors (GPCRs), and their functional diversity has led us to consider how visual/non-visual systems developed during evolution to utilize light as an external signaling source [8,9,10]. In addition, structural and molecular analyses of Opsins in invertebrates such as tunicates and jellyfish have led us to consider the evolution of photoreceptor proteins [11, 12]. However, it is still difficult to precisely compare the functions of and predict the evolution of the light-dependent system in deuterostomes because we do not have experimental data about the precise function of the Opsin family in Ambulacraria, a sister group of chordates, although the evolutionary comparisons based on the primary structures have been performed [13].

Sea urchins are echinoderms and members of one of the phyla of Ambulacraria, and their embryos/larvae have been used as model organisms in developmental and cell biology for more than a century, but scientifically reproducible light-response data from embryos/larvae under genetic modifications have never been reported thus far, although they are under debate [13,14,15], and it has been reported that Ambulacrarian adults, including sea urchins, express a wide variety of Opsins and show light-response behaviors [16,17,18]. Even so, none of the Opsin family functions has been reported, although the presence of 6 Opsin genes (Opsin1, 2, 3.1, 3.2, 4, and 5) has been identified in the genomes of Strongylocentrotus purpuratus [13, 19] and Hemicentrotus pulcherrimus [20] and that the diversity and evolution of the Opsin family based on their gene structures were well discussed [21]. In addition, the functions of sea urchin nervous systems, which are presumptive mediators between photoreception and larval/adult behaviors, have been reported only in a limited number of papers. For example, serotonergic neurons, which are present exclusively in the anterior neuroectoderm, are required for gravity-dependent swimming behaviors [22], and a nitric oxide (NO) neuron in the pylorus regulates the pyloric opening [23]. On the other hand, a functional analysis of ciliary band neurons, including recently identified cholinergic neurons [24], has never been performed [25], and we do not have any experimental data about the relationship between the nervous system and light stimuli in sea urchins. In this study, we report that the sea urchins utilize light for regulating their pyloric opening with the function of serotonergic neurons and nitric oxide. Because the serotonergic neurons and Go-Opsin-expressing cells are present in/near the larval anterior neuroectoderm [26, 27], which is suggested as developmentally homologous to chordate brains [28,29,30], these data show the clear evidence for the presence of light-dependent brain-gut regulatory system in Ambulacraria and suggest that one of the primitive functions of deuterostome brains is regulating the digestive tract function.

Results

To investigate the responses of sea urchins to light, we observed the living larvae of H. pulcherrimus, which were transferred to a bright field just after incubation under dark conditions overnight, under a microscope for several minutes. As a result, we discovered that the pylori of some larvae opened in response to photoirradiation (Fig. 1a, Additional file 1: Fig. S1, Additional file 2: Movie 1). Since the opening time and frequency were variable depending on the incubation conditions prior to the light stimulus (Additional file 1: Fig. S2a, b), to estimate the opening ratio precisely, we set a constant light/dark cycle (Fig. 1b; 10 min of light, 16 h of dark, 0–10 min of photoirradiation [photon flux density, 1000 μmol m−2 s−1]) and checked pyloric opening/closing with immunohistochemistry using anti-TroponinI, which detects the pyloric sphincter [31], in the fixed larvae. The ratio of pyloric opening/closing events in all the subsequent experiments was measured in fixed and immunostained larvae (see the “Methods” section). The pylori of approximately 20% of the larval population were opened 2 min after photoirradiation, and by 6 min, most of them were closed under the set light conditions. The average pyloric opening rates were 4.0% (0 min), 8.9% (1 min), 17.9% (2 min), 16.1% (3 min), 10.9% (4 min), 7.0% (5 min), 1.5% (6 min), 1.7% (7 min), 3.4% (8 min), 2.2% (9 min), and 2.7% (10 min). The average pyloric opening time among individuals was approximately 1.5 min (n = 8, Additional file 1: Fig. S1), and the pylori of more than 40% of the larval population opened in response to the light stimulus (Fig. 1b; see the figure legend for the calculation). The reason why only 40% of the larvae responded to light will be elucidated in the future. This ratio jumped to more than 80% with food (mono-cellular algae) (Additional file 1: Fig. S3), indicating that food stimulation represents another regulatory pathway that activates the digestive tract, although we did not focus on this pathway in this paper. On the other hand, the timing of the response did not vary with or without food, suggesting that the light-dependent regulatory pathway that mediates pyloric opening is very stable. Based on these observations, the light>pylorus pathway is present in sea urchin larvae. We checked pyloric opening/closing 2 min after photoirradiation following 10 min of light and 16 h of dark in all of the following experiments (Fig. 1b).

Fig. 1
figure 1

Photoirradiation opens the pylorus in sea urchin larvae. a Photoirradiation drives the pyloric opening. Images of 5 s and 1 min 30 s after photoirradiation were captured from Supplementary Movie 1. b The set light condition was 10 min of light, overnight in darkness, and photoirradiation by high-intensity light. The graph shows the pyloric opening rates between 0 and 10 min after photoirradiation. The fluorescence images show pyloric closing (left) and opening (right), as visualized by TnI immunohistochemistry. N = 4; n (0 min) = 60, 39, 29, 43; n (1 min) = 45, 21, 30, 32; n (2 min) = 45, 26, 33, 54; n (3 min) = 43, 28, 23, 31; n (4 min) = 38, 28, 37, 24; n (5 min) = 68, 39, 35, 37; n (6 min) = 41, 28, 40, 26; n (7 min) = 72, 27, 23, 39; n (8 min) = 43, 27, 31, 13; n (9 min) = 84, 29, 29, 29; n (10 min) = 76, 38, 24, 21. The increase in the opening rate between 0 and 2 min after photoirradiation was assessed by the Welch two-sample t-test. Error bars show SE. More than 40% of larvae opened their pylorus in response to the photoirradiation (= sum of open rate from 1 to 5 min (8.9% (1 min) + 17.9% (2 min) + 16.1% (3 min) + 10.9% (4 min) + 7.0% (5 min)) was divided by 1.5 min (average open time)). c The region around the anterior neuroectoderm was necessary for pyloric opening induced by photoirradiation. The images in the upper row show the location at which the anterior neuroectoderm and pre-oral arms (− ANE and pre-oral arms) and post-oral arms (− post-oral arms) were removed. The lower images show open and closed pylori 2 min after photoirradiation. The inset of each image shows the bright-field images. The graph shows the opening rate of the pylorus in control larvae, larvae without the AP and pre-oral arms, and larvae without post-oral arms 2 min after photoirradiation. N = 2–4; n (control) = 19, 20, 24, 45; n (ANE and pre-oral arms less) = 13, 15, 17, 17; n (post-oral arms less) = 13, 12. We used one-way ANOVA followed by Tukey’s post hoc test. Error bars show SE. Scale bars in a and b = 20 μm and in c = 50 μm

Full size image

It is reasonably expected that this light>pylorus pathway is managed by the nervous and photoreception systems, which are intensively present in the anterior region of the larva [26, 32]. To examine which body part plays the central role in the light>pylorus pathway, we initially removed the anterior neuroectoderm/pre-oral arms or post-oral arms and checked the pyloric opening/closing under photoirradiation (Fig. 1c). The average pyloric opening rates were 21.7% (control), 4.4% (without ANE and pre-oral arms), and 19.9% (without post-oral arms). The pylori in larvae without anterior neuroectoderm/pre-oral arms did not open, whereas they did open in larvae without post-oral arms, indicating that the anterior neuroectoderm/pre-oral arms are necessary for this pathway (Fig. 1c). Within the anterior neuroectoderm and the adjacent regions, serotonergic neurons and Go-Opsin-expressing cells, respectively, are exclusively present [26, 32] (Fig. 2a, d). Therefore, we examined whether these cells are involved in the light>pylorus pathway. First, we applied serotonin and checked pyloric opening/closing without photoirradiation. Intriguingly, the pylori of serotonin+ larvae opened in a concentration-dependent manner, even without photoirradiation, whereas pylori of the seawater-applied larvae did not (Fig. 2c, Additional file 1: Fig. S4a, b). In addition, serotonin opened the pylori of larvae without anterior neuroectoderm/pre-oral arms (Fig. 2b, c). The average pyloric opening rates were 2.5% (+seawater), 72.2% (+serotonin), and 86.7% (without the ANE and pre-oral arms +serotonin). In contrast, in tryptophan 5-hydroxylase (TPH; serotonin synthase) morphants (in which the TPH function is knock downed using a specific morpholino), the pylorus did not open under photoirradiation (Additional file 1: Fig. S4c-e). These data suggest that serotonin, which is produced exclusively in the anterior neuroectoderm, is essential for the mechanisms of pyloric opening under photoirradiation in sea urchin larvae. Next, we knocked down the function of Go-Opsin, which is expressed in neurons adjacent to the anterior neuroectoderm in H. pulcherrimus, as shown in S. purpuratus [26, 27] (Fig. 2d, Additional file 1: Fig. S5), with morpholino oligos, because the other members of the Opsin family are not expressed around this region during the larval stages [13]. In Go-Opsin morphants, in which endoderm activity seemed to be normal (Fig. 2e), the pylorus did not open even under photoirradiation, but serotonin rescued this effect (Fig. 2f, Additional file 1: Fig. S6). The average pyloric opening rates were 1.2% (control; no treatment), 19.7% (control; photoirradiation), 63.9% (control; +serotonin), 0% (Go-Opsin morphants with no treatment), 1.0% (Go-Opsin morphants with photoirradiation), and 50.4% (Go-Opsin morphants with serotonin). This suggests that serotonin functions in pyloric opening downstream of the light>Go-Opsin pathway. It is very intriguing that the timing of the peak of pyloric opening in serotonin-supplied larvae was the same as that in photoirradiated larvae (Additional file 1: Fig. S4); this supports the idea that serotonin functions downstream of photoirradiation.

Fig. 2
figure 2

The anterior neuroectoderm plays a crucial role in opening the pylorus. a Serotonergic neurons and pan-neurons recognized by Synaptotagmin B (synB) in sea urchin larvae. Green, serotonin; magenta, SynB; blue (DAPI), nuclei. b Serotonin induced pyloric opening in both control larvae and larvae without the ANE and pre-oral arms. The inset of each image shows the bright-field images. c The graph shows the opening rate of the pylorus in control larvae treated with seawater, in control larvae 2 min after the addition of serotonin, and in larvae without the ANE and pre-oral arms 2 min after the addition of serotonin. N = 3; n (with seawater) = 55, 79, 74; n (with serotonin) = 23, 11, 10; n (without the ANE and pre-oral arms and with serotonin) = 5, 8, 4. Error bars show SE. d The expression pattern of Go-Opsin (opn3.2) in Hemicentrotus pulcherrimus (arrows). e The activity of alkaline phosphatase in the gut was invariant in control and Go-Opsin morphants. f The graph shows the opening rate of the pylorus in control larvae and Go-Opsin morphants 2 min after photoirradiation or the addition of serotonin. N = 3; n (control; no treatment) = 45, 28, 22; n (control; photoirradiation) = 78, 63, 54; n (control; +serotonin) = 25, 49, 38; n (Go-Opsin morphants with no treatment) = 16, 14, 14; n (Go-Opsin morphants with photoirradiation) = 33, 32, 11; n (Go-Opsin morphants with serotonin) = 9, 51, 28. Error bars show SE. Scale bars in a, b, d, and e = 50 μm

Full size image

Because the sea urchin neurons do not tend to form synaptic structures in a manner similar to serotonergic neurons in mammalian brains [33, 34], it is expected that serotonin, which is secreted from the anterior neuroectoderm, will be dispersed through the entire body of the larvae and activate locally present receptors. To examine how the serotonin pathway regulates the pylorus, we pharmacologically inhibited serotonin receptors and checked pyloric opening/closing. When the wide-ranging monoamine/serotonin receptor inhibitors, methysergide maleate [35] and asenapine maleate [36], were applied, the pylorus did not open even under photoirradiation (Additional file 1: Fig. S7a). In sea urchin genomes, 4 types of serotonin receptors, 5HT1, 5HT2, 5HT4/5/6, and 5HT7, were identified, and their phylogenetic positions among bilaterians were confirmed [19, 20, 37]. Because 5HT2 and 5HT7 are mainly expressed during the embryonic and larval stages based on their temporal expression patterns [38, 39] and because methysergide maleate works as an agonist of 5HT1 in mammals, we inhibited 5HT2 and 5HT7 with their specific antagonists, melperone [40] and ketanserin tartrate [40] (for 5HT2) and SB269970 [41] (for 5HT7). The results suggest that 5HT2, but not 5HT7, is involved in the serotonin>pylorus pathway under photoirradiation (Fig. 3a, Additional file 1: Fig. S7a, b). The average pyloric opening rates were 13.8% (control), 0% (melperone hydrochloride), and 1.0% (ketanserin tartrate). In addition, when the 5HT2 receptor was knocked down (Fig. 3c), the pylori of the morphants responded to neither photoirradiation nor serotonin but did respond to the nitric oxide (NO) donor S-nitroso-N-acetyl-D,L-penicillamine (SNAP) [23] (Fig. 3d, Additional file 1: Fig. S7d). This SNAP experiment was performed because we had recently reported that NO is involved in the pyloric opening in sea urchin larvae, similar to mammals [23]. The average pyloric opening rates were 4% (control with no treatment), 21.9% (control with photoirradiation), 81.6% (control with serotonin), 93.5% (control with SNAP), 0% (5HT2 morphants with no treatment), 2.1% (5HT2 morphants with photoirradiation), 7.5% (5HT2 morphants with serotonin), and 84.7% (5HT2 morphants with SNAP). These data suggest that 5HT2 mediates the pathway between serotonin and NO-dependent pyloric opening. In situ hybridization did not detect the expression of 5HT2 mRNA in our hands, but the microinjection of a DNA reporter construct, in which the putative cis-regulatory elements of 5HT2 and DNA encoding fluorescent protein Venus were fused, drove the expression of the reporter Venus in the stomach (Fig. 3b, and Additional file 1: Fig. S8; see the “Methods” section [microinjection of morpholino anti-sense oligonucleotides (MO), mRNAs, and DNA]), suggesting that 5HT2 is expressed in the stomach. Because the Venus signal representing 5HT2 expression did not localize to the pylorus but localized everywhere in the stomach, it is likely that serotonin from the anterior neuroectoderm indirectly activates the NO enteric neurons (hereafter referred to as sEN [23]; Fig. 4a) and the pyloric sphincter. To confirm whether the sEN functions in the pathway between photoirradiation and pyloric opening, we knocked down neuronal nitric oxide synthase (nNOS), which is expressed in the sEN, and checked pyloric opening/closing. When nNOS was attenuated, neither photoirradiation nor exogenous serotonin opened the pylorus, whereas SNAP did, indicating that NO release from the sEN is essential for opening the pylorus downstream of the light>serotonin pathway (Fig. 4b). The average pyloric opening rates were 4.5% (control with no treatment), 24.6% (control with photoirradiation), 66.1% (control with serotonin), 60.9% (control with SNAP), 3.3% (nNOS morphants with no treatment), 3.6% (nNOS morphants with photoirradiation), 14.8% (nNOS morphants with serotonin), and 60.4% (nNOS morphants with SNAP).

Fig. 3
figure 3

The 5HT2 receptor mediates the light>pylorus signaling pathway. a The opening rate of the pylorus under photoirradiation was extremely reduced by the addition of melperone hydrochloride and ketanserin tartrate (5-HT2 receptor antagonists). N = 4; n (control) = 54, 26, 42, 66; n (melperone hydrochloride) = 15, 17, 27, 36; n (ketanserin tartrate) = 17, 17, 27, 25. Error bars show SE. b The putative cis-regulatory elements of the 5HT2 receptor drove Venus signaling in the stomach. The rate of Venus expression in the stomach was 81.4% (57/70) in all larvae that had Venus signals. s, stomach; i, intestine. c The activity of alkaline phosphatase in the gut was invariant in control and 5HT2 morphants. d The graph shows that the 5HT2 receptor was required for the light>pylorus signaling pathway. N = 2–3; n (control with no treatment) = 25, 37; n (control with photoirradiation) = 21, 11, 55; n (control with serotonin) = 25, 19, 22; n (control with SNAP) = 28, 14, 19; n (5HT2 morphants with no treatment) = 13, 16; n (5HT2 morphants with photoirradiation) = 16, 16, 17; n (5HT2 morphants with serotonin) = 35, 45, 23; n (5HT2 morphants with SNAP) = 28, 28, 37. Error bars show SE. Scale bars in b and c = 50 μm

Full size image
Fig. 4.
figure 4

The light>serotonin pathway activates enteric neurons in the stomach (sENs), which express nNOS and induce pyloric opening. a A schematic image of neuronal nitric oxide synthase (nNOS)-expressing enteric neurons (sENs; magenta) with axon-like structures around the pyloric sphincter (green) [23]. b The graph shows the pyloric opening rate in control and nNOS morphants under photoirradiation, the addition of serotonin, and the addition of SNAP. N = 3; n (control with no treatment) = 10, 21, 26; n (control with photoirradiation) = 14, 21, 28; n (control with serotonin) = 15, 12, 30; n (control with SNAP) = 21, 19, 13; n (nNOS morphants with no treatment) = 13, 11, 10; n (nNOS morphants with photoirradiation) = 21, 17, 20; n (nNOS morphants with serotonin) = 19, 26, 12; n (nNOS morphants with SNAP) = 63, 12, 29. Error bars show SE. c A schematic diagram of the light>pylorus signaling pathway in sea urchin larvae

Full size image

Discussion

Taken together, our data suggest that sea urchin larvae utilize light to open the pylorus through the photoirradiation>Go-Opsin>anterior neuroectoderm>serotonin>5HT2>sEN/NO>pylorus pathway (Fig. 4c). Two points shown here are especially intriguing: (1) light behaves as a tool to regulate the activity of the digestive tract, and (2) the light-dependent signaling pathway in echinoderm larvae was revealed under the modification of genetic function. Because neither stable pigmented shade nor sophisticated neural integration was observed around Go-Opsin-expressing cells, the photoreceptors of sea urchin larvae work as a non-visual and non-directional system [26]. Light stimulates several non-visual activities in a wide variety of organisms, as a number of previous scientific works have proven. For example, non-visual Opsin3, which is expressed in human epidermal melanocytes, mediates pigmentation [42], and Opsin5, which is expressed in the bird brain, is essential for seasonal reproduction [6]. In sea urchin larvae, light, which is received around the anterior neuroectoderm, likely drives digestive activity. Although it has been well reported that vertebrates utilize light-dark cycles, i.e., circadian rhythms, for their digestive activities, such as gut motility, gene expression, and mucosal production [43, 44], these activities are relatively slow since they change over a 24-h cycle. In addition, to drive established regulatory pathways, individuals need to prepare the execution unit, such as the photoreception system, products of clock genes, signal receivers in the gut, and mediators between each. Compared with these pathways, what we describe here happens within a relatively short time during early developmental stages when embryos/larvae are still establishing tissues and organs and requires only a simple signaling pathway involving a few factors, meaning that this phenomenon is likely a reflexive movement and not a circadian rhythm. In contrast, compared with the conventional light reflex carried out by the combination of photoreceptors, neurons, and muscles in many kinds of animals, including their embryos/larvae [45], pyloric opening is relatively slow. It would be of interest to see the timescale of each response step of the light > pylorus pathway in the future.

This reflex seems to be important not only for gut function but also for gut development. It is well known that the precise repetitive contraction and relaxation of the enteric muscle is essential for its development [46, 47]. Because sea urchin larvae have to swallow and digest algae from the beginning of their larval lives and because, during this period, there is no liquid or baby food such as milk to induce the maturation of the digestive tract, they need to train their gut before swallowing food for the first time; the light reflex we describe here might represent this training.

Light reflex-based pyloric opening is also important for the daily activities of sea urchins because it has been reported that phytoplankton stay at the surface range all day long but that zooplankton sink to a relatively deep region during the daytime and float to the surface at night [14], implying that sea urchin larvae need to swallow and store algae in the stomach as much as possible at night, when they are near algae. Then, when the larvae are exposed to strong sunlight, their pylori begin to open to pass the digested algae to the intestine. Simultaneously, this system, by which the pylorus tends to remain closed in the dark (Additional file 1: Fig. S3), is likely important because larvae should not pass undigested algae from the stomach to the intestine at night; therefore, it is speculated that the light-dependent pyloric opening system has been acquired and developed along with daily migration during evolution.

It is still unclear why the pylorus opened in response to light stimuli in only 40% of larvae. This might be because the light used for the experiments was simply not enough to induce pyloric opening; we used half of the number of photons as direct sunlight in our experiments, which corresponded to a depth of a few meters in the ocean. However, even when the larvae were exposed to a photon flux density corresponding to direct sunlight, the pyloric opening rate was similar to that of the experiment shown in Fig. 1b (Additional file 1: Fig. S9), leaving the question unsolved. In addition, we could not find a Go-Opsin-specific wavelength in this study, but in Platynereis, the blue-cyan light, which is the main component of the LED wavelength, is specific for Go-Opsin absorbance [48], suggesting that a similar range of wavelengths is effective for the pyloric opening in sea urchin larvae. This 40% opening rate increased to close to 80% after the exposure to serotonin or food algae (Figs. 2c and S3). Although future works will elucidate how food stimuli are involved in the light>serotonin>pylorus pathway, our data clearly indicates that anterior neuroectodermal serotonin is a key neurotransmitter that regulates pyloric opening and that food can stimulate the neural pathway. It is also still unclear why the pylorus opened 2 min after stimulation. It is very intriguing that the timing of pyloric opening under photoirradiation was similar with and without food stimuli. This suggests that light plays a part of the central and initial roles in the pathway of pyloric opening. In addition, because larvae have to collect the stomach contents in the pylorus to push them forward, the serotonergic system might function to control ciliary beating in the stomach for 2 min after photoreception since the 5HT2 receptor is likely expressed throughout the stomach. Understanding the biochemical and biophysical characteristics of Go-Opsin and the serotonergic nervous system will elucidate the regulation of the timing of pyloric opening.

Since the sea urchin genome was sequenced, the gene structure and expression patterns of Opsin family members have been reported [13, 26]. However, their functions in larval behavior have been unclear. In this study, we revealed the pathway between Go-Opsin and pyloric opening, which helps larvae pass food through the gut. Sea urchin Go-Opsin is a member of the Go(/RGR)-opsin family [21, 27] and is similar to Opsin expressed in ciliary photoreceptors in scallops and in rhabdomeric photoreceptors in Platynereis [48,49,50]. In tunicates, the ciliary photoreceptor in larvae is associated with pigmented cells [51], indicating that ciliary photoreceptors in chordates are conservatively directional [52]. Therefore, future analyses of sea urchin larval and adult ciliary photoreceptor cells will elucidate how these photoreceptors were acquired and diversified in deuterostome evolution.

Conclusions

Our data show that the neurons at the anterior neuroectoderm regulate the function of the digestive tract in response to light stimuli in sea urchin larvae. Because the anterior neuroectoderm of sea urchin embryos/larvae is homologous to the brain region in vertebrates based on gene expression profiles and gene functions during its formation [28,29,30], it is suggested that the regulatory pathway between the brain and the gut was already present in the common ancestor of deuterostomes although the primitive system is not speculated yet since the vertebrates’ system, in which the vagus nerves and neural crest cells manage the brain-gut interactions, is too unique to compare with those in other systems. As the body size increased during metazoan evolution, organisms developed a sophisticated gut to digest food and obtain nutrients efficiently. Simultaneously, it is speculated that the brain-dependent gut control systems have been developed since the brain can integrate both body internal and external information and can transfer them to the digestive tract to reflect the information in regulating gut activity.

Methods

Animal collection and embryonic/larval culture

Adult Hemicentrotus pulcherrimus were collected around Shimoda Marine Research Center, University of Tsukuba, and around the Marine and Coastal Research Center, Ochanomizu University. Adult sea urchins were collected under the special harvest permission of prefectures and Japan Fishery cooperatives. Gametes were collected by the intrablastocoelic injection of 0.5 M KCl, and the embryos/larvae were cultured at 15 °C in glass beakers or plastic dishes that contained filtered natural seawater (FSW) with 50 μg/ml kanamycin. In some experiments, we fed 3.3 μl/ml SunCulture algae (Chaetoceros calcitrans, Marinetech, Aichi, Japan, approx. 30,000 cells/μl) to the larvae as forage.

Photoirradiation experiments

White LED beam light (PLATA Inc., Osaka, Japan), a general LED light irradiating a broad range of visual light wavelengths, was used for the photoirradiation experiments, and the photon flux density was measured with a quantum sensor (Apogee Instruments, Logan, UT, USA). The distance between the light and sea urchin samples was adjusted to make the photon flux density to be 1000 μmol m−2 s−1. The details of the photoirradiation are shown in Additional file 1: Fig. S10. Since 6 wells in a 24-well plate can receive the LED beam light equally, all photoirradiation experiments were carried out simultaneously for as many as 6 wells. The light source was placed 6 cm above the 24-well plate to irradiate one-half the photons of sunlight (approximately 1000 μmol m−2 s−1). To obtain the sunlight-level photons (approximately 2000 μmol m−2 s−1) (Fig. S9), the light source was placed 2 cm above the 24-well plate. The samples were prepared in 900 μl or 1000 μl SW per well, and the 24-well plates were wrapped in the aluminum foil and maintained in dark incubators until use. When the experiment required neither photoirradiation nor reagent treatment, the samples were fixed with 1000 μl 7.4% formaldehyde-SW (final concentration 3.7%) within 10 s after removing the aluminum foil (Additional file 1: Fig. S10a). When the experiment required light stimulation for a certain period of time, photoirradiation was started within 10 s after removing the aluminum foil (Additional file 1: Fig. S10b). When the experiment required reagent treatment, the reagent was added within 10 s after removing the aluminum foil (Additional file 1: Fig. S10c). When the experiment required both photoirradiation and reagent treatment, we added reagents first within 5 s after removing the aluminum foil and then exposed the light to the larvae within 5 s after reagent treatment. To make the reagents or fixative diffuse throughout in the well, the plate was hand-tapped immediately after the reagents or formaldehyde-SW was added to the well.

Microsurgery

Larvae were transferred to new 10-cm plastic dishes filled with FSW, and a part of the body of each larva was dissected under a dissecting microscope. After surgery, the larvae were transferred to new 6-cm plastic dishes filled with FSW containing 50 μg/ml kanamycin and cultured in the dark until the next day.

Chemical treatments

Melperone hydrochloride (FUJIFILM Wako Pure Chemical Co., Osaka, Japan), ketanserin tartrate (FUJIFILM), methysergide maleate (Sigma-Aldrich, St. Louis, MO, USA), asenapine maleate (Sigma-Aldrich), and SB269970 (Sigma-Aldrich) were added to 10 μM FSW to inhibit the serotonin/monoamine pathway. The inhibitors were applied 10 s before photoirradiation. S-Nitroso-N-acetyl-D,L-penicillamine (SNAP; FUJIFILM) was used as a nitric oxide (NO) donor (final concentration of 100 μM) [23]. SNAP was applied to the larvae 5 min before observation. 3,5-Difluorophenyl-acetyl-L-alanyl-L-S-phenylglycine T-butyl ester (DAPT; Sigma-Aldrich) was used as a γ-secretase inhibitor (final 20 mM). Serotonin (Tokyo Chemical Industry, Tokyo Japan) was dissolved to distilled water just before use and applied to culture (final 10 μM). The same volume of DMSO or seawater was applied as controls for chemical treatments.

Whole-mount in situ hybridization and immunohistochemistry

Whole-mount in situ hybridization was performed as described previously [53] with some modifications. cDNA mix from several embryonic stages was used to make RNA probes based on the H. pulcherrimus genome and transcriptome [20]. The samples were incubated with digoxygenin (Dig)-labeled RNA probes for Go-Opsin (HPU_20590) and tryptophan 5-hydroxylase (tph; HPU_21307) [22] at a final concentration of 1.2 ng/μl at 50 °C for 5 days. The Dig-labeled probes were detected with an anti-Dig POD-conjugated antibody (Roche, Basel, Switzerland) and treated with the Tyramide Signal Amplification Plus System (TSA; PerkinElmer, Waltham, MA, USA) for 8 min at room temperature (RT). When observed, the samples were incubated in MOPS buffer containing 2.5% 1,4-diazabicyclo-2-2-2-octane (DABCO; Wako Pure Chemical Co., Osaka, Japan) to prevent photobleaching.

Whole-mount immunohistochemistry was also performed as described previously [53] with some modifications. The samples were blocked with 1% skim milk in PBST for 1 h at RT and incubated with primary antibodies (dilutions: mouse anti-Synaptotagmin B (SynB) [25], 1:100; rabbit anti-Troponin-I (TnI) [31], 1:200; rabbit anti-serotonin (#S5545, RRID; AB_477522, Sigma-Aldrich), 1:1000) overnight at 4 °C. The primary antibodies were detected with a goat anti-mouse IgG Alexa Fluor Plus 555-conjugated (#A32727, RRID; AB_2633276, Thermo Fisher Scientific, Waltham, MA, USA) or a goat anti-rabbit IgG Alexa Fluor Plus 488-conjugated (#A32731, RRID; AB_2633280, Thermo Fisher Scientific) antibodies diluted 1:2000.

Double staining for SynB protein and Go-Opsin mRNA was performed as described previously [22] with some modifications. Samples were fixed at 4 °C for 5 h and were blocked with 1% bovine serum albumin prior to incubation with the primary antibody (1:100 dilution of mouse anti-SynB [25]) at ambient temperature for 1 h. The primary antibody was detected with a goat anti-mouse IgG HRP-conjugated antibody (#405306, RRID; AB_315009, BioLegend, San Diego, CA, USA) diluted 1:2000 and TSA treatment. After SynB detection by TSA-based immunohistochemistry, whole-mount in situ hybridization was performed to detect Go-Opsin under dark conditions as described above.

Microinjection of morpholino anti-sense oligonucleotides (MO), mRNAs, and DNA

Microinjection was performed according to a previously described method [54] with injection buffer (24% glycerol, 20 mM HEPES pH 8.0 and 120 mM KCl). The morpholino (Gene Tools, Philomath, OR, USA) sequences and the in-needle concentrations in injection buffer were as follows:

Go-Opsin MO1 (0.8–1.0 mM): 5′-ATCTTCTTGAATATGCTTCCGCGCC-3′,

Go-Opsin MO2 (1.0–1.5 mM): 5′-ACGAATTCATTGTGGTCAAATCCGC-3′,

5HT2 MO1 (0.5–1.0 mM): 5′-GGAAAGGAACATCTCAGATCGGCCT-3′,

5HT2 MO2 (0.5 mM): 5′-GATGTCCTTATGGTATGTGCA-3′,

nNOS MO1 (1.0–1.5 mM): 5′-GGAAAGGAACATCTCAGATCGGCCT-3′ (previously characterized) [23], and

TPH MO (1.2 mM): 5′-ACAGAGTAGGACGTTGATGATCTAT-3′ (the specificity was checked by immunohistochemistry for serotonin (Additional file 1: Fig. S4D)).

Two non-overlapping translation-blocking morpholinos for Go-Opsin and 5HT2 were used to confirm the specificity of their function (Additional file 1: Fig. S6b, 7c). For negative control experiments, we injected random MO (1.0–1.5 mM, Gene Tools, Additional file 1: Fig. S6a) or injection buffer only.

The DNA construct for the putative cis-regulatory element of 5HT2 was prepared and injected as previously described [55]. Five thousand base-pairs of the genomic DNA of H. pulcherrimus were isolated and combined with a DNA sequence encoding Venus.

Detection of alkaline phosphatase

To observe the stomach and intestine under clearer conditions, we detected alkaline phosphatase (AP) activity in the digestive tract. Larvae were fixed with cold 100% ethanol (− 20 °C) for 5 min and washed 3 times with PBST (PBS [Nippon Gene Co., Tokyo, Japan], 0.1% Tween-20). The samples were washed 3 times with AP buffer (100 mM Tris pH 9.5, 50 mM MgCl2, 100 mM NaCl, 1.0 mM levamisole, and 0.1% Tween-20), and the AP signal was detected with NBT/BCIP (Promega, Madison, WI, USA).

Microscopy and image analysis

Live samples (Figs. 1a and S1 only) were observed under a light/fluorescence microscope (IX70, Olympus, Tokyo, Japan). The fixed and stained specimens (all samples other than those shown in Figs. 1a and S1) were observed using a light/fluorescence microscope (IX70, Olympus) and a confocal laser scanning microscope (FV10i, Olympus). All transmission images were taken with an IX70 microscope. We set the pyloric opening rate as crucial for the response to the light because the pylorus is rarely open when the larvae are maintained in the dark. All of the pyloric opening rates were judged and counted under immunohistochemically stained larvae with anti-TnI antibody as explained above (Fig. 1b, see the “Methods” section [whole-mount in situ hybridization and immunohistochemistry]). We judged that the pylorus was closed when a strong TnI signal was observed at the center of the pylorus (Fig. 1b). Each sample size (n) was variable because the survival rates in normal and experimental larvae were changed in each batch. The figure panels and drawings for the figures were made using Adobe Photoshop and Microsoft PowerPoint.

Statistical analysis

No statistical methods were used to predetermine the sample sizes. All sample sizes and p-values are provided in the figure legends. To compare the two groups of data shown in Figs. 1b, S4e, S6a, b, and S7b, c, we used Welch’s t-test (two-tailed) with a significance level of 0.01 or 0.05; the t-values for the data in these figures were 4.6500, 8.3187, 0.26576, 3.5544, 0.42644, and 7.4837, respectively, and the degrees of freedom (d.f.) were 5.8326, 4.1031, 3.9592, 3.9315, 2.1605, and 3.427, respectively. To compare more than two groups, we used one-way ANOVA followed by Tukey’s post hoc test with a significance level of 0.01 or 0.05; the following F values (F) and d.f. were used: in Fig. 1c, f = 12.4959 and d.f. = 2; in Fig. 2c, F = 33.804 and d.f. = 2; in Fig. 2f, F = 143.1067 and d.f. = 2 for control and F = 10.1249 and d.f. = 2 for Go-Opsin MO1; in Fig. 3a, F = 69.4826 and d.f. = 2; in Fig. 3d, F = 11.6514 and d.f. = 3 for photoirradiation and F = 32.985 and d.f. = 5 for chemical treatments; in Fig. 4c, F = 6.0302 and d.f. = 3 for photoirradiation and F = 15.3079 and d.f. = 5 for chemical treatments; in Additional file 1: Fig. S2a, F = 10.2301 and d.f. = 3; in Additional file 1: Fig. S2b, F = 3.724 and d.f. = 3; in Additional file 1: Fig. S4a, F = 21.9395 and d.f. = 3; in Additional file 1: Fig. S7a, F = 49.6755 and d.f. = 2; in Additional file 1: Fig. S8a, F = 12.3052 and d.f. = 2 for the digestive tract.