Initial report of γ-aminobutyric acidergic locomotion regulatory system and its 3-mercaptopropionic acid-sensitivity in metamorphic juvenile of sea urchin, Hemicentrotus pulcherrimus

The γ-aminobutyric acid (GABA) signal transmission system (GSTS) contributes to larval swimming through the regulation of ciliary beating. However, whether this system also contributes to the primary podia (PP)-generated motility of juveniles remained unclear. The present study aimed to elucidate the involvement of the GSTS in the motility of metamorphic juveniles (juveniles) (1) by immunohistochemically elucidating the location of molecular constituents of the PP, and (2) by inhibiting the activity of GΑΒΑ decarboxylase (GAD) with 3-mercaptopropionic acid (3-MPA). During metamorphosis, the echinus rudiment protrudes its PP out of the body surface in 8-arm plutei. The PP expresses immunopositive signal (-IS) of GAD, GABA, GABAA receptor and tropomyosin, and is constituted with the GABA-IS negative distal tip and the GABA/GAD-IS gaiter region. The latter radiates distal projections to the disc that contains a GAD-IS cellular network. The juvenile body cavity houses a GABA/βIII-tubulin-IS Penta-radial ring (PrR) that extends branches into each PP and several bridges to the GAD/GABA-IS Penta-radial plate (PrP) on the oral side but does not reach to the gaiter region. 3-MPA reversibly inhibits the juvenile motility and GABA-IS expression in the PrR/PrP complex. This indicates that the complex is the major contributor to the GABAergic motility in juveniles.

Whole-mount immunohistochemistry. The larvae that reached the developmental stages described in the text, and the 1-day post-metamorphosis (1-dpm) to 5-dpm juveniles were fixed in 4% paraformaldehyde (diluted in FSW) for 15 to 20 min at ambient temperature (AmT) and dehydrated in a series of increasing concentration of ethanol starting from 30% (v/v) and stored in 70% ethanol at 4 °C until use. For WMIHC, the samples were hydrated in decreasing concentration of ethanol to 30% and transferred to 0.1 M phosphate buffered saline with 1% (v/v) Tween-20 (PBST; Medicago AB, Uppsala, Sweden).
The samples were blocked with 1% (w/v) bovine serum albumin in PBST for 1 h and exposed to the following primary antibodies (Abs) in PBST for 24 to 48 h at 4 °C (Table 1) in various combinations for multi-stained WMIHC. These Abs were then washed in PBST three times (10 min each) and were visualized with the following secondary Abs that were diluted in PBST. The secondary antibodies used were Alexa Fluor 488-, 568-or 594-tagged goat anti-rabbit or -mouse IgG Abs (Invitrogen, Carlsbad, CA. USA; diluted 1:200 to 1:500), Alexa Fluor 647-tagged goat anti-mouse IgG Ab (Abcam plc. Cambridge, USA; diluted 1:200 to 1:1,000) or Alexa Fluor 488-tagged goat anti-chicken IgY Ab (Abcam; diluted 1:500) and they were incubated with the samples for 2 h at AmT. Most of the samples were counterstained with 1-2 μg/ml 4′, 6-diamidino-2-phenylindole (DAPI) and cleared in Olympus Scaleview-A2 optical clearing solution (Wako Pure Chemical Ind. Ltd. Osaka, Japan) or CUBIC 17 for a few hours (for 8-arm pluteus larva in early metamorphosis period) or for 3 days (for juveniles) at 4 °C according to the manufacturer's instructions. They were examined under a Micro-Radiance confocal laser scanning microscope (CLSM; Bio-Rad Microscience, Hemel Hempstead, UK) or a TCS SP8 CLSM (Leica Microsystems, Co. Japan, Tokyo, Japan); otherwise, they were analyzed as described in the text. Images were analyzed with ImageJ (National Institutes of Health, NIH; http://rsbweb.nih.gov/ij/) or Adobe Photoshop CS2 software (ver. 9.02, Adobe Systems Inc., San Jose, CA, USA).

3-D image construction.
To clarify the spatial relationship of the images obtained with CLSM, the optical sections were reconstructed three-dimensionally with the following 3-D visualization and analysis softwares. They are an Avizo software ver.6.1.1 (FEI Visualization Sciences Group, Bordeaux, France), an Amira software (FEI Visualization Sciences Group, Burlington, MA 01803, USA) or an ImageJ 1.45 s (National Institutes of Health, USA).
Video recording of the metamorphosis process. The late-8-arm pluteus larvae that were in the process of metamorphosis were transferred to FSW in 4 cm diameter plastic dishes, and the behavior of larvae or juveniles at AmT was recorded for 2 to 9 min with an EOS Kiss X3 digital camera (Canon Inc. Tokyo). www.nature.com/scientificreports www.nature.com/scientificreports/ Bioassay for movement of juveniles with or without 3-MPA. A total of 171 3-dpm juveniles were separated into the following three groups: (1) plain FSW (39 juveniles; Control), (2) 1 μM 3-MPA in FSW (88 juveniles; 3-MPA), and (3) 1 μM 3-MPA + 100 μM GABA in FSW (44 juveniles; 3-MPA+GABA). Prior to the first assay period, each group was further split into five wells of the 24-well plate (Coster, Corning, NY14831, USA). Each well contained seven to eleven juveniles. The juveniles were incubated for 1 h at AmT. Then, the control group was replaced with fresh plain FSW, and the 3-MPA group was split into the following two groups: one group was treated with replaced fresh 1 μM of 3-MPA, and the other was washed three times (5 min each) with fresh FSW and was further incubated in FSW; the 3-MPA+GABA group was replaced with a fresh 3-MPA+GABA. They were further incubated for 2 h with recording as stated above for the second assay period (Fig. 1A), while their respective positions in the wells at every 15 min by an image scanner GT-8200 UF (Seiko Epson Co., Suwa, Japan) at 300 dpi, 768 × 1024 pixel, 16 bit gray. The assay was repeated three times. The motile behavior of juveniles was examined with an animation made by ImageJ1.47v (National Institutes of Health) with the aid of Java 1.6.0_65 (32-bit) (Oracle Co., Redwood Shores, CA 94065. USA).
An initial analysis of the motile behavior was conducted based on the proportion of juveniles that changed their position in a well from the initial settled place for 1 h for "before washing" and, from the initial place after FSW exchange for 2 h for "after washing" after 30 min rest. The statistical significance of average proportions of juveniles that changed from initial positions between the two groups ("before FSW replacement" and "after FSW replacement") were analyzed using an unpaired t-test using the public-domain software QuickCalcs GraphPad (http://www.graphpad.com/quickcalcs/ttest1.cfm). Two-tailed P values of less than 0.05 were regarded as statistically significant differences.
A statistical analysis of the number of juveniles that expressed the GABA-IS PrR/PrP complex was conducted using samples that were fixed soon after the 3-MPA assay test, as described above. A total of 22 juveniles (7 for the control, 9 for 3-MPA, 3 for 3-MPA+GABA and 3 for washed after 3-MPA) were triple stained with mouse anti-GABA monoclonal antibody (mAb) (Sigma), anti-TM rabbit antibody 18 and DAPI.

Results
Primary podia formation from the late 8-arm pluteus larva stage to the end of metamorphosis.
Consistent with our previous report 7 , PP were detected in the larva as Syt-IS small protrusions in EcR ( Fig. 2A, pp). GAD-IS network constituted the basal surface of the blastocoel (the Blastocoel basal GAD-IS-network) and connected to the ciliary band-associated strand in some places [ Fig. 2A, arrows, inset (a), arrowhead; 2B, arrow]. According to the optical cross-sections, some of the basal GAD-IS network surrounded EcR (Fig. 2B) and reached to the surface of PP (Fig. 2C, arrow).
In slightly older plutei than that mentioned above, the EcR bulged slightly toward the left side of the larval body surface, and TM-IS and GABA A R-IS were faintly detected in all five PP (Fig. 2D,D1,D2). In those whose EcR was more enlarged, GABA A R-IS remained weak (Fig. 2E,E1), while the PP presented more intensive TM-IS ( Fig. 2E; pp, E2). In further developed plutei, the GAD-IS PP (Fig. 2F) asynchronously expressed GABA-IS among some of the PP (Fig. 2F1,F2). Both ISs were weak at the distal tip of the PP (Fig. 2F-F2; arrowheads). In the plutei that close to the end of the metamorphosis, the TM-IS PP (Fig. 2G1,G2) extended further, and a body surface opening was seen immediately above the EcR on the anterior side of the anterior epaulette (Fig. 2G1,G2, asterisks). According to optical cross-sections of a PP [box (h) in Fig. 2G1], it also was GAD-IS ( Fig. 2H-H2).
The video images of the metamorphosis of 36-dpf plutei indicated that all five PP were extended from the larval body surface and utilized for the creeping movement that resulted in shifting the visual focusing level of PP during the observation (Fig. 3A-D). However, occasionally they detached from the substrate and swam actively with the ciliary beating at both anterior and posterior epaulettes. The occurrence of the creeping movement by PP and swimming by the ciliary beating at the epaulets occurred with mixed combinations, which suggested little or no coordination systems presented between these two motile activities during metamorphosis. The larval arms were retained but were considerably shortened accompanied by apical exposure of internal arm spicules (Fig. 3A, green arrow). By the 37-dpf stage, the larval arms were further absorbed (Fig. 3E, red arrow), and a trace of the arms was left with long spicules (Fig. 3E, green arrow). They did not show the ciliary beating activity on their body surface.

GABA/GAD-iS expression pattern in pp of early juveniles.
A GABA-IS pattern in the PP of EcR in the late 8-arm pluteus larva stage ( Fig. 2F1) was inherited by juveniles (Fig. 4A1). GAD-IS on the EcR surface ( Fig. 2G) was also inherited by the juvenile body surface (Fig. 4A2,A3,B for negative control). In the PP, GABA-IS was detected accompanied by a spiral ring pattern around the central lumen that radiated distal spiny projections at the distal end into the disc region (Fig. 4C1, arrows). GAD-IS was also detected in the luminal wall of the PP (Fig. 4A2,A3,C2,C3 arrowheads) and in the disc associated with spots around the tips of distal spiny projections (Fig. 4C2, arrows). A merged image localized GAD-IS spots around the tips of GABA-IS distal spiny projections (Fig. 4C3, arrows). The higher magnification image of the GABA-IS distal spiny projection region (Fig. 4D1) depicted apparent partial overlapping with the GAD-IS spots (Fig. 4D2, arrows; D3, arrows). 3D images of the distal spiny projection that were reconstructed based on the above CLSM images indicated that GAD-IS spots actually constituted a network in the disc region around the distal end of PP (Fig. 4E). Each tip of the GABA-IS distal spiny projection was enveloped by the GAD-IS network (Fig. 4F, arrowheads). This could be the structural basis of interaction between the GAD/GABA-IS areas at the distal spiny projection of the PP.
In the relaxed PP, GABA/GAD-IS spiral rings were eminently detected ( In the 5-dpm juveniles GABA-IS was detected on the TM-IS layer of the PP (Fig. 5A). A more detailed examination of the PP using higher magnification images detected a gap between the GABA-IS layer and the luminal epithelium (Fig. 5B1, arrow). The same PP area indicated that TM-IS layer immediately on the luminal epithelium with a clear IS-negative space between the surrounding area ( Fig. 5B2, arrow). The merged image between GABA-IS and TM-IS images positioned the GABA-IS precisely at the previous gap over the TM-IS area (Fig. 5B3, arrow), which indicated that both IS-areas were closely associated with the luminal epithelium of the PP.
To examine the possible involvement of GABA A R in close association with the GABA-IS area with the TM-IS layer, the PP of the 5-dpm larvae was double-stained with these antibodies, and the merged image was 3-D reconstructed (Fig. 5C,D). The TM-IS and GABA A R-IS area were localized together at PP. The resultant GABA A R-IS area was also accompanied to the disc at the distal end (Fig. 5C, arrows), which resembles the structure detected as GABA-IS distal spiny projection of the PP (Fig. 4C1,C3). A 3-D image of an optical cross-section of PP located GABA A R-IS layer immediately on the outer surface of the TM-IS layer on the luminal epithelium (Fig. 5D, arrow). A CLSM image of the PP's optical cross-section that was double-stained for GABA-IS and GABA A R-IS detected both ISs negative space immediately on the luminal epithelium (Fig. 5E, white arrow), which implicated that the muscle layer resided there. Thus, the spatial location of PP-constituting proteins indicated a concentric distribution as summarized by Fig. 5F. the GABAergic prR/prp complex in the juvenile body cavity. A GABA-IS Penta-radial structure in the middle of the body cavity of the juvenile radiated five branches toward each of PP (Fig. 4B, arrowheads). Thus, the next question about the creeping movement regulation was to specify the molecular constituents of that potentially involved in the signal transmission to PP.

Scientific RepoRtS |
(2020) 10:778 | https://doi.org/10.1038/s41598-020-57567-w www.nature.com/scientificreports www.nature.com/scientificreports/ www.nature.com/scientificreports www.nature.com/scientificreports/ A series of optical cross-sections of a juvenile localized a Penta-radial GABA-IS ring (Fig. 6A, PrR) and a Penta-radial plexus at the center of the ring (Fig. 6A, PrP) on the aboral side which radiated GABA-IS branches into each of the five PP in 83.4 ± 16.7% of the 8 juveniles examined. Five corners of the central PrP (Fig. 6B, arrowheads) and of the peripheral PrR intersected at an angle of approximately 35° (Fig. 6A,B, triangle), which suggested that apparent ventral fragmental GABA-IS ring may not be a part of the dorsal PrR. The next optical cross-section on the further oral side detected five TM-IS clumps at the proximal end of the PP (Fig. 6B, ppm) with GABA-IS dots on the proximal side (Fig. 6B, arrows). They surrounded a central GABA-IS fragmental ring of the PrP (Fig. 6B, arrowheads). In the very oral end of the optical cross-section, five TM-IS triangles (Fig. 6C, red arrows) and five small GABA-IS dots (Fig. 6C, white arrows) were detected around the mouth.
To elucidate the spatial relationship between the PP constituents and the surrounding tissues, an animation of the 3-D image of a PP of the same CLSM data as in Fig. 6A (box "d-g" region) was reconstructed (Fig. 6D-G). According to the 3-D animation, TM-IS constituted a central column of the PP (Fig. 6E-G, ppm; Supplementary Video S1) all the way to the distal tip, which reached to the disc (Fig. 6D, disc). TM-IS area also was connected to the muscular ring on the oral side (Fig. 6C, red arrows). The GABA-IS area encircled to form a gaiter region on the distal half of the PP (PP gaiter; Fig. 6D-G, gp) but was disconnected from the proximal GABA-IS ring via proximal TF muscle region (Supplementary Video S1). The latter muscle region retained the connection to PrR (Fig. 6D-G, PrR).
The aboral view of the ring was shown to be embedded in TM-IS muscles of PP that orally extended (Fig. 6H, pp). Its side view indicated the GABA-IS aboral layer (Fig. 6I, PrR) and the oral layer (Fig. 6I, PrP). The oral side view revealed five discs of the PP and a GABA-IS plate (Fig. 6J, PrP). Thus, two GABA-IS structures were detected there, and their spatial relationship with PP was shown by a schematic summary (Fig. 6K).
To ensure the positional relationship of these GABA-IS Penta-radial structures, a 3-D animation was reconstructed using above same CLSM data as shown in Fig. 6A-C. The side view revealed several bridges between the PrR and the PrP (Fig. 7A, arrows). Then, the PrR/PrP complex was digitally split horizontally at the middle of two GABA-IS layers [  (Fig. 7B). In the oral slice [ Fig. 7A, box (c)] the PrP was detected only between the TM-IS PP muscles (Fig. 7C, Supplementary Video S2B). The oral-aboral positioning was confirmed also by using an optical cross-section made at the middle of a PrP, and an aboral PrR and an oral PrP (Fig. 7D). This cross-section also detected a conical dome that raised toward the aboral side at the center of the PrP (Fig. 7D, red arrow). Thus, the PrR and the PrP constituted a structurally united complex.

Juvenile mobility inhibition by 3-MPA.
The aforementioned data suggest that the GABAergic composition of PP reveals neurotransmitter involvement in the PP-dependent movement of the juvenile. To ensure that idea physiologically, the movement of 3-dpm juveniles was assayed under the presence or absence of 3-MPA. The assay procedure was constituted with two periods, as described in the Materials and Methods, and summarized by Fig. 1A. After the initial 60 min of the assay period, the motility of the three groups were severely inhibited, except for the control juveniles ( Fig. 1B-1; red circles are juveniles that did not move), which indicated that the inhibitory effect of 3-MPA was not rescued by the exogenous GABA. After the second 120 min of the assay period, while the control juveniles crept actively, those that replaced with fresh seawater after the initial 3-MPA treatment (3-MPA-W) restored the active creeping, indicated that the 3-MPA effect was invalidated by the washing. However, the juveniles that constantly incubated with 3-MPA and those incubated in a mixture of 3-MPA and 1 μM GABA (3-MPA+GABA) retained inactive motility, indicating exogenous GABA did not restore the 3-MPA-induced suppression. (Fig. 1B-2; red circles are those that did not move). Similarly, the average proportion of juveniles that moved throughout the assay period was similar before and after the washing in the control at 67.51 ± 2.46 (Fig. 1C, "before"; blue columns) and 64.14 ± 7.3 (Fig. 1C, "after"; red columns), 19.22 ± 6.96 (before) and 12.83 ± 4.44 (after) in 3-MPA treated (before) that was incubated with fresh 3-MPA (after). Addition of GABA to the 3-MPA medium did not significantly restore the movement from 13.05 ± 3.3 www.nature.com/scientificreports www.nature.com/scientificreports/ (before) to 7.2 ± 2.62 (after). However, washing out the inhibitor with normal seawater significantly restored the motility (two-tailed P value = 0.0013) from the initial 4.13 ± 2.16 (before) to 36.57 ± 14.6 (after). Thus, the creeping movement was found to be dependent on the GABAergic signal transmission system.

3-MPA-independent expression of GABA-IS at PP and -dependent expression at PrR/PrP
complex. Above 3-MPA-induced inhibition of the juvenile motility implicated a decline of GABA-IS in the PP. However, the present stacked image of the CLSM detected little decline of GABA-IS at gaiter region of PP. (Fig. 8B). This revealed that GABA-IS in gaiter region of PP is little or not involved in the creeping movement of the juveniles. However, GABA-IS was also detected in the PrR/PrP complex of the juvenile (Figs. 6 and 8A).  (Fig. 8B), and the GABA-IS was scarcely restored even by the addition of exogenous GABA (Fig. 8C, PrR). Washing the juveniles with plain FSW for 1 to 2 h significantly restored the GABA-IS (Fig. 8D), although the immunosignal intensity was not fully recovered to the level that was shown by the control. This could contribute to the incomplete restoration of motility (Fig. 1C). The representative WMIHC images of the resultant of the above four experiments were summarized by a frequency distribution graph (Fig. 8E) that indicated considerable similarity with the above swimming activity (Fig. 1C).
Next, to examine what provided GABA to the PrR/PrP complex in the absence of the exogenous GABA source, the presence of the endogenous GABA supply system, a GAD-IS site, was examined using 3-D image reconstructed CLSM images. The 3-D image detected the GABA-IS PrR/PrP complex being accompanied by the GAD-IS area (Fig. 8F, red arrow). An optical 26 μm thick cross-section that was dissected horizontally between GABA-IS PrR and PrP, which was consistent with above Fig. 7B, had only GABA-IS PrR (Fig. 8G), while the other side of the cross-section contained GAD-IS region (Fig. 8H, red arrow), which was closely associated with GABA-IS PrP (Fig. 8H). The single filter analysis of the oral PrP region detected the GAD-IS cell plexus (Fig. 8I, inset arrow). To further analyze the spatial relationship between the GABA-IS region and the GAD-IS region of PrP, an optical cross-section was cut perpendicularly through the middle of PrP was produced (Fig. 8H, dotted line j1-j3). The optical cross-section showed the GAD-IS region throughout the oral-aboral axis of PrP (Fig. 8J1) and a little thinner GABA-IS region (Fig. 8J2). The merged image between these two revealed a slightly wider GAD-IS region to the oral side of the PrP (Fig. 8J3, arrow). This suggested the occurrence of GAD-IS/GABA-IS area interactions in PrP and, thus, a possible GABA supply from PrP to PrR and then to the proximal region of the PP. This further verified that the GABAergic regulatory system of juvenile motility resides at the PrR/PrP complex but may not be in the PP.

Discussion
The "basal blastocoel GAD-IS-network" contacts to the ciliary band-associated strand ( Fig. 2A,B), which suggests some signal transmission between them 1,20,21 . A part of the network reaches also to the EcR surface and PP. Thus, the body surface GAD-IS cells that contact the distal spiny projections of PP in juvenile may be derived from the GAD-IS network in plutei 22 . Although the biological role of these body surface GAD-IS cells other than the GABA supply needs to be examined in the future, some other molecules could contribute to the 3-MPA resistant GABA supply to the gaiter region of PP, such as aldehyde dehydrogenase1a1 (ALDH1a1) 23 and the similar molecule is reported in sea urchins, such as ALDH1 in H. pulcherrimus by HpBase (http://cell-innovation.nig.ac.jp/ cgi-bin/Hpul_public/Hpul_annot_search_output.cgi) 24 and in S. purpuratus by SpBase (https://urchin.nidcr.nih. gov/cgi-bin/exp.plx). This will be discussed later.  25 ], and its formation may include the organ rearrangement mechanism, such as apoptosis 26,27 .
During the final stage of metamorphosis, the 8-arm pluteus larva occasionally extends PP to anchor or creeping on the substrate. This is interrupted frequently by swimming with the ciliary beating at the epaulettes, which implies that their respective motilities are not interrupted by each other. Thus, both neuroregulatory systems are valid, and the larval GABAergic system remains active despite the ongoing absorption of the larval organs, as has been reported for the remained partial larval organs in H. pulcherrimus 6 . A part of the mechanism of the irrespective movement of these motile organs may be due to disconnection of adult and larval nerve fiber 27 at the node nearby EcR 7 .
The pharmacological response of TF to GABA in major echinoderm species 12,28,29 , and its immunohistochemical detection was reported in starfish 11 and sea cucumber 30 , but not in sea urchins to date. The present striking GABA-IS site was the gaiter region of the juvenile PP. Along with the GABA A R-IS layer on the muscle layer, this resembles GABA A R involvement in the neuromuscular junctions as has been reported in C. elegans 31 . It also suggests GABA involvement in PP motility-related osmotic regulation via GABA A R signaling 32 , such as that reported in GABA-induced depolarization of echinoderm TF muscle preparations 12,28 .
The GAD-IS was seen accompanied by GABA-IS in PP from the later stages of metamorphosis to juvenile. The distal tip of the PP radiated numerous projections toward the GAD-IS network of the body-surface, and they are apparently in contact with each other. This may indicate a morphological GABA transmission pathway to the PP. In larvae, ciliary band-associated strand that contains GAD transmits GABA to CB via synaptophysin-involved mechanism 14 . Thus, synaptophysin mediation could be predicted in GABA transmission to the gaiter region of the PP.
The present study detected 3-MPA-insensitive or low-sensitive GABA-IS at the gaiter region of the PP, which reminds inhibitor-insensitive GAD activity in mice and the fish, Salmo irideus 33 or GAD independent GABA syntheses, such as glial monoamine oxidase B in the cerebellum and the striatum of adult mice 33 . Regarding our previous reports of the GAD expression in ciliary band-associated strand 13,14 and the dopamine expression at nearby CB 6 , involvement of the GAD-independent ALDH1a1-dependent GABA synthetic pathway in Xenopus tadpole and mammalian cells 23 may also be a possible interpretation. Regarding that ALDH1 and the likes are reported in sea urchin genome databases (HpBase; http://cell-innovation.nig.ac.jp/cgi-bin/Hpul_public/Hpul_ annot_search_output.cgi, SpBase; https://urchin.nidcr.nih.gov/cgi-bin/exp.plx?ind=1amp;cname=Sp-Aldh1am p;glean=GLEAN3_01700), such possibility can be considered.
In the PrR/PrP complex, both components are connected via several intra-structural GABA-IS cables (Fig. 7A), which suggests that GABA is synthesized at the GAD-IS cells in/near PrP and transmitted to PrR through the cables, and the neurotransmitter is delivered from there to PP to regulate the creeping movement. Furthermore, the neurotransmitter may first reach the GABA A R-IS layer on the outer surface of the muscle layer of PP, which suggests the occurrence of the receptor-mediated signal transmission in the regulation of muscle contraction for the creeping movement. The presence of GABA A R on the muscle layer of PP is consistent with an electrophysiological report on the holothurian neuromuscular junction 29 . This suggests the occurrence of the excitatory responses to PP, as has been predicted in echinoderms 30 and is apparently consistent with the present inhibition of the creeping movement of juveniles by 3-MPA application. This will be further discussed in the next section. A similar GABA-IS radial nerve cord and that in PP were reported in the asteroid Asterias rubens 11 . However, such organization as the PrR/PrP complex has never been reported to date in other echinoderms.
The reversible GABA-IS vanishing at the PrR/PrP complex following the administration of 3-MPA, which resulted in the inhibition of creeping motility of juveniles, while that in the gaiter region of PP has remained unaffected. This implicates that there are different effects of the inhibitor on GAD activity between in the PrR/PrP complex that contributes PP activity and the other near the body surface and the gaiter region that is not involved www.nature.com/scientificreports www.nature.com/scientificreports/ in PP motility. The present inhibitory effect of 3-MPA for juvenile motility implicates that GABA acts as an excitatory neurotransmitter on the PP musculature, as has been widely reported for the echinoderm such as the sea urchin Strongylocentrotus franciscanus 12 and the starfish Asterias amurensis 28 . Involvement of GABA A R in GABA signal transmission has also been reported to be excitatory in the sea cucumber, Holothuria glaberrima 30 . Unlike the GAD-IS plexus of the PrP/PrR complex for the Penta-radial ring, the GABA source for the gaiter region of PP could not solely depend on GAD activity, as has been suggested the GAD-independent GABA synthesis pathway, such as putrescine, as an initial substrate reported in mouse and fish brain 33,34 and sea urchin 35 . Although >90% of GABA synthesis activity by GAD67 subunit against GAD65 subunit has also been reported 36 , no such heterogenous subunit complex has been reported in the sea urchin genome (HpBase: http://cell-innovation.nig.ac.jp/ cgi-bin/Hpul_public/Hpul_annot_search_output.cgi 24 ) to date. Thus, the molecular basis by which to interpret the different responsiveness of GAD in sea urchins needs to be examined in the future. It also would be important to extend this investigation to the definitive juvenile.