Adult X. laevis was purchased from Watanabe Yosyoku (Hyogo, Japan) and embryos were obtained after artificial fertilization. Oocytes were obtained from female injected with 500 IU human chorionic gonadotropin (Sigma-Aldrich) 15 hours earlier, and fertilized with minced testis. Jelly coat was removed by 3% cysteine (pH 8.0) treatment, and embryos were maintained in 10% Steinberg’s solution (1x Steinberg’s solution: 58 mM NaCl, 0.67 mM KCl, 0.34 mM Ca(NO₃)₂・4H₂O, 0.83 mM MgSO₄・7H₂O, 10 mM HEPES, pH 7.3 at 23˚C) until the described stages. The developmental stages were determined according to [11] . For microinjection, mRNAs and MOs were injected into the dorsal animal blastomere of eight-cell-stage embryos. β-Galactosidase (β-Gal) (400 pg) and yellow fluorescent protein (YFP) (400 pg) mRNAs were co-injected to confirm the injected region.

MOs were designed (Gene Tools) for the P2Y1 or P2Y11 transcript. The P2Y1 MO sequence is 5’-gagagaaagacttctgtcatgatct-3’ [6] ; the P2Y11a MO sequence is 5’-ttgcagacggaggaagccatttatt-3’; the P2Y11b MO sequence is 5’-ttgcaggtgatgga- agccattcttc-3’; and the COMO sequence is 5’-CCTCTTACCTCAGTTA- CAATTTATA-3’. Each MO (2 mM stock in sterilized water) was adjusted to the concentration of 0.4 mM, and 4 nl of the solution was injected. P2Y11a and P2Y11b likely represented alternative copies of the same genes. P2Y11 MO is a 1:1 mixture of P2Y11a MO and P2Y11b MO. To confirm the injected region, β-Gal or YFP mRNA was co-injected. β-Gal activity of cytoplasmic lacZ was visualized with 5-bromo-4-chloro-3-indolyl-β-D-galactopyrano- side (X-gal, Wako).

Embryos were collected and fixed in MEMFA [12] , whole-mount in situ hybridization (WISH) was performed as previously described [13] with minor modifications. For the chromogenic reaction, nitro blue tetrazolium/5-bromo-4- chloro-3-indolyl phosphate (NBT/BCIP, Roche) was used as the substrate of alkaline phosphatase. Embryo images were collected on a LEICA M2 FLⅢ microscope (Leica Microsystems). Antisense RNA probes for in situ hybridization were prepared from a template encoding anf1 [14] , en2 [15] , pax6 [16] , foxE3 [17] , l-maf [18] , rx1 [19] , dlx5 [20] , bf1 [21] , six6 [22] , krox20 [23] , n-cam [24] and otx2 [25] . pCS2AT+ containing P2Y1 or P2Y11 was constructed with the following primers: P2Y1-F (5’-CCCCCTCGAGACCATGACAGAAGTCTTTC- TCTCAGCT-3’); P2Y1-R (5’-CCCCGGCGCGCCTCACAAGCTTGTGTCCCC- ATTCTG-3’); P2Y11-F (5’-GGGATCGATCCACCATGGCTTCCTCCGTCTGC- A-3’) and P2Y11-R (5’-GGGCTCGAGTTAAAGTCTACTCTCTCCTTGG-3’). Digoxigenin-labeled-antisense RNAs were generated by in vitro transcription with a MAXIscript Kit (Ambion) and an RNA Labeling Kit (Roche).

Embryos at stage 45 were fixed in MEMFA and stained overnight in 0.2% alcian blue/30% acetic acid in EtOH. After that, the embryos were washed with 80% glycerol/2% KOH.

Fixed embryos were rehydrated in PBS containing 0.1% Tween 20 and washed with terminal deoxynucleotidyltransferase (TdT) buffer (Takara) for 30 min. End labeling was carried out at room temperature overnight in TdT buffer containing 2 uM digoxigenin-11-dUTP (Roche) and 150 U/ml TdT (Takara). Embryos were washed with 1 mM EDTA in PBS for 1 hour and detected with NBT/BCIP.

Embryos were fixed with MEMFA overnight and dehydrated with 100% MeOH. The samples were embedded in paraffin (Pathoprep 568, Wako) and sectioned to 10 um thickness. The sections were deparaffinized and stained with hematoxylin and eosin.

To analyze the spatial expression patterns of P2Y1 and P2Y11 during early development in Xenopus embryos, we performed whole-mount in situ hybridization. P2Y1 was first detected in the animal hemisphere and the dorsal blastopore lip at early gastrula stage (Figure 1(A) & Figure 1(B)), and the expression was maintained in the prospective neuroectoderm and the involuting mesoderm at mid-gastrula stage (Figure 1(C)). In neurulae, P2Y1 expression was observed in the neural fold (Figure 1(D) & Figure 1(G) arrowheads) and the anterior neural plate (Figure 1(E), Figure 1(F), Figure 1(H)), and P2Y1 transcripts were also found in the notochord at mid-neurula stage (Figure 1(I)). At late neurula stage, P2Y1 was detected in the central nervous system (Figure 1(J), Figure 1(K)). Further, transverse sections showed that P2Y1 localized to the dorsal structures, including the notochord and the area around the neural tube (Figure 1(L)). P2Y1 localized to various head regions, including eye, otic vesicle, anterior neural tissues and migrating cranial neural crest cells (Figure 1(M)). Thereafter, P2Y1 was detected in the pallium, the dorsal domain of telencephalon, dorsal diencephalon, mesencephalon and metencephalon at tadpole stage (Figure 1(N), Figure 1(O)). The earliest P2Y11 expression was also observed in the dorsal blastopore lip at early gastrula stage (Figure 2(A)), and a sagittal hemisection showed an expression pattern similar to P2Y1 expression (Figure 2(B), Figure 2(C)). The P2Y11 expression observed in the prospective neuroectoderm, the involuting mesoderm, and the central nervous system from gastrula to neurula stage had been reported previously [7] . P2Y11 was expressed in the neural fold (Figure 2(D), Figure 2(G), arrowheads), the anterior neural plate (Figure 2(D), Figure 2(F), Figure 2(G), Figure 2(H)), and the posterior mesoderm (Figure 2(F)) at early-mid neurula stage, and in the entire central nervous system at late neurula stage (Figure 2(J), Figure 2(K)). Further, transverse

sections showed that P2Y11 was expressed in the notochord and the peripheral area of the neural tube (Figure 2(I), Figure 2(L)). Thereafter, P2Y11 was detected in the migrating cranial neural crest cells (Figure 2(M), arrowheads), the otic vesicle, and the anterior neural tissue at early tailbud stage. At tadpole stage, P2Y11 localized to restricted parts of the brain, the telencephalon, the dorsal diencephalon, the mesencephalon, and the anterior metencephalon (Figure 2(N)). Thus, the expression of P2Y1 and P2Y11 during early development showed similar patterns, namely, they were predominantly detected in the head region and the notochord. Although little is known about the role of these genes in early development in X. laevis, it had been shown that E-NTDP2 and P2Y1 synergistically regulated eye development [6] [26] . Moreover, P2Y11 was expressed in the notochord and was important for convergent extension, coordinated cell polarization, and elevation of intracellular calcium in chordamesoderm cells [8] . These expression patterns indicated that P2Y1 and P2Y11 were expressed in vital areas and the head formation stages, suggesting that P2Y1 and P2Y11 might have the same functions, assisting each other in the same tissues, particularly in the head region.

To investigate the role of P2Y1 and P2Y11 in head formation, we knocked down P2Y1 or P2Y11 by injecting MOs. P2Y1 MO (13.6 ng), P2Y11 MO (13.6 ng) or control MO (COMO: 29.9 ng) was injected with β-gal mRNA lineage tracer (400 pg) into one dorsal animal blastomere of 8-cell stage embryos. Embryos injected with P2Y1 MO or P2Y11 MO exhibited head malformation with small eyes (Figure 3(B), Figure 3(C), arrowheads), brain atrophy (Figure 3(H), Figure 3(I), arrowheads), and cartilage tissue defect (Figure 3(K), Figure 3(L), arrowheads), whereas those injected with COMO did not show head malformation (Figure 3(A), Figure 3(G), Figure 3(J), arrowheads). Further, histological sections of abnormal eyes in P2Y1 MO or P2Y11 MO injected embryos showed eye defects (Figure 3(E), Figure 3(F) arrowheads). These phenotypes could be divided into two groups: the severe group that showed serious eye defects, and the moderate group that showed slight malformation (P2Y1 MO: severe, 68.3%, moderate, 29.3%, n = 41, P2Y11 MO: severe, 64.8%, moderate, 29.6%, n = 54). P2Y1 or P2Y11 knockdown resulted in similar head malformation at tadpole stage, suggesting that P2Y1 and P2Y11 could be involved in the same normal head formation process.

To analyze the effects of P2Y1 or P2Y11 depletion on Xenopus head formation in detail, we examined the expression of anterior neural marker genes (Figure 4). Embryos injected with P2Y1 MO or P2Y11 MO exhibited a significant decrease in the expression of fore to midbrain marker otx2 at mid-neurula stage (reduction in P2Y1 MO: 60%, n = 20, P2Y11 MO: 66.7%, n = 18, Figure 4(B)

and Figure 4(C), arrowheads) and at late neurula stage (P2Y1 MO: 40%, n = 20, P2Y11 MO: 52.9%, n = 17, Figure 4(E) and Figure 4(F), arrowheads). The expression of retinal maker pax6 was not significantly affected at mid-neurula stage (Figure 4(G), Figure 4(H), Figure 4(I)), but was decreased at late neurula stage (reduction in P2Y1 MO: 69.7%, n = 33, P2Y11 MO: 51.6%, n = 31, Figure 4(K) and Figure 4(L), arrowheads). The expression of another retinal marker, rx1, was slightly affected by the depletion of P2Y1 or P2Y11 (P2Y1 MO: 33.3%, n = 21, P2Y11 MO: 64%, n = 25, Figure 4(N) and Figure 4(O)) at stage 20. Further, at mid-neurula stage, the expression of hindbrain marker krox20, midbrain- hindbrain boundary marker en2, and telencephalon marker bf1 was markedly reduced in embryos injected with P2Y1 or P2Y11 MO (P2Y1 MO: 63%, n =

11, P2Y11 MO: 56%, n = 16, Figure 4(Q), Figure 4(R), arrowheads). These data indicated that P2Y1 or P2Y11 knockdown affected the expression of anterior neural markers at neurula stage, consistent with the morphological phenotypes shown in Figure 3 (Figure 3(B), Figure 3(C), Figure 3(E), Figure 3(F), Figure 3(H), Figure 3(I)). As discussed earlier, P2Y1 and P2Y11 were expressed in the prospective dorsal region of the anterior neural plate (Figure 1(H), Figure 2(H)) and localized to the dorsal diencephalon at tadpole stage (Figure 1(O), Figure 2(N)). It is known that otx2 is expressed in the prospective dorsal diencephalon and pax6 is expressed in the ventral diencephalon at neurula stage [27] . Therefore, the results may be reasonable in that the depletion of P2Y1 or P2Y11 causes loss of otx2 expression but has no effect on pax6 expression at mid-neu- rula stage. In the anterior neuroectoderm, anf is the earliest transcriptional repressor gene and plays a key role in rostral forebrain development [14] [28] . It is essential to inhibit the expression of two key regulators, otx2 and pax6, in the posterior forebrain [27] . Based on these reports, we attempted to investigate the effects of P2Y1 or P2Y11 MO on anf1 expression. anf1 was normally expressed in the prospective forebrain and the lens placode region at early neurula stage, and gradually localized to the prospective anterior pituitary at the end of neurulation [29] . anf1 expression in embryos injected with P2Y1 or P2Y11 MO at early neurula stage did not show any significant difference (P2Y1 MO: n = 9, P2Y11 MO: n = 12, Figures 5(A)-(C)), but anf1 expression increased laterally in the injected side at late neurula stage (P2Y1 MO: 87.5%, n = 16, P2Y11 MO: 76.5%, n = 17, Figure 5(E) and Figure 5(F), arrowheads). At early tailbud stage, anf1 localized to the anterior pituitary upon P2Y1 knockdown (n = 45, Figure 5(H)), although it did not localize to the pituitary (93.3%, n = 15, Figure 5(I), arrowhead) in the P2Y11 MO injected side. In Xenopus embryo, anf overexpression can lead to enlargement of the rostral forebrain and inhibit genes that normally express in more posterior regions [14] . Moreover, anf functions as a

suppressor of posterior forebrain fate within the presumptive rostral forebrain and directly inhibits the expression of otx2 and pax6 in this area [27] . Thus, we speculated that ectopic anf1 expression would repress otx2 and pax6 expression during neurulation. At early tailbud stage, anf1 localized to the pituitary in P2Y1 morphants, whereas it did not localize to the pituitary in P2Y11-depleted embryos (Figure 5(H), Figure 5(I)). Although P2Y1 and P2Y11 may play a similar role in head formation, it is likely that they have different regulation mechanisms. The expression domains of en2 and krox20 do not overlap with the anf1 expression area, and anf1 cannot directly regulate the expression of bf1, which is co-expressed in a more anterior region of the neural plate [27] . Thus, we considered that the decrease in en2, krox20, and bf1 expression might be independent of ectopic anf1 expression. Together, our results demonstrated that the depletion of P2Y1 or P2Y11 led to loss of neural development as shown by the down-reg- ulation of some neural marker genes.

As the depletion of P2Y1 or P2Y11 resulted in lens malformation (Figure 3(E), Figure 3(F)), we examined the expression of lens placode marker genes, foxe3 and l-maf, at early tailbud stage (Figures 6(A)-(F)). The morphants showed reduced expression of foxe3 (P2Y1 MO: 71.9%, n = 32, P2Y11 MO: 65%, n = 13, Figure 6(B) and Figure 6(C), arrowheads) and l-maf (P2Y1 MO: 7.1%, n = 14, P2Y11 MO: 73.3%, n = 15, Figure 6(E) and Figure 6(F), arrowheads). Embryos injected with P2Y1 or P2Y11 MO showed significant reduction of pax6 expression in the lens placode region at late neurula stage (Figure 4(K), Figure 4(L), arrowheads). The lens field is lens-biased ectoderm within preplacodal ectoderm and is formed at mid-neurula stage in response to signals emitted from the adjacent anterior neural plate [30] . The reduction of pax6 and rx1 expression suggested that insufficient induction from retina might bring about a defect in lens placode at neurula stage. In addition, we examined the expression of dlx5, which marked the most cranial placode but not the lens placode, and detected a slight decrease in dlx5 expression in P2Y1 or P2Y11 knockdown embryos (P2Y1 MO: 55.6%, n = 18, P2Y11 MO: 42.9%, n = 14, Figure 6(H) and Figure 6(I), arrowheads). Furthermore, we examined the effect of P2Y1 or P2Y11 depletion on the expression of neural crest marker genes, snail1 (Figures 7(A)-(L)) and foxD3 (data not shown), at neurula stage, because P2Y1 or P2Y11 knockdown resulted in a severe defect in cartilage tissues of neural crest derivatives. P2Y1 or P2Y11 knockdown rexpressednail1 expression at early (P2Y1 MO: 75%, n = 8, P2Y11 MO: 80%, n = 10, Figure 7(B), Figure 7(C), arrowheads) and late (P2Y1 MO: 66%, n = 12, P2Y11 MO: 88%, n = 9, Figure 7(E), Figure 7(F), arrowheads) neurula stages, thereby resulting in the reduction of migrating neural crest cells (P2Y1 MO: 83%, n = 8, P2Y11 MO: 80%, n = 10, Figure 7(J), Figure 7(L), asterisk). These results indicated that the loss of neural crest derivatives was caused

by early hypoplasia of the neural crest region. Together, these data suggested that P2Y1 or P2Y11 might also be required for the entire head components, such as the placode and the neural crest cells.

P2Y1 or P2Y11 was detected in the prospective neuroectoderm and the dorsal marginal zone from early gastrula stage (Figure 1, Figure 2), and snail1 expression was affected by the depletion of P2Y1 or P2Y11 at early neurula stage (Figure 7(B), Figure 7(C)). Therefore, we speculated that morphants affected developmental events occurring earlier than neurula stage. To identify the effects of P2Y1 or P2Y11 knockdown on neural fate at early neurula stage, we examined the expression of epidermal marker epi-keratin (Figures 8(A)-(C)) and neural plate marker NCAM (Figures 8(D)-(F)) upon MO injection. We observed an enlargement of the area of epi-keratin expression (P2Y1 MO: 90.9%, n = 11, P2Y11 MO: 76.2%, n = 21, Figure 8(B) and Figure 8(C), yellow arrowheads) coupled with a decrease in the area of NCAM expression (P2Y1 MO: 54.5%, n =

22, P2Y11 MO: 28.6%, n = 28, Figure 8(E) and Figure 8(F), red arrowheads) at the P2Y1 or P2Y11 MO injected side. P2Y1 or P2Y11 morphant also showed the “salt and pepper” pattern for epidermal marker expression in the neural plate. Taken together, these data suggest that the depletion of P2Y1 or P2Y11 may affect the induction and/or maintenance of neural tissue. Bone morphogenetic protein (BMP) signaling plays a role in epidermal fate establishment at the presumptive ectoderm, and BMP antagonists, such as Noggin, Chordin and Follistatin, induce neural tissue by antagonizing BMP activities [31] [32] [33] . P2Y1 and P2Y11 were expressed in the neural ectoderm, the endomesoderm and the involuting mesoderm (Figure 1, Figure 2). Thus, we speculated that P2Y1 and P2Y11 might be required for the establishment of the organizer and/or the competence of the ectoderm for neural induction.

Furthermore, we examined whether the observed effects of P2Y1 or P2Y11 knockdown might be attributed to an increase in the number of apoptotic cells. We performed TUNEL assay and found significant enhancement of apoptosis at neurula stage in embryos injected with P2Y1 or P2Y11 MO (P2Y1 MO: 42.9%, n = 14, P2Y11 MO: 100%, n = 10, Figure 8(H) and Figure 8(I); P2Y1 MO: 53.3%, n = 15, P2Y11 MO: 50%, n = 10, Figure 8(K) and Figure 8(L)), whereas COMO did not result in a significant enhancement of apoptosis in the injected side (n = 11, Figure 8(G); n = 12, Figure 8(J)). However, the number of apoptotic cells was

not sufficient to allow interpretation of the observed phenotypes, and minimal apoptosis was detected at early neurulastage (data not shown). P2Y receptors increase intracellular Ca²⁺ concentration and modulate the intracellular environment as second messengers [34] . The increase in intracellular Ca²⁺ concentration is known to regulate cell proliferation, migration and differentiation in early development [35] . In fact, intracellular Ca²⁺ increase is induced in SCs of MOE through P2Y4 receptors in X. laevis, and is involved in cell turnover [9] [10] [36] . It is also known that cell proliferation is essential for the induction and maintenance of neural crest cells [37] , and P2Y1 and P2Y11 may maintain neural crest cells in proliferating and undifferentiated states. Although we could not entirely exclude the possibility that the depletion of P2Y1 or P2Y11 might directly induce apoptosis, we considered that the significant increase in apoptosis at mid and late neurula stages might be due to reduced cell survival as development progressed.

In the present study, we revealed that P2Y1 and P2Y11 receptors were involved in X. laevis head formation. The knockdown approach revealed loss of neural development, suggesting that these receptors would be required for the induction and/or maintenance of neural tissues. As P2Y receptors in X. laevis have already been classified into six subtypes, the cellular and molecular functions of these receptors, including P2Y1 or P2Y11, should be studied further to fully understand the roles they play. The fact that P2Y receptors exist in only vertebrates and are highly conserved, and that P2Y1 or P2Y11 is involved in head formation has led us to expect that P2Y receptors participate in the acquisition of vertebrate head during evolution. As only the physiological and pathophysiological functions of P2Y receptors in the nervous system have been investigated, it is important to study further the roles of these receptors and their related molecules in early development.