Nonreproductive role of gonadotropin‐releasing hormone in the control of ascidian metamorphosis
Abstract
Background: Gonadotropin‐releasing hormones (GnRHs) are neuropeptides that play central roles in the reproduction of vertebrates. In the ascidian Ciona intestinalis, GnRHs and their receptors are expressed in the nervous systems at the larval stage, when animals are not yet capable of reproduction, suggesting that the hormones have non‐reproductive roles. Results: We showed that GnRHs in Ciona are involved in the animal′s metamorphosis by regulating tail absorption and adult organ growth. Absorption of the larval tail and growth of the adult organs are two major events in the metamorphosis of ascidians. When larvae were treated with GnRHs, they completed tail absorption more frequently than control larvae. cAMP was suggested to be a second messenger for the induction of tail absorption by GnRHs. tGnRH‐3 and tGnRH‐5 (the “t” indicates “tunicate”) inhibited the growth of adult organs by arresting cell cycle progression in parallel with the promotion of tail absorption. Conclusions: This study provides new insights into the molecular mechanisms of ascidian metamorphosis conducted by non‐reproductive GnRHs. Developmental Dynamics 243:1524–1535, 2014. © 2014 Wiley Periodicals, Inc.
Introduction
Ascidians are members of the subphylum Tunicata of the phylum Chordata (Lemaire, 2011), and Tunicata is the closest relative to Vertebrata (Satoh, 2003; Delsuc et al., 2006). As this phylogenetic position suggests, swimming larvae of ascidians are typical tadpoles and possess characteristics of chordates as exemplified by a dorsal neural tube and notochord. Ascidians are known to metamorphose dramatically to adjust their bodies to be suitable for sessile adult life (Cloney, 1982). Ascidian swimming larvae become attached to substrates by means of their papillae, which are adhesive organs located at the anterior‐most part of the larval trunk. The papillae contain epidermal neurons (Takamura, 1998), and attachment is thought to stimulate the adhesive papillae to trigger the initiation of metamorphosis through the nervous systems (Nakayama‐Ishimura et al., 2009). After metamorphosis, ascidians lose their tails and some other larval organs while developing adult organs that are responsible for feeding and settlement (Fig. 1A–E). Ascidians that have metamorphosed do not exhibit some of the characteristics specific to chordates, such as the notochord; however, ascidians acquire other characteristics of chordates such as the gill at the pharynx and the endostyle that is homologous to the thyroid gland (Ogasawara et al., 1999), suggesting that ascidian metamorphosis does not represent an escape from the tadpole body but can be considered a highly divergent modification of the tadpole structure.
Gonadotropin‐releasing hormone (GnRH) induces tail absorption. A‐D: Metamorphosis of Ciona intestinalis. A: A larva, lateral view. Anterior is toward the left. Pa, adhesive papilla; Po, preoral lobe; Tr, trunk; Ta, tail. B: An animal absorbing its tail. C: A juvenile that has completed tail absorption. AT, absorbed tail. D: A juvenile that has developed adult tissues. Es, endostyle; Gi, gill. E: Induction of tail absorption by GnRHs at a concentration of 10 µM. Blue and red bars respectively illustrate the ratio of tail‐absorbed juveniles at 1 day and 2 days after hatching (dph). The experiments were repeated three times. Scores whose significance was not supported by the Student′s t‐test compared to the results of Ci‐LFs are indicated by asterisks. Black bars represent standard errors. no treatment, animals cultured in seawater. LFs, animals cultured in seawater containing the peptide LF‐1, 2, 3, 4, 5, 6, 7 or 8. N indicates the number of counted animals. F: Induction of tail absorption by cAMP at a concentration of 1 mM. “NaOH” indicates treatment with seawater containing sodium hydroxide as the solvent of 8‐Br‐cAMP and 8‐Br‐cGMP. N indicates the number of counted animals.
Ten typical events that constitute ascidian metamorphosis have been listed (Cloney, 1982). We classified these metamorphic events into three groups in the previous study (Nakayama‐Ishimura et al., 2009). The groups include cellulose‐sensitive events, destructive events and cell‐division‐requiring events. Cellulose‐sensitive events include morphogenesis during metamorphosis, namely rotation of the body axis; the name is derived from the phenotypes of cellulose‐absent mutants (Sasakura et al., 2005). Destructive events are related to the loss of larval structures including the adhesive papillae and tail. Cell‐division‐requiring events are events involved in the growth of adult organs that are in an immature state in the larval body.
Although adhesion serves as the initial cue for the onset of all metamorphic events, there are several different downstream pathways, each of which triggers a subgroup of metamorphic events (Nakayama‐Ishimura et al., 2009). The presence of downstream pathways is supported by observations regarding the phenotypes of mutants. In the ascidian Ciona intestinalis, two mutants that show defects in metamorphic events have been isolated (Sasakura et al., 2005; Nakayama‐Ishimura et al., 2009). The common characteristic of the two mutants′ phenotypes is that some metamorphic events occur but others are not initiated, indicating that the causative genes for the mutants are responsible for the initiation of a subset of metamorphic events. The causative gene of one of the mutants, swimming juvenile (sj), encodes cellulose synthase, suggesting that cellulose in the tunic is involved in the regulation of metamorphic events.
The initiation of ascidian metamorphosis is not controlled solely by mechanisms that activate metamorphic events; the presence of a mechanism inhibiting the initiation of metamorphic events has also been suggested (Nakayama‐Ishimura et al., 2009). Ascidian larvae acquire the competence for metamorphosis several hours after hatching. Having acquired the competence for metamorphosis, larvae can begin to metamorphose when they attach to substrates with their papillae. Even if larvae have the competence for metamorphosis, they do not start to metamorphose without attachment, and they keep their larval shape during swimming. When the larval preoral lobe, the anterior part of the trunk containing the adhesive papillae, is surgically removed from larvae at the pre‐competent stage, these larvae autonomously start the growth of adult organs without other metamorphic events including adhesion and tail absorption (Nakayama‐Ishimura et al., 2009). Therefore, the preoral lobe and particularly the adhesive papillae have a role in suppressing the initiation of the growth of adult organs until settlement. The molecular mechanism that inhibits the initiation of adult organ growth by the organs is not known.
The gonadotropin‐releasing hormones (GnRHs) are a group of peptide hormones that function in the maturation of reproductive organs (Millar et al., 2008). In C. intestinalis, six GnRHs (tGnRH‐3 to tGnRH‐8; the “t” indicates “tunicate”) have been reported (Adams et al., 2003). Ascidian GnRHs have a unique characteristic in that they share a transcription unit: three GnRH peptides including tGnRH‐3, tGnRH‐5 and tGnRH‐6 are encoded by the single gene Ci‐GnRH1, and the remaining three are encoded by another single gene, Ci‐GnRH2. Transcripts of these genes are thought to be translated into single polypeptides, which are then cleaved to form mature forms of GnRH peptides. GnRHs are received by receptors that are seven‐pass transmembrane proteins. Four receptors of ascidian GnRHs including Ci‐GnRHR1 to Ci‐GnRHR4 have been identified from C. intestinalis (Kusakabe et al., 2003; Tello et al., 2005; Sakai et al., 2010). Ascidian GnRHs are known to induce spawning of gametes (Terakado, 2001).
Recently, the expression patterns of GnRHs and their receptors have been reported in larvae of C. intestinalis (Kusakabe et al., 2012). Genes encoding GnRHs and their receptors are strongly expressed in the peripheral and central nervous systems including the papillae, suggesting that they play a role in the larval nervous system. Because larvae of C. intestinalis do not have gonads (Okada and Yamamoto, 1999) and they are in a sexually immature state (Shirae‐Kurabayashi et al., 2006), the functions of larval GnRHs may be unrelated to reproduction. As GnRHs are expressed in the adhesive papillae, we assumed that larval GnRHs may have roles in metamorphosis. In the present study, we found that GnRHs are responsible for the control of two major metamorphic events, namely tail absorption and adult organ growth, in both the initiative and inhibitory modes.
Results
GnRHs Promote Initiation of Metamorphosis
To examine whether GnRHs play roles in ascidian metamorphosis, we administered synthetic GnRHs to larvae of C. intestinalis. Ascidian larvae are surrounded by a tunic layer which could be a barrier inhibiting the uptake of molecules from seawater. To facilitate the uptake of GnRH peptides by larvae, we cut off the posterior portion of their tails. The previous study showed that tail‐cut larvae rarely start to metamorphose under natural conditions because vigorous swimming is necessary to keep larvae attached to substrates; however, the tail‐cut larvae can complete their metamorphosis once they receive a certain stimulus that is sufficient for initiating metamorphosis (Nakayama‐Ishimura et al., 2009). Therefore, we were able to address whether GnRHs can induce the initiation of metamorphosis in tail‐cut larvae. We identified the occurrence of metamorphosis by the completion of tail absorption (Fig. 1C). As a negative control we used another peptide group, namely Ci‐LF‐1 to Ci‐LF‐8 (Kawada et al., 2011), whose expression was also detected in the larval central nervous system in the previous study (Hamada et al., 2011).
As shown in Figure 1E, all of the GnRHs induced tail absorption much more frequently than controls, when the peptides were administered to larvae at a concentration of 10 µM. The efficiency with which they induced metamorphosis was different among GnRHs. tGnRH‐3 showed the highest frequency of induction (83%, n = 65) at 48 hr after administration, while tGnRH‐7 showed the lowest activity (43%, n = 62).
It is known that cyclic AMP (cAMP) is the major second messenger of GnRH receptors in Ciona (Tello et al., 2005; Sakai et al., 2010). Indeed, all Ciona GnRH receptors can up‐regulate cAMP (Tello et al., 2005; Sakai et al., 2010, 2012). Therefore, it is possible that cAMP triggers tail absorption as the downstream factor of GnRH receptors. To show this, we administered a membrane‐permeable cAMP analog (8‐Br‐cAMP) to tail‐cut larvae at the concentration of 1.0 mM. As a result, most of these cAMP‐administered larvae completed the process of tail absorption (Fig. 1F). We next examined the effect of a cGMP analog (8‐Br‐cGMP) to tail absorption and found that cGMP did not induce tail absorption (Fig. 1F). We concluded that cAMP is a strong candidate to be the second messenger of the GnRH signaling pathway that induces the initiation of tail absorption.
We investigated whether lower concentration of GnRHs can induce tail absorption. When GnRHs were administered at a concentration of 1.0 µM, their induction of tail absorption was generally not significant compared to controls (Fig. 2A); tGnRH‐8 was the only GnRH that showed significant induction of tail absorption. We next investigated whether lower concentrations of cAMP can induce tail absorption. As shown in Figure 2B, cAMP at a concentration of 100 µM was unable to induce tail absorption significantly. The requirement of a relatively higher concentration of GnRH ligands and cAMP for inducing tail absorption can be explained by inefficient penetration of these molecules into larval bodies due to the reasons mentioned above.
Lower concentration of gonadotropin‐releasing hormones (GnRHs) and cAMP do not induce tail absorption effectively. A: Induction of tail absorption by GnRHs at a concentration of 1.0 µM. The blue and red bars illustrate the ratio of tail‐absorbed juveniles at 1 day and 2 days after hatching (dph), respectively. The experiments were repeated three times. The black bars represent standard errors. No treatment, animals treated with the solvent of GnRHs (H2O). The Student′s t‐test supported the significant induction of tail absorption by tGnRH‐8 (the “t” indicates “tunicate”), but did not support significance for other GnRHs. B: Induction of tail absorption by cAMP at a concentration of 100 µM. “NaOH” indicates treatment with seawater containing sodium hydroxide as a solvent of 8‐Br‐cAMP and 8‐Br‐cGMP. The experiments were repeated twice. The Student′s t‐test did not support the significance of the scores.
tGnRH‐3 and tGnRH‐5 Inhibit Growth of Adult Organs During Metamorphosis
We next examined whether GnRHs can regulate the growth of adult organs. For this purpose, we observed the morphology of juveniles at 48 hr after the administration of GnRHs. The larvae that were administered tGnRH‐4, tGnRH‐6, tGnRH‐7 and tGnRH‐8 had well‐grown bodies with developed adult organs (Fig. 3A,E). The larvae that were administered tGnRH‐8 had the highest percentage of animals with adult organ growth (90%; n = 42), and the larvae administered tGnRH‐7 had the lowest percentage (71%; n = 27). Seventy percent of the larvae treated with 8‐Br‐cAMP developed adult organs (n = 46; Fig. 3D,E).
Effects of gonadotropin‐releasing hormones (GnRHs) on adult organ growth. A: A 2‐days after hatching (dph) ‐old juvenile cultured in seawater containing 10 µM tGnRH‐6 (the “t” indicates “tunicate”). This animal had well‐developed adult organs such as the endostyle (Es) and gill (Gi). B: A 2‐dph‐old juvenile cultured in seawater containing 10 µM tGnRH‐3. Adult organs were poorly developed. C: A 2‐dph‐old juvenile cultured in seawater containing 10 µM tGnRH‐5. Adult organs were poorly developed. D: A 2‐dph‐old juvenile cultured in seawater containing 1 mM cAMP. This animal had developed adult organs. E: Effects of GnRHs on adult organ growth. Blue bars illustrate % of juveniles that had developed adult organs among juveniles that had completed tail absorption. GnRHs and cAMP were administered at concentrations of 10 µM and 1 mM, respectively. The experiments were repeated three times. Black bars represent standard errors. N indicates the number of counted animals. F: tGnRH‐3 and tGnRH‐5 canceled the adult organ growth induced by tGnRH‐8. GnRHs were administered at a concentration of 10 µM. This experiment was repeated three times. N indicates the number of counted animals.
In contrast to the case of the four GnRHs and cAMP, the animals that were administered tGnRH‐3 and tGnRH‐5 did not show any development of adult organs (Fig. 3B,C,E). There are two possible interpretations of this result. The first is that tGnRH‐3 and tGnRH‐5 do not promote adult organ growth. The other is that tGnRH‐3 and tGnRH‐5 each play a role in inhibiting the growth of adult organs. To determine which of these two possibilities is correct, we treated larvae with tGnRH‐3 and tGnRH‐5 in addition to tGnRH‐8, which had resulted in the highest rate of animals that grew adult organs. The results showed that tGnRH‐3 and tGnRH‐5 can override the adult organ growth that is initiated by tGnRH‐8 (Fig. 3F). These results suggest that tGnRH‐3 and tGnRH‐5 have inhibitory activity against the growth of adult organs.
tGnRH‐3 and tGnRH‐5 inhibit adult organ growth by transiently or permanently suppressing cell proliferation. To examine which mechanism is used, we removed tGnRH‐3 and tGnRH‐5 from seawater after 48 hr of their administration. When tGnRH‐3 and tGnRH‐5 were washed from the seawater, the juveniles restarted their growth and developed well‐grown adult organs (Fig. 4A–D). Therefore, tGnRH‐3 and tGnRH‐5 can arrest the growth of adult organs by a reversible mechanism.
Tunicate gonadotropin‐releasing hormone‐3 (tGnRH‐3) and ‐5 affect adult organ growth in a reversible manner. Numbers at the left‐bottom corner indicate the proportion of animals with developed adult organs. A: A 3‐days after hatching (dph) ‐old juvenile cultured in seawater containing 10 µM tGnRH‐3. The animal had poorly developed adult organs. B: A 3‐dph juvenile cultured in seawater containing 10 µM tGnRH‐3 until 2‐dph and then cultured in seawater. Adult organs were developed well. C: A 3‐dph juvenile cultured in seawater containing 10 µM tGnRH‐5. The animal had poorly developed adult organs. D: A 3‐dph juvenile cultured in seawater containing 10 µM tGnRH‐5 until 2‐dph and then cultured in seawater. Adult organs were well developed.
Two Phases of Cellular Proliferation in the Larval to Metamorphic Stages
We were interested in the mechanisms by which tGnRH‐3 and tGnRH‐5 negatively regulate the growth of adult organs, because these mechanisms may be correlated with the previous finding that the adhesive papillae and preoral lobe of Ciona larva inhibit adult organ growth (Nakayama‐Ishimura et al., 2009). The regulation of cell proliferation was considered as a possible mechanism because previous studies have shown that adult organ growth during metamorphosis is accomplished by massive cell proliferation (Nakayama et al., 2005). Prior to that, we carried out in‐depth observation of cell proliferation from the larval to metamorphic stages by detecting cells at the S‐phase with EdU, a kind of deoxynucleotide analog that can label the nuclei of cells at the S‐phase. We found that there are two phases of cell proliferation in the larval to adult stages. The first is the proliferation seen in the larval trunk (Fig. 5A–D). We named the proliferation “larval cell proliferation.” Larval cell proliferation does not require the initiation of metamorphosis; it is observed in larvae after hatching, and it ceases around 2 days after hatching when the larvae fail to settle. We named these larvae that continue swimming for a long period “aged larvae.” The second cell proliferation is seen after the initiation of metamorphosis (Fig. 5E–L). We named the proliferation “metamorphic cell proliferation.” The presence of metamorphic cell proliferation was discovered through the following experiment. The aged larvae, which ceased larval cell proliferation, were forced to start metamorphosing by injuring their adhesive papillae. When the incorporation of EdU in these animals was examined, the animals started incorporating EdU in many cells 3 to 4 hr after the induction of metamorphosis, and cells at the S‐phase increased over successive hours (Fig. 5H–L). This result suggests that cell proliferation is reactivated by the initiation of metamorphosis. When the metamorphosis of aged larvae was induced by treatment with tGnRH‐6 and cAMP, the animals had many EdU‐incorporating cells (Fig. 5M,N), confirming that the metamorphic cell proliferation is initiated during the metamorphosis induced by these agents.
Larval and metamorphic cell proliferation. A‐D: Larval cell proliferation, as revealed by the incorporation of EdU. Nuclei labeled by EdU are shown by green spots. A: A larva at 1.5 hours posthatching (hph), to which EdU was administered from 0.5 to 1.5 hph. Many EdU‐positive cells are seen in the endodermal tissue (En). B: A larva at 5.5 hph, to which EdU was administered from 4.5 to 5.5 hph. Many EdU‐positive cells are seen in the endodermal tissue. C: A larva at 9.5 hph, to which EdU was administered from 8.5 to 9.5 hph. Many EdU‐positive cells are seen in the endodermal tissues. D: An aged larva to which EdU was administered for 1 hr at 2 days posthatching (dph). While some background signals were seen, massive cell proliferation was undetected. E‐L: Metamorphic cell proliferation. The aged larvae at 2 dph were forced to start metamorphosis through injuries to their papillae, and they were then administered EdU for 1 hr at the corresponding timing given in the panels. After EdU was washed out by exchanging the seawater, juveniles were further cultured for 1 day and then fixed for the detection of EdU‐incorporated cells. H‐L: Many EdU‐positive cells were detected when EdU administration was performed at 3 hr after the induction of metamorphosis. M,N: Cell proliferation in animals treated with tunicate gonadotropin‐releasing hormone‐6 (tGnRH‐6) and cAMP. The numbers in the lower left‐hand corner indicate the proportion of animals with many EdU‐incorporated cells. GnRHs and cAMP were administered at concentrations of 10 µM and 1 mM, respectively. M: The effect of tGnRH‐6 on metamorphic cell proliferation, as revealed by EdU incorporation. Green, EdU‐incorporated cells; blue, DAPI (4′,6‐diamidine‐2‐phenylidole‐dihydrochloride) counterstaining. N: The effect of cAMP on metamorphic cell proliferation.
tGnRH‐3 and tGnRH‐5 Specifically Inhibit Metamorphic Cell Proliferation
We examined whether tGnRH‐3 and tGnRH‐5 inhibit larval and/or metamorphic cell proliferation. First, the effect of these peptides on larval cell proliferation was observed. Larvae, at 1 hr after hatching, were administered tGnRH‐3 or tGnRH‐5 for 5 hr. When their incorporation of EdU was detected, tGnRH‐3/5‐treated larvae had many cells at the proliferative phase, as did the control larvae (Fig. 6A–C). This suggests that tGnRH‐3 and tGnRH‐5 do not inhibit larval cell proliferation. Next, we examined the effects of tGnRH‐3 and tGnRH‐5 on metamorphic cell proliferation. To clearly distinguish between larval and metamorphic cell proliferation, we utilized aged larvae. The aged larvae were forced to start their metamorphosis by injuring their adhesive papillae, and they were then treated with tGnRH‐3 or tGnRH‐5. Their cell proliferation was examined by EdU incorporation from 4 to 5 hr after the induction of metamorphosis. As a result, the juveniles that were treated with tGnRH‐3 or tGnRH‐5 had few EdU‐positive cells in their bodies compared to control juveniles at the same time point (Fig. 6D–F), suggesting that metamorphic cell proliferation is inhibited by tGnRH‐3 and tGnRH‐5. From these results, we concluded that the inhibitory effect of tGnRH‐3 and tGnRH‐5 is specific to the cell proliferation initiated along with metamorphosis.
Tunicate gonadotropin‐releasing hormone‐3 (tGnRH‐3) and ‐5 specifically inhibit metamorphic cell proliferation. Numbers at the left‐bottom corner indicate the proportion of animals with many EdU‐incorporated cells. A‐C: The effect of tGnRH‐3 and ‐5 on larval cell proliferation, as revealed by EdU incorporation. Green, EdU‐incorporated cells; blue, DAPI counterstaining. EdU‐positive proliferating cells are seen in the larvae to which 10 µM tGnRH‐3 and −5 were administered. Control, a larva cultured in seawater. D‐F: The effect of tGnRH‐3 and ‐5 on metamorphic cell proliferation, as revealed by EdU incorporation. Few cells incorporated EdU in juveniles to which 10 µM tGnRH‐3 and tGnRH‐5 were administered. G: Semiquantitative reverse transcription polymerase chain reaction (RT‐PCR) suggests that expression of Ci‐GnRH1 in the adhesive organ decreases during metamorphosis. Total RNA isolated from the larval anterior region containing the adhesive organ (Larva) and from the region containing the ampullae of juveniles (Juve) was subjected to RT‐PCR. EF1α was used as a positive control in RT‐PCR and quantification of cDNA. M, marker lane; RT−, lanes are negative controls without reverse transcription.
In order for Ciona to restart cell proliferation during metamorphosis, it must cancel the inhibitory function of tGnRH‐3 and tGnRH‐5 against cell proliferation. Ci‐GnRH1, which encodes tGnRH‐3 and tGnRH‐5 with tGnRH‐6, is expressed in the adhesive papillae that play an inhibitory role in adult organ growth (Nakayama‐Ishimura et al., 2009; Kusakabe et al., 2012). We hypothesized that Ci‐GnRH1 expression in the papillae is diminished during metamorphosis, and that this decrease may release Ciona from the inhibition of cell proliferation mediated by tGnRH‐3 and tGnRH‐5. To show this, we compared the expression level of Ci‐GnRH1 in the adhesive organs before and after metamorphosis by semiquantitative reverse transcription polymerase chain reaction (RT‐PCR). We found that the expression of Ci‐GnRH1 dramatically decreased after metamorphosis compared to that during the larval stage (Fig. 6G). These results suggest that the levels of tGnRH‐3 and tGnRH‐5 decrease during metamorphosis, which may cause Ciona to allow metamorphic cell proliferation to restart.
tGnRH‐3 and tGnRH‐5 Inhibit Cell Cycle Progression by Arresting the G1‐S Phase Transition
Cell proliferation is regulated by cell cycle progression. To investigate how tGnRH‐3 and tGnRH‐5 affect the cell cycle, we utilized the fluorescent cell cycle indicator Fucci (Sakaue‐Sawano et al., 2008). In the Fucci system, cells at the G1 phase are labeled with red fluorescence derived from the G1 Fucci probe, while cells at the proliferative phases (S, G2, and M phases) are labeled with green fluorescence derived from the S/G2/M Fucci probe. We already reported that the Fucci probes can chase cell cycle progression in C. intestinalis (Ogura et al., 2011). We created a pair of transgenic lines expressing G1 and S/G2/M Fucci probes in the whole body with the cis element of the house‐keeping gene EF1α (Sasakura et al., 2010). A double‐transgenic line of the two Fucci lines was established, and the larvae of the double‐transgenic line were administered tGnRH‐3/5 to see the cell cycle progression in the animals. As a result, the cells with red fluorescence increased in tGnRH‐3/5‐treated animals compared to control juveniles (Fig. 7), suggesting that more cells were in the G1 phase when treated with tGnRH‐3/5. The absence of cells in the S/G2/M phases was evident in cells at zone B of the gill slits (Shimazaki et al., 2006). The endostyle, which shows very massive cell proliferation, decreased in cells at the S/G2/M phases in tGnRH‐3/5‐administered juveniles, although some proliferating cells remained. We concluded that tGnRH‐3 and tGnRH‐5 inhibit cell proliferation by arresting the G1 to S phase transition.
Tunicate gonadotropin‐releasing hormone‐3 (tGnRH‐3) and ‐5 inhibit the G1 to S transition, as revealed by the expression of Fucci probes. Green and red respectively represent cells in the S/G2/M and G1/G0 phases. mVenus‐hGem(1/110), S/G2/M Fucci; mKO2‐hCdt1(1/100), G1/G0 Fucci; Merge, merged images of green and red fluorescence. GnRHs were administered at a concentration of 10 µM. A: A 48‐hours posthatching control juvenile. Many cells in the endostyle (Es) and cells at zone B of the gill (Gi) exhibit green fluorescence, suggesting that they are in the proliferative phase. The green color in the absorbed tail (AT) and the green/red colors in the central nervous system (CNS) represent autofluorescence. B: A juvenile cultured in seawater containing tGnRH‐3. While many cells exhibit red fluorescence, few cells in the endostyle exhibit green fluorescence. C: A juvenile cultured in seawater containing tGnRH‐5. While many cells exhibit red fluorescence, few cells exhibit green fluorescence. D: Box‐and‐whisker plots of the ratio between mVenus‐hGem(1/110) positive pixel counts and mKO2‐hCdt1(1/100) pixel counts in the endostyle (n=8 animals for each condition). Administration of tGnRH‐3 or tGnRH‐5 significantly reduced the ratio. The Wilcoxon rank sum test shows that the reduction is statistically significant (p‐value is shown in the figure). The upper and lower quartiles of the datasets are shown using boxes. The line between the quartiles represents the median value. Whiskers represent the difference between the maximum and the upper quartile (or the lower quartile and the minimum).
Discussion
In the present study, we showed that GnRH peptides have two roles during metamorphosis in C. intestinalis. First, all GnRHs induce tail absorption. Second, tGnRH‐3 and tGnRH‐5 inhibit the growth of adult organs. Previous studies have shown that the metamorphosis of C. intestinalis is triggered by the single stimulus of the settlement, which then initiates each metamorphic event by activating different downstream pathways (Nakayama‐Ishimura et al., 2009). This study shed light on the molecular characterization of the pathways in which GnRHs are involved.
GnRHs Induce Tail Absorption
This study has shown that all GnRHs have the ability to induce tail absorption. Because GnRHs are expressed in the nervous system (Kusakabe et al., 2012), GnRH peptides are secreted from GnRH neurons. Ciona GnRH receptors are expressed in the tissues that make up the tail in addition to the nervous system (Kusakabe et al., 2012), suggesting that GnRHs can be received by other neurons and/or directly by tail cells, and they then regulate tail absorption. Previous studies have reported that GnRHs show different affinities to their receptors; however, the cAMP signaling response is dominant in the GnRH receptors of C. intestinalis, and indeed all of the GnRH receptors can up‐regulate cAMP (Tello et al., 2005; Sakai et al., 2010). We showed that cAMP can induce tail absorption like GnRHs. Therefore, cAMP probably acts as the second messenger of GnRHs and their receptors for tail absorption. This finding suggests that the mechanism by which tail absorption is induced is shared among GnRHs. The redundant activity of GnRHs in metamorphosis may be a system that assures the initiation of metamorphosis with the proper timing. For ascidians, metamorphosis is not a simple conversion of the body from larval to the adult state; ascidians change their lifestyle through metamorphosis from free‐swimming to sessile modes. In addition, ascidians have to complete their metamorphosis quickly because ascidian larvae do not take food. In this respect, ascidian metamorphosis differs from metamorphosis in other animals. For ascidians, metamorphosis failure or starting metamorphosis with inappropriate timing has a serious effect on survival. By using several peptides for the same functions, ascidians may have evolved a metamorphosis system that is highly resistant to alternations caused by the accumulation of mutations. The unique character of ascidian GnRHs such that three of them are encoded by the single gene (Tello et al., 2005) may also be helpful in acquiring the redundant system, because the three peptides share temporal and spatial expression patterns.
tGnRH‐3 and tGnRH‐5 Inhibit the Growth of Adult Organs
Our previous study has shown the presence of an inhibitory mechanism of adult organ growth at the larval stage of C. intestinalis (Nakayama‐Ishimura et al., 2009). This mechanism is thought to contribute to the restriction of adult organ growth to the post‐settlement stage. Such restriction may be important for the swimming life style of ascidian larvae. If growth starts during the larval period, the trunk becomes so large that it may interfere with swimming. tGnRH‐3 and tGnRH‐5 probably share the mechanisms of the inhibitory activity of cell proliferation for the following reasons. First, both tGnRH‐3 and tGnRH‐5 inhibit adult organ growth by arresting cell proliferation at the G1 phase. Second, the expression pattern of tGnRH‐3 and tGnRH‐5 is completely identical because both peptides are encoded by Ci‐GnRH1. Third, tGnRH‐3 and tGnRH‐5 have identical properties of receptor binding (Sakai et al., 2010). Because GnRH receptors are not expressed in the tissues that show massive cell proliferation during metamorphosis such as the epidermis, mesenchyme and endoderm (Nakayama et al., 2005; Kusakabe et al., 2012), tGnRH‐3 and tGnRH‐5 may regulate the cell proliferation of these tissues and organs in an indirect fashion. Indeed, GnRHs and their receptors are expressed in the nervous system, suggesting that tGnRH‐3 and tGnRH‐5 stimulate the neurons, and the neurons then transmit the stimuli to the adult tissues/organs to inhibit their proliferation.
Curiously, the downstream cascade of the receptors of tGnRH‐3 and tGnRH‐5, namely Ci‐GnRHR1 and Ci‐GnRHR3, is mediated by cAMP, as with the other Ci‐GnRHRs (Tello et al., 2005; Sakai et al., 2010; Kusakabe et al., 2012). This might contradict our data that animals treated with cAMP showed growth of adult organs. This contradiction can be explained by the expression pattern of GnRH receptors. Ci‐GnRHR1 and Ci‐GnRHR3 are expressed in subsets of neurons (Kusakabe et al., 2012). Some neurons expressing Ci‐GnRHR1 and/or Ci‐GnRHR3 may receive tGnRH‐3 and tGnRH‐5 exclusively and synthesize cAMP to exert inhibitory activity over cell proliferation, whereas neurons that do not express these receptors can activate cell proliferation mediated by cAMP. The two different roles of tGnRH‐3 and tGnRH‐5, namely the activation of tail absorption and the inhibition of adult organ growth, may involve different subsets of neurons expressing tGnRH‐3 and tGnRH‐5, as is suggested by the broad expression of Ci‐GnRH1 that encodes both tGnRH‐3 and tGnRH‐5 (Kusakabe et al., 2012). The presence of two receptors (Ci‐GnRHR1 and Ci‐GnRHR3) for tGnRH‐3 and tGnRH‐5 may also be responsible for the two different activities of tGnRH‐3 and tGnRH‐5. Although tGnRH‐6 is encoded by Ci‐GnRH1 together with tGnRH‐3 and tGnRH‐5, cell proliferation was observed in the animals treated with tGnRH‐6, unlike those treated with tGnRH‐3 and tGnRH‐5. This different activity can be explained by a different receptor preference as shown in a previous study (Sakai et al., 2010).
The inhibition of cell proliferation by tGnRH‐3 and tGnRH‐5 should be canceled during metamorphosis in order for Ciona to restart cell proliferation to construct adult organs. We have shown that expression of Ci‐GnRH1 in the adhesive organs dramatically decreases during metamorphosis. This reduction of transcription level can explain the release of juveniles from inhibition by tGnRH‐3 and tGnRH‐5. Previous studies have suggested that the adhesive papillae that express Ci‐GnRH1 degenerate during metamorphosis in ascidians including Ciona (Cloney, 1982; Sasakura et al., 2005). This degeneration could be a mechanism of the down‐regulation of Ci‐GnRH1. Even though the degeneration of adhesive papillae is one of the earliest events in ascidian metamorphosis, there may be a time gap between the initiation of metamorphosis and the complete loss of expressed tGnRH‐3 and tGnRH‐5, since both RNA and peptides should disappear. This gap can explain the time that is taken from the initiation of metamorphosis to the initiation of metamorphic cell proliferation.
Possible Conservation of Nonreproductive Roles of GnRHs Among Chordates
In vertebrates, the best‐known role of GnRH is the regulation of reproduction (Kah et al., 2007; Millar et al., 2008). However, recent studies have shown that GnRH has broader functions. In mammalian cell culture, GnRH functions in the suppression of cancer cell proliferation via the Gi protein (Imai et al., 1996; Yu et al., 2011). This is similar to the role of Ciona tGnRH‐3 and tGnRH‐5 in arresting cell proliferation. In the case of vertebrate cancers, GnRH decreases the cAMP concentration in the cells, which results in the inactivation of MAP kinase activity (Limonta et al., 2003). The mechanism is completely different from the deduced cascade of tGnRH‐3/5 in that tGnRH‐3 and tGnRH‐5 usually increase the cAMP concentration by binding to their receptors (Tello et al., 2005; Sakai et al., 2010). This difference may reflect the direct and indirect regulation of cell proliferation in vertebrates and Ciona, respectively. In addition to the inhibition of proliferation, mammalian GnRHs can up‐regulate cell proliferation in a cell‐type‐specific manner (Park et al., 2013), suggesting that GnRHs play multiple roles in mammals like Ciona.
The expression of GnRHs in younger developmental stages is seen in vertebrates. In medaka fish and zebrafish, GnRHs and their receptors are expressed in the central nervous system of sexually immature embryos (Marin and Rubenstein, 2003; Abraham et al., 2010; Kusakabe et al., 2012). In zebrafish, knocking down GnRHs results in an abnormal projection of axon fibers of GnRH neurons and a disturbance of brain regionalization (Wu et al., 2006; Abraham et al., 2008). Considering the close phylogenetic relationship between ascidians and vertebrates (Delsuc et al., 2006), it is possible that the chordate GnRHs have functions in the developmental processes that are shared among ascidians and vertebrates. On the other hand, fish do not undergo a typical metamorphosis like ascidians, indicating that the function of GnRH in fish embryos may be different from inducing metamorphosis. Because cell proliferation occurs in fish embryos, it is possible that fish GnRHs function in the regulation of cell proliferation like in Ciona.
Conclusions
This study provides evidence that GnRHs of C. intestinalis play pivotal roles in the regulation of metamorphosis. They play major roles in the activation of tail absorption and the regulation of adult organ growth. These findings shed light on the characterization of the molecular mechanisms of ascidian metamorphosis. It has been shown that many genes encoding neuropeptides are expressed in the larval nervous system of C. intestinalis (Hamada et al., 2011); however, the roles of these neuropeptides remain unknown. It is possible that these neuropeptides also function in the regulation of metamorphosis in concert with GnRHs. Numerous studies have supported the presence of the signaling that initiates the metamorphosis of ascidians (Cloney, 1982; Eri et al., 1999; Chambon et al., 2002; Kimura et al., 2003; Zega et al., 2005; Nakayama‐Ishimura et al., 2009). In contrast, the inhibitory mechanism of metamorphosis has not been described well. This study is the first report to address the neuronal molecules that inhibit the metamorphic event. Finally, this study has suggested that GnRHs play roles in animals in a sexually premature state. Because GnRHs are neuropeptides that are widely conserved among metazoans, GnRHs may also function in immature stages in other animals. Future studies will elucidate the evolution of the non‐reproductive functions of GnRHs in addition to their major roles in reproductive systems.
Experimental Procedures
Animals
Wild‐type C. intestinalis were collected from or cultivated at Maizuru (Kyoto), Mukaishima (Hiroshima), Misaki (Kanagawa), and Usa (Kochi). Eggs and sperm were surgically collected. After fertilization, the embryos and larvae were cultured at 18°C. Transgenic lines were cultured according to the procedures given in a previous report (Joly et al., 2007).
Constructs and Transgenic Lines
pMiCiEF1ahGem(1/110) and pMiCiEF1ahCdt1(1/100) were created with Gateway technology (Invitrogen) based on pMiDestF (Sasakura et al., 2010) and Fucci vectors described previously (Ogura et al., 2011). Transgenic lines of pMiCiEF1ahGem(1/110) and pMiCiEF1ahCdt1(1/100) were created by electroporation‐mediated transgenesis as described previously (Matsuoka et al., 2005). One transgenic line was established for each construct, and they were mated to obtain double‐transgenic animals harboring two Fucci probes. Image analyses were performed using ImageJ software (http://imagej.nih.gov/ij/), and statistical tests (Wilcoxon rank sum test) were performed using R software (http://www.r‐project.org/).
Peptide Administration
All peptides used in this study were synthesized using an ABI 430A solid‐phase peptide synthesizer (Life Technologies) according to the manufacturer's instructions. A part of the larval tail was cut off with fine razors around 17.5 hours post‐fertilization (hpf) to prevent metamorphosis and to induce efficient incorporation of GnRHs by the larvae. GnRHs were administered to the tail‐cut larvae at concentrations of 10 µM and 1.0 µM. cAMP and cGMP analogs 8‐bromoadenosine 3′:5′‐cyclic monophosphate (8‐Br‐cAMP; Sigma) and 8‐bromoguanosine 3′:5′‐cyclic monophosphate (8‐Br‐cGMP; Sigma) were dissolved in the seawater containing 6 mM NaOH at the concentration of 10 mM. When the reagents were incompletely dissolved in the solvent under this condition, their solutions were diluted with seawater to 1–2 mM. The cAMP and cGMP analogs were administered to the tail‐cut larvae at concentrations of 1 mM and 100 µM. Animals were cultured at 18°C, and their metamorphosis was observed at approximately 24 hr and 48 hr after administration.
Aged Larvae
In this study, we defined the aged larvae as the larvae which continued swimming more than 48 hr after hatching. Most intact larvae had started their metamorphosis by that time in Petri dishes due to their highly frequent settlement, and thus the appearance of aged larvae was not common. As a countermeasure for this issue, we cut off the tails of larvae as mentioned above to inhibit their settlement and subsequent metamorphosis.
Detection of Proliferating Cells
Larvae and juveniles were treated in seawater containing 10 µM 5‐ethynyl‐2′‐deoxyuridine (EdU) for 1 hr, followed by fixation with formaldehyde. Detection of cells that had incorporated EdU was carried out with the Click‐iT EdU Alexa Fluor Imaging Kit (Invitrogen) according to the method described in a previous report (Ogura et al., 2011).
RT‐PCR
The anterior tip of 1‐dpf‐old larvae that contained adhesive papillae plus a preoral lobe and a small portion of the endostyle primordium was isolated with a razor. The body of 2‐dpf‐old juveniles that contained the ampulla and a small portion of the endostyle was also isolated with a razor. Approximately 30 specimens were collected, and total RNA was isolated with Isogen (Wako). Residual genome DNA was digested with DNase I (Invitrogen). Reverse transcription was performed using Superscript III reverse transcriptase (Invitrogen) and oligo(dT) primers. The quantity of cDNA in each sample was estimated with SYBR Premix DimerEraser (Takara Bio) and a Thermal Cycler Dice Real Time System TP800 (Takara Bio) following the procedure described in our previous report (Sasakura et al., 2010). EF1α was used for quantification with primers, 5′‐CATGTCACGGAC AGCGAAACG‐3′ and 5′‐CAATGTGTGTTGAGGCATTCCAAG‐3′. Depending on the relative quantity of cDNA in the samples, we adjusted the concentration of cDNA solution from each sample in order to start PCR with the same amount of cDNA to detect Ci‐GnRH1. PCR was carried out with Takara Ex Taq Hot Start Version (Takara Bio). The primers used to amplify Ci‐GnRH1 were 5′‐CATCTCGATCTTGGTTTAGGAG‐3′ and 5′‐TTCTACGTCAGATC CTTTGTCG‐3′. As positive controls of the PCR, EF1α was again amplified with primers 5′‐TTGGACAAACTTAAGGCCGAGC‐3′ and 5′‐GTCTCCAGCAACATAACCTCTC‐3′. The PCR conditions were 10 min at 94°C for denaturing genome DNA, then 35–40 cycles of 30 s at 94°C, 30 s at 55°C, and 30 s at 72°C, followed by a final extension for 10 min at 72°C.
Acknowledgments
We thank the members of the Shimoda Marine Research Center at the University of Tsukuba for their cooperation with this study. We also thank Drs. Shigeki Fujiwara, Nobuo Yamaguchi, Kunifumi Tagawa, and all of the members of the Department of Zoology, Kyoto University, Misaki Marine Biological Station, University of Tokyo, and the Maizuru Fishery Research Station at Kyoto University for collecting the C. intestinalis adults. We are grateful to Dr. Atsushi Miyawaki for his kind provision of Fucci vectors. We thank Dr. Kazuo Inaba for helpful discussions. This study was supported by Grants‐in‐Aid for Scientific Research from the Japan Society for the Promotion of Science (JSPS) and the Ministry of Education, Culture, Sports, Science and Technology (MEXT) to Y.S. Y.S. was supported by a Toray Science and Technology Grant. Further support was provided by grants from the National Bioresource Project.
