Abstract
Larvae of the ascidian Ciona initiate metamorphosis tens of minutes after adhesion to a substratum via its adhesive organ. The gap between adhesion and metamorphosis initiation is suggested to ensure the rigidity of adhesion, allowing Ciona to maintain settlement after losing locomotive activity through metamorphosis. The mechanism producing the gap is unknown. Here, by combining gene functional analyses, pharmacological analyses, and live imaging, we propose that the gap represents the time required for sufficient cAMP accumulation to trigger metamorphosis. Not only the Gs pathway but also the Gi and Gq pathways are involved in the initiation of metamorphosis in the downstream signaling cascade of the neurotransmitter GABA, the known initiator of Ciona metamorphosis. The mutual crosstalk of stimulatory and inhibitory G-proteins functions as the accelerator and brake for cAMP production, ensuring the faithful initiation of metamorphosis at an appropriate time and in the right situation.
eLife assessment
This important work substantially advances our understanding of the molecular mechanisms underlying the timing of the initiation of metamorphosis of the Ciona ascidian tadpole larva. Through the combination of gene knockdown experiments and fluorescent molecular reporters the authors provide compelling evidence about a crosstalk between different G protein mediated signalling pathways and are able to place different signalling molecules within a signalling network. The work will be of interest to molecular, developmental and marine biologists and to scientists working on animal metamorphosis.
Introduction
Metamorphosis is a widespread feature of animal development that allows them to have different functions between larval and adult stages 1. As larvae become adults, the shape and various characteristics of their physiology, gene expression, behavior, and lifestyle change. With these changes, animals can focus on a limited number of biological activities at each stage to increase feeding, growth, dispersal, and reproduction efficiencies. This biphasic mode of life is a widely conserved and ancient trait of animals, as evidenced by its presence in groups that maintain primitive states 2, indicating that this long-conserved feature has contributed to animals’ flourishment. Therefore, to understand animal evolution, it is essential to elucidate the mechanisms underlying metamorphosis.
A key characteristic of metamorphosis is that both internal and external conditions determine its initiation. Sufficiently matured (or metamorphically competent3) larvae start metamorphosis only when they meet an appropriate external condition to be adults. Various organic and inorganic external stimuli suited to the lifestyle of adulthood trigger metamorphosis. Many marine invertebrates exhibit a benthic lifestyle at the adult stage4. Their planktonic larvae have an adhesive organ that secretes adhesives and triggers metamorphosis when they adhere to a substratum. Upon starting metamorphosis, their larvae lose locomotive organs and transition into benthic adult forms. Determining the timing of metamorphosis is important for benthic animals, particularly sessile ones, to ensure their future survival and reproduction, as they become unable to move after metamorphosis. Therefore, sessile animals are suspected to have elaborate mechanisms to start metamorphosis only when a firm adhesion is achieved at an appropriate place. Although several reports describe this kind of phenomenon5–7, the mechanism by which the external condition is turned into an internal mechanism to trigger metamorphosis at the right timing remains elusive.
Ascidians are marine invertebrate chordates that are the closest living relatives to vertebrates8,9. The chordate features of ascidians are best represented by their tadpole larval shape. Like vertebrate tadpoles, ascidian larvae swim by beating their tail through neuromuscular activities10,11. However, ascidians lose the tadpole shape during metamorphosis, and they exhibit a sessile lifestyle at the adult stage12,13. Ascidian metamorphosis comprises several key events, such as tail regression, body axis rotation, and adult organ growth14 (Figure 1A).
![](https://prod--epp.elifesciences.org/iiif/2/99825%2Fv1%2Fcontent%2F588679v3_fig1.tif/full/max/0/default.jpg)
Gq and Gs pathways are required for Ciona metamorphosis.
(A) Schematic illustration of Ciona metamorphosis.
(B) A larva knocked down with Gαq using the antisense morpholino oligonucleotide (MO). Metamorphosis did not initiate at 2 days post-fertilization (2 dpf).
(C) A Gαs knockdown larva.
(D) A control animal injected with the standard (STD) MO. Metamorphosis initiated, as indicated by the completion of tail regression.
(E) Effect of Gαq knockdown on the percentage of metamorphosis initiation (indicated by the initiation of tail regression), shown as box-and-whisker plots. Dots indicate experiment replicates. ***, p<0.001 (Fisher’s exact test). n, number of examined larvae in total.
(F) Effect of Gαs knockdown.
(G-I) Effects of PLCβ1/2/3, PLCβ4, and IP3R knockdowns.
(J-M) Results of the experiments where adhesion was prevented by tail amputation and laying larvae on agar-coated plates.
(J) A control larva overexpressing the Kaede reporter in the entire nervous system using the PC2 cis element.
(K) A larva overexpressing constitutive active (ca)PLCβ1/2/3.
(L) Effect of caGαq overexpression.
(M) Effect of caPLCβ1/2/3 overexpression.
(N-O) Live imaging of Ca2+ transient, monitored by injecting GCaMP8 mRNA.
(N) Snapshot of the control larva injected with GCaMP8 mRNA. An increase in the fluorescence intensity in the adhesive papillae (arrow) was observed. Rod, the position of the glass rod used to stimulate papillae. The percentage and number exhibit the rate of animals showing Ca2+ transient in the papillae.
(O) A larva injected with Gαq MO plus GCaMP8 mRNA. An increase in the fluorescence intensity in the papillae (arrow) was not observed even though the brightness of the whole body was raised. See also Figure S1.
Adhesion to a substratum triggers ascidian metamorphosis14. Ascidian larvae have adhesive organs called adhesive papillae, usually triangular protrusions at the anterior end. Adhesive papillae secrete a lectin-type mucus substance that is suspected to aid in binding to the substratum15. Moreover, adhesive papillae serve as mechanical sensors16. The papilla is a neuronal organ that includes epidermal neurons positive for vesicular glutamate transporter (vGLUT) and GABA, which innervate axons toward the sensory vesicle (larval brain)17–20. The papilla neurons are responsible for initiating metamorphosis16,21–23. Moreover, our recent study suggests the involvement of a mechanoreceptor channel (the TRP channel) in the responsiveness to mechanical stimuli22. However, a simple physical adhesion is not sufficient to trigger metamorphosis. In our study using the model ascidian Ciona intestinalis Type A (this species has been suggested to be renamed C. robusta 24–26, and hereafter we call it Ciona), continuous adhesion for about 30 min is necessary before starting metamorphosis27. When larvae detach before reaching the critical period, they must adhere for 30 min again for metamorphosis. Therefore, ascidian larvae are suspected of sensing the duration of adhesion and initiating metamorphosis only when adhesion is firm enough to be maintained for the required period. Ciona larvae continue tail beating during adhesion to push their body to the substratum. We showed that the strength of force generated by this swimming activity influences the timing of metamorphosis initiation22. In this phenomenon, abolishment of swimming activity elongates the time from settlement until metamorphosis is initiated. The mechanisms for measuring the duration of adhesion and the strength of the force generated by adhesion are unknown.
To elucidate the mechanisms triggering ascidian metamorphosis, the signaling pathways must be characterized. Many studies have discovered signaling molecules as possible inducers of ascidian metamorphosis28–35. These molecules include neurotransmitters, suggesting that transmission of excitation in the nervous system, starting from the adhesive papillae, is crucial for metamorphosis. Recently, our group reported that the inhibitory neurotransmitter GABA plays a pivotal role in initiating Ciona metamorphosis36. Knockout and knockdown of the genes encoding GABA synthase (GAD), vesicular inhibitory amino acid transporter (VIAAT or VGAT), and metabotropic GABA receptor (GABABR) resulted in the perturbation of metamorphosis. Moreover, GABA administration induces metamorphosis without adhesion. Our previous studies demonstrated that GABA induces the secretion of gonadotropin-releasing hormone (GnRH)36,37; however, it remains unknown how the inhibitory neurotransmitter activates neuronal functions for initiating metamorphosis.
In this study, we addressed the characterization of the downstream cascade stimulated by GABA. Because GABABR is a G-protein-coupled receptor (GPCR)38, we searched for trimeric G-proteins necessary for metamorphosis. These G-proteins are heterotrimers of an α β and γ subunit39. Upon ligand binding, GPCR exchanges GDP of the α subunit to GTP, then the GTP-bound α subunit and the βγ complex are released from the GPCR. Although both the GTP-bound α subunit and the βγ complex have activities, the α subunit mainly determines the reaction specific to the G-protein type. We found that three G-proteins are activated in the downstream GABA cascade, resulting in cyclic adenosine monophosphate (cAMP) elevation. Because the signaling includes stimulating and inhibiting cAMP synthesis, Ciona initiates metamorphosis only when a sufficient quantity of cAMP is accumulated due to sustained adhesion. Our results revealed the ingenious mechanism that permits Ciona to start metamorphosis only when achieving an appropriate adhesion firm enough to relinquish swimming ability and commence sessile adult life.
Results
Characterization of G-proteins necessary for metamorphosis initiation
We knocked down the genes encoding G alpha (Gα) proteins40 using antisense morpholino oligonucleotides (MOs). We found that two genes corresponding to Gαq and Gαs are necessary for metamorphosis. Neither morphant showed any signature of metamorphosis even though both were allowed to adhere to the base of culture dishes, which was sufficient for control larvae to initiate metamorphosis (Figure 1A-F).
Gαq activates phospholipase C beta (PLCβ) to produce inositol triphosphate (IP3)41. IP3 is received by its receptor on the endoplasmic reticulum (ER) and releases calcium ion (Ca2+). The Ciona genome encodes two PLCβ (PLCβ1/2/3, PLCβ4) and one IP3 receptor (IP3R) (Figure S1 and Table S1). We knocked down PLCβ1/2/3, PLCβ4, and IP3R genes. The knockdown larvae of these three genes failed to start metamorphosis (Figure 1G-I), suggesting that Gαq initiates metamorphosis by the conventional Ca2+ pathway mediated by PLCβ and IP3/IP3R. To confirm this, we overexpressed a constitutively active form of Gαq (caGαq)42 and of caPLCβ1/2/343 in the entire nervous system with the cis element of the gene encoding prohormone convertase 2 (PC2)44–46. When the microinjected animals reached the larval stage, they were cultured on agar-coated dishes after tail amputation to prevent adhesion. In this condition, the control larvae rarely started metamorphosis because of the absence of adhesion (Figure 1J). In contrast, caGαq- or caPLCβ1/2/3-overexpressed larvae initiated metamorphosis without adhesion (Figure 1K-M).
Adhesive papillae exhibit a Ca2+ increase soon after sensing an adhesive stimulus16,22. We examined whether this Ca2+ transient is dependent on the Gq pathway. Compared to controls, significantly fewer larvae injected with Gαq MO plus GCaMP8 mRNA exhibited GCaMP8 fluorescence elevation in the adhesive papilla upon stimulation (Figure 1N, O). This result suggests that the Gq pathway is activated upon adhesion to cause a Ca2+ transient in the adhesive papilla.
Gs pathway initiates metamorphosis by activating cAMP synthesis
The involvement of Gs in metamorphosis was confirmed by the overexpression of a constitutively active form of Gαs (caGαs)47 in the nervous system. This overexpression resulted in the initiation of metamorphosis without adhesion (Figure 2A-C). The Gs pathway activates adenylate cyclase (AC) to produce cAMP48. We previously reported that cAMP can induce metamorphosis37. Because externally added cAMP is not a strong inducer of metamorphosis, we attempted to confirm this hypothesis through another experiment. Theophylline increases cAMP by inhibiting the cAMP-degrading enzyme phosphodiesterase (PDE)49. We treated wild-type larvae with theophylline after tail amputation to prevent larvae from adhesion. Most theophylline-treated larvae completed tail regression without adhesion (Figure 2D-F). Theophylline has several target proteins in addition to PDE50. To further confirm that cAMP is responsible for the initiation of metamorphosis, we overexpressed photo-activating AC (bPAC)51 in the nervous system. The bPAC-overexpressed larvae regressed their tails without adhesion, suggesting that a cAMP increase triggers metamorphosis (Figure 2G).
![](https://prod--epp.elifesciences.org/iiif/2/99825%2Fv1%2Fcontent%2F588679v3_fig2.tif/full/max/0/default.jpg)
cAMP initiates Ciona metamorphosis.
(A) An adhesion-prevented control (uninjected) larva at 2 dpf. Metamorphosis did not initiate.
(B) An animal with adhesion prevention and overexpression of caGαs in the entire nervous system. Metamorphosis initiated.
(C) Effect of caGαs overexpression.
(D) A control (uninjected) larva.
(E) A theophylline-treated animal. Metamorphosis initiated.
(F) Effect of theophylline treatment.
(G) Effect of bPAC overexpression in the nervous system.
(H) Effect of bPAC overexpression over time. In this experiment, adhesion was permitted. Control groups (overexpressed with mCherry) exhibited a slower initiation of metamorphosis than bPAC-overexpressed groups.
Using bPAC, we addressed whether enhanced cAMP production facilitates the initiation of metamorphosis. At 24 hours post-fertilization (hpf), only a few animals in both the bPAC-overexpressed and control groups initiated metamorphosis because 24 hpf is somewhat too early for Ciona larvae to be metamorphically competent27 (Figure 2H). Even in this condition, the bPAC-overexpressed group exhibited a statistically higher rate of metamorphosis-initiated larvae. The proportion of animals initiating metamorphosis increased over successive time points, with the bPAC-overexpressed group consistently showing a statistically higher rate of metamorphosis compared to controls. These results support the idea that cAMP plays a role as a timer for Ciona metamorphosis; accumulation of this molecule to exceed a threshold could trigger metamorphosis. Because many of the bPAC-overexpressed larvae did not initiate metamorphosis at 24 hpf, similar to control larvae, cAMP accumulation does not alter the timing of acquiring metamorphic competence.
Gq-Gs pathways work in the adhesive papillae for metamorphosis
The above results showed that activation of the Gq and Gs pathways are the key events in initiating metamorphosis. Gαq is necessary to induce Ca2+ transients in the adhesive papillae, suggesting that the Gq pathway functions in this region. If both Gq and Gs function in the papillae, they should be expressed in the adhesive papilla. The transcriptome analysis of the larval papilla region showed that the genes encoding Gq and Gs pathway proteins are expressed in this region (Figure S2 and Table S1). Therefore, Gq and Gs could function in the papilla to initiate metamorphosis.
As shown above, theophylline induced metamorphosis without settlement. However, when papillae were removed from larvae, theophylline failed to induce metamorphosis (Figure 3A-C). This suggests that cAMP elevation in the adhesive papillae is essential for starting metamorphosis. The overexpression of caGαq, caPLCβ1/2/3, caGαs, and bPAC by the PC2 cis element resulted in the initiation of metamorphosis without adhesion. Like theophylline, amputation of the papillae inhibited them from starting metamorphosis (Figure 3D), confirming that activation of the Gq and Gs pathways in the adhesive papillae triggers metamorphosis.
![](https://prod--epp.elifesciences.org/iiif/2/99825%2Fv1%2Fcontent%2F588679v3_fig3.tif/full/max/0/default.jpg)
Gq and Gs pathways are activated in the adhesive papillae to initiate metamorphosis.
(A) A theophylline-treated animal. Metamorphosis initiated.
(B) A papilla-amputated larva treated with theophylline. Metamorphosis did not initiate.
(C) Effect of papilla amputation on theophylline treatment.
(D) Effects of papilla amputation on caGαq, caPLCβ1/2/3, caGαs, and bPAC overexpression.
(E) Measurement of cAMP concentration, as indicated by the fluorescence of Pink Flamindo. Left, the quantification of Pink Flamindo fluorescent intensity is shown for a larva that initiated metamorphosis at the time indicated by an arrowhead. The time of stimulation of the adhesive papillae with a glass rod is denoted by a bar. Right, the position of the region of interest (ROI). See also Figure S3.
If the Gs pathway is activated in the adhesive papillae, a cAMP increase should be observed in this region upon adhesion. We examined this possibility using a fluorescent cAMP indicator called Pink Flamindo (PF)52. After stimulation, the fluorescence of PF in the papillae temporarily decreased on average by 0.83-fold (n=5) from the initial intensity, followed by a gradual increase to 1.17-fold over successive time points (Figure 3E). Such an increase in fluorescence intensity was not observed in the adhesive papillae of the larvae that had failed to initiate metamorphosis following stimulation (n=4, Figure S3). Therefore, increased cAMP in the papillae serves to indicate that sufficient stimuli of adhesion have been received to induce metamorphosis. This strengthens our hypothesis that cAMP accumulation in the adhesive papillae determines the initiation of metamorphosis.
GABA in the adhesive papillae is responsible for metamorphosis
Our previous study demonstrated that GABA is the chief neurotransmitter that induces metamorphosis36. To gain insight into the relationships between the GABA, Gq, and Gs pathways, we addressed whether GABA functions in the adhesive papillae for initiating metamorphosis, similar to Gq and Gs.
The previous studies and our transcriptome data suggest that GAD is expressed in the adhesive papillae53 (Table S1). Moreover, GABA-immunopositive signals have been detected in the papillae19,53. Therefore, papillae can provide GABA to stimulate themselves. Next, we examined whether the adhesive papillae can receive GABA to initiate metamorphosis. We showed that the metabotropic GABA receptor is responsible for the initiation of metamorphosis36. Using whole-mount in situ hybridization, the previous study did not detect the expression of two GABAB receptor (GABABR) genes in the papillae53; however, our transcriptome data detected the low-level expression of three genes encoding GABABR (Table S1) in the papillae, suggesting that adhesive papillae could receive GABA. GABA can induce metamorphosis without adhesion (Figure S4A)36. However, amputation of adhesive papillae eliminated this activity (Figure S4B-C), suggesting that GABA reception by the papillae is responsible for starting metamorphosis.
The larval brain is the major domain expressing GABABR53. Therefore, it remains possible that GABA signaling in the brain stimulates the papillae in a retrograde manner to initiate metamorphosis. The connectome analyses of the larval nervous system did not suggest a nerve that inputs into the papillae from the brain54,55; however, retrograde stimulation would be possible through the volume transmission of a signaling molecule. To examine this possibility, we isolated larval fragments anterior to the brain (Figure S4D), followed by the administration of GABA and theophylline. Through the application of these chemicals, anterior fragments exhibited increased clarity, elongation, and retraction of papillae, which are the signatures of metamorphosis (Figure S4E-G). Because their responses to GABA were weaker than those observed in the theophylline treatment, we further examined whether the anterior fragments can respond to GABA by monitoring the expression of GABA-responsive genes. By comparing expression levels between control and GABA-administered larvae, we made a list of the genes upregulated by GABA in the papillae (Table S2). Among them, we compared the expression levels of two genes (KY21.Chr5.240 and KY21.Chr8.489) between GABA-administered and control anterior fragments. GABA-treated anterior fragments expressed these genes at levels at least three times higher than controls (Figure S4H). We concluded that GABA is secreted from and stimulates the papillae to start metamorphosis.
Gs/cAMP is downstream of the Gq pathway
Our next question is: What are the upstream or downstream relationships between GABA, Gq/Ca2+, and Gs/cAMP in the adhesive papillae? We first examined the relationships between the Gq and Gs pathways. Theophylline ameliorated metamorphosis-failed phenotypes in Gαq, PLCβ, and IP3R knockdowns (Figure 4A-E). Moreover, Gαq morphants initiated metamorphosis when caGαs was overexpressed in the nervous system (Figure 4F). These results suggest that the Gq pathway is upstream of the Gs/cAMP pathway. Gαs knockdown larvae exhibited Ca2+ transients in the adhesive papillae upon stimulation (Figure 4G). Because this Ca2+ transient is Gq-dependent (Figure 1O), this result confirms that the Gs pathway is downstream of the Gq-dependent Ca2+ increase.
![](https://prod--epp.elifesciences.org/iiif/2/99825%2Fv1%2Fcontent%2F588679v3_fig4.tif/full/max/0/default.jpg)
Relationships between Gq and Gs pathways in metamorphosis.
(A) A Gαq knockdown larva.
(B) Theophylline ameliorated the effect of Gαq knockdown.
(C) Effect of theophylline on Gαq knockdown.
(D) Effect of theophylline on PLCβ knockdown. In this experiment, PLCβ1/2/3 and PLCβ4 were knocked down simultaneously.
(E) Effect of theophylline on IP3R knockdown.
(F) Effect of caGαs overexpression on Gαq knockdown.
(G) Gαs is not necessary for Ca2+ transient in the adhesive papillae (arrow) upon adhesion.
(H) A Gαq knockdown and Kaede-overexpressed larva.
(I) A Gαq knockdown and caPLCβ1/2/3-overexpressed larva.
(J) A Gαs knockdown and caPLCβ1/2/3-overexpressed larva.
(K) Effect of caPLCβ1/2/3 overexpression on Gαq knockdown.
(L) Effect of caPLCβ1/2/3 overexpression on Gαs knockdown.
(M) Effect of caGαq overexpression on Gαs knockdown.
(N) Effect of theophylline treatment on Gαs knockdown.
To gain further insight into the epistatic order of the Gq and Gs pathways, we overexpressed constitutive active forms of Gq pathway proteins in Gαs morphants with the PC2 cis element. If Gq is upstream of the Gs pathway, forced activation of the Gq pathway by caGαq or caPLCβ1/2/3 would not ameliorate the Gαs MO effect. Indeed, caPLCβ1/2/3, which rescued Gαq morphants well, failed to rescue Gαs morphants (Figure 4H-L). In contrast, caGαq significantly ameliorated the metamorphosis-failed phenocopies of Gαs morphants (Figure 4M), contradicting the hypothesis that Gq is upstream of the Gs pathway. One possibility is that the Gq pathway stimulates cAMP synthesis through a massive Ca2+ increase and protein kinase C (PKC) activation56,57. Indeed, Ciona larvae can synthesize cAMP without Gs, as suggested by the partial rescue of Gαs morphants by theophylline treatment (Figure 4N). These data suggest that Gαs-independent AC could be a target of the Gq pathway. We concluded that the Gq pathway is upstream of the Gs pathway in the signaling cascade initiating Ciona metamorphosis.
Interaction between GABA and Gq pathways
We next investigated the relationships between the GABA and Gq/Gs pathways. As our previous study showed, GABA pathway knockdown by GAD or GABABR1 MO disrupted the initiation of metamorphosis (Figure 5A)36. Overexpression of caPLCβ1/2/3 and theophylline treatment ameliorated the metamorphosis-failed phenocopies of GAD/GABABR1 morphants (Figure 5B-D). These results suggest that the GABA pathway is upstream of the Gq and Gs pathways. We observed Ca2+ transients in the adhesive papillae of GAD knockdown larvae. GAD morphants rarely exhibited a Ca2+ increase after stimulation (Figure 5E). Because the Ca2+ transient in the papillae is Gq-dependent, this result confirms that the GABA pathway is upstream of, or at least in parallel with, the Gq pathway.
![](https://prod--epp.elifesciences.org/iiif/2/99825%2Fv1%2Fcontent%2F588679v3_fig5.tif/full/max/0/default.jpg)
GABA functions in the adhesive papillae to induce metamorphosis.
(A) A GABABR1 knockdown and mCherry-overexpressed larva.
(B) A GABABR1 knockdown and caPLCβ1/2/3-overexpressed larva. Metamorphosis initiated.
(C) Effect of caPLCβ1/2/3 overexpression on GABABR1 knockdown.
(D) Effect of theophylline on GABABR1 and GAD knockdowns.
(E) GABA is necessary for Ca2+ transient in the adhesive papillae (arrow).
(F) Effect of GABA on Gαq knockdown.
(G) Effect of GABA on PLCβ knockdown.
(H) Effect of GABA on Gαs knockdown. See also Figure S4.
However, puzzling results were obtained when we administered Gαq- and Gαs-knockdown larvae with GABA. If GABA is upstream of the Gq and Gs pathways, this chemical does not induce metamorphosis when the Gq or Gs pathway is disrupted. However, GABA significantly ameliorated the metamorphosis-failed phenocopies of Gαq, PLCβ, and Gαs morphants (Figure 5F-H). These results could be explained by assuming enhancement of the Gq pathway by GABA through PLCβ and another GABA-mediated metamorphic pathway bypassing Gq components. These possibilities are examined in the next section.
Contribution of Gi to metamorphosis
The function of GABA as the potential upstream factor of Gq could be easily explained if GABABR is coupled with Gq. Although a few studies suggest this possibility58, Gi is regarded as the major G-protein coupled with GABABR59. A previous study suggested that the Ciona genome encodes one typical Gαi protein40. The gene encoding this Gαi is strongly expressed in the entire larval nervous system including adhesive papillae60. The knockdown of this gene exhibited a significantly reduced rate of metamorphosis (Figure 6A); however, the reduction was not conspicuous, suggesting the presence of another Gαi regulating metamorphosis.
![](https://prod--epp.elifesciences.org/iiif/2/99825%2Fv1%2Fcontent%2F588679v3_fig6.tif/full/max/0/default.jpg)
Gi is required for metamorphosis initiation.
(A) Effect of Gαi knockdown.
(B) A control animal injected with STD MO.
(C) A dvGαi_Chr2 knockdown larva.
(D) A dvGαi_Chr2 knockdown and theophylline-treated animal.
(E) Effects of GABA and theophylline on dvGαi_Chr2 knockdown.
(F) Effects of wtGαi and wtGαs overexpression.
(G) Effect of dnGαi overexpression.
(H) Effect of GABABR1 and Gαq simultaneous knockdown.
(I) Effects of caPLCβ1/2/3 and caGαs overexpressions on GABABR1 and Gαq simultaneous knockdown.
(J) GIRK knockdown promoted metamorphosis initiation.
(K) Effect of U0126 on GABA treatment.
(L) Effect of U0126 on theophylline treatment. See also Figure S5.
A previous study showed that Ciona has four Ciona-specific Gα proteins40. We found three gene models (KY21.Chr2.875, KY21.Chr4.943, and KY21.Chr9.455) encoding these divergent Gα proteins from the newest genome database61. Our phylogenetic analyses showed that the divergent Gα proteins have a strong affinity to the Gαi/o family proteins (Figure S5A). Moreover, the amino acid residues at the C-terminal end of KY21.Chr2.875 are more similar to human Gαi than the other Gα proteins (Figure S5B). The C-terminal residues, particularly glycine at position 3, are essential for determining the G-protein partner of GPCR62. We tentatively name this protein dvGαi_Chr2. The transcriptome data showed that the gene encoding dvGαi_Chr2 is strongly expressed in the papillae (Table S1). The dvGαi_Chr2 knockdown larvae failed to initiate metamorphosis (Figure 6B-C), and this phenocopy was rescued by theophylline and GABA (Figure 6D-E), suggesting that dvGαi_Chr2 is necessary for metamorphosis. This strengthens the hypothesis that GABABR regulates metamorphosis through Gi. The rescue of dvGαi_Chr2 morphants by GABA could be attributable to a redundant function of the typical Gαi protein in the papillae.
We further explored the involvement of Gi upon metamorphosis as the upstream factor of the Gq-Gs pathways. Gi is known to activate PLCβ through the Gβγi complex63,64. Released βγ is inactivated by overexpressing normal (usually GDP-bound) Gα subunits because GDP-Gα quenches the βγ complex65. Overexpressing wild-type Gαi and Gαs in the nervous system suppressed metamorphosis (Figure 6F), suggesting that the activated βγ complex is necessary for initiating metamorphosis. To confirm that the wild-type Gαi exerts its effect through the sequestration of the βγ complex, we overexpressed a dominant-negative form of Gαi (dnGαi)66,67 which has reduced GDP/GTP affinity while maintaining βγ binding activity. dnGαi significantly reduced the occurrence of metamorphosis (Figure 6G), suggesting that the negative effect of Gαi on metamorphosis occurs through interaction with the βγ complex.
If Gβγ-mediated PLCβ activation is used in Ciona metamorphosis, PLCβ receives two independent inputs (Gαq and Gβγi) for its activation. These pathways could compensate for each other. We noticed that the MOs for GABA pathway genes and the Gαq MO did not strongly disrupt metamorphosis compared to the Gαs MO (Figure 1E-F; 5C-D). The compensatory role could explain this phenomenon. Indeed, the simultaneous knockdown of GABABR1 and Gαq resulted in a strong impairment of metamorphosis (Figure 6H), and these morphants initiated metamorphosis by overexpression of caPLCβ1/2/3 or caGαs (Figure 6I).
If the role of the GABA/Gi pathway relies specifically on PLCβ activation in the mechanism of metamorphosis, GABA administration would not rescue PLCβ morphants. However, GABA weakly but significantly induced metamorphosis in PLCβ1/2/3 plus PLCβ4 double-knockdown larvae (Figure 5G). Therefore, GABA/Gi is likely to activate another pathway that bypasses PLCβ. Gβγi is known to activate group III AC56. Its Ciona counterpart (AC5/6) is expressed in the adhesive papillae (Figure S2B), which could explain the results of the rescue experiments. In addition, there are two other major targets of Gβγi. One is the G-protein-activated inwardly rectifying potassium (GIRK) channel68. Our RNA-seq data on the papillae indicated the expression of two GIRK channel genes (Table S1). The GIRK channel negatively regulates the excitation of neurons through hyperpolarization. If the metamorphosis of Ciona is induced by the excitation of papilla neurons as suggested previously16,22, the GIRK channel is likely to regulate metamorphosis negatively. Indeed, the knockdown of one GIRK channel gene weakly enhanced the initiation of metamorphosis without settlement (Figure 6J). Although the negative role of the GIRK channel supports the involvement of Gβγi in the pathway of metamorphosis, this does not explain the PLCβ-independent activation of the metamorphic pathway by GABA/Gi.
The third function of Gβγi is to activate MAPK signaling through positive regulation of MEK69. PLCβ does not mediate this pathway. It has been suggested that MAPK signaling is required to induce metamorphosis70,71. To show that MEK activation is necessary for inducing metamorphosis mediated by GABA, we treated adhesion-prevented larvae with GABA plus U0126, a potent MEK1/2 inhibitor repeatedly used in ascidians72,73. U0126 reduced the rate of metamorphosis induction by GABA (Figure 6K).
cAMP is also known to activate the pathway that involves MEK1/274. We found that U0126 antagonized the effect of theophylline on metamorphosis (Figure 6L). Therefore, GABA and cAMP have the same target (MEK1/2) to initiate metamorphosis, suggesting their compensatory function and/or that cAMP synthesis is upregulated by GABA-Gβγi. These bypassing pathways could explain the amelioration of the phenocopies of PLCβ and Gαs morphants by GABA (Figure 5G and H).
The constitutive function of wild-type Gαq
We found that Ciona wild-type Gαq (wtGαq) has constitutive activity. In contrast to wtGαs and wtGαi, overexpression of wtGαq did not arrest metamorphosis (Figure S6A). Rather, its overexpression induced metamorphosis without settlement (Figure S6B). The constitutive wtGαq activity was confirmed by the rescue of Gαs morphants (Figure S6C). As mentioned above, caGαq also has these activities (Figure 1L). caGαq-overexpressed larvae had a somewhat rounder trunk shape, abnormal tail bending, and frequent tail twitching, perhaps due to a constitutive increase in the cytosolic Ca2+ concentration (Figure S6D). Overexpression of wtGαq did not show such abnormalities (Figure S6E), suggesting that wtGαq has milder activity than, or works through a different mechanism from, caGαq. To gain further insight into the constitutive effect of wtGαq, we constructed a dominant-negative form of Gαq66 that has reduced GDP/GTP binding activity while maintaining βγ binding activity. Overexpression of dnGαq weakly inhibited metamorphosis (Figure S6F), suggesting that GDP/GTP exchange is important for the constitutive activity of wtGαq.
Discussion
Among chordates, the tunicate ascidian is the only group that exhibits a sessile lifestyle at the adult stage. Understanding how ascidians acquired the mechanism to metamorphose into sessile adults is important for elucidating the evolution of chordates75,76. Through extensive molecular, physiological, and pharmacological analyses, we identified signaling molecules responsible for the initiation of metamorphosis of the ascidian Ciona. Combining the results with previous knowledge, we described a schematic that best represents the cascades running in the adhesive papillae upon adhesion to initiate metamorphosis (Figure 7). This working hypothesis will be the basis for future research on ascidian/tunicate metamorphosis. The hypothesis explains the characteristics of Ciona metamorphosis, the mechanisms of which have not yet been elucidated. The genes, proteins, and signaling pathways in this schematic are essential targets for elucidating how the ancestor of ascidians acquired the metamorphosis system peculiar to this group during evolution.
![](https://prod--epp.elifesciences.org/iiif/2/99825%2Fv1%2Fcontent%2F588679v3_fig7.tif/full/max/0/default.jpg)
Schematic illustration of the signaling cascades for initiating Ciona metamorphosis. AC, adenylate cyclase; ER, endoplasmic reticulum; GPCR, G-protein coupled receptor; PKA, protein kinase A.
Mechanisms measuring duration and strength of adhesions
One mystery of Ciona metamorphosis is that its initiation requires continuous adhesion with the adhesive organ for a certain period. This requirement is suggested for the faithful achievement of metamorphosis only when adhesion is firm enough to allow the transition out of the free-living larval stage. Our previous study suggested that approximately 30 min of adhesion is necessary27. A shorter duration of adhesion does not trigger metamorphosis. Larvae that experience short-term adhesion (such as detaching before 30 min) need to adhere again for 30 min to initiate metamorphosis, suggesting that the experience of temporal adhesion is erased. Our recent study showed that the force given to the papillae by adhesion also affects metamorphosis initiation: weaker force extends the time requirement22. Therefore, Ciona larvae likely possess a mechanism that somehow measures the duration and strength of adhesion and the ability to cancel transient excitations of the adhesive organ. Because fewer than 300 neurons constitute the larval nervous system of Ciona54,55,77, how the larva measures the strength and duration of adhesion with its simple nervous system is an important question to address the mechanisms of metamorphosis.
This study showed that cAMP is an essential second messenger molecule for triggering metamorphosis. Because PDE constitutively degrades cAMP, accumulation of this molecule requires persistent and/or strong activation of the metamorphic pathway that overcomes PDE activity. In other words, the duration and strength of adhesion could be converted into the quantity of cAMP, and when its quantity reaches a threshold, metamorphosis can be initiated. The time between the start of adhesion and metamorphosis initiation may be the period necessary to accumulate sufficient cAMP. This “cAMP timer” mechanism does not demand a complicated neuronal network system and could work in the simple nervous system of Ciona.
G-protein signaling relay for metamorphosis
This study showed that three major G-proteins are involved in the initiation of metamorphosis (Figure 7). This complicated mechanism contrasts the simple determination of metamorphosis initiation by cAMP accumulation. What factor promoted the ancestor of ascidians to acquire this complicated signaling network of metamorphosis? We suspect that incorporating multiple components increases the chance of generating the crosstalk between molecules, providing rigidity and flexibility to the system. Particularly, the involvement of Gi may have been important for achieving metamorphosis at the correct time and condition. GABA and Gi have inhibitory and excitatory functions78,79. As we discussed, Ciona larvae must ignore transient adhesion or stimulation on the papilla, which may occur frequently during swimming by colliding with an object or by strong water flows. At the same time, larvae must maintain sensing ability and start metamorphosis when they encounter an appropriate stimulus. Through the GIRK channel and Gαi, GABA can suppress the excitation and cAMP accumulation in the papilla. The GIRK channel could increase the excitation threshold of papilla neurons, allowing them to ignore weak stimuli. The transient reduction of cAMP (Figure 3E) after papilla stimulation guarantees the requirement of long-time adhesion irrespective of the initial quantity of cAMP. This cAMP reduction could be partly explained by the inhibitory function of Gαi on ACs. PDE1 is known to be activated by Ca2+/calmodulin80. Because PDE1 is abundantly expressed in the papillae, the increase in Ca2+ upon adhesion could also trigger sudden cAMP reduction through PDE1.
Together with its inhibitory function, GABA can activate the Gq pathway, which promotes metamorphosis through Gβγi. Group III AC and MEK could also be the targets of Gβγi. When adhesion is long enough, their activations somehow overcome the inhibitory functions of GABA. The GABA pathway is an ideal player that fulfills multiple requirements in the regulation of metamorphosis through its excitatory and inhibitory functions. We suspect that papillae themselves are the source of GABA for metamorphosis initiation. Because VIAAT/VGAT is not expressed in the papillae, the secretion of GABA from the papilla neurons requires an atypical mechanism. GABA is detected in the cell body of the neurons19. The release of cytosolic GABA may be triggered by the reception of mechanical stimuli using the TRP channel22. A VIAAT-independent cytosolic GABA release is observed in pancreatic beta cells81. Curiously, papilla neurons express transcription factor Islet82.
GPCRs for initiating metamorphosis and atypical Gαq activity
Future studies need to specify the mechanisms underlying Gαi, Gαq, and Gαs activation as well as the interaction between them more precisely. The activation of trimeric G-proteins relies on GPCRs, suggesting the presence of the receptors coupled with Gi, Gq, and Gs for metamorphosis initiation (Figure 7). Among them, Gi is likely to be coupled with GABABR; however, the expression of GABABR in the papillae is weak, and their interaction needs to be clarified in the future. We found that wild-type Gαq exhibits constitutive activity that does not require activation through adhesion. The constitutive Gαq activity requires GTP binding. If a GPCR owns this Gαq activation, its ligand should be supplied constitutively. Wild-type Ciona does not initiate metamorphosis without adhesion even though Gαq is expressed in the adhesive papillae, suggesting that the quantity of Gαq is strictly regulated to prevent autonomous metamorphosis in normal conditions.
We did not identify the GPCR coupled with Gs in the metamorphic mechanism. We suspect that GnRH receptors (GnRHRs) are strong candidates for this role. Our previous studies showed that GnRHs can induce metamorphosis36,37 as a downstream factor of GABA. Among the four genes encoding GnRHR proteins, GnRHR1 and GnRHR2 are expressed in the adhesive papillae83. These GnRHRs stimulate cAMP signaling, and the ligand GnRHs for GnRHR1 are also expressed in the papillae83–85. Therefore, GnRHs could be secreted from adhesive papillae as downstream molecules of GABA and stimulate the Gs pathway through GnRHR1. Future studies targeting the regulation of the secretion and reception of GnRH peptides after stimulation of papillae will improve our understanding of the signaling cascade conducting Ciona metamorphosis.
Understanding whether the G-protein signaling relay occurs in cell-autonomous or non-autonomous fashions is crucial for deepening our understanding of Ciona metamorphosis. Each papilla comprises approximately 20 cells, including four papilla neurons (PNs). We showed that activation of Gq and Gs pathways in PNs is sufficient for initiating metamorphosis. This coincides with the previous reports22,23 that the loss of PNs by disrupting the POUIV transcription factor caused the complete arrest of metamorphosis. However, we do not think PNs are the only cells that activate G-proteins, because Ca2+ and cAMP imaging showed upregulation of fluorescence in the entire papillae16,22. A recent study also reports Ca2+ transient in another cell type than PNs34. Other papilla cells are likely responsible for activating G-proteins by directly responding to adhesion and by receiving signal input from adjacent cells, thereby enhancing the signal strength to accumulate a sufficient quantity of cAMP in the PNs to initiate metamorphosis. In this study, technical limitations prevented us from characterizing cells exhibiting increases in Ca2+ and cAMP; we need to address this question in future studies.
Evolutionary implications
Sessile or benthic marine invertebrates lose locomotive activity when they undergo metamorphosis triggered by adhesion to a substratum86. These animals are suspected of having a system to control the adhesion state, allowing them to repeat attachment and detachment before meeting the appropriate conditions for metamorphosis. For example, barnacle larvae “walk” on the substratum before adhering firmly by secreting cement87. Like Ciona, the larvae of sessile/benthic animals may have a system to initiate metamorphosis only when appropriate adhesion is provided, while erasing stimuli from transient and inappropriate adhesion. GABA serves as the metamorphosis inducer of some benthic invertebrates, including mollusks and echinoderms88–90. Moreover, GPCR is implicated as the mediator of settlement and metamorphosis induction in hydrozoans, mollusks, and barnacles88,91,92. In the abalone and barnacle, the dual use of Gq and Gs pathways in metamorphosis has been suggested93,94. Supported by these shared features in the metamorphic mechanisms, our working hypothesis about the initiation of Ciona metamorphosis (Figure 7) will serve as a cue to elucidate how marine benthic invertebrates regulate their metamorphosis.
Materials and Methods
Animals
Ciona intestinalis Type A wild types collected from Onagawa Bay (Miyagi, Japan) and Onahama Bay (Fukushima, Japan) were cultivated in closed colonies by the staff of the National BioResource Project, Japan. They were kept under a constant light condition to prevent gamete release. Eggs and sperm were collected surgically from gonadal ducts, and insemination was carried out in dishes. To prevent larvae from initiating metamorphosis, the tail’s posterior half was manually cut with a scalpel. Removing the tail prevents larvae from swimming efficiently, and these larvae are usually unable to adhere to a substrate. These adhesion-prevented larvae do not metamorphose, since adhesion is required to initiate metamorphosis. Larvae developed from dechorionated eggs were cultured on a 2% agar-coated dish after tail amputation to prevent metamorphosis. Because larvae stick to plastics, tail amputation is insufficient to avoid their adhesion to the culture dish.
Pharmacological treatment
Larvae or larval anterior fragments isolated with a scalpel were administered overnight with 700 μM GABA (Fujifilm Wako #010-02441) or 1 mM theophylline (Sigma-Aldrich #T1633) dissolved in seawater, or 4 μM U0126 (Promega #V1121) dissolved in DMSO.
Plasmids
The open reading frame (ORF) of Kaede was removed from pSPCiPC2Kaede by inverse PCR with PrimeStar GxL DNA polymerase (Takara-bio #R050). The ORFs of Gαq, Gαs, Gαi, and PLCβ1/2/3 were PCR amplified using full-insert cDNAs95. These PCR fragments were fused with the In-Fusion HD Cloning kit (Clontech #639650). The region encoding XY linker43 was deleted from PLCβ1/2/3 cDNA by inverse PCR to create a constitutively active form. The mutations were introduced by inverse PCR to create constitutive active or dominant-negative forms of Gα cDNAs. The introduced mutations are as follows: caGαq, corresponding to Q223L, abolishes GTPase activity42; caGαs, corresponding to Q227L, abolishes GTPase activity47; dnGαi, corresponding to S47C66,67; and dnGαq, corresponding to S54N66. The ORF of bPAC51 was PCR amplified and inserted between the BamHI and EcoRI restriction sites of pSP-Kaede96 to create pSPbPAC. The PCR-amplified PC2 cis-element46 was inserted into the BamHI site of pSPbPAC. The plasmid DNAs were linearized by a restriction enzyme and purified with the Qiaquick Gel Extraction Kit (Qiagen #28706) before microinjection. The ORF of Pink Flamindo52 was PCR amplified and inserted into the EcoRV restriction sites of pBS-HTB97,98 to create pHTBPinkFlamindo. The plasmids pHTBGCaMP822 and pHTBPinkFlamindo were linearized using XhoI for subsequent in vitro synthesis of GCaMP8 and Pink Flamindo mRNA, respectively. GCaMP8 mRNA was synthesized with the MEGAscript T3 kit (Thermo Fisher Scientific #AM1338), the Poly (A) Tailing kit (Thermo Fisher Scientific #AM1350), and the Cap structure analog (New England Biolabs #S1404). Pink Flamindo mRNA was synthesized with the mMESSAGE mMACHINE T3 Transcription kit (Thermo Fisher Scientific #AM1348).
Microinjection
Unfertilized eggs were dechorionated in sterilized seawater containing 1% sodium thioglycolate (Fujifilm Wako #590-11762) and 0.05% actinase E (Kaken Pharmaceutical #650164) as previously described99. The microinjection solution included 2 mg/ml of Fast Green FCF (Fujifilm Wako #061-00031), 0.5-1.0 mM of morpholino oligonucleotide(s) (MOs)100, 5 ng/ul of plasmids, and/or 1 mg/ml of GCaMP8 mRNA for imaging22. For cAMP imaging, 1 mg/ml Pink Flamindo mRNA was dissolved in water without Fast Green, since this chemical emits fluorescence. Microinjected unfertilized eggs were inseminated in a gelatin-coated plastic dish. After the seawater was exchanged to remove excess sperm, the fertilized eggs were cultured at 18 °C or 20 °C overnight until the larval stage. MOs are listed below. Standard control MO, 5’-CCTCTTACCTCAGTTACAATTTATA-3’; GAD ATG MO36, 5’-ACCTCCAAGCCGATTGTTTCTGCAT-3’; GABABR1 ATG MO36, 5’-GCTTACGACTTTACATAACCTTACA-3’; Gαq ATG MO, 5’-GGCATATTTGTGACTATAATGACG-3’; Gαs ATG MO, 5’-AAAGCAACCCATTGGCATTATCGAC-3’; PLCβ1/2/3 splicing MO, 5’-GTGTTACTTACGCTTTCTCTA-3’; PLCβ4 splicing MO, 5’-AACCACCAACCACCAACCTTTTG-3’; IP3R splicing MO, 5’-AATGATGGTTTAAAATTGCCACCTG-3’; Gαi ATG MO, 5’-GTGGAGACTGTGCAACCCATGATTC-3’; dvGai_Chr2 ATG MO, 5’-CCATCTTGAGTAATCCAGGCTTTTA-3’; GIRK channel ATG MO, 5’-TCTGCTGGTTCAGTAATAGACATAG-3’.
Photographing and imaging
Photographs were taken with an AxioImager Z1 and AxioObserver Z1 (Carl Zeiss). The images were treated with AxioVision Rel.4.6 or Zen (Carl Zeiss) and Photoshop 2021 (Adobe) software. Imaging of the Ca2+ transient was carried out according to previous reports16,22. The tails of GCaMP8 mRNA-injected larvae were removed by a scalpel at 15-17 hours after fertilization (hpf). At 30-40 hpf, the larvae were mounted onto a plastic dish under an M165FC stereo microscope (Leica Microsystems), and their adhesive papillae were stimulated by a glass rod using a three-axis coarse manipulator M-152 (Narishige Japan). Immediately after the tip of the rod had contact with the papillae, visible light was shut off and the fluorescence of GCaMP8 was taken with a DFC 310FX digital camera (Leica), Las version 4.12 (Leica), and the movie-capturing function of Windows 10. Images were treated with Photoshop 2021.
Imaging of cAMP was taken by confocal laser scanning microscopy (FV1000, Olympus). The movies were treated with ImageJ (National Institutes of Health). The Pink Flamindo mRNA-injected larvae were immobilized on Poly L lysine-coated glass-bottom dishes at 20-21 hpf. For stimulation of their adhesive papillae with a glass rod, an electric manipulator MM-89 (Narishige) and hydraulic micromanipulator MMO-202ND (Narishige) were used.
RNA sequencing
Larval anterior portions, including adhesive papillae and the other trunk region, were separately isolated by a scalpel around 22-25 hpf at 18°C and collected in Isogen (Nippon Gene #317-02503) soon after isolation. To identify GABA-responsive genes, half of the tail was amputated from larvae around 21-24 hpf, followed by GABA administration for 2 hours before treatment with Isogen. In each experiment, approximately 50 fragments were collected, and two biological replicates were taken. Total RNA was isolated according to the manufacturer’s instructions. Ribosomal RNA was depleted from the total RNA using the NEBNext rRNA Depletion kit (New England Biolabs #E6310), followed by the conversion of the remaining RNA into an Illumina sequencing library using the NEBNext Ultra Directional RNA Library Prep kit (New England Biolabs #E7760). Following library preparation, validation was performed using the Bioanalyzer system (Agilent Technologies) to assess size distribution and concentration. Subsequently, sequencing was carried out on the NextSeq 500 platform (Illumina) employing the paired-end 36-base read option. Reads were mapped on the Ciona intestinalis Type A reference genome (HT model; http://ghost.zool.kyoto-u.ac.jp/default_ht.html) and quantified using CLC Genomic Workbench version 22.0 (Qiagen). RNA-seq data sets were deposited in the NCBI Gene Expression Omnibus (GEO) under accession numbers SAMN40712866 to SAMN40712873, which will appear online upon publication of this paper.
The read counts were normalized by calculating the number of reads per kilobase per million (RPKM) for each transcript in each sample using CLC Genomic Workbench version 22.0 (Qiagen). Eight candidate genes showing a minimum 1.5-fold increase following GABA administration, along with notably higher expression in the papillae in comparison to the trunk (an approximately 4-fold increase in a sample), were chosen for quantitative PCR analysis to assess their response to GABA.
Quantitative PCR
Total RNA was isolated from the anterior tips of larvae treated with the chemical using Isogen (Nippon Gene), following the manufacturer’s instructions. After isopropanol extraction, genomic DNA removal and reverse transcription were carried out using the PrimeScript™ RT reagent kit with gDNA Eraser (Takara Bio #RR047). Quantitative RT-PCR (qRT-PCR) was done using a SYBR Premix Ex TaqII Dimer Eraser (Takara Bio #RR091) and a Thermal Cycler Dice Real Time System III (Takara Bio). The gene encoding elongation factor 1α (EF1α) was used to normalize RNA quantity according to a previous study98. qPCR primers are listed below. KY21.Chr5.240, 5’-TCTTCTCAAAGTTGCACATTCC-3’ and 5’-CAGCAGCAACCAAACGATAAAC-3’; KY21.Chr8.489, 5’-CAATGCAACTTTGACTGCATAC-3’ and 5’-TCCAAACTGCATTCCACATATC-3’; EF1α, 5’-CATGTCACGGACAGCGAAACG-3’ and 5’-CAATGTGTGTTGAGGCATTCCAAG-3’.
Statistical analysis
Differences between conditions were evaluated by Fisher’s exact test. Statistical analyses and most graph visualizations were conducted utilizing R x64.4.1.2 and RStudio software.
Phylogenetic analysis
We performed a tBlastn search101 against the HT gene model of Ciona61 using the amino acid sequences of human PDE, PLCβ, and Ciona Gα proteins as queries, followed by a reciprocal Blast search. We aligned the amino acid sequences from the HDc domain of PDE, the EF-hand_like, PLCXc, PLCYc, and C2 domains of PLCβ, and the whole open reading frame of Gα proteins using the M-COFFEE program102,103. After the removal of unnecessary residues, maximum likelihood trees were constructed using MEGA software version 11104, employing the WAG amino acid substitution matrix105. The trees were assessed with 1,000 bootstrap pseudoreplicates.
Acknowledgements
We would like to express our earnest gratitude to Drs. Kazuki Horikawa, Atsuo Nishino, Ryusuke Niwa, and Takahiro Yamashita for their kind material provision, helpful comments, and general support of our study. We thank Dr. Kogiku Shiba for instructing us with the pharmacological analyses. We also thank Dr. Masafumi Muratani and members of i-Laboratory at the University of Tsukuba for their support for RNA sequencing. We are grateful to the past and present members of the Shimoda Marine Research Center at the University of Tsukuba for their contributions to the initial step of this study and the maintenance of the animals. We thank Drs. Shigeki Fujiwara, Manabu Yoshida, Yutaka Satou, and all members of the Department of Zoology, Kyoto University, the Misaki Marine Biological Station, the University of Tokyo, the Maizuru Fishery Research Station of Kyoto University, and the National BioResource Project (NBRP) for the cultivation and provision of Ciona adults and experimental materials. This study was supported by grants from the Japan Society for the Promotion of Science to K.H. (21H00440, 23H04717), T.H. (19H03204, 21K19249, 21H05239), and Y.S. (19H03262). Y.S. was further supported by a Takeda Bioscience Research Grant. T.H. was supported by the Collaborative Research in Computational Neuroscience program (CRCNS2021). N.M.T was supported by JST SPRING (JPMJSP2123).
Supplementary Figures
![](https://prod--epp.elifesciences.org/iiif/2/99825%2Fv1%2Fcontent%2F588679v3_figs1.tif/full/max/0/default.jpg)
Ciona phospholipase β proteins
Phylogenetic relationships between PLCβ proteins, as revealed by the maximum likelihood method. The number beside each branch indicates the percentage of times that a node was supported in 1,000 bootstrap pseudoreplications. Scores equal to or exceeding 50% are shown. Ciona PLCβs are indicated by black circles. Human homologs are indicated by accession numbers in the Uniprot database (https://www.uniprot.org) and “HS” for Homo sapiens.
![](https://prod--epp.elifesciences.org/iiif/2/99825%2Fv1%2Fcontent%2F588679v3_figs2.tif/full/max/0/default.jpg)
Ciona phosphodiesterase repertoire
(A) Phylogenetic relationships between PDEs. Black circles and gene models indicate Ciona protein models in the Ghost database (http://ghost.zool.kyoto-u.ac.jp/default_ht.html). The numbers and letters in parentheses indicate the PDE groups and their suspected substrates. A, cAMP; G, cGMP; D, dual substrate. Human and insect proteins are shown by Uniprot (https://www.uniprot.org) accession numbers, as well as “HS” for Homo sapiens and “DM” for Drosophila melanogaster proteins.
(B) Expression profiles of the genes related to Gq and Gs pathways, as revealed by RNA-seq of the papilla region. RPKM, reads per kilobase of exon per million mapped reads. See also Table S1.
![](https://prod--epp.elifesciences.org/iiif/2/99825%2Fv1%2Fcontent%2F588679v3_figs3.tif/full/max/0/default.jpg)
cAMP increase in the papillae is coupled with metamorphosis initiation
Measurement of cAMP concentration in the larva that did not initiate metamorphosis upon stimulation of the papillae, as indicated by the fluorescence of Pink Flamindo. Left, the quantification of Pink Flamindo fluorescent intensity of a larva that did not initiate metamorphosis; right, the position of the region of interest (ROI).
![](https://prod--epp.elifesciences.org/iiif/2/99825%2Fv1%2Fcontent%2F588679v3_figs4.tif/full/max/0/default.jpg)
GABA stimulates the adhesive papillae to initiate metamorphosis.
(A) A GABA-treated animal initiated metamorphosis.
(B) A papilla-amputated larva treated with GABA. Metamorphosis was not initiated.
(C) Effect of papilla amputation on GABA treatment.
(D-H) The sensory vesicle is not required for the anterior region to respond to GABA.
(D) Schematic of the experiment.
(E) A control anterior fragment at 2 dpf.
(F) Anterior fragment treated with GABA. The percentage indicates the proportion of the fragments exhibiting the morphological signature of metamorphosis.
(G) Anterior fragment treated with theophylline.
(H) Expression levels of two GABA-responsive genes in the control- and GABA-treated anterior fragments, as revealed by quantitative RT-PCR.
![](https://prod--epp.elifesciences.org/iiif/2/99825%2Fv1%2Fcontent%2F588679v3_figs5.tif/full/max/0/default.jpg)
Ciona has atypical Gαi proteins.
(A) Phylogenetic relationships between Gα proteins, as revealed by the maximum likelihood method. The three atypical Gα proteins are highlighted in red.
(B) Comparison of the five C-terminal amino acid sequences of Gα proteins that are shown to be essential for their functions, including two additional amino acids for dvGαi_Chr2. Residues conserved with human Gαi1 are shown in red.
![](https://prod--epp.elifesciences.org/iiif/2/99825%2Fv1%2Fcontent%2F588679v3_figs6.tif/full/max/0/default.jpg)
Autonomous activity of wild-type Gαq.
(A) Effect of wtGαq overexpression in the adhesion-permitted condition.
(B) Effect of wtGαq overexpression in the adhesion-prevented condition.
(C) Effect of wtGαq overexpression on Gαs knockdown.
(D) Morphology of caGαq-overexpressed larva at 1 dpf (before initiation of metamorphosis).
(E) Morphology of wtGαq-overexpressed larva at 1 dpf.
(F) Effect of dnGαq overexpression in the adhesion-permitted condition.
![](https://prod--epp.elifesciences.org/iiif/2/99825%2Fv1%2Fcontent%2F588679v3_tbls1.tif/full/max/0/default.jpg)
Expression levels of genes related to G-protein signaling during metamorphosis, as revealed by RNA-seq of the larval anterior tip including papillae
![](https://prod--epp.elifesciences.org/iiif/2/99825%2Fv1%2Fcontent%2F588679v3_tbls2.tif/full/max/0/default.jpg)
List of candidate genes upregulated by GABA
References
- 1.Developmental Biology
- 2.Indirect development, transdifferentiation and the macroregulatory evolution of metazoansPhilos. Trans. R. Soc. Lond. B Biol. Sci 365:653–669
- 3.Metamorphic Competence, a Major Adaptive Convergence in Marine Invertebrate Larvae1A MER. ZOOL 41:1123–1131
- 4.The initiation of metamorphosis as an ancient polyphenic trait and its role in metazoan life-cycle evolutionPhilos. Trans. R. Soc. Lond. B Biol. Sci 365:641–651
- 5.Effective induction of abalone settlement and metamorphosis corresponds to biomolecular composition of natural cuesCommun. Integr. Biol 2:347–349
- 6.Molecular and behavioural evidence that interdependent photo - and chemosensory systems regulate larval settlement in a marine spongeMol. Ecol 29:247–261
- 7.Host specificity and adaptive evolution in settlement behaviour of coral-associated barnacle larvae (Cirripedia: Pyrgomatidae)Sci. Rep 13
- 8.Tunicates and not cephalochordates are the closest living relatives of vertebratesNature 439:965–968
- 9.The ascidian tadpole larva: comparative molecular development and genomicsNat. Rev. Genet 4:285–295
- 10.A mechanism for graded motor control encoded in the channel properties of the muscle ACh receptorProc. Natl. Acad. Sci. U. S. A 108:2599–2604
- 11.Simple motor system of the ascidian larva: neuronal complex comprising putative cholinergic and GABAergic/glycinergic neuronsZoolog. Sci 27:181–190
- 12.Metamorphosis in solitary ascidiansGenesis 53:34–47
- 13.The ontology of the anatomy and development of the solitary ascidian Ciona: the swimming larva and its metamorphosisSci. Rep 10
- 14.Ascidian larvae and the events of metamorphosisAm. Zool 22:817–826
- 15.Papillae revisited and the nature of the adhesive secreting collocytesDev. Biol 448:183–198
- 16.Two-Round Ca2+ transient in papillae by mechanical stimulation induces metamorphosis in the ascidian Ciona intestinalis type AProceedings of the Royal Society B 288https://doi.org/10.1098/rspb.2020.3207
- 17.Preprint
- 18.Glutamatergic networks in the Ciona intestinalis larvaJ. Comp. Neurol 508:249–263
- 19.GABAergic synaptic transmission modulates swimming in the ascidian larvaEur. J. Neurosci 22:2541–2548
- 20.Neuronal identities derived by misexpression of the POU IV sensory determinant in a protovertebrateProc. Natl. Acad. Sci. U. S. A 119https://doi.org/10.1073/pnas.2118817119
- 21.Delineating metamorphic pathways in the ascidian Ciona intestinalisDev. Biol 326:357–367
- 22.The TRP channel PKD2 is involved in sensing the mechanical stimulus of adhesion for initiating metamorphosis in the chordate CionaDev. Growth Differ 64:395–408
- 23.Using CRISPR/Cas9 to identify genes required for mechanosensory neuron development and functionBiol. Open 12https://doi.org/10.1242/bio.060002
- 24.Cryptic speciation in a model invertebrate chordateProc. Natl. Acad. Sci. U. S. A 104:9364–9369
- 25.Morphological Differences between Larvae of the Ciona intestinalis Species Complex: Hints for a Valid Taxonomic Definition of Distinct SpeciesPLoS One 10
- 26.Morphological evidence that the molecularly determinedCiona intestinalistype A and type B are different species:Ciona robustaandCiona intestinalisJ. Zoolog. Syst. Evol. Res 53:186–193
- 27.Time course for tail regression during metamorphosis of the ascidian Ciona intestinalisDev. Biol 405:71–81
- 28.Presence of thyroid hormones in ascidian larvae and their involvement in metamorphosisJ. Exp. Zool 290:426–430
- 29.Novel gene containing multiple epidermal growth factor–like motifs transiently expressed in the papillae of the ascidian tadpole larvaeDev. Dyn 210https://doi.org/10.1002/(SICI)1097-0177(199711)210:3<264::AID-AJA7>3.0.CO;2-E
- 30.Cholinergic activation of settlement in Ciona intestinalis metamorphosing larvaeJ. Exp. Zool 280:314–320
- 31.Hemps, a novel EGF-like protein, plays a central role in ascidian metamorphosisDevelopment 126:5809–5818
- 32.Interaction between noradrenaline or adrenaline and the beta 1-adrenergic receptor in the nervous system triggers early metamorphosis of larvae in the ascidian, Ciona savignyiDev. Biol 258:129–140
- 33.Dopamine and serotonin modulate the onset of metamorphosis in the ascidian Phallusia mammillataDev. Biol 282:246–256
- 34.Polymodal sensory perception drives settlement and metamorphosis of Ciona larvaeCurr. Biol 34:1168–1182
- 35.The effect of l-thyroxine on the metamorphosis of Ascidia malacaCell Tissue Res 214:289–301
- 36.GABA-Induced GnRH Release Triggers Chordate MetamorphosisCurr. Biol 30:1555–1561
- 37.Nonreproductive role of gonadotropin-releasing hormone in the control of ascidian metamorphosisDev. Dyn 243:1524–1535
- 38.Molecular mechanisms of metabotropic GABAB receptor functionSci. Adv 7
- 39.Molecular biology of the cell 6th edGarland Science
- 40.The repertoire of heterotrimeric G proteins and RGS proteins in Ciona intestinalisPLoS One 4
- 41.Mechanism of activation and inactivation of Gq/phospholipase C-β signaling nodesChem. Rev 111:6120–6129
- 42.Study of the constitutively active form of the alpha subunit of rice heterotrimeric G proteinsPlant Cell Physiol 46:381–386
- 43.Membrane-induced Allosteric Control of Phospholipase C-β Isozymes*J. Biol. Chem 289:29545–29557
- 44.Comprehensive morphological analysis of individual peripheral neuron dendritic arbors in ascidian larvae using the photoconvertible protein KaedeDev. Dyn 243:1362–1373
- 45.The ventral peptidergic system of the adult ascidian Ciona robusta (Ciona intestinalis Type A) insights from a transgenic animal modelSci. Rep 10
- 46.The nervous system of the adult ascidian Ciona intestinalis Type A (Ciona robusta): Insights from transgenic animal modelsPLoS One 12
- 47.Constitutive activation of Galphas within forebrain neurons causes deficits in sensorimotor gating because of PKA-dependent decreases in cAMPNeuropsychopharmacology 32:577–588
- 48.G protein pathwaysScience 296:1636–1639
- 49.Cyclic nucleotide phosphodiesterasesJ. Allergy Clin. Immunol 108:671–680
- 50.Analogues of AdenosineTheophylline, and Caffeine: Selective Interactions with A1 and A2 Adenosine Receptors. In Topics and Perspectives in Adenosine Research (Springer Berlin Heidelberg) :23–36
- 51.Light modulation of cellular cAMP by a small bacterial photoactivated adenylyl cyclase, bPAC, of the soil bacterium BeggiatoaJ. Biol. Chem 286:1181–1188
- 52.Red fluorescent protein-based cAMP indicator applicable to optogenetics and in vivo imagingSci. Rep 7
- 53.Developmental expression of glutamic acid decarboxylase and of gamma-aminobutyric acid type B receptors in the ascidian Ciona intestinalisJ. Comp. Neurol 506:489–505
- 54.The CNS connectome of a tadpole larva of Ciona intestinalis (L.) highlights sidedness in the brain of a chordate siblingElife 5https://doi.org/10.7554/eLife.16962
- 55.The peripheral nervous system of the ascidian tadpole larva: Types of neurons and their synaptic networksJ. Comp. Neurol 526:583–608
- 56.Physiological roles for G protein-regulated adenylyl cyclase isoforms: insights from knockout and overexpression studiesNeurosignals 17:5–22
- 57.Store-operated Ca2+-entry and adenylyl cyclaseCell Calcium 58:368–375
- 58.GABA(B) receptors couple to Gαq to mediate increases in voltage-dependent calcium current during developmentJ. Neurochem 135:88–100
- 59.Structural basis of GABAB receptor-Gi protein couplingNature 594:594–598
- 60.Identification of neuron-specific promoters in Ciona intestinalisGenesis 39:130–140
- 61.A Manually Curated Gene Model Set for an Ascidian, Ciona robusta (Ciona intestinalis Type A)Zoolog. Sci 39https://doi.org/10.2108/zs210102
- 62.Substitution of three amino acids switches receptor specificity of Gqα to that of GiαNature 363:274–276
- 63.GABABreceptor activation protects neurons from apoptosis via IGF-1 receptor transactivationJ. Neurosci 30:749–759
- 64.Gi-coupled γ-aminobutyric acid-B receptors cross-regulate phospholipase C and calcium in airway smooth muscleAm. J. Respir. Cell Mol. Biol 45:1232–1238
- 65.Identification of a Structural Element in Phospholipase C β2 That Interacts with G Protein βγ Subunits*J. Biol. Chem 273:7148–7154
- 66.Mechanisms of dominant negative G- protein α subunitsJ. Neurosci. Res 85:3505–3514
- 67.Functional analysis of a dominant negative mutant of Gαi2J. Biol. Chem 270:4037–4041
- 68.The roles of Gβγ and Gα in gating and regulation of GIRK channelsIn International Review of Neurobiology International review of neurobiology. (Elsevier :27–85
- 69.G protein-coupled receptor-mediated mitogen-activated protein kinase activation through cooperation of Galpha(q) and Galpha(i) signalsMol. Cell. Biol 20:6837–6848
- 70.Tail regression in Ciona intestinalis (Prochordate) involves a Caspase-dependent apoptosis event associated with ERK activationDevelopment 129:3105–3114
- 71.ERK- and JNK-signalling regulate gene networks that stimulate metamorphosis and apoptosis in tail tissues of ascidian tadpolesDevelopment 134:1203–1219
- 72.Role of the FGF and MEK signaling pathway in the ascidian embryoDev. Growth Differ 43:521–533
- 73.A conserved role for the MEK signalling pathway in neural tissue specification and posteriorisation in the invertebrate chordate, the ascidian Ciona intestinalisDevelopment 130:147–159
- 74.G protein regulation of MAPK networksOncogene 26:3122–3142
- 75.Cardiopharyngeal deconstruction and ancestral tunicate sessilityNature 599:431–435
- 76.A mid-Cambrian tunicate and the deep origin of the ascidiacean body planNat. Commun 14
- 77.Neuronal identity: the neuron types of a simple chordate sibling, the tadpole larva of Ciona intestinalisCurr. Opin. Neurobiol 56:47–60
- 78.GABA: a pioneer transmitter that excites immature neurons and generates primitive oscillationsPhysiol. Rev 87:1215–1284
- 79.Heterotrimeric G protein subunit Gαq is a master switch for Gβγ-mediated calcium mobilization by GI-coupled GPCRsMol. Cells https://doi.org/10.2139/ssrn.3578140
- 80.Clinical and molecular genetics of the phosphodiesterases (PDEs)Endocr. Rev 35:195–233
- 81.Mechanism and effects of pulsatile GABA secretion from cytosolic pools in the human beta cellNat Metab 1:1110–1126
- 82.Islet is a key determinant of ascidian palp morphogenesisDevelopment 141:3084–3092
- 83.A conserved non-reproductive GnRH system in chordatesPLoS One 7
- 84.Tunicate gonadotropin-releasing hormone (GnRH) peptides selectively activate Ciona intestinalis GnRH receptors and the green monkey type II GnRH receptorEndocrinology 146:4061–4073
- 85.Invertebrate Gonadotropin-Releasing Hormone Receptor Signaling and Its Relevant Biological ActionsInt. J. Mol. Sci 21https://doi.org/10.3390/ijms21228544
- 86.Ecological regulation of development: induction of marine invertebrate metamorphosisInt. J. Dev. Biol 46:679–686
- 87.Settlement behavior and antennulary biomechanics in cypris larvae of Balanus amphitrite (Crustacea: Thecostraca: Cirripedia)Mar. Biol 141:513–526
- 88.GABA induces metamorphosis in Haliotis, V: Stereochemical specificityBrain Res. Bull 5:381–387
- 89.Induction of metamorphosis and substratum preference in four sympatric and closely related species of sea urchins (genus Echinometra) in OkinawaZool. Stud 40:29–43
- 90.Induction of metamorphosis of larvae of the Green Sea urchin, Strongylocentrotus droebachiensis, by coralline red algaeBiol. Bull 179:304–311
- 91.Evidence for the involvement of PI-signaling and diacylglycerol second messengers in the initiation of metamorphosis in the hydroid Hydractinia echinata FlemingDev. Biol 121:82–89
- 92.Ion and neuropharmacological studies of barnacle settlementNeth. J. Sea Res 20:269–275
- 93.Signal transduction in barnacle settlement: Calcium re-visitedBiofouling 10:141–159
- 94.Metamorphic-Signal Transduction in Hydroides elegans (Polychaeta: Serpulidae) Is Not Mediated by a G ProteinBiol. Bull 195:21–29
- 95.A multicassette Gateway vector set for high throughput and comparative analyses in ciona and vertebrate embryosPLoS One 2
- 96.Efficient transposition of a single Minos transposon copy in the genome of the ascidian Ciona intestinalis with a transgenic line expressing transposase in eggsDev. Dyn 239:1076–1088
- 97.Notch signaling is involved in nervous system formation in ascidian embryosDev. Genes Evol 212:459–472
- 98.Maternal factor-mediated epigenetic gene silencing in the ascidian Ciona intestinalisMol. Genet. Genomics 283:99–110
- 99.Microinjection of exogenous nucleic acids into eggs: Ciona speciesAdv. Exp. Med. Biol 1029:5–13
- 100.Action of morpholinos in Ciona embryosGenesis 30:103–106
- 101.Gapped BLAST and PSI-BLAST: a new generation of protein database search programsNucleic Acids Res 25:3389–3402
- 102.T-Coffee: A novel method for fast and accurate multiple sequence alignmentJ. Mol. Biol 302:205–217
- 103.M-Coffee: combining multiple sequence alignment methods with T-CoffeeNucleic Acids Res 34:1692–1699
- 104.MEGA11: Molecular Evolutionary Genetics Analysis version 11Mol. Biol. Evol 38:3022–3027
- 105.A general empirical model of protein evolution derived from multiple protein families using a maximum-likelihood approachMol. Biol. Evol 18:691–699
Article and author information
Version history
- Preprint posted:
- Sent for peer review:
- Reviewed Preprint version 1:
Copyright
© 2024, Hozumi et al.
This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.