Introduction

Scleractinian corals are characterized by a calcium carbonate skeleton produced by the epidermis, and by symbiotic dinoflagellates (zooxanthellae, Symbiodinium spp.) harbored in the gastrodermal cells. Corals begin life as soft-bodied planktonic planula larvae and undergo metamorphosis to primary polyps (Harrison and Wallace 1990). Coral planulae develop some tissues but not tentacles or skeletons. In particular, the formation of the mouth and coelenteron is necessary to feed and acquire the Symbiodinium for non-symbiotic planula.

Two distinct modes of sexual reproduction occur among scleractinian corals: gamete release followed by external fertilization and development (spawning); and brooding of planulae within the polyp (brooding). Brooding planulae are usually released at an advanced stage of development and possess a mouth, coelenteron and some mesenteries, and tend to already contain Symbiodinium when they are released from the parent colony (Harrison and Wallace 1990; Isomura and Nishihira 2001). On the other hand, the planulae of spawning corals usually require up to a week after fertilization to attach to the substrate and then metamorphose into primary polyps (Babcock and Heyward 1986; Harrison and Wallace 1990). Most gametes of spawning coral species do not contain symbionts (Harrison and Wallace 1990). The structure of planulae is different between different species of spawning corals (e.g., Szmant-Froelich et al. 1980; Babcock and Heyward 1986; Hayashibara et al. 1997; Hirose et al. 2000; Hirose and Hidaka 2006).

Acropora is the dominant and most speciose genus of scleractinian corals in the Indo-Pacific (Veron and Wallace 1984), and all Acropora species studied thus far have been spawners, releasing eggs that do not contain Symbiodinium. New generations of Acropora must, therefore, acquire Symbiodinium from the environment. Hayashibara et al. (1997) described the early development of four Acropora species (Acropora hyacinthus, Acropora nasuta, Acropora florida, and Acropora secale). The planulae of these species had an oral pore and mesogleal layer, but no Symbiodinium were present. Since Hayashibara et al. (1997) did not examine the process of metamorphosis, they were unable to show the order of development of a functional mouth, coelenteron, mesenteries, tentacles, and skeleton, and it is still unclear exactly when Symbiodinium are acquired in Acropora species.

Most coral planulae select a site of permanent attachment using external chemical cues that induce the metamorphosis from planulae into polyps (Morse et al. 1988, 1994, 1996; Heyward and Negri 1999; Iwao et al. 2002; Webster et al. 2004). Iwao et al. (2002) reported that Hym-248 (EPLPIGLWa), a cnidarian neuropeptide of the GLWamide family, induces the metamorphosis of planulae of the genus Acropora. Hym-248 appears to act as an internal mediator, after release from neurons following an external stimulus, triggering metamorphosis. Hym-248 induces synchronous metamorphosis and settlement of Acropora planulae, and is, therefore, a useful tool for the study of Acropora larval metamorphosis.

The aims of this study were (1) to examine the process of metamorphosis from planula larva into the primary polyp in the scleractinian corals Acropora nobilis and Acropora microphthalma using the cnidarian neuropeptide Hym-248, and (2) to describe the process of acquisition of the freshly isolated Symbiodinium (zooxanthellae) (FIZ) by the non-symbiotic polyps.

Material and methods

Preparation of planula larvae

Colonies of Acropora spp. (A. nobilis and/or A. microphthalma) were maintained in an outdoor tank (4 m × 5 m × 1.5 m) supplied with unfiltered running seawater at Okinawa Churaumi Aquarium. Several colonies of both coral species released egg-sperm bundles on 12 June 2006 between 2100 and 2200 hours. The two species of Acropora were kept in the same tank and spawned simultaneously, as a result, the embryos could not be differentiated into individual species. However, from preliminary observations, the embryonic development and course of metamorphosis for these two species is known to be very similar. Released gametes were collected and mixed for 1–2 h to allow fertilization, and then transferred to plastic containers filled with seawater filtered through a 0.45 μm-pore (FSW). Embryos were maintained in the containers at room temperature (27–29°C) in FSW changed daily.

Hym-248 treatment

Hym-248 (EPLPIGLWamide) was purchased from Genenet Co. Ltd, Fukuoka, Japan. Swimming planulae 2–16 days after fertilization were washed several times with FSW, and one or two planulae were then incubated in a 20–30 μl drop of peptide solution (4 × 10−6 M Hym-248 in FSW) on waterproof parafilm sheets (Parafilm, Alcan Inc., Wisconsin). The droplets containing planulae were incubated on the sheet in a moist-chamber at room temperature (27–29°C) for 24 h. The parafilm sheets were then fixed with a clamp made of silicon tube and nylon thread and incubated in small containers containing FSW.

Symbiodinium acquisition by primary polyps

In previous experiments, non-symbiotic primary polyps of A. nobilis initially took up freshly isolated Symbiodinium (zooxanthellae) (FIZ) from the bivalve Tridacna crocea more readily than FIZ from A. nobilis (parent coral colonies), the soft coral Xenia sp. or the sea anemone Aiptasia sp. (Y. Higuchi, personal communication). Consequently, in the present study, in order to examine the course of Symbiodinium acquisition, primary polyps of A. nobilis and A. microphthalma were exposed to FIZ from T. crocea.

A single specimen of T. crocea was collected from the shallow reef lagoon (<1 m) at Bise (26°42′32″N, 127°52′45″E) for Symbiodinium isolation. Restriction fragment length polymorphism (RFLP) patterns of 18S rDNA indicated that the Symbiodinium from T. crocea belonged to clade A (sensu Rowan and Powers 1991). The animal tissue (mantle) of T. crocea was cut out and homogenized. The Symbiodinium were cleaned several times with FSW and concentrated by centrifugation. Primary polyps of both Acropora spp. (1–2 days after Hym-248 treatment) were exposed to FIZ (105 cells ml−1) at 27°C under a 12:12 hours light (100 μmol m−2 s−1): dark cycle. Six hours after exposure to FIZ, the seawater in the polyp cultures was replaced with FSW. Polyp development was monitored and seawater was changed daily.

Histological observation and electron microscopy

Live planulae and polyps were fixed, dehydrated, embedded and sectioned (Hirose and Hidaka 2006). The skeleton was decalcified with 1% ascorbic acid—0.15M NaCl for 2 days. For scanning electron microscope (SEM), specimens were immersed in t-butanol and freeze-dried. Dried specimens were sputter-coated with gold-palladium, and examined under a scanning electron microscope (SEM; JEM-6060LV, JEOL Ltd, Tokyo).

Results

Embryonic development and swimming planulae

The eggs of both Acropora species (diameter ∼400 μm) did not contain any Symbiodinium. About 2 h after spawning (=fertilization), more than 90% of embryo began to cleave. Forty-eight hours after fertilization, embryos became elongated and started to swim using cilia. Gastric cavities (coelenteron) were not present in non-Hym-248 treated swimming planulae until 16 days after fertilization (Fig. 1a, b). These planulae consisted of two layers of cells; epidermis (ectoderm) and gastrodermis (endoderm), which were separated by an extracellular matrix (ECM) called the mesoglea (Fig. 1a, inset). The planula surface was completely covered with cilia surrounded by microvilli (Fig. 1c). Epidermal cells of planulae were not arranged into a monolayer. Most epidermal cells were columnar and contained round granules (ca. 1-μm diameter) in the apical cytoplasm (Fig. 1d). Some round cells were found in the basal half of the epidermis, attached to the mesoglea (Fig. 1d). In contrast, gastrodermal cells were spherical in shape (Fig. 1e) and contained many lipid granules (Fig. 1a).

Fig. 1
figure 1

Swimming planula and initial stages of metamorphosis of Acropora spp. a Histological section of swimming planula, square indicates the area enlarged in inset; inset: arrows indicate mesogleal layer (bar = 10 μm), b SEM micrograph of fractured swimming planula, c highly magnified SEM micrograph of epidermal surface of planula. Cilia surrounded by microvilli (arrows). d SEM micrograph of a fractured epidermal layer of planula. Arrows indicate mesogleal layer, double arrows indicate round cells in the epidermis, arrowheads indicate columnar cells in the epidermis, and double arrowheads indicate granules in columnar cell, e SEM micrograph of a fractured gastrodermis of planula showing spherical gastrodermal cells, f elongated planula attached to parafilm (just after exposure to Hym-248), g polyp with six primary mesenteries (6 h after exposure to Hym-248). Arrowheads indicate mesenteries, h histological section of metamorphosing polyp (6 h after exposure to Hym-248), square indicates the area enlarged in inset; inset: arrow indicates the just-extended mesogleal layer (bar = 10 μm) (ci cilium, co coelenteron, ep epidermis, ga gastrodermis, mi microvilli). Bars = 100 μm in a, b, f, g and h, 10 μm in d, e, 1 μm in c

Full size image

Metamorphosis of planula larvae from planula to primary polyp

When swimming planulae were exposed to the cnidarian neuropeptide Hym-248 at 4 × 10−6 M, they first moved actively for a few minutes, then, the elongated planulae attached themselves perpendicularly to the waterproof parafilm sheets and stood upright (Fig. 1f). Six to nine hours after exposure to Hym-248, metamorphosing planulae spread out on the sheets and formed six primary mesenteries, which we designated the “pumpkin-stage” (Fig. 1g). These planulae metamorphosis experiments were performed a total of 12 times for 2–16 days after fertilization, and more than 1,000 planulae were treated with Hym-248. In all cases, more than 90% of the planulae metamorphosed into the “pumpkin-stage” within 9 h. In the “pumpkin-stage” a cavity (coelenteron) could be seen in the larvae (Fig. 1h). This coelenteron appeared to form as a result of secondary delamination during metamorphosis.

The mesenteries were primary gastrodermal structures composed of the gastrodermis and mesoglea, and began to form from the anti-substrate side of the polyp (Fig. 1h). In the mesentery forming area, the epidermal layer was concave and gastrodermal cells were rearranged to form mesenteries (Fig. 1h, inset). About 12 h after Hym-248 treatment, secondary mesenteries appeared between each primary mesentery (Fig. 2a). As the gastrodermis was organized and the coelenteron extended, mesenteries grew inward from the body wall (Fig. 2b, c). The columnar epidermal cells tightly adhered laterally and formed a simple epidermal layer, while the granular inclusions were inconspicuous in the apical cytoplasm (Fig. 2c). The gastrodermal surface of the bottom of coelenteron was covered with cilia and there were small spheres present which were ∼2–4 μm in diameter (Fig. 2d).

Fig. 2
figure 2

Further stages of metamorphosis of Acropora spp. a Polyp initiating secondary mesentery formation (9 h after exposure to Hym-248). Arrowheads indicate mesenteries. b Histological section of metamorphosing polyp (9 h after exposure to Hym-248). Arrows indicate mesogleal layer and double arrowheads indicate the site approximately corresponding to d. c SEM micrograph of a fractured polyp with mesentery, d SEM micrograph of the gastrodermal surface of the bottom of coelenteron. Arrows indicate spherical objects, e polyp with six tentacles and mouth (20 h after exposure to Hym-248). Double arrowheads indicate tentacular grooves. f Histological section of metamorphosing polyp (24 h after exposure to Hym-248), g highly magnified histological section of blastopore of early primary polyp, h SEM micrograph of oral region of early primary polyp, i histological section of mature primary polyp (3 days after exposure to Hym-248). Double arrowheads indicate the site approximately corresponding to k. j SEM micrograph of marginal region of polyp, k SEM micrograph of epidermis of the substratum side (co coelenteron, ep epidermis, ga gastrodermis, me mesentery (=gastrodermal tissue), mo mouth, ms mesentery, sk skeleton). Bars = 100 μm in a, b, e, f and i, 10 μm in c, d, g, h and j, 1 μm in k

Full size image

Twelve to twenty-four hours after Hym-248 treatment, the polyps had six tentacular grooves around their mouth and skeletons developed (Fig. 2e, f). The oral pores of two species Acropora were formed as simple openings through invagination of the epidermal layer after the formation of the coelenteron in the polyps (Fig. 2g). Some small particles (2–4 μm in diameter) were discharged from the coelenteron through the oral pore (Fig. 2g, h).

Three days after Hym-248 treatment, the gastrodermal layer became thinner and the mesenteries developed (Fig. 2i). Many small particles (2–5-μm diameter) appeared in marginal parts of the coelenteron (Fig. 2j). Although the epidermal surfaces of polyps were covered with cilia and microvilli, the epidermis of the side towards the substrate had a smooth surface without any cilia and microvilli (Fig. 2k).

Symbiodinium acquisition by primary polyp

When non-symbiotic primary polyps (2–3 days after metamorphosis) were exposed to freshly isolated Symbiodinium (zooxanthellae) (FIZ) from T. crocea (105 cells ml−1), FIZ were taken into the coelenteron through the mouth by water currents produced by the cilia on the tentacles (Fig. 3a). Six hours after exposure to FIZ, infected polyps were transferred to FSW. One–two days after exposure to FIZ, many brown cell masses were present in the coelenteron (Fig. 3b). Histological examination showed these brown cell masses to comprise intact Symbiodinium and cell debris from both T. crocea and Symbiodinium (Fig. 3c, d). Some Symbiodinium were trapped by cytoplasmic processes from the gastrodermal cells (Fig. 3e). The cilia of the gastrodermal cells were elongated, and sometimes caught some of the Symbiodinium in the coelenteron (Fig. 3f).

Fig. 3
figure 3

Symbiodinium aqcuistion of primary polyp. a Polyp taking up freshly isolated zooxanthellae (FIZ) (10 min after exposure to FIZ). Arrows indicate Symbiodinium. b Polyp showing a clumped cell mass (arrows) in the coelenteron (1 day after exposure to FIZ), c histological section of primary polyp containing cell mass (arrow) in its coelenteron, d enlargement of c, the cell mass was composed mainly of degraded Symbiodinium and other materials, e high magnification of histological section of coelenteron, f SEM micrograph of fractured polyp. Arrow indicates elongated cilia of gastrodermal cells. g Polyp with several Symbiodinium in basal tissue of coelenteron (3 days after exposure to FIZ), h Histological section of primary polyp with Symbiodinium in gastrodermal cells. Arrowheads indicate Symbiodinium, i high magnification of histological section of oral regional body wall, j high magnification of histological section of aboral region, k polyp with many Symbiodinium in tentacles (5 days after exposure to FIZ), l polyp with Symbiodinium in body wall and initiating skeletal formation (9 days after exposure to FIZ). m Histological section of polyp. Arrowheads indicate Symbiodinium. n SEM micrograph of fractured mesentery with Symbiodinium (co coelenteron, ep epidermis, ga gastrodermis, me mesentery (=gastrodermal tissue), mo mouth, ms mesogleal layer, sk skeleton, te tentacle, zo zooxanthella(e) (=Symbiodinium). Bars = 100 μm in a, b, c, g, h, k, l and m, 50 μm in f, 10 μm in d, e, i, j and n

Full size image

Two–three days after exposure to FIZ, the cell masses in the coelenteron disappeared, and Symbiodinium appeared in the mesenteries and at the bottom of the gastrodermal layer (Fig. 3g–j). The gastrodermis of the aboral region of polyps was thickened and one gastrodermal cell contained Symbiodinium (Fig. 3j).

Four–five days after exposure to FIZ, many Symbiodinium were visible in the tentacles (Fig. 3k). Four days after exposure to FIZ, the percentage of Symbiodinium undergoing cell division (MI) in the both Acropora species polyps were 1.4 ± 0.3% (mean ± SD, n = 6). Seven–ten days after exposure to FIZ, Symbiodinium were distributed in gastrodermal cells throughout the polyps (Fig. 3l–n). Eight days after exposure to FIZ, the percentage of MI in the both Acropora species polyps was 1.3 ± 0.3% (n = 9). The number of symbionts were counted at a longitudinal section (including mouth and mesenteries) of the polyps. Four days after exposure to FIZ, 36.5 ± 11.9 algal cells were seen in polyp sections (n = 4); 21.8, 68.3, and 9.9% of the symbionts were distributed in the gastrodermis of the aboral region of polyps, the mesenteries, and the gastrodermis of the oral region (=body wall) of the polyps (n = 4), respectively. Eight days after exposure to FIZ, 82.3 ± 18.8 algal cells were seen in polyp sections (n = 3); 16.2, 47.6, 0.3, and 35.8% of symbionts were distributed in the gastrodermis of the aboral region of polyps, the mesenteries, the gastrodermis of tentacles, and the gastrodermis of the oral region (=body wall) of the polyp (n = 3), respectively.

Discussion

Metamorphosis of Acropora spp. from planula to primary polyp

Observing larval settlement in nature is inherently difficult, as a result, direct information is very limited. In the laboratory, Babcock and Heyward (1986) observed Acropora millepora in culture jars starting to metamorphose 4.5 days after fertilization. Hayashibara et al. (1997) reported that first settlement on plastic walls began 6 days after fertilization in A. nasuta, 7 days in A. hyacinthus, and 21 days in A. florida. In the present study, non-Hym-248 treated planulae from A. nobilis and A. microphthalma in the culture container began to metamorphose on plastic walls or the water surface 5 days after fertilization. However, metamorphosis in both species of the planulae could be induced by Hym-248 treatment after 48 h, continuing until at least 16 days after fertilization. This observation indicates that planulae of Acropora spp. acquire the ability to settle and metamorphose just after beginning to swim.

The coelenteron is formed during the planula stage in Fungia scutaria (Krupp 1983; Schwarz et al. 1999), Montipora verrucosa (Máte et al. 1998), Pocillopora verrucosa, Pocillopora eydouxi (Hirose et al. 2000) and Porites cylindrica (Hirose and Hidaka 2006). In contrast, Hayashibara et al. (1997) reported that the coelenteron or gut did not develop in the planulae of four species of Acropora (A. hyacinthus, A. nasuta, A. florida, and A. secale) until 30 days after insemination, and they found a coelenteron only in metamorphosing planulae. In the present study, the non-Hym-248 treated swimming planulae did not have a coelenteron until 16 days after fertilization. However, following the induction of metamorphosis by Hym-248, the polyps formed coelenterons and mesenteries. This suggests that settlement may be a cue for the formation of coelenteron and mesenteries in Acropora spp.

In planulae from Favia fragum (Szmant-Froelich et al. 1985), Porites porites (Tomascik and Sander 1987), and Montipora digitata (Hirose and Hidaka 2006), the formation of the oral pore occurs by invagination of part of the epidermal layer before the coelenteron (gastric cavity) starts to appear. In the present study, the oral pore of two species Acropora was formed by invagination after the formation of the coelenteron in polyps. This suggests that the oral pore or mouth does not need to be psesent in order for the coelenteron to form in Acropora spp.

In polyps, the cilia were found in the gastrodermis (Fig. 2d). This is similar to the findings of Benayahu et al. (1988) in planulae and polyps of the octocorallia, Xenia unbellata. These ciliated gastrodermal cells probably produce water currents to move material into the coelenteron. The mesenteries consisted of the elongated endodermal layer together with the mesoglea. Although little is known about the syntheses of mesoglea components in cnidarians, the result of this study indicates that the mesoglea in Acropora may be synthesized by gastrodermal cells forming mesenteries during metamorphosis.

Acquisition of Symbiodinium by Acropora spp.

Non-Hym-248 treated swimming planulae of the two species of Acropora did not establish symbioses with Symbiodinium until at least 16 days after fertilization even if the swimming planulae were exposed to freshly isolated Symbiodinium (zooxanthellae) (FIZ). Hayashibara et al. (1997) also reported in four species of Acropora that no Symbiodinium were acquired during the planula period until 30 days after formation. However, van Oppen (2001) found that cultured A. millepora planulae established stable symbioses and that these symbiotic larvae survived for more than 4 months, although she did not show any histological data or describe the culture conditions. Whether there is in fact variation between acroporids in the timing of initial infection, or whether reported differences in timing of infection reflect differences in experimentation or culturing, remains to be determined.

Intact Symbiodinium and cell debris from both T. crocea and Symbiodinium were found in the coelenteron during this study. However, most Symbiodinium in the gastrodermal cells of the polyps were intact. Lesser (1997) reported that damaged Symbiodinium were discharged selectively by adult corals. Young polyps may also have the ability to differentiate between intact and damaged Symbiodinium. The percentage of Symbiodinium MI in both Acropora species polyps was 1.3–1.4%, which is consistent with previous reports of MI indexes of 0.5–7.0% in Acropora spp. (e.g., Wilkerson et al. 1988; Jones and Yellowlees 1997; Cervino et al. 2003), and 0.8–11.5% in Tridacna spp. (Ambariyanto and Hoegh-Guldberg 1997; Maruyama and Heslinga 1997). Most algal symbioses are characterized by relatively constant densities of symbionts (e.g., Muscatine and Pool 1979; Jones and Yellowlees 1997; Smith and Muscatine 1999; Fitt and Cook 2001; Titlyanov et al. 2004), giving rise to the hypothesis that symbiont numbers are regulated by their hosts. Young polyps may also regulate the density of Symbiodinium. Recently, Yuyama et al. (2005) reported host genes involved in the establishment or maintenance of Symbiodinium symbiosis in Acropora tenuis. These molecular tools might assist in future analyses of ontogenic establishment of symbioses between coral and Symbiodinium.