In the early juveniles of Ciona intestinalis, primordial germ cells arise on the degenerated mass of the resorbed tadpole tail, and assemble to form a discrete gonad rudiment. The present study elucidated the morphological sequences during differentiation of the gonad rudiment into the testis and ovary. In 11- to 12-day juveniles, the gonad rudiment, an elongate sac, divided into the testicular and ovarian rudiments. The testicular rudiment separated as a round vesicle from the thickened wall of the elongate sac. The original sac, after separation of the round vesicle, developed into the ovary. In the testicular rudiment, germ cells formed a continuous central mass without association of somatic cells, while in the ovarian rudiment, each germ cell was associated with somatic cells within the epithelium composing the wall of the rudiment. In 13- to 15-day juveniles the testicular rudiment changed into branched tubes ending in club-shaped follicles. Cells characterized by many flattened cisternae of rough endoplasmic reticulum (distal cells) constituted the distal wall of each follicle. Spermatogenic cells were freely present in the follicular lumen, but the largest spermatogonia were in contact with the distal cells. Both in the testicular and ovarian rudiments, germ cells entered meiosis in 18-day juveniles. A novel body (periesophageal body) was found just beneath the ventral margin of the esophageal opening. It comprised irregular follicles made up of one cell type whose cytoplasm, filled with round vesicles and Golgi complexes, was suggestive of an endocrine function. Fragments derived from the periesophageal body were present around the developing ovary.
Ascidians are typically hermaphroditic animals. In early light microscopic studies on ascidian gonadogenesis, it has been reported that the gonad arises as a single rudiment consisting of a small body of undifferentiated cells in the juvenile after metamorphosis. This body divides into the ovary and testis at a later stage ( De Sélys-Longchamps & Damas 1900; Simkins 1924; H˘uus 1937; also see reviews by Berrill (1975) and Ishikawa et al. (1988) ). However, the processes have been described only by light microscopy and the morphological details during differentiation of the gonad rudiment into the ovary and testis are not fully revealed.
In the just-metamorphosed juveniles of Ciona intestinalis, a mass of tissue debris derived from the resorbed tadpole tail is situated in the space around the ventral side of the esophagus. We have shown in a preceding paper ( Yamamoto & Okada 1999) that round cells characterized by a large nucleus and nuage occur singly on the surface of the degenerated tissue mass. We presumed these cells to be primordial germ cells from morphological continuity to obvious gonads in later stages. In a few days the primordial germ cells assemble to form a solid slender body (gonad rudiment) together with flattened somatic cells. The gonad rudiment gradually changes the position from the space around the esophagus toward the stomach, and becomes an oval vesicle with an eccentric cavity in juveniles 7–8 days after settlement. In the preceding paper ( Yamamoto & Okada 1999), however, we have not explicitly presented sufficient evidence for the presumption that the oval vesicle is the gonad rudiment. In the present paper we show that the presumed gonad rudiment actually differentiates into the ovary and testis on the basis of ultrastructural tracing of the differentiating gonad rudiments in juveniles from 8 to 21 days after settlement.
Materials and Methods
Adult C. intestinalis were collected from the subtidal zone near the Ushimado Marine Laboratory. Eggs were removed from the oviduct and fertilized in plastic dishes (9 cm in diameter) with a suspension of non-self sperm. Fertilized eggs kept at 20°C hatched about 15 h after fertilization. Juveniles that settled and metamorphosed on the plastic dishes within 48 h after fertilization were used. Twenty to 30 juveniles were kept in a vessel containing gently aerated seawater (~3 L) at 20°C, and sufficiently fed with the diatom Chaetoceros gracilis. The seawater was changed once a day. The early stages of juvenile C. intestinalis were monitored by the formation of the pharyngeal basket according to Berrill (1947). The morphological data on the progression of juvenile development have been presented in the previous paper ( Yamamoto & Okada 1999). On observing living materials by light microscopy, a juvenile was placed in the shallow well of a depression slide with the ventral side (the side of the endostyle) towards the right, and pressed lightly with a coverslip. In the current paper the ages of the juveniles are designated by the time after settlement, for example, 10-day juveniles are juveniles between 9 and 10 days after settlement.
Posterior halves of juveniles were fixed for 2 h at room temperature with 2.5% glutaraldehyde dissolved in a sucrose-containing buffer solution (0.4 M sucrose and 0.1 M sodium cacodylate, pH 7.4). After a short rinse with the same buffer solution, the specimens were postfixed for 1 h with 1% OsO4 dissolved in a mixture of 1% potassium-ferrocyanide, 0.4 M NaCl and 0.1 M sodium cacodylate (pH 7.4). The specimens were dehydrated through a graded series of ethanol and embedded in epoxy resin (TAAB embedding resin). Thin sections cut on a Porter-Blum MT-2 ultramicrotome were stained with 1% alcoholic uranyl acetate and 0.1% lead citrate, and examined with a Hitachi H-500H electron microscope. Semi-thin sections of about 1 μm were cut from the epoxy-embedded blocks, stained with 1% toluidine blue and examined by light microscopy.
Light microscopic observation
We have reported in the previous paper ( Yamamoto & Okada 1999) that the gonad rudiment is first discernible as a solid slender body in the 2-day juveniles, and becomes an oval vesicle in the 6- to 7-day juveniles. In sections of the 8-day juveniles, the gonad rudiment was found in the narrow space between the epidermis and peritoneal membrane ( Fig. 1a). The structure designated as the peritoneal membrane in the present paper corresponds to the pericardium in Berrill (1947). The cross-section of the gonad rudiment showed an eccentric cavity enclosed by a thin roof and a thick bottom. In living materials, the gonad rudiment was always visible near the opening of the duct of the pyloric gland on the right surface of the stomach ( Fig. 1e,f). The gonad rudiment gradually increased in size, and in the 10-day juveniles it was an elongate sac with a thickened wall toward the ventral end ( Fig. 1e). In 11-day juveniles, a round vesicle appeared in association with the elongate sac. In the living materials, a round vesicle was seen superimposed on the elongate sac ( Fig. 1f). In sectioned materials, the round vesicle was seen arising on the thickened wall of the elongate sac ( Fig. 1c). As described later, ultrastructural tracing revealed that this vesicular structure was the rudiment of the testis, and that the elongate sac left after separation of the round follicle developed into the ovary. In 12-day juveniles, the vesicular rudiment of the testis separated from the original sac, entering into the perivisceral space between the peritoneal membrane and the gastric epithelium ( Fig. 1d). The ovarian rudiment stayed for a while in the space between the epidermis and the peritoneal membrane. The thickened wall of the ovarian rudiment gradually showed a lobulated appearance separated by branches of the lumen ( Fig. 1d,g), and the dorsal end of the rudiment formed a thin-walled tubular prolongation toward the atrial siphon (future oviduct; Fig. 1g). In 13- to 14-day juveniles, the testicular rudiment changed in shape from a large round vesicle into many branched tubes each ending in a small follicle. In the living 14-day juveniles, alveolar follicles were seen joined with each other by branched thin tubes around the ovarian rudiment ( Fig. 1g). In 15- to 16-day juveniles, the follicles became club-shaped (cf. Fig. 4a). The branched tubes converged to form a single duct (sperm duct) that ran parallel to the oviduct. Thus a system of the adult testis consisting of branched tubes with club-shaped follicular endings ( Burighel & Cloney 1997) was accomplished by 18 days after settlement. The ovarian rudiment gradually changed position, finally reaching the space within the ventral loop of the intestine by 18 days after settlement.
Ultrastructure of the gonad rudiment separating into testis and ovary
The ultrastructural feature of the gonad rudiment in the 10-day juveniles was basically identical to that already described in 8-day juveniles ( Yamamoto & Okada 1999). In the thickened wall of the saccular gonad rudiment, germ cells were mingled with somatic cells. Germ cells were characterized by their large size (7–9 μm in diameter), a round nucleus with prominent nucleolus, nuage on the outer surface of the nucleus, and homogeneous cytoplasm containing densely scattered ribosomes. The somatic cells were smaller in size than germ cells, usually flattened and somewhat irregular in shape. The thin roof of the sac consisted of a single layer of flattened or cuboidal somatic cells. In 11-day juveniles, a group of germ cells began to form a continuous mass within the thickened wall of the gonad rudiment; no somatic cells were associated with the germ cells in the mass ( Fig. 2a). A small cavity arose in association with the continuous mass of germ cells. The round vesicle separating from the original gonad rudiment in 12-day juveniles consisted of this continuous mass of germ cells associated with a small cavity and somatic cells surrounding the cavity, and the mass of germ cells ( Fig. 2b). Differences in cellular arrangement were apparent between the separating round vesicle (testicular rudiment) and the large sac left after separation of the round vesicle (ovarian rudiment). In the testicular rudiment, germ cells did not show any epithelial arrangement but formed a cluster in the center of the vesicle; somatic cells were not present among the germ cells but constituted a single layer of epithelium encircling the vesicle ( Fig. 2b). In the ovarian rudiment, the germ cells remained within the thickened wall of the sac; somatic cells were associated with each germ cell. The thickened wall of the ovarian rudiment was separated into complicated folds with branches of the lumen of the sac, and the germ cells together with somatic cells became arranged in an epithelial structure ( Fig. 2b). Germ cells themselves were not different in appearance between the ovarian and testicular rudiments at this stage.
Ultrastructure of the developing testis
We recognized the presence of characteristic cells whose cytoplasm was full of cisternae of endoplasmic reticulum (ER) at an end of the testicular vesicle just separated from the ovarian rudiment ( Fig. 2b). In 13- to 14-day juveniles, where the testicular rudiment was forming branched tubes ending in swollen follicles, many ER-rich cells appeared in a cluster at the distal blind end of each follicle ( Fig. 3a). We will refer to the ER-rich cells as the distal cells. The distal cells were polyhedral in outline, showing many flattened cisternae of rough-surfaced ER in the cytoplasm, and a somewhat irregular nucleus with a prominent nucleolus ( Fig. 3a–c). Free cells similar in appearance to the distal cells were often encountered in the space surrounding the testicular rudiment, especially in the immediate vicinity of the distal end of the testicular follicles ( Fig. 3a). After the follicles became club-shaped, a single layer of the distal cells always constituted the wall of the distal end of each follicle ( Figs 3c,4a,c). The wall of the follicle, except for the distal end, consisted of a simple epithelium of flattened cells without well-developed ER. Germ cells (spermatogonia) in the developing testis in 12- to 16-day juveniles were all similar in appearance to the oogonia at the earliest stage ( Okada & Yamamoto 1993). Spermatogonia were characterized by their large size (7–9 μm in diameter), round shape, electron-lucent cytoplasm and large nucleus with one or two prominent nucleoli ( Figs 3c,4a). Clouds of an electron-dense material (nuage) were present on the outer surface of the nucleus of the spermatogonia. A single cilium was sometimes observed arising from the spermatogonia. In the club-shaped follicles of 15- to 17-day juveniles ( Fig. 4a), many spermatogonia were floating singly or in loose aggregations in the lumen, preferentially in the distal part. However, some spermatogonia that were a little larger in size than the floating ones were in contact with the distal cells. Spermatogonia undergoing mitosis were frequently encountered. Some spermatogonia were connected to each other by a narrow cytoplasmic bridge. Germ cells undergoing meiosis (spermatocytes) were first found in the 18-day juveniles, and those after meiotic division (spermatids) appeared first in 21-day juveniles ( Fig. 4c). In the lumen of each testicular follicle in 21-day juveniles, spermatogenic cells were arranged following the order of the spermatogenetic stages from the distal blind end toward the open exit to the connecting tube ( Fig. 4c). The largest spermatogonia were always present in close contact with the distal cells at the distal end of the follicle.
Ultrastructure of the developing ovary
In sections of the ovarian sac from older than 13-day juveniles, thick stratified epithelia appeared to alternate with thin simple epithelia ( Fig. 4b). The germ cells (oogonia) were always present within the stratified part. Oogonia were surrounded by flattened somatic cells, never exposing their free cell surface from the epithelium. Ultrastructural tracing revealed that the stratified epithelium was the direct precursor of the germinal epithelium described by Kessel (1983) and Sugino et al. (1987 , 1990). The simple epithelia were made up of flattened or cuboidal cells measuring about 3 μm in thickness. We have already described the morphlogical sequences of early oogenesis in Ciona and have defined three stages in oogonial development ( Okada & Yamamoto 1993). Here we describe briefly the progression of oogenetic stages in developing ovaries. Germ cells found in the ovary of 12- to 14-day juveniles were all at the first oogonial stage (type A oogonia); they were large in size (7–9 μm in diameter) with clear cytoplasm ( Fig. 4b). In the ovary of 15-day juveniles some oogonia were in the proliferation phase (type B and C oogonia); were smaller in size (4–6 μm in diameter) and showed darker cytoplasm than the type A oogonia. Germ cells in the meiotic prophase (oocytes) were first found in the ovary of 18-day juveniles ( Fig. 4d). Oocytes entering the growth phase (early diplotene) were first found in the ovary of 21-day juveniles.
A novel body
In living 1-day juveniles, we found the presence of a discrete transparent body attaching to the degenerated mass of the resorbed tadpole tail. It was a small oval-shaped body (~15 × 30 μm in diameter), situated just beneath the ventral margin of the funnel-shaped opening of the esophagus. The body was at first solid but soon (in 2-day juveniles) showed follicular appearance with a small central cavity ( Figs 1h,5a). It gradually increased in size ( Fig. 1i,j) and became irregular in outline ( Fig. 1j,k), staying near the original position for at least 20 days after settlement. We will refer to this body as a periesophageal body. In living juveniles, the periesophageal body was easily found at the fixed position near the periphery of the esophageal funnel from 3 to 10 days after settlement, but became gradually obscured by decreasing transparency in the body wall. Electron microscopically, the periesophageal body was made up of cells of one type ( Fig. 5b); they were wedge-shaped cells measuring 6–8 μm in height with a broad base and a narrow distal end facing the central cavity. The cytoplasm was full of various-sized (0.1–0.5 μm in diameter) round vesicles of rough-surfaced ER that contained an electron-lucent fluffy material. Many Golgi complexes were also scattered in the cytoplasm. The nucleus was large (~3 μm in diameter) with a large prominent nucleolus. In 8- to 10-day juveniles, the periesophageal body became a large complex of irregularly shaped follicular structures with many irregular cavities ( Figs 1k,5c). In 11- to 13-day juveniles, where the gonad rudiment was undergoing differentiation into testis and ovary, solid or follicular masses of cells were isolated from the periesophageal body. They migrated into the narrow space between the epidermis and the peritoneal membrane, and further into the perivisceral space surrounding the stomach ( Fig. 5d). Isolated fragments of the periesophageal body were also visible in the vicinity of the gonad rudiment dividing into testicular and ovarian rudiments ( Fig. 2a,b), and around the developing ovary after separation. We could not confirm the presence of the periesophageal body in fully grown adults.
In the previous paper ( Yamamoto & Okada 1999), we described the presence of presumed primordial germ cells in the just-metamorphosed juveniles of C. intestinalis, and the process of formation of a presumed gonad rudiment through assemblage of the presumed germ cells. From the present results of ultrastructural tracing of the presumed gonad rudiment to the ovary and testis, it is evident that the presumed cells and structures described in the previous paper are truly primordial germ cells and gonad rudiments, respectively. Thus the story of gonadal development in C. intestinalis can be summarized as follows. The earliest germ cells occur singly on the surface of the degenerated tissue mass of the resorbed tadpole tail in just-metamorphosed juveniles. The cell lineage to primordial germ cells is unknown at present. Germ cells and somatic cells assemble to form an oval vesicle (gonad rudiment) in the space between the epidermis and the peritoneal membrane. The gonad rudiment increases in size on the right side of the stomach, and divides into a round vesicle (testicular rudiment) and an elongate sac (ovarian rudiment) about 11 to 12 days after settlement (20°C). The testicular rudiment migrates into the perivisceral space surrounding the stomach and forms a system consisting of branched tubes with follicular endings. Germ cells enter meiosis at the same time in the developing testis and ovary (~18 days after settlement).
We confirmed results of early light microscopic studies on ascidian gonadogenesis that the gonad arises as a sexually undifferentiated mass, which later differentiates into the testis and ovary ( De Sélys-Longchamps & Damas 1900; Simkins 1924; H˘uus 1937). It is, however, reported in the blastozooid of a compound ascidian, Distomus, that the rudiments of the testis and ovary occur separately ( Newberry 1968). In the adult ovary, the oogonia and oocytes are present within a stratified epithelium (germinal epithelium) in association with somatic cells ( Kessel 1983; Sugino et al. 1987 , 1990; Okada & Yamamoto 1993). In the adult testis, the spermatogenic cells are scattered singly or in clusters in the lumen of the follicles without forming epithelial structure ( Cavey 1994). Such structural differences between the ovary and the testis were discernible in the earliest rudiments just after separation. In the ovarian rudiment, the germ cells were always associated with somatic cells within the wall of the sac, and were gradually arranged into a stratified epithelium corresponding to the germinal epithelium by Kessel (1983) and Sugino et al. (1987 , 1990). In the just separated testicular rudiment, the germ cells showed no epithelial arrangement but formed a continuous mass in the center of the vesicular rudiment. It remains to be solved what brings about the structural differences between the ovary and the testis at the early stage of differentiation. Cavey (1994) has reported that in Boltenia villosa the accessory (somatic) cells project into the testicular lumen and contact the spermatogenic cells. In the present material, no components of somatic cells could be found among spermatogenic cells in the testicular lumen at any stage of testicular development.
We found characteristic cells showing well- developed ER (distal cells) at the distal blind end of each testicular follicle. The distal cells appeared after the testicular rudiment was separated from the ovarian rudiment. The origin of the distal cells is unclear but the fact that free cells similar to the distal cells were frequently encountered in the vicinity of the distal end of each follicle suggests that the distal cells derive from a kind of blood cells. Two possible functions are conceivable in the distal cells. When the testicular rudiment began to change in shape from a round vesicle into branched tubes with follicular endings, distal cells appeared in a cluster at the distal blind end of each follicle. This suggests a possibility that the distal cells organize the morphogenesis of the tubular system of the testis through an inductive function. The other possibility is that the distal cells play a role in coordinating the differentiation of spermatogenic cells. In the lumen of the club-shaped follicles spermatogonia occupied the distal part, and the largest spermatogonia, possibly spermatogenic stem cells, were in contact with the distal cells. Mitotic spermatogonia were often encountered near the distal end, but meiotic cells (spermatocytes) were present some distance apart from the distal end. In the tubular gonad of Caenorhabditis elegans it has been reported ( Kimble & Hirsch 1979) that a ‘distal tip cell’ is present, and that the spermatogenic cells show a similar spatial arrangement to that seen in the present material. It has been conjectured from destruction experiments of the ‘distal tip cell’ by laser beams that the cell secretes a substance that maintains the germ cells in mitosis and inhibits them from entering meiosis ( Kimble & White 1981).
We found a discrete body (periesophageal body) comprising a follicle-like aggregation of cells of only one type in a fixed position in the juveniles (the ventral margin of the esophageal opening). The cells composing the body were characterized by a large nucleus with a prominent nucleolus, abundant round vesicles of ER with a fluffy content and many well-developed Golgi complexes. Such ultrastructural features suggest that the cells undergo active synthesis of protein for secretion. To our knowledge there has been no report on an organ corresponding to this body. When the gonad rudiment began to differentiate into the ovary and testis, fragments derived from the periesophageal body migrated to the vicinity of the differentiating gonads and remained around the developing ovary. The cells composing the fragments of the periesophageal body were distinct in ultrastructure from the distal cells. The structure and behavior of the periesophageal body suggest the possibility that it controls gonadal development through an endocrine function.