On The Hormonal and Neural Control of the Release of Gametes in Ascidians

The homology and function of the neural gland of ascidians have been subjects of controversy for nearly a century since the first description of the organ by Hancock in 1868. The most recent critical discussion of these questions in any language is that by Huus (1937), and the most recent in English is that by Butcher (1930). The various suggestions made can be summarized as follows :

Function:

Mucus gland (Fol, 1876; Roule, 1884).

Excretory organ (Julin, 1881 b).

Lymphatic organ (Cuénot, 1891).

Endocrine organ (Butcher, 1930).

Organ of phagocytic excretion (Pérès, 1943).

Organ of special sense (Hancock, 1868; Huus, 1937).

Homology:

Whole pituitary (Julin, 1881 a).

Posterior lobe only of pituitary (most recent authors, e.g. Huus, 1937).

Not pituitary (Pérès, 1943).

The older views have mostly been abandoned or disproved and the two main questions of interest now are :

1. Is the neural gland of ascidians the homologue of part or of the whole of the pituitary of vertebrates ?

2. If it is an endocrine organ, does it combine some other function with this?

The investigations reported here elucidate these questions to some extent.

These investigations, on ascidians, are part of a wider study of the mechanisms whereby chordate pituitary hormones are elaborated, and the route by which they elicit their responses. All the experiments reported here are concerned with one such response—the release of gametes—regarded merely as a response, since I am chiefly interested in the route of elicitation.

The animals used were Ciona intestinalis L. and Phallusia mammilata (Cuv.), which are both simple ascidians.

The chorionic gonadotrophin used in certain of the experiments was extracted by Mr W. M. S. Russell and myself from human pregnancy urine. It possessed an activity of c. 110 i.u./mg.

The experiments fall into two series—’injection experiments’ and ‘feeding experiments’—each with their appropriate controls.

1. I call ‘injection experiments’ those in which the stimulus to gamete release was provided by means of a single injection of 0·5 or 1·0 mg. (55 or no i.u.) of chorionic gonadotrophin dissolved in 1·0 ml. aq. dest.

2. I call ‘feeding experiments’ those in which the stimulus to gamete release was provided by feeding to the animal eggs and sperm of its own species. This follows the postulate of Huus (1937) that such a stimulus might be effective (cf. also the demonstration that this is so in sea-urchins by Lindahl & Runnstrom, 1929).

In all experiments the animals were isolated in individual jars, and the normal controls were not observed to release gametes spontaneously in the aquarium. In ‘feeding experiments ‘the water was changed 2 hr. after feeding so as to remove any remaining eggs and sperm which might have provided a false positive response. The eggs and sperm for feeding were mixed from several individuals, and the mixture was kept for half an hour before pipetting it into the open inhalant siphons of the animals. It contained active sperm, and fertilized and unfertilized eggs at the time when it was added to the jars containing the animals. At present it is unknown which of the three constituents is the active agent. Dosage was large and indefinite.

In some animals in both series of experiments the hearts were removed or nerves sectioned. These various surgical manipulations were performed without anaesthesia or aseptic precautions. In a very few early experiments with Ciona sulphanilamide was sprinkled on the water to combat infection; this proved to be unnecessary and was discontinued. The operations were performed by paring away the tunic of Phallusia in the appropriate region until the necessary detail was visible; in Ciona this was not done since the tunic is thin. Nerves were cut with fine scissors. Blood was removed from the heart with a hypodermic syringe until the flow stopped and then the main mass of the heart and pericardium dissected out.

In other animals the neural gland was destroyed with relatively little damage to the ganglion. This operation was only performed in Phallusia. The test over the neural region was pared away and a slit cut through the mantle and pharyngeal wall at one side of the ganglion. The neural gland was then scraped with the tip of a fine scalpel inserted through this slit. After a little practice this operation was quite simple, and the degree of scraping needed to destroy the gland with little damage to the ganglion could be judged. The continued functioning of the ganglion was checked 24 hr. after the operation and at the close of the experiment, by means of the so-called ‘crossed reflex’ whereby delicate stimulation of one siphon causes closure of the other. According to Prof. Z. M. Bacq (personal communication) this reflex requires an intact and functioning ganglion. In addition, histological checks were performed after conclusion of the experiments.

The positive response in all the experiments consisted in the release of eggs and/or sperm into the surrounding water and cessation of feeding. In all neural glands which were examined histologically this was accompanied by the discharge of the secretion; in those Ciona where the ovaries were examined histologically corpora lútea were observed after ovulation and never before (Carlisle, 1951). The positive response appeared within about 20 hr. at 16°C.

Eleven normal Ciona and eleven Phallusia all gave positive responses to an injection of chorionic gonadotrophin. The same number of each gave negative responses to an injection of distilled water.

Of nine Phallusia which received between them the extract prepared from the neural complexes of 1000 Ciona (by the method of Carlisle, 1950 c), six gave positive responses and three negative.

When the nerves running back from the ganglion to the region of the gonads are cut near the ganglion this positive response to gonadotrophin is prevented. Three Ciona and five Phallusia gave a negative response to gonadotrophin under these conditions. This is not merely due to trauma. The operation does not interfere with feeding or defaecation (cessation of feeding is an invariable part of the positive response); this therefore cannot be a false negative response. Equivalent damage to the neural region and section of all the nerves which run in other directions than towards the gonad does not prevent the positive response; three Ciona and six Phallusia all gave positive responses when such control operations were performed prior to injection with gonadotrophin. In one Phallusia which gave a positive response upon injection of chorionic gonadotrophin after section of the nerves to the region of the gonads, examination showed that the denervation had been only unilateral.

When the heart of Phallusia is destroyed and the blood removed the animal continues to live and eventually regenerates these tissues. When an animal which has undergone this operation so recently that regeneration has not begun is injected with chorionic gonadotrophin direct into the region of the gonad, there is no response. Nine specimens were so treated, the injection being made 24 hr. after the operation; of these two died and seven gave a negative response and continued feeding and defaecating. Post-mortem examination showed that they were sexually ripe.

When chorionic gonadotrophin is injected into a similar animal direct into the neural region, the animal gives a positive response. Of eight specimens so injected 24 hr. after the operation six gave positive responses and two negative. When in an animal without a heart the nerves to the region of the gonads are cut, the response to an injection of chorionic gonadotrophin into the neural region is negative. Of eight animals, in which I cut these nerves near the ganglion 24 hr. after destruction of the heart and made the injection into the neural region 2 hr. later, seven gave a negative response and one died. When the other nerves were cut under like conditions, but the nerves to the region of the gonads left intact, the response was positive. Eight animals had the heart destroyed, then after 24 hr. I cut nerves (except those to the region of the gonads) near the ganglion, and after another 2 hr. I injected the animals with chorionic gonadotrophin into the neural region. All gave positive responses.

Eight Phallusia had the neural gland destroyed and the ganglion left intact; after 24 hr. each received an injection of chorionic gonadotrophin. All gave positive responses.

Five Ciona fed with Ciona eggs and sperm all gave a positive response. Five Ciona fed with Phallusia eggs and sperm gave a negative response. Eight Phallusia fed with eggs and sperm of Phallusia all gave a positive response. Two Phallusia fed with eggs and sperm of Ciona gave a negative response.

Ten Ciona fed with carmipe or indian ink particles gave a negative response. A number of these particles was always to be found subsequently in the neural gland.

As in injection experiments it was found that the positive response was prevented by section of the nerves to the region of the gonad (eight Phallusia) but not by section of the other nerves (eight Phallusia) nor by destruction of the heart and removal of the blood (eight Phallusia). Many Ciona were killed 2−3 hr. after feeding with eggs and the neural gland examined. Eggs and sperm were always found in the neural gland.

Eight Phallusia had the neural gland destroyed and the ganglion left intact, and then after 24 hr. were fed with eggs and sperm of Phallusia. All gave a negative response.

In the introduction two questions were posed, but before discussing these it might be well to consider the nature of the’ ‘positive response’. Ascidians are highly efficient filter-feeders and filter a large volume of water. (It is estimated that a medium-sized specimen of Phallusia nigra filters 173 l. in 24 hr.) One of their difficulties perhaps is to avoid eating their own faeces and gametes. This may be the reason (speaking teleologically) for the heavy mucoid faeces which will not divide into flocculent particles. The more important purpose of avoiding eating their own gametes, which are, of course, floating, is attained by the method of cessation of feeding for long enough for the gametes to be dispersed. The release of gametes in vertebrates is always immediately followed by the discharge of secretion by the anterior pituitary (vide discussion by Burrows, 1945). So also in Ciona and Phallusia the neural gland discharges its secretion immediately after the release of gametes. Corpora lútea have been observed in the ovaries of Ciona after ovulation (Carlisle, 1951).

Is the neural gland homologous with the pituitary? Pérès (1943) belives it is not. He maintains that the various activities interpreted as hormonal by other authors are all due to histamine. This view is inadmissible. The activities concerned are an oxytocic activity (Butcher, 1930; confirmed by Bacq & Florkin, 1935), a vasopressor activity (Bacq & Florkin, 1935) and a chromatophorotrophic activity (Bacq & Florkin, 1935; Abramowitz, 1937). No further evidence has been produced with regard to the first two of these, but even at the time when Pérès wrote, it had been established that histamine has no chromatophorotrophic effect (vide Hogben, 1924). Carlisle (1950b) tested this once more and found that dilute extracts of the eye-stalk of Leander, extracts of the neural gland of Ciona and of mammalian pituitary posterior lobe (commercial preparation) all expanded the xanthophores of Leander, whereas histamine did not. At. the same time he found that the chromatophorotrophic principle is confined to the neural gland and is not found even in the adjacent organs and certainly not in the rest of the body, whereas histamine (according to Pérès) is distributed equally over the whole body. The chromatophorotrophic activity at least cannot be due to histamine. Nor can the gonadotrophic activity which was demonstrated to be present in extracts of the neural gland by Hogg (1937) and confirmed by Carlisle (1950c); Carlisle found that the activity was not present in the rest of the body.

The endocrinological evidence, then, suggests that the neural gland is the homologue of the pituitary of vertebrates and confirms the anatomical evidence. Until the discovery by Hogg (1937) of the gonadotrophic activity of the neural complex it was believed that this homology was restricted to the posterior lobe only of the pituitary and that the anterior lobe was not represented in the ascidians (vide Huus, 1937). Hogg’s discovery, however, suggested that this was not the case. His view that the neural gland represents the whole pituitary can be supported by several slight pieces of evidence :

1. The method of secretion of the neural gland is comparable only with that of the anterior lobe of selachians and is nowhere found in the posterior lobe (vide Butcher, 1930).

2. The discharge of secretion by the neural gland after ovulation is strictly comparable with that by the anterior pituitary of vertebrates under like conditions.

3. A mammalian gonadotrophin will stimulate ascidians to ovulate and discharge sperm.

4. An extract of the neural gland will stimulate an ascidian to discharge gametes.

5. The chromatophorotrophic principle of the vertebrate pituitary is believed to derive from the intermediate lobe and therefore to be of hypophysial origin, not infundibular (vide Landgrebe & Waring, 1950). This is likely also to apply to the chromatophorotrophic principle of ascidians.

6. Van Beneden & Julin (1884) showed that a stomodaeal element entered into the formation of the neural gland which thus had a dual origin, stomodaeal as well as neural. This was confirmed by Metcalf (1895).

The weight of evidence would seem to suggest that the neural gland (including ciliated pit) of Ascidians represents the entire pituitary of Vertebrates, adenohypophysis as well as neurohypophysis.

Huus (1937) postulated that the neural gland of ascidians acted as a sense organ to detect the presence of gametes of the animal’s own species in the imbibed water. By this means, he suggests, it is ensured that all animals in one neighbourhood discharge their gametes simultaneously, and thus a great measure of cross-fertilization ensues. This postulate is discussed below.

From the ‘injection experiments’ it appears that gonadotrophin does not act directly on the gonads. Injection of chorionic gonadotrophin or of neural-complex extract direct into a normal animal is followed by the release of gametes whatever the site of injection; the blood carries the hormone to the site of its action. After the heart has been destroyed and the blood removed, then the only site at which injection is effective in stimulating release of gametes is the neural region, which accordingly must be the site of action. Section of, the nerves from the ganalion to the region of the gonads prevents the positive response so that the immediate stimulus to the gonads to release gametes is presumably nervous. Destruction of the neural gland, so long as the ganglion is left intact does not interfere with the action of gonadotrophin in stimulating the discharge of gametes.

I conclude that when chorionic gonadotrophin is injected into an ascidian, it acts upon the ganglion. When the ganglion has been stimulated in this way, in its own turn it stimulates the gonads by means of nervous pathways, so that they discharge their gametes. The same is likely to be true of the route of stimulation when a gonadotrophic extract of the neural gland is used instead of chorionic gonadotrophin, though this extract has only been used in the intact Phallusia.

When an ascidian is fed with eggs of its own species it releases gametes—Huus’s hypothesis is confirmed on this point. The sequence of events following feeding on eggs is the same as that following injection of chorionic gonadotrophin, i.e. the positive response is prevented by section of the nerves to the region of the gonads but not by destruction of the heart and removal of the blood. The stimulus from ingested gametes then must act on the neural region. And for the positive response to follow ingestion of gametes the neural gland must be intact, a condition which is not necessary for the positive response to follow gonadotrophin injections. The stimulus, then, must be through the neural gland and must reach the gonads by nervous, not humoral, pathways from the ganglion. The problem then is how it reaches the neural gland and the ganglion.

Van Weel (1940) has shown that the food stream in ascidians, carries particles to the mouth of the ciliated pit. Carlisle (1950 a) has observed the same in a transparent tunicate, Salpa, and has reported that a proportion of particles always lodges in the ciliated pit, and it is postulated that this organ is the sense organ controlling (via the central nervous system) the rate and mode of feeding. In the course of the present investigations I have observed that a proportion of carmine or Indian ink particles fed to Ciona is always to be recovered in the ciliated pit or the neural gland. Pérès (1943) would interpret these as in the process of a kind of phagocytic excretion. This cannot be true since :

(1) Carmine and Indian ink are defaecated, and there is no sign of their particles ever being absorbed by the gut.

(2) They appear in the neural gland within a few minutes of first feeding.

(3) Direct observation in Salpa shows that there they are carried into the ciliated pit by the food current.

(4) Eggs are similarly found in the neural gland after feeding animals with them; these could not survive digestion and excretion by phagocytes in the form of entire particles. It is certain that the particles found in the neural gland are carried there by the food stream, and that they represent a sample of the ingested particles.

The neural gland and ciliated pit are. reported not to be innervated in ascidians (Julin, 1881 a; Metcalf, 1895, who records one species with innervation of the ciliated pit, viz. Boltenia, where he says that the fibres do not cross the basement membrane. He remarks that no case of innervation had been observed previously; nor have there been any records since). The neural gland contains gonadotrophin; and gonadotrophin stimulates the ganglion as described above. Moreover, the neural gland is needed to mediate between the stimulus provided by the gametes and the ganglion. The obvious hypothesis to be induced from these observations is that the neural gland receives a stimulus from the ingested eggs and in response produces gonadotrophin. The gonadotrophin passes over to the adjacent ganglion by a non- vascular and non-nervous route and causes it to stimulate the gonads by nervous pathways as discussed above. Such a passage of gonadotrophin to the ganglion may be compared with the passage of colloid material by non-vascular paths from the anterior pituitary to the hypothalamus in a fish reported by Bretschneider & de Wit (1947).

If this hypothesis is correct then gonadotrophin in ascidians is playing the part of a sensory pathway in a reflex arc :