Cyclical Changes in the Distribution of the Testis Lipids of a Teleost Fish, Esox lucius

With two plates (figs, 1 and 2)

There exists an extensive literature dealing with the seasonal changes in A the gonads of birds, but those of fish are still little understood. In the Rtale bird there has been demonstrated a cyclical regeneration and exhaustion of the cholesterol-positive Leydig cells of the testis interstitium, a regular post-spermatogenetic tubule steatogenesis (also involving the production of cholesterol), and a post-nuptial rehabilitation of the tunica albuginea (Marshall, 1949). These events follow an almost invariable time sequence under average environmental conditions although the rates and events are subject to profound disruption during periods of climatic fluctuation (Marshall, 1955).

The purpose of the present study was threefold :

1. To determine whether in the seminiferous and endocrine elements of the testis of a fish there occur events comparable with those in birds.

2. To discover whether fish (which also undergo gross seasonal tubule expansion) annually renew their old, stretched, and presumably weakened tunica albuginea as do birds.

3. To correlate, as far as possible, observed events in the testis cycle with external fluctuations in photoperiodicity and temperature.

The present study is based on the examination of 71 adult male pike (Esox lucius L.) collected at Lake Windermere and nearby Tarn Hows from October 1954 to September 1955, inclusive. The testes were weighed, sliced, and immersed in formaldehyde-calcium. In addition, 42 char (Salvelinus willughbii) were netted whilst migrating into shallow water to spawn in October and November 1954. Their gonads were removed and preserved for histological comparison with those of the pike. The material was treated in three ways:

Portions of each organ were (1) embedded in wax, sectioned at 6μ, and stained with iron haematoxylin and orange G for routine examination of spermatogenetic stages; (2) embedded in gelatine, sectioned at 8μ. on the freezing microtome, coloured with Sudan black and stained with haemalum for the investigation of tubule and interstitial lipids; and (3) sections were subjected to the Schultz test for the detection of cholesterol. All measurements in μ refer to material embedded in wax.

The testes of the pike are elongated bodies situated below the swim-bladder. They extend along the whole length of the abdomen. Seasonal volumetric variation is great. Before the pike begin their spawning run, during January to March inclusive, the gonads are large and often lobulate. The shedding of the spermatozoa, however, results in a striking reduction in volume and weight (table 1).

The testis differs in internal structure from the characteristic vertebrate pattern of seminiferous tubules and interstitium. A piece of teased testis is seen to consist of a mass of elongated, branching, blindly ending structures applied closely together. Turner (1919) described such structures radiating from a connective tissue core in the perch (Perea flavescens) and called them lobules. No such core was found in pike, but otherwise the basic structure is similar. Hann (1927) used the term tubule to describe structures in Cottus bairdii that apparently correspond with the lobules in perch and pike. The canals of the stickleback (Gasterosteus pungitius) described by van Oordt (1924) are obviously the same structures, and so too are the crypts mentioned by Jones (1940) in Salmo salar. Since teleosts lack tubules with permanent germinal epithelium like that of tetrapods, the term lobule will be used in this contribution.

The lobule wall in pike is composed of fibrous connective tissue. There are no interstitial Leydig cells. During the seasonal cycle, however, most, if not all, of the fibroblasts of the lobule wall become remarkably modified into glandular ‘lobule boundary cells’ which resemble Leydig cells (Marshall and Lofts, 1956). As they expand during spermatogenesis the lobules come closer together. The small inter-lobular spaces become occupied by blood-vessels and lymph spaces. What at first appears to be endocrine interstitial tissue in odd isolated inter-lobular patches is, in fact, the blind ends of sectioned lobules. A few are usually seen in any given section.

After spawning has taken place at the end of April, primary germ cells migrate into the collapsed lobules and initiate the new season’s spermatogenesis. No general agreement has been reached concerning the origin of these cells. Turner (1919) described a ‘cord’ of germ cells outside the body of the testis of the perch. He claimed that from this the testis is periodically supplied by cell-migration into the lobules. These conclusions, however, were based on e vidence from only one specimen and no mention was made of precisely how the germ cells got from the cord into the testis. Primary germ cells observed by us in the inter-lobular tissue of the pike during the post-nuptial period ‘‘vere similar to those described by Turner, but we found no evidence of an extra-testicular cord. In Cottus bairdii some germ cells lie dormant within the cysts during maturation and spawning, and these divide during the following summer to provide the next generation (Hann, 1927). There was no evidence of this in the pike. In Gambusia affinis the germ cells are said to develop from inconspicuous cells lying in the interlobular areas of the testis (Geiser, 1922). In the present species we were unable to determine their point of origin.

During summer and early autumn spermatogenesis reaches a peak of activity and the lobules are filled with thin-walled cysts. Spermatogonia divide and give rise to clusters of primary spermatocytes. The cells inside the cysts all divide together so that no cyst contains more than one spermatogenetic stage. A cyst with primary spermatocytes contains cells dividing together and will thus come to contain exclusively secondary spermatocytes, and next entirely spermatids, and finally spermatozoa. The cyst then ruptures and liberates its contents into the lumen of the lobule.

The entire organ is enveloped in a tunica albuginea of fibrous connective tissue, the thickness of which varies seasonally.

The tunic is about 20μ thick. The lobules measure from 130 to 180μ in width and contain cysts of developing primary and secondary spermatocytes and spermatids. Most peripheral cysts are composed of spermatocytes, whereas the more central areas are occupied chiefly but not exclusively by free spermatozoa (fig. 1, A). A few isolated, free spermatogonia persist in some lobules.

Surrounding each lobule, and incorporated in its substance, is a ring of modified fibroblasts one cell thick. Each individual ‘lobule boundary cell’ contains numerous cytoplasmic lipid droplets (fig. 1, B). These dissolve out during wax-embedding and leave a series of vacuoles with the appearance of a cytoplasmic reticulum (fig. 1, c). The average size of the nucleus is 5×2·5 μ. There is no lipid within the lobules. Many of the lipid droplets of the boundary cells are cholesterol-positive.

The average width of the lobule is 180 μ and the tunic is 20 μ thick. Cysts of all spermatogenetic stages except spermatogonia still occur but in smaller numbers than in October. The peripheral zone of developing cysts has disappeared. All lobules contain large masses of free spermatozoa.

The cytoplasmic lipids of the lobule boundary cells have increased so that in many of the cells individual droplets are no longer distinguishable. In such cells the densely sudanophil cytoplasm resembles that of the Leydig cells of birds in winter and early spring. This peri-lobular lipid is still cholesterolpositive.

Lobule width averages 190 p. and the tunic is 15 p. thick. The cysts are gradually reduced in number. The few remaining contain secondary spermatocytes and spermatids. The lobules have become distended with spermatozoa and are tightly packed together within the testis tunic. There is no change in the lipid or cholesterol content of the boundary cells.

The average lobule width is ijop,. The testis coat is now stretched to a thickness of only 12 p. All cysts have matured and discharged and the lobules are packed exclusively with mature spermatozoa (fig. 2, A). Lobule boundary cells are still densely lipoidal and cholesterol-positive.

The lobule diameter averages 180/z but there is otherwise no change from January.

As in January.

Lobules still remain identical in size and appearance with those of January, but the cholesterol-positive boundary cells appear to be discharging their cytoplasmic lipids and cholesterol. Individual lipid globules can be again distinguished. Spermatozoa are discharged at the end of April or the beginning of May.

With the shedding of spermatozoa the average lobule is sharply reduced from 180μ to a width of only 40 μ. The testis tunic has become weakened and appears to be disintegrating. Beneath the old tunic an upsurge of fibroblasts is forming a new one.

The size of the whole testis is sharply decreased (fig. 2, B). There is at the same time an increase in the true volume of inter-lobular tissue. In it occur large primary germ cells; these begin also to appear within the lobules (fig. 2, c). They seem, in fact, to migrate between the cells of the lobule wall into the lumina of the lobules. Such migratory germ cells are about 15 μ in diameter and contain a nucleus 10 μ in diameter. They are probably the same as those described by Turner (1919) and Essenburg (1923). Geiser (1922) and van Oordt (1925) used the term primary spermatogonia for similar cells in Gambusia affinis and Xiphorphorus helleri respectively.

The lobule wall is now composed of unmodified and lipid-free boundary cells (presumably of a new generation), but the lumen is full of a dense, amorphous mass of densely sudanophil material (fig. 2, D) such as occurs in birds at a comparable stage of the sexual cycle (Marshall, 1955). This is cholesterol-positive. Some individuals show the first stages of post-nuptial lobule clearance in that the central areas of their lumina are already free of lipids.

The lobules average 50 μ in diameter and the testis tunic is 10 μ thick. The old weakened tunic has disappeared and has been replaced by the pro_T liferation of new fibroblasts from within. Intralobular lipid clearance has continued. Many of the lobules now have their central regions free of lipid (fig. 2, E). Primary germ cells have given rise to spermatogonia. These line the lobules but have not yet multiplied and formed cysts. The fibroblasts surrounding the lobules remain unmodified and lipid-free, having spindleshaped nuclei measuring 8 to 10 μ in length. Cholesterol remains present in the lobule lipids but none has yet arisen in the potentially glandular boundary cells.

The lobules average 50 μ in diameter and the testis tunic is 30 μ thick. Little lipid now remains in the lobules and that which persists is wedged between germ cells. Spermatogenesis is advancing: cysts of primary spermatocytes have arisen. There is still no evidence of glandular boundary cells but an appreciable increase in the number of blood-vessels in the interlobular tissue has occurred. Weak indications of cholesterol still persist in some lobules.

The lobules have swollen to a diameter of 120 μ and the testis tunic has stretched to 20 μ in diameter. The lobules are now almost cleared of lipid. The fibroblasts in the lobule borders still remain unmodified and in the July condition. Spermatogenesis, however, is proceeding actively. Every lobule is filled with cysts of dividing germ cells with primary spermatocytes most abundant, although secondary spermatocytes and spermatids are also present. In a few lobules free spermatozoa have appeared. The last traces of cholesterol have disappeared from within the lobules but none has yet arisen in their walls.

The average lobule diameter is still 120μ and the testis tunic has a width of 20 μ. Most lobules now contain at least some spermatozoa. Cysts of primary and secondary spermatocytes and spermatids are numerous ; few spermatogonia remain. The fibroblasts of the lobule wall have started to become lipoidal. Individual cytoplasmic lipid droplets are cholesterol-positive.

The internal events described above fall into the following phases :

After the shedding of spermatozoa the remaining lobule-elements undergo steatogenesis ; the lumina become filled with cholesterol-positive post-nuptial lipids. This is part of the internal rhythm of the testis. Such lipids now gradually disappear. Meanwhile primary germ cells appear and multiply rapidly to produce spermatogonia which in turn divide and form cysts of primary spermatocytes. The lobule wall consists of unmodified fibroblasts, none of which has begun a seasonal accumulation of lipid. The shrivelled testis tunic is disintegrating. Beneath the old coat new fibroblasts are building up a new tunic.

These rehabilitative processes begin while day-length is increasing and the temperature of the water is rising. On 2 May the temperature at the surface was 9·35° C and this steadily increased to 20·25° C by 31 August. At greater depths the temperature is lower but nevertheless steadily rising (fig. 3). On

1 May day-length is 15 h 5 min and increasing by 4 min a day. These increments extend day-length to 17 h 9 min by 21 June. Day-length then begins to decrease and by 31 August it is reduced to 13 h 5 min. Spermatogenesis, continues unabated during this period of shorter days.

The last traces of post-nuptial lobule lipids disappear and the cytoplasm of the lobule boundary-cells now accumulates cholesterol-positive lipids. There is a gradual maturation of the cysts into spermatozoa which are discharged mto the lobule lumina. The previous season’s testis tunic has disintegrated.

During this maturation water temperature starts falling but on 31 December it is still about 6·75° C. Day-length is also declining and reaches a minimum of 7 h 22 min at the winter solstice (21 December). Throughout September the average daily decrease is 4 min. After 21 December day-length begins to increase but during the rest of December the daily increments are only about 30 sec per day.

The boundary-cell cytoplasm is now heavily charged with cholesterolpositive lipids. Lobules are at their maximum size and are densely packed with free spermatozoa. This stretches the testis tunic to its minimum thickness. Meanwhile, the temperature of the water is gradually decreasing and day-length increasing.

For a while the lobules remain unchanged but the boundary cells begin to discharge their lipids and cholesterol. Spermatozoa are shed at the end of April and early in May. These culminating events occur at a period of rising temperature. Such an increase begins about the third week of March and by the end of April has increased to 8-9° C (fig. 3). On 1 April the day-length is 13 h r min and increasing by 5 min a day.

In the testis of the pike there occur cyclical events not unlike those of seasonal birds. These involve a post-nuptial steatogenesis of the lobules and a subsequent gradual expenditure of recently arisen cholesterol. In birds, however, no more than primary spermatocytes are formed while tubule lipids remain. The last traces of tubule lipids disappear before, or with, the main upsurge of the next season’s spermatogenesis, whereas in the pike, as in Vipera berus and Rana temporaria (Marshall, unpublished), spermatogenesis is well under way whilst appreciable quantities of lobule or tubule lipid still remain.

The cycle of the boundary cells is essentially similar to that of the interstitial cycle in birds. The unmodified fibroblasts which arise around the ‘spent’ lobules grow and accumulate cholesterol-positive lipids in their cytoplasm. This occurs also in the teleosts Salvelinus and Labeo (Marshall and Lofts, 1956).

Two distinct endocrine arrangements, therefore, occur in the testes of teleost fish. One arrangement is here shown in the pike and the char : these lack a true interstitium. The second arrangement conforms to the typical vertebrate pattern (of Leydig cells) and occurs in the stickleback (Gasterosteus), the sprat (Clupea), and many other species of teleosts and also the elasmobranchs Scylliorhinus and Chimaera. A boundary-cell cycle occurs also in certain urodeles. Miller and Robbins (1954) described in Taricha (= Trituras) tortosa large peri-tubular connective tissue cells with an osmiophil cytoplasm with an apparently glandular function.

In the pike, as in birds, a new testis wall is formed after each breeding season.

A consideration of seasonal fluctuations in day-length and water temperature shows that rehabilitative changes, beginning immediately after the discharge of spermatozoa in late April and early May, occur whilst both day-length and temperature are increasing. Maturation of germ cells (i.e. the transformation of primary germ cells to spermatozoa) occurs from June to December. Thus, decreasing day-lengths after the summer solstice do not inhibit generative processes. By the winter solstice, and at a period when the surface is sometimes covered with ice, the lobules appear to be ‘ripe’. They are now distended with spermatozoa and the lobule boundary cells have become heavily charged with cholesterol-positive lipids which are, however, still unexpended. The impression is gained that, whatever the subsequent function of such a cholesterol reservoir, it is not primarily concerned with manufacture of hormone material concerned with spermatogenesis.

During the coldest period of the year, and despite increasing day-length, there ensues a period of inactivity. This lasts until April. Later that month, with the temperature of the water elevated to 8·9° C, the fish undergo an annual pre-spawning assembly and, as their boundary cells suddenly become denuded of cholesterol (probably signifying the manufacture and discharge of sex hormone), spermatozoa are shed.

In the absence of experimental data we cannot evaluate fully the relative importance of temperature and day-length in the control of the events outlined above. Fabricius (1950) has suggested that the spring increase in daylength and light intensity are the external stimuli critical in the pike. That day-length can indeed affect the reproductive cycle of fish has been demonstrated by Hoover and Hubbard (1937), who subjected brook trout (Salvelinns fontinalis) to a gradual daily increase and then decrease in illumination, and observed the production of ripe eggs and mature spermatozoa several months earlier than usual. Light may possibly be an effective stimulus when the pike have already assembled in shallow water, but it seems unlikely that it initiates the neuro-hormonal events that lead to such assembly from the main body of the lake. Light intensity is greatly reduced in deep water, and furthermore, assembly occurs in many areas when the water is still covered by ice (Fabricius, 1950; Moen and Lindquist, 1954). Carbine (1941) recorded 18 inches of ice on Haughton Lake when the pike showed the first signs of assembling to spawn.

Increase in water temperature, on the other hand, has been shown to accelerate spermatogenesis in a number of teleosts (Bullough, 1939; Burger, 1939; Matthews, 1939)-Individuals of the three-spined stickleback (Gasterosteos aculeatus) from different habitats showed a variation in spermatogenetic rates apparently owing to the variation in the rates of temperature reduction in autumn (Craig-Bennett, 1930). When such reduction was rapid, the rate of spermatogenesis was delayed and maturation continued into the breeding season. When slow, spermatogenesis was completed two to three months before the breeding season as in the pike.

We wish to thank Mr. G. J. Thompson of the Fresh-Water Biological Association Laboratory, Windermere, for his help in obtaining material.