Introduction
The study of spermatogenesis began in the middle of the 19th century following the foundational discovery in 1839 by the German team of Theodor Schwann and Matthias Schleiden of ‘cell theory’. Cell theory was first applied to the study of male reproduction following observations by Rudolph Wagner and a comprehensive study published in 1841 by Albert von Kölliker, in which he provided compelling evidence using invertebrate models (37 species in total) that spermatozoa were actually formed in the testis (Koelliker, 1841). The next ~100 years of research on spermatogenesis (~1850–1950) were aptly described by Edward Roosen-Runge (Roosen-Runge, 1977) as the ‘histological era’. It was during this time that scientists used the light microscope to image fresh and fixed unstained and stained testes from a wide variety of species to painstakingly describe cellular morphology and associations. This era has been largely unexamined by later reproductive biologists, with few manuscripts having been translated to English. As a result, many fundamental discoveries made during this era have been attributed to scientists working in the 1950s.
Scientists studying mammalian spermatogenesis during the first 30 years of the histological era had an apparent fundamental disagreement regarding the origin of spermatozoa. One group of scientists (e.g. Sertoli, Kölliker, Henle, Schweigger-Seidel, La Valette St-George, and Merkel) recognized the full spectrum of subtle progressive changes during male germ cell development (spermatogonia at the basement membrane of the seminiferous tubules give rise to spermatocytes and then spermatids, which became spermatozoa when released from the seminiferous epithelium) to correctly conclude that spermatozoa originated from the germline. This view was opposed principally by another group of scientists (e.g. Ebner, Mihalkowics, Neumann, and Blumberg), who believed that spermatozoa originated instead from Sertoli cells.
We became interested in this controversy after translating a manuscript by Enrico Sertoli (Geyer, 2018; Sertoli, 1878), in which he detailed his extraordinarily comprehensive histological studies of the rat testis. That report provided first detailed descriptions of the blood-testis-barrier, intercellular bridges, spermatogonial cytoplasmic projections, as well as a detailed and accurate characterization of the cellular associations during the stages of the seminiferous epithelium. In Sertoli’s discussion of his results, Sertoli was sharply critical of the views of his scientific rivals, principally Victor von Ebner. Therefore, to assess the controversy from Ebner’s point-of-view, we translated this manuscript describing Ebner’s studies on spermatogenesis, which he submitted for what is analogous to today’s promotion from assistant to associate professor with conferral of tenure. We were impressed by Ebner’s clear writing style, careful descriptions, and precisely-drawn figures. Ebner clearly recognized that spermatogenesis followed a developmental trajectory within the seminiferous epithelium and occurred asynchronously along the tubules, and he used illustrations to delineate the process into eight stages. Ebner also took commendable efforts to compare spermatogenesis between multiple mammal species (rat, mice, cat, dog, guinea pig, rabbit, hedgehog, and human), and reported the remarkable conservation of cells and structures.
We struggled to understand how Ebner failed to recognize that spermatozoa were the final product of germline development. As best as we can tell, several things contributed to his incorrect conclusions, all relating to aspects of spermatid development (now termed spermiogenesis). First, Ebner noted difficulty in discerning cell boundaries (e.g. “Thus far, I have not been able to represent cell borders in the Keimnetz. It seems that we are dealing with a fusion of the cells, similar as, for example, in the so-called cloudy epithelium in the tortuous tubules or in certain developmental stages of the embryonic connective tissue”); this technical issue likely resulted from his choice of fixative, stain, tissue sectioning and mounting techniques and/or lack of microscope resolution. Indeed, Sertoli seems to have had an advantage in one or more of these areas; Sertoli’s ability to discern so many of the intermediate stages during spermatid development led him to the correct conclusion that round spermatids (Ebner’s ‘spermatoblasts’) transform into elongating and then condensing spermatids before becoming spermatozoa at spermiation. Second, after a largely accurate description of the cellular development of the germline up to what appears to be the secondary spermatocyte stage (Table 1), Ebner did not observe their division at the end of meiosis II to become “spermatoblasts” (round spermatids). Third, it seems he could not commit to the concept that the highly modified heads of spermatozoa were indeed nuclei; by doing so, he would have appreciated the gradual changes that the spermatid nuclei undergo during spermiogenesis. Fourth, Ebner seemed to have been influenced by the incorrect findings of contemporary scientists. For example, he noted that Letzerich reported spermatoblasts were only seen in mammals, and that “spermatozoa originate independently of cell nuclei as separate formations”. Ebner wrote, on page 29 that “[a]ll these observers consider the cells of the seminiferous tubules as precursors of spermatozoa. The Keimnetz (Sertoli cell) was either overlooked or completely misunderstood”. Thus, he ended the manuscript with the conclusion that Sertoli cells produced spermatozoa, while the germline likely contributed to “formation of fluid contained in seminiferous tubules and perhaps also for the supply of nutritional material to the spermatozoa”.
Table 1.
Growth of cells of the seminiferous tubules
|
Main Text
Although the development of spermatozoa has been studied frequently in recent years, the various results by different scientists rarely coincide, at times even contradicting one another.
When looking for the reason for this disagreement, the answer seems obvious: a reliable investigative method is lacking. There is only one completely certain way to solve histogenic problems: that is the direct observation of the gradual transformation of one and the same object into another under a microscope. Unfortunately, this method is not applicable in our case. Thus, the development must be deduced from observations, which have been made on different objects.
To understand spatially separated shapes as subsequent stages of development is only valid to a point. The degree of certainty depends in part on the way in which the specimens have been prepared. Already 14 years ago Heinle criticized scientists of cell genesis2, who “order cells cut from any tissue and transferred to a slide chronologically, rather than deducing their chronological age from the place they occupied in the tissue.” Studies about the development of spermatozoa have been carried out using approximately the same method. When examining a scrape preparation of a fresh testis sample it becomes clear that the great diversity of shapes – and illusions – give free reign to the imagination of the observer, to the point to which it would be miraculous if two scientists were to come to the same results using this method. Heinle has thus tried to determine the spatial arrangement of differently shaped elements in the seminiferous tubules in order to obtain more reliable conclusions about the spermatogenesis than previously existed3.
Heinle examined cross-cut sections of hardened testes. Though he did not find a specific spatial arrangement of the cells, he did note indications of it, that is rows of cells in radial direction. Thus the use of this little tried method constitutes progress and it is regrettable that recent investigators of our subject have not followed Heinle’s path. This can only be explained by the assumption that some seem certain about the absence of a specific spatial arrangement of the developmental stages of the spermatozoa4.
As a result of the above considerations, I have mainly worked with fine cross-sections and only used a scrape preparation when studying details or to verify results.
Before presenting the results of my study, a few remarks about preparations are in order. As a hardening agent I almost exclusively used Müller solution, in which the examined objects were soaked for several weeks or even several months. The pieces of testis intended for the preparations were put into strong alcohol for 24 hours and then melted on a glass plate into a mass of Peremeshko5. The connections between the seminiferous tubules are only very loose, especially in certain animals, because of the numerous wide lymphatic columns6. If making thin cross-sections, it is easy for the whole specimen to break up into the transverse and longitudinal sections of individual seminiferous tubules. To counter this, I proceed as follows:
After the testis piece, from which the alcohol has been largely removed with blotting paper, is about half melted down, I heat the glass rod in a gas flame and use it to apply heated oil-wax drops. This heats the alcohol within the specimen to boiling, which then escapes in the form of bubbles through the mass. The dripping of the heated mass is continued until no or only a few steam bubbles escape. Using this procedure, the oil-wax mass penetrates throughout the entire tissue, resulting in sections, which, though extremely thin, do not fall apart when handled with care. The heating of the glass rod must be carefully regulated so that the steam does not develop too rapidly. Otherwise the excessively heated mass may tear or destroy the tissue.
The sections are stained with hematoxylin and then usually rendered transparent with clove oil or with glycerol. Hematoxylin was chosen as a staining agent, because it stains nuclei much more reliably and intensively in preparations of Müller solution than carmine, but especially because it tinges the heads of spermatozoa particularly strikingly.
Of particular importance was the choice of the subject of the study. In the rat, I was fortunate to find an animal, in which the structure of the seminiferous tubules and the development of the spermatozoa is relatively easy to isolate. There are two different features in rats, which drew my attention to the animal from the very beginning. First, the spermatozoa of this animal, are among the largest ever found in mammals, a fact well-known since R. Wagner’s studies. Second, the heads of the spermatozoa are of such a peculiar shape that they are not easily confused with anything else, e.g. the nuclei7.
In addition, there are some peculiarities in the structure of the rat’s testis, which greatly facilitate their study. The seminiferous tubules are relatively large in diameter and arranged in such a way that one can freely make any longitudinal or transverse cross-sections of them, which is not possible with other animals.
In the rat, the seminiferous tubules do not show as numerous irregular convolutions as occur in the testis of humans, dogs, etc., but they are placed in multiple, regular loops, whose almost perfectly straight legs run parallel to each other and usually have a length of 5–7 μm8. However, even very short loops, and – especially on the surface near the rete testis – also irregular convolutions of the seminiferous tubules, can be observed. The loops are usually arranged in such a way that a cross-section perpendicular to the axis of the testis will show primarily sections of seminiferous tubules. Connective tissue partitions, which separate the testes of man and many mammals into lobules, are completely absent in the rat. The connective tissue, or rather the filling material located between the seminiferous tubules, is altogether peculiar, and although considering the title of this study I should not expand on its structure, I think it proper to offer some remarks about it.
In the testis of man, of the dog, the rabbit, etc., the septa of the lobules are composed of fibrillar connective tissue, while the interior of the lobules consists of tracts of connective tissue, which extend through the lymphatic spaces containing the seminiferous tubules. Clusters of connective tissue often pass through the lymphatic space and attach themselves to the tunica propria9 of the seminiferous tubules. In the connective tissue, especially around the vessels, there are now peculiar cell clusters, which at times present as strands and have received very different interpretations from the histologists. Leydig considers them a special form of connective tissue cells analogous to the fat and pigment cells. Similarly, Kölliker considers them to be cells of the connective tissue, while Heinle expresses uncertainty about their function, but points out their morphological similarity with nerve cells.
It appears that these are the same cell strands that Boll refers to when he claims that the capillaries of the testes might be confused with epithelial tubes.
Such peculiar cell clusters (Leydig cells), which, in addition to the vessels and nerves, occupy the interstitia between the seminiferous tubules almost exclusively in the rat. Fibrilliary connective tissue occurs only as adventitia of the larger vessels.
If a moderately thin longitudinal cross-section is made from a testis hardened in Müller solution, the seminiferous tubules can be easily removed by shaking in water, and then only the above-mentioned supporting substance remains. It is composed of fairly strong strands running parallel to the seminiferous tubules, which are interconnected by more loosely-spread cell clusters. The latter are often interspersed with fairly regular, roundish or elliptical gaps10, so that two cell strands lying next to each other, together with the substance connecting them, might be compared with a ladder. Sometimes, however, the gaps are missing for longer stretches. Larger blood vessels are embedded in the cell strands, although sometimes the strands contain no vessels.
Numerous capillaries run through the connecting bridges of the strands, while here and there a large vessel also crisscrosses from one cell strand into another.
A cross-section of a testis part melted down in the previously mentioned manner shows the supporting substance arranged in a very regular manner (Fig. 1). The seminiferous tubules, which appear as circles on the cross-section, leave a triangular gap where they meet in groups of three. In each of these gaps lies the cross-section of a cell strand, which takes the form of a triangle with inflected sides. When the cross-section goes through the connecting bridges mentioned earlier, fine bands of substance now pass from the corners of this triangle to the corners of the adjacent triangles, or, when the cut goes through a gap, there remain two disconnected triangles here and there.
Fig. 1.
Part of a testis cross-section at 100X magnification. This drawing shows the seminiferous tubules with the different developmental stages of the spermatozoa, which are distributed irregularly on the cross-section. The arrangement of the seminiferous tubules, the interstitial tissue and the contents of the seminiferous tubules, etc. are drawn from nature. 1–8 seminiferous tubules with spermatozoa in I.-VIII. stage of development; a. Tunica propria (PTMs); b. Wandschicht (Sertoli cells, spermatogonia, and early primary spermatocytes); c. Cells resting on the Wandschicht (spermatocytes); d. Spermatoblasts (spermatids) with the rudiments of spermatozoa (these letters are only indicated in Fig. 1). Here and there the interstitial strands show transverse and longitudinal cross-sections of vessels. There is a longitudinal artery between 1 and 2.
Each seminiferous tubule, therefore, appears surrounded by a hexagon with heavily thickened corners, which is occasionally missing one side. The supporting substance thus represents a system of six-sided prismatic tubes. The surfaces of the prisms are punctured by numerous openings, so that the individual tubes often communicate with each other.
The supporting substance is only very loosely connected to the seminiferous tubules. The wax mass was able to penetrate nearly everywhere between the supporting substance and tunica propria. Thicker cross-sections of testes hardened in Müller solution demonstrate that this is not a violent detachment. Immediately after immersing the section in the fluid, the seminiferous tubules mostly fall out of the framework, leaving primarily the delicate hexagons with the thickened corners.
Rarely, fine transverse and longitudinal cross-sections show delicate trabeculae, which pass either from the thick strands or from the connecting bridges to the tunica propria of the seminiferous tubules and attach themselves to it.
With regard to the finer construction of the supporting substance it is worth noting that it differs considerably in the cell strands and in the connecting bridges. The cells of the strands are largely irregularly polyhedral, approximately 18–21 μm in size, usually slightly elongated in the direction of the strands. They are very grainy, so that the nuclei are often barely visible or invisible altogether.
The nucleus represents on hardened preparations a roundish or elliptical sharply contoured mass of 6–8 μm diameter, which includes one or more grains. Not infrequently, however, you will also find nuclei that show a coarse-grained appearance and an indistinct border. Multi-nuclear cells are not uncommon. Here and there you can also see cells that anastomose through thick extensions.
There is a very coarse-grained substance in the interstices between the cells, which at times appears formed in discrete, sausage-shaped or irregular masses. Occasionally, these masses stretch into undulating fibers, which have little resemblance to connective tissue fibers, however. The picture changes towards the connecting bridges. The cells show very irregular shapes and often have longer extensions on one side, which follow the course of the capillaries. The granular masses also stretch out considerably and turn into fibrous, sometimes nodular swollen structures. Finally, there are cells with long tails, which carry a strongly oval smooth nucleus.
The cells are often nothing more than proof of the blood vessels, both larger ones, which have a strong adventitia consisting of fibrillar connective tissue, as well as smaller ones, as well as of capillaries.
Looking at fresh or non-tinged specimens of the latter, which do not show the structural elements of the vessel wall, can lead to the assumption that the cell strands themselves are vessels. However, looking at stained preparations it is always possible to observe either the nuclei of capillaries or endothelial nuclei as well as transverse and longitudinal nuclei of muscle as elements, which cover the vascular lumen. In the connecting bridges (the prism surfaces), the cells are usually arranged in a single position, so that the capillaries connect only to 2 sides of supporting cells, while the capillaries are not covered. In addition to these vessels containing strands, however, there are also those which do not hold any vessels.
The supporting substance of the testis of the mouse behaves essentially as that of the rat as well. In other animals, e.g. the dog, the cat, and the rabbit, as well as in humans, the fibrillar connective tissue plays a prominent role, and the cells are formed in strands or nests in the interstitial connective tissue, as well as in the septa of the lobules. In these animals, as well as in humans, cross-sections of triangular cell strands can be found in particular at the places where the cross-sections of 5 seminiferous tubules meet. Here, however, the cells do not show the same diversity in shapes as in the rat; they are usually irregularly polyhedral, very coarse-grained, and often contain yellowish pigment grains, especially in older men. The nuclei are sharply contoured, roundish or elliptical, and almost always show a clear nucleus. Incidentally, the rabbit also has a similar diversity in cell shapes, as does the rat. In the dog, I have observed that all vessels show the normal structure, and that the cell strands partly cover vessels and partly represent independent masses. I examined a testis of a dog whose vessels, which had been dissolved in Berlin blue-colored11 water, still showed the structure of its wall, especially on stained preparations. According to the above observations, Boll’s assertion that the testis contained capillaries of blood looking like epithelial tubes is only correct in so far as those peculiar cells, which comprise the interstitial tissue of the testes in some animals in part and in others completely, can also enclose the capillaries.
I believe that the findings regarding the testis of the rat are of great importance for the interpretation of the discussed tissue. They strongly suggest that the strands and bridges of cells are nothing but a peculiar form of connective tissue. I thus maintain that my results confirm what Leydig succinctly and clearly asserted in the above cited location.
In passing from these remarks to the very subject of my investigations, I will first describe the pictures, which present the seminiferous tubules of the rat on thin cross-sections stained with blue-wood-extract. The tunica propria of all seminiferous tubules is constructed in the same way. On the cross-sections it appears as a ring of about 1.2–1.9 μm thick (compare Fig. 1 and 2 and Fig. 4–10), which contain rod-shaped nuclei here and there12. These nuclei have an elliptical outline when viewed from the surface, are smooth, 8–12 μm long and 3.5–9 μm wide, and positioned partly parallel to the direction of the longitudinal axis, partly obliquely or transversely to the same. The tunica propria can be isolated in preparations of Müller solution and then, apart from the nuclei, presents as a seemingly structureless, at most finely granulated skin. Examining preparations that are treated with silver, an image depicting polyhedral shapes emerges. If these preparations are stained with hematoxylin, which is easily done, the same elliptical nuclei appear within the shapes, which can be observed on stained sections and isolation specimens of testes preserved in Müller solution. There is usually one nucleus in each shape, though at times two or three can be observed. I did not see any silver lines crossing with a nucleus. When observed at strong magnifications, nuclei and silver lines appear to be in the same plane, which does not prove much, however, given the thin nature of the tunica propria. However, the above-mentioned seems to clearly suggest that the tunica propria of rat seminiferous tubules consists of a single layer of polyhedric flat cells (PTMs). A fibrous adventitia, which, according to the authors13, is said to occur in men and in some animals, is definitely not found in the seminiferous tubules of the rat14. As I mentioned earlier, delicate trabeculae appear only rarely on the seminiferous tubules.
Fig. 2.
Longitudinal cross-section of a seminiferous tubule with spermatozoa in VI. stage of development. The curvature of spermatozoa tails becomes visible. At 100X magnification.
Fig. 4. I. Development Stage.
a. Tunica propria (PTMs); b Wandschicht with the pale nuclei of the Keimnetz (Sertoli cells) and the dark granulated cells (early primary spermatocytes, likely at leptotene stage); c. Cells resting on the Wandschicht (pachytene spermatocytes); d. Spermatoblasts with spermatozoa (elongating spermatids).
Figure 10. VIII. Developmental Stage.
The contrast between the three cell generations (c, c’ and the granulated cells of the Wandschicht b) is now most pronounced. The extensions of the Keimnetz vanish between the cells of the 1st generation.
When looking at the contents of seminiferous tubules on cross-sections of the testis, it is striking that the individual seminiferous tubules do not all show the same picture. In addition to seminiferous tubules, which are filled with mature spermatozoa in the center, there are others, which seem void of spermatozoa, and which show a center that is almost free of morphotic elements. In addition to these two extremes there are seminiferous tubules, which represent unmistakable developmental stages of spermatozoa (spermatids).
When trying to differentiate the present developmental stages, we can distinguish about eight different images15, which, of course, are not immediately adjacent to one another.
First, these eight developmental stages (I-VIII) will be characterized, while their interpretation shall follow in the subsequent sections.
I will strive to be as objective as possible in the description, always allowing the reader to distinguish between observations and conclusions. For the purpose of an orderly and reasonably clear presentation, I will later explain and substantiate much of what I will consider as established for the time being.
I. Stage (Fig. 1,1 and Fig. 4)
The first rudiments of the spermatozoa heads are recognizable as roundish nucleus-like structures.
Inward from the tunica propria, the seminiferous tubules first show a peculiar layer, which seems quite sharply demarcated from the remainder of the seminiferous tubules (Fig. 4, b). This layer was formerly called the epithelium of the seminiferous tubules, because it was thought to be composed of regularly arranged polyhedral cells. This is not the case, and I will therefore use the more innocuous term “Wandschicht”16. On the cross-section, it borders on the tunica propria on the outside, but it is clearly delimited from it, while inwards its border does not appear as a straight line, but rather jagged17. On a superficial level, it gives the impression that it consists of a granulated mass containing a layer of fairly round nuclei within. Upon closer examination, one notices, but only on very thin cross-sections, that not all nuclei are the same. A distinction can be made between round coarse-grained dark nuclei18 and round or elliptic sharply-contoured pale nuclei, which usually include a distinct strongly blue-colored nucleolus19. Strikingly, the latter (Sertoli cell) nuclei occur precisely at the location of the pointed edges of the jagged line. At times it can be observed on particularly thin cross-sections that the jags can protrude against the lumen of the tubules, and that the extensions often also have smooth, nucleoli-containing elongated nuclei arranged in radial direction, which differ from the similarly formed nuclei of the actual Wandschicht in that they appear pointed at their inner end20. Larger magnifications also often give the impression that the coarsely granulated nuclei do not belong to the mass carrying the pale nuclei (Sertoli cells), but rather to the roundish cells, which are almost entirely filled by the nuclei21. If a tinged seminiferous tubule is dissected, it is possible to easily isolate pieces of the membrana propria which still contain parts of the Wandschicht, while the remaining contents of the seminiferous tubule are removed. It can then be ascertained that it consists of a mass, which contains numerous larger and smaller very shiny elements, and that this mass forms a coarse network (Fig. 13). For reasons that will become clear later, I call this network a “Keimnetz”22. In it the previously mentioned nuclei containing nucleoli (of Sertoli cells) can be found, often at the nodes of the network. Roundish or polyhedral cells with the previously mentioned granulated nuclei, which almost completely fill the cells, lie in the gaps within the network. The two nucleus forms also differ in their size: while the granulated nuclei measure only 5.2–5.5 μm, the pale nuclei of the Keimnetz (referring to Sertoli cell nuclei) reach a size of 7–7.5 μm. The long nuclei that appear in some extensions can even be 16 μm long.
Fig. 13.
A piece of the Keimnetz has been isolated. This as well as the following figures were drawn according to preparations immersed in glycerol. In a we see three granulated cells of the Wandschicht. At 300X magnification.
The cross-section shows one or two layers of about 19–24 μm granulated, usually irregularly polyhedral cells following the Wandschicht (Fig. 4, c), which cover a granular round nucleus of 8–10 μm diameter (pachytene spermatocytes). If the section is very thin and successfully executed, it shows that these cells lie in the interstices between extensions23 that go inward from the Keimnetz.
Dinuclear cells of this kind24, whose nuclei are arranged in radial direction over each other, as well as dinuclear cells whose cell body thins in the middle, can be frequently observed. Inwardly from this large-celled layer follows a pale granulated mass, comprising numerous strongly blue-colored, sharply-contoured, granular, nucleus-like structures, which are only 4.3–5.2 μm in diameter (round spermatids).
Sometimes it seems that portions of the finely granulated mass surround some of these nucleus-like formations. It can further be noted that in some places this layer protrudes outwards and that individual nucleus-like grains reach the extensions of the Keimnetz. Oftentimes, the connection of this innermost layer with the extensions of the Keimnetz can be determined by means of sections. If a scrape preparation is made, it is easy to isolate extensions of the Keimnetz, which grow a strongly widened end, containing several (Figs. 4, 8–12) of the former nucleus-like structures (round spermatids) (Figs. 4, d and 12). Usually, the widened end is lobed in the shape of a hand, and the nucleus-like structures (round spermatids) then sit at the base of the lobes of differing sizes. In the majority of cases, however, only fragments of these structures can be obtained, and the individual lobes appear isolated like cells with a small, walled nucleus. Even easier than by scrape preparation, one can obtain the mentioned structures, which I will refer to as spermatoblasts25, by crushing a piece of hardened seminiferous tubule in the first developmental stage through moderate pressure on the coverslip. It is not uncommon to see pieces of the burst tunica propria (PTMs), which still contain the Wandschicht and one or more spermatoblasts, while the large cells (pachytene spermatocytes), which often obscure everything on the cross-sections, were removed by the pressure. It thus follows, that the finely granulated mass interspersed with nucleus-like structures, which follows the large-celled layer inwardly on the cross-section, consists of the widened, lobed ends of the extensions of the Keimnetz (spermatoblasts). It is not always possible to decide whether cells similar in size and appearance to the isolated lobes of spermatoblasts exist in addition to these spermatoblasts. This does happen at times, however, as we will explain when describing stage VIII of the development.
Fig. 8. V. Development Stage.
The Wandschicht b now shows dark, granulated cells (third cell generation). The cells of the second generation (c’) have enlarged, while those of the first (c) have shrunk and multiplied88. The spermatoblasts (d) already show the tails of the spermatozoa, while the pointed nuclei seen in Figs. 4, 5, and 6 at the base of some spermatoblasts, are no longer present.
Fig. 12.
An isolated spermatoblast in the first stage of development bearing a pointed nucleus at the base (of a Sertoli cell). At 300X magnification.
II. State (Fig. 1,2 and Fig. 526)
Fig. 5. II. Development Stage.
The rudiments of spermatozoa take on the nail shape. The letters in this figure refer to the same parts as in the following 5 figures.
The spermatozoa heads have a nail-shaped shape.
This stage differs from the former primarily by metamorphoses involving the spermatoblasts containing nucleus-like structures, but the Wandschicht also shows a somewhat different appearance. The nucleus-like structures, which are nothing other than the rudiments of spermatozoa27 (elongating spermatid) heads, have now taken the form of small round-headed nails, whose tip is directed towards the periphery of the seminiferous tubules. At the same time, all spermatoblasts are now clearly lobed, or the lobes, already visible in the first stage of development, have become longer. The strongly elongated nuclei at the base of the spermatoblasts are still present, but they also may be missing here or there, as in the previous stage.
The Wandschicht (Figure 3, b) is beginning to separate into two layers. Whereas in the first stage, the nuclei of the Keimnetz lay in the same layer as the coarsely granulated nuclei, the cross-section shows that the latter have now moved slightly more inward. The large cells (pachytene spermatocytes) between the spermatoblasts show the same behavior as in the 1st stage. It should be noted, however, that the images indicating division processes have become more frequent.
Fig. 3.
Tangential cross-section of a seminiferous tubule with spermatozoa in IV. or V. development stage. The groups of dark dots correspond to the cross-sections of spermatoblasts (spermatids), while the individual dark spots correspond to the heads of the spermatozoa. In between are cells of the middle layer. At 300X magnification.
The empty lumen of the seminiferous tubules behaves as in the previous stage.
III. Stage. (Fig 1,3.; Fig. 5, 7, and 9)
Fig. 7. III. Developmental Stage.
The Wandschicht b shows the nuclei of the Keimnetz (Sertoli cells). Near and on the tunica propria we notice dark nodules in two places (likely nuclei of PTMs). The cells, which were still in the Wandschicht in the previous figure now rest on the Wandschicht on c’ (second cell generation), whereas the cells which formerly rested on the Wandschicht c (first cell generation) multiplied by division and have become smaller.
Fig. 9. VI. Developmental Stage.
The spermatozoa are detached, the cells of the first generation (c) (spermatids) have become even smaller, while those of the second generation (c’) (pachytene spermatocytes) have grown in size89.
The spermatozoa heads have a hook-shaped form.
The tips of the nail-shaped spermatozoa (condensing spermatids) heads begin to curve, simultaneously losing the spherical shape which the nail head had in the previous stage, and the whole structure already has an unmistakable resemblance to the fully formed head of a rat spermatozoa. The head, which is still without midpiece and tail28, can now be best compared to a petal of the Aquilegia viewed from the side, which has a little curved spur. Furthermore, the whole spermatozoa (condensing spermatid) head has become longer. While it was only 7–9 μm long in the previous stage, it now measures 12–14 μm. At the same time, the lobes, which contain the spermatozoa (condensing spermatid) heads at their bases, have considerably lengthened and narrowed. Finally, it is striking that the spermatozoa (condensing spermatid) heads have moved significantly further outward, so that in some places they already approach the Wandschicht, which is only rarely the case in the earlier stages. The strongly elongated nuclei29 at the base of the spermatoblasts are hardly noticeable, but there are almost always similar nuclei to the ones occurring in the Keimnetz.
The separation process, which started at the Wandschicht in the previous stage, has now been largely completed. The coarse granulated cells (early primary spermatocytes) have risen above the Wandschicht, so that they are now in the same plane as the base of the spermatoblasts30. The Wandschicht is now almost exclusively formed from the Keimnetz. Only close to the tunica propria (PTMs) we occasionally observe granulated, strongly imbibing nodules, which are similar to the cells that moved from the Keimnetz inward, but they are smaller in size.
The large cells (pachytene spermatocytes) also present a different picture. Whereas before they lay only in 1–2, at most 3 rows of the Wandschicht, they can now be observed in 3–4 rows. Moreover, various stages of cell division31, even entire cell strings, are present. Furthermore, in addition to 1 and 2 nucleated cells we find – although very rarely – cells that contain 3–4 nuclei32. However, division stages of nuclei cannot be observed. At the same time, the size of the cells has decreased somewhat, and the nuclei have also changed, especially in the inner layers. The innermost cells often look swollen, and pale seedless spheres33 frequently fill the free lumen of the seminiferous tubules.
IV. Stage (Fig. 1,4 and Fig. 15)
Fig. 15.
A spermatozoa in the IV. developmental stage. The midpiece has not yet formed, the tail is already visible. At 570X magnification.
The first traces of spermatozoa (spermatid) tails develop out of the strongly extended lobes of the spermatoblasts.
This stage of development differs from the one discussed above, in that the lobes of the spermatoblasts have grown out so much that they almost reach the center of the seminiferous tubule. It can further be noted that these lobes become pointed at their ends, so that they form a thread, which does not typically appear sharply detached from the wide lobe, but rather gradually blends into it. At the same time, the spermatozoa (condensing spermatid) heads have moved even further outwards towards the Wandschicht, so that they now touch the Wandschicht almost everywhere. It should be noted that the spermatozoa (condensing spermatid) heads, which appear in the spermatoblasts in groups of 8 to 12, are still not located in the same layer, as was the case in the first stage. Some heads are always closer to the lumen of the tubule than the others. The behavior of the Wandschicht, of the coarsely granulated cells (spermatogonia and early primary spermatocytes), etc., has not changed significantly in comparison with the previous stage.
V. Stage (Fig. 1,5; Fig. 8 and 16)
Fig. 16.
Spermatozoa in the V. developmental stage. The midpiece is already clearly differentiated at its head end. At 570X magnification.
The tails of the spermatozoa (spermatids) are clearly differentiated, while simultaneously the midpiece is formed.
At this stage, the lobes of the spermatoblasts, which had been greatly elongated at the previous stage, have narrowed significantly or nearly disappeared, while the spermatozoa (condensing spermatid) midpieces have taken their place. The spermatoblasts now represent a sheaf of 8–12 spermatozoa (spermatids), whose heads are planted in a granular mass protruding only slightly from the Keimnetz of the Wandschicht. Stretched out in a straight line, the long tails of the spermatozoa (spermatids) would now occupy far more than the radius of the seminiferous tubules. However, they are curved in flat arcs, so that their ends all come to lie approximately in the longitudinal axis of the seminiferous tubules (see Fig. 2). Cross-sections in the middle of the seminiferous tubules therefore show numerous cross-sectioned tails of spermatozoa (spermatids).
The cells we have seen multiplying in the previous stages show little changes in appearance. The coarsely granulated cells, however, which we saw move out of the Wandschicht in the previous stages, are markedly enlarged in size, and are approximately equal in size to the outermost of the multiplying cells, although they differ slightly from these in their more granular nuclei, which take on a very dark color when imbibing. In addition to the nuclei of the Keimnetz we now see again smaller, roundish, coarsely granulated cells (likely early primary spermatocytes) in the Wandschicht. Therefore, it resembles completely the Wandschicht of the first stage of development.
VI. Stage (Fig. 1,6; Fig. 2, 9 and 17)
Fig. 17.
Spermatozoa in the VII. developmental stage. At the head and at the connection of the midpiece with the tail it still shows an appendage. At 570X magnification.
Start of spermatozoa (spermatid) detachment.
The spermatozoa (spermatids) have now nearly fully formed, to the degree that this takes place in the testes at all34. The lobes of the spermatoblasts have almost completely disappeared, but there is often an appendage of granulated appearance35 on the midpiece, especially where connects to the tail. The heads themselves are still located in a coarsely granulated mass. The detachment is now initiated by the Keimnetz as it begins to drive out extensions at the points where the spermatozoa (spermatid) sheaves sit. Having emerged each from one spermatoblast, these extensions now push the spermatozoa (spermatid) sheaves towards the center of the seminiferous tubules. At the base of these extensions we now find again the same kind of elongated, partly pointed nuclei (likely of Sertoli cells), as we observed in the first stages at the base of the spermatoblasts. The extensions also show the same appearance as the spermatoblasts and can often be tracked along the lumen of the tubules up to a length of 40–50 μm. They pass into the very coarsely granulated mass surrounding the heads of the spermatozoa (spermatid) sheaves. The Wandschicht has not changed since the previous stage, whereas the outer cell layer, consisting of the more developed coarsely granulated cells (early primary spermatocytes), which were formerly in the Wandschicht, shows a noticeable enlargement of its elements. The inner cells (likely round spermatids) are usually arranged in 3–4 layers and now have a size of 9–10.5 μm and nuclei of about 5 μm in diameter. On the inside of these cells one can see numerous spermatozoa heads that are already completely separated from their matrix and still contain coarse-grained masses.
VII. Stage (Fig. 1,7 and 7’)
The detachment of spermatozoa (spermatids) is complete.
The seminiferous tubules at this stage provide a very delicate and regular image.
Following the tunica propria are the unchanged Wandschicht, then the single-row outer cell layer, containing cells (pachytene spermatocytes), which are now 15–19 μm large with nuclei of 7–9 μm diameter.
This is followed by the inner cell layer usually consisting of three, rarely more, rows. The cells (spermatids) of this layer only have 8–9 μm diameter and 4–5 μm nuclei.
These cells are now followed by the spermatozoa (condensing spermatids), whose heads are all directed outwards, and whose tails are usually grouped into spiral shapes, as shown in Figs. 1,7 and 7’. The small sheaves, in which the spermatozoa (spermatids) were organized before the detachment, are dissolved, and the spermatozoa (condensing spermatids) are now bundled together in large clusters. The heads, especially their hook-shaped ends, are still surrounded by coarse-grained appendages.
The extensions, which, as we have explained in the description of the VI. stage, push the spermatozoa (spermatid) sheaves towards the center of the seminiferous tubule, are still present. On good sections one can see them blend into the coarse-grained mass, which sticks to the spermatozoa (spermatid) heads36. The pointed nuclei (of Sertoli cells) in the base of these extensions have grown larger, for while in the VI. stage they were only about 7–8 μm long, they are now 9 and 15 μm long and 6–7 μm wide at the base.
VIII. Stage (Fig. 1,8 and Fig. 10)
The detached spermatozoa are discarded, while the seminiferous tubule shows neither mature nor developing spermatozoa37 (spermatids).
The spermatozoa (spermatid) sheaves, which filled the whole lumen in the VII. stage of development, have now disappeared. There are only isolated spermatozoa (likely testicular spermatozoa) in the center of the seminiferous tubules, which is nearly free of any formed elements, but rather filled only with a coagula on hardened specimens. In the very center one can sometimes see a small sheaf of cross-sectioned spermatozoa.
Otherwise, this picture is not significantly different from that described under VII. The Wandschicht shows no difference, neither have the outer cells grown larger, although they are sometimes already arranged in two layers and narrowed at times, etc., in short, images that point to division processes.
The inner small cells (elongating spermatids) also have nearly the same size and arrangement as in the previous stage, but it is noticeable that the nuclei often strongly imbibe hematoxylin, much like the spermatozoa (spermatid) heads in stage I. The cells often appear in elongated shape, and the nuclei are then usually located eccentrically at one end38. Such cells are indistinguishable from lacerated lobes of spermatoblasts from seminiferous tubules in stage I39.
It is not uncommon to see cells whose nuclei are crescent-shaped and shrunken, cells that look swollen, and numerous Eiweisskugeln40.
The Keimnetz extensions also behave quite similar to the previous stage. Their ends, which disappear between the small inner cells, often look torn (see Fig. 11) and at times it seems that they are somewhat swollen and very coarse-grained.
Fig. 11.
A spermatoblast from a seminiferous tubule with spermatozoa in the VII. development stage. The split end shows two nucleus-like structures (new spermatozoa?). The base of the spermatoblasts is covered by a cell and also connects with a piece of the Wandschicht. At 300X magnification90.
A look at the drawings (Figs. 1, 1 and 8 as well as Figs. 4 and 10) suggests that the stage in question is easily confused with Stage I. Cross-sections that are not very thin may leave in doubt whether the objects under consideration are spermatoblasts showing first rudiments of spermatozoa (spermatid) heads or only extensions of the Keimnetz, which disappear between cells, whose nuclei are not easily distinguishable from spermatozoa (spermatid) heads by their size, shape, and appearance.
The I. stage immediately follows this VIII. stage, a point we shall return in more detail. [The reader will now also understand the remark made on p. 20941], that in the first stage, in addition to the spermatoblasts containing the first detectable traces of the spermatozoa (condensing spermatid) heads, the smaller cells (round spermatids) from the VIII stage may also still be partly present.
We have now reached the end of the description of the pictures, which can be distinguished from the cross-sections of testes of rats in heat whose seminiferous tubules develop spermatozoa everywhere. I believe, that I have provided ample details in outlining the stages, so that the numerical succession of pictures observed on different cross-sections of seminiferous tubules can be fairly safely concluded.
However, we can provide further important evidence for the genetic relationship of the described stages in the order as we have listed them. For it is possible to observe that the developmental stages follow one another in the course of one and the same seminiferous tubule in the order given. The tracts of seminiferous tubules exposed in longitudinal sections are too short to track how the various stages of development are distributed. However, it is easy to isolate individual loops of seminiferous tubules of testes hardened in Müller solution and alcohol.
When spread out, these loops are mostly 10–15 mm long. I stain the seminiferous tubules in blue wood extract, put them in clove oil or glycerin on the microscope slide and cover them with the coverslip. Since the seminiferous tubule is too opaque, I crush it by gradually increasing pressure on the coverslip. The tunica propria bursts open, the contents of the seminiferous tubules spread out flat, and the details are now discernible enough that one can decide what stage of development is present. It is natural that this raw process oftentimes causes the tissue elements to be displaced, but they do not move far away from where they originally lay, and the seminiferous tubule still constitutes a continuous band-shaped mass. Only at the margin of the specimen we find numerous completely isolated cells floating, such as spermatoblasts, parts of the Keimnetz, etc., which may be washed away further from their original location by the fluid.
Seminiferous tubules treated in this way show that the developmental stages, connected by gradual transitions, follow each other in comparatively short tubule sections. One can observe not only how the spermatoblasts with nuclei like spermatozoa (condensing spermatids) gradually transition through the nail and hook form into fully-formed spermatozoa sheaves, but also, and this seems particularly important, that seminiferous tubules filled with mature spermatozoa are immediately followed by the stages VIII, I, II, III, etc.
Of course, based on the method discussed above, I cannot give any more precise information regarding the absolute lengths of the individual developmental stages in the seminiferous tubules. However, it may be of interest to indicate how many stages occur within a certain section and to what length each individual developmental stage may grow.
I was able to determine the following:
A tubule of 10–14 mm in length may have 2–7 different stages of development. Stage VIII takes up the shortest length; in fact, this stage may be completely absent, so that stage I immediately follows stage VII, and that therefore a new spermatozoa (condensing spermatid) generation immediately follows a previous one. The tract lengths occupied by the other stages of development vary greatly, though no developmental stage seems to extend beyond 8 mm.
Now that we have described the development of spermatozoa (spermatids) in their main features, we want to take a closer look at the finer details.
We have seen that the first nucleus-like structure of the spermatozoa (spermatid) heads can be recognized in peculiar structures that emerge from the Keimnetz of the seminiferous tubules, the spermatoblasts. The question emerges where the nucleus-like spermatozoa (condensing spermatid) heads come from; whether they truly develop immediately as discrete rudiments in the spermatoblasts, or whether they result from division of a larger nucleus-like structure42. I strongly argue in favor of the former since there is no evidence to suggest the latter. In the beginning of my investigations, I assumed that the strikingly pointed nuclei43, which can be seen at the base of the spermatoblasts in the VI., VII., and VIII. stages of development, are directly related to the development of the spermatozoa (condensing spermatids). However, these nuclei do not change; they are still there when a new generation of spermatozoa (condensing spermatids) is already established, and then disappear only in the IV. and V. developmental stage, where the spermatozoa (condensing spermatid) heads move towards the base of the spermatoblasts. There is never a large nucleus at the end of the spermatoblasts, but rather several denser areas of the roundish protoplasm of 3–4 in diameter. I am not likely to have missed an earlier stage in the large number of preparations that I examined. In general, nucleus divisions are becoming a matter of concern44, as the number of relevant credible observations is decreasing with every passing year.
Once, I observed rudiments of spermatozoa (condensing spermatid) heads on a spermatoblast of a seminiferous tubule section in the VII stage of development45. As mentioned before, the connecting tissue between the now almost fully-formed spermatozoa (condensing spermatid) sheaf and the Keimnetz extends again, creating a new spermatoblast, which pushes the fully-formed spermatozoa (condensing spermatid) sheaf in front of it. The end of this new spermatoblast is now frayed into fine-grained threads, each holding a spermatozoa (condensed spermatid) ripe for detachment. On such finely split spermatoblasts, near the branching point of the threads, I now saw roundish nucleus-like formations of the appearance and size of the heads of the spermatozoa (condensing spermatids), as typically occur in the first stage of development (see Fig. 14). There was no sign of lobing of spermatoblasts, which occurs only after the threads connecting the mature spermatozoa with the new spermatoblasts have disappeared.
Fig. 14.
A torn lobe of a spermatoblast in the III. stage of development. At 570X magnification.
Thus, the first we see of the spermatozoa (condensing spermatids), are the rudiments of the heads as roundish or oval densifications in the empty protoplasm of the spermatoblasts. Then, in the first developmental stage, a number of hand-shaped lobes corresponding to the number of heads appears on the spermatoblasts.
Then the head transitions through the nail shape into the shape of the Aquilegia petal, whereby the future hook grows from the base of the head. A contrast between the different parts of the head emerges: Towards the pointed end it is shiny and imbibes strongly, while the roundish part is paler, comparatively large and imbibes less strongly (see Fig. 14). At the same time, the lobes of the spermatoblasts grow significantly longer and narrower.
When they have reached slightly more than half the length of the future midpieces of the spermatozoa (condensing spermatids), the development of the tails begins46, while the midpieces start differentiating shortly thereafter (see Fig. 15). As mentioned above, the tails grow out from the upper end of the lobes of the spermatoblasts. On isolated lobes one can frequently see the tail growing out from the end of the lobe, while inside the lobe itself there are only larger and smaller granules, with no sign of a tail.
Sometimes the thread47 does not grow out from the end of the lobe, but from the side, and the end of the lobe may well partially cover the thread. This may be related to the fact that the spermatozoa (condensing spermatids) do not develop exactly under the same conditions. Their position towards the spermatoblasts may differ and lobes may branch out higher up or further down, etc.
The differentiation of the midpiece occurs within the lobe itself, while it narrows and finally disappears. The insertion site on the head becomes visible first as a shiny stripe (Fig. 16), which gradually disappears towards the tail end of the lobe. This stripe grows progressively longer, until it finally merges with the tail, while simultaneously the granular substance, within which the development occurs, disappears except for some remains found especially at the end of the midpiece. However, such appendages can also exist elsewhere in the midpiece, especially near the head (Fig. 17). They may remain attached to the spermatozoa (condensing spermatids) until after the detachment. Indeed, they are sometimes found in spermatozoa from the epididymis and vas deferens (Fig. 18).
Fig. 18.
Spermatozoa from the epididymis of the rat. The midpiece and tail are barely separated from each other, still showing an appendage at the point of connection. At 570X magnification.
Just before detachment of the spermatozoa (condensing spermatids), the hook of the head is still heavily surrounded by a granular substance. Likewise, the extension of the spermatozoa (condensing spermatid) head and the midpiece are often still connected by a granular mass, which disappears only after the detachment (Fig. 17).
According to these observations, we must imagine that the spermatozoa (condensing spermatid) head and midpiece are formed through internal densification of the protoplasm of the spermatoblasts, while the tails develop from outgrowths of the surface layers of the individual lobes.
We will now explain the processes involved in the detachment itself in more detail. It is initiated by the growth of the spermatoblasts, destined to become a new generation of spermatozoa (condensing spermatids), and which push the sheaf of mature spermatozoa in front of them. As it grows, the sheaf dissolves somewhat from the base, with the new spermatoblasts dividing into granular filaments, each carrying a spermatozoon. The ends of the tails, which, as already mentioned, always run in the direction of the longitudinal axis of the seminiferous tubules, are united, so that during the detachment the sheaf represents a curved conical figure whose base consists of the heads, while the tip is made up of the tail ends of the spermatozoa (condensing spermatids). During detachment, the tails of the spermatozoa (condensed spermatids) precede the rest (see Fig. 2) and are always directed against the descending developmental series. It must be emphasized here that the developmental stages of the rete testis always follow in ascending order towards the ends of the seminiferous tubules. This is confirmed by examining the seminiferous tubules isolated from the rete testis in the manner previously indicated. Therefore, the detached spermatozoa, which are finally positioned completely parallel to the longitudinal axis, must move in between the spermatozoa (condensed spermatids) in the process of detachment, and pass through them in the center of the seminiferous tubules. This is probably the reason for the rat’s spermatozoa regularly appearing as spiral-shaped figures on cross-sections (Figs. 1, 7 and 7’). For if one imagines that a group of detached spermatozoa cones pushes helicoidally forward – as it must if the detachment does not occur completely simultaneously and symmetrically at the different points of a cross-sectional plane – one can easily understand of the origin of the spirals, which are present in cross-sectional images from the stages V through VII.
It may seem conspicuous that parts of the seminiferous tubules carrying spermatozoa (condensing spermatids) often contain no or only few mature spermatozoa (spermatids) in the early stages of development. This can be explained by considering that the detached spermatozoa first always fall into the oldest stages of development, which also extend the furthest. Thus, while the spermatozoa (condensing spermatids) move through the older stages, the younger spermatozoa (condensing spermatids) can move forward according to their own development. This adaptation allowing the detached spermatozoa into the older stages of development, probably also has a mechanical significance. Namely, the tails of the developing spermatozoa (condensing spermatids) (see Fig. 2), which are all pointing downwards in relation to the course of the seminiferous tubule, may act in the manner of a tube valve. They probably allow the downward movement of the free contents of the seminiferous tubules, but would immediately stand up and create a seal, if the movement were to occur upwards. Analogous to valve-guiding vessels, therefore, any pressure exerted externally on the seminiferous tubules can effect the removal of the contents48.
I do not consider these considerations to be superfluous, because it is difficult to imagine how spermatozoa, immobile in the testes and always preceded by the tail, emerge from the muscle-free seminiferous tubules49. It is possible that the changing filling of the vessel can have an effect on the transport of its content.
There is one more element influencing the detachment of spermatozoa (condensed spermatids), namely the cells found in the seminiferous tubules, at which we now want to take a closer look.
It has already been mentioned that the Wandschicht of the seminiferous tubules consists of two substantially different components: the Keimnetz and the granular cells50 found in the gaps of the same, which are almost completely filled with their nuclei on stained preparations, so that one might be inclined to also consider them as free nuclei. As for the Keimnetz itself, it looks the same in all stages of its development, apart from its extensions, the spermatoblasts. It is a network of short, fairly broad, irregular bars, which often become very thin at their edges. The bars contain nuclei, which are usually somewhat elliptical, often quite round, and are – even on very fresh specimens – sharply contoured, and they typically include a very shiny grain51. Inside the bars of the network, these nuclei are usually located at the nodal points. The bar substance itself appears fresh from the shine typical for the animal protoplasm and contains numerous granules, which can reach a significant size. Large shiny grains are often clustered around the nuclei so densely that it reminds of an embryo cell filled with yolk granules. The granules do not dissolve in acetic acid, but dissolve rather easily in diluted sodium hydroxide solution.
Thus far, I have not been able to represent cell borders in the Keimnetz. It seems that we are dealing with a fusion of the cells, similar as, for example, in the so-called cloudy epithelium in the tortuous tubules or in certain developmental stages of the embryonic connective tissue.
The Keimnetz usually covers the inner surface of the tunica propria as a single layer in rats and mice. Only rarely do the anastomosing bars go inward and form a second layer that, when observed on a cross-section, rests on the first one like an arcade. This structure of the Keimnetz, which is rare and only vaguely present in rats and mice, is much more pronounced in other animals and in humans. Here, the network routinely represents a sponge-like tissue projecting far into the interior of the seminiferous tubules, which we will return to later.
The spermatoblasts grow out of this Keimnetz, usually, but not always, in places where it shows nuclei. As mentioned, the nuclei (of Sertoli cells) change in a peculiar manner during growth: They are often strongly elliptical and even pointed at their central end. I do not know the significance of this – this phenomenon is not directly related to the development of spermatozoa (condensing spermatids). The occurrence or absence of the strongly elongated or pointed nuclei seems to be related to the following circumstances: As has been mentioned, the new spermatoblasts push the spermatozoa (condensing spermatids) in front of them and in the process lengthen significantly. However, not all spermatoblasts are of the same length. There are also very short ones, where the spermatozoa are positioned near the Wandschicht right from the beginning. These latter ones usually do not contain elongated nuclei. As far as the presence of long and short spermatoblasts is concerned, I believe that it can be explained by the fact that the latter grew out of places where there were no spermatoblasts and therefore no spermatozoa sheaf was detached.
The Keimnetz is quite stable in form and shows no significant variations in shape apart from the changing pictures of its extensions, the spermatoblasts. It is worth noting that the Keimnetz generally appears less granular and thinner in places where the spermatozoa (condensing spermatids) are close to their formation (stage V).
The cells located in the gaps of the Keimnetz (spermatogonia and primary spermatocytes) undergo a series of metamorphoses. We have seen that by the time the nail-shaped rudiments of the spermatozoa (condensing spermatid) head (stages II and III) develop a curved tip, the coarsely granulated cells gradually rise above the Keimnetz and grow rather rapidly (see Figs. 1, 5, 6, and 7).
Fig. 6. Begin of the III. Development Stage.
In the Wandschicht b, the dark granulated cells (c’) (primary spermatocytes) begin to rise above the Keimnetz (Sertoli cells), the cells (c’) have become smaller and more numerous compared to the previous stage.
We can assume with great certainty that the coarsely granulated cells (early primary spermatocytes) of the Wandschicht do gradually transition into the large cells (pachytene spermatocytes) of the middle layer and then continue to divide and change into the small cells (round spermatids), which we find in the seminiferous tubules with almost fully-formed spermatozoa (condensed spermatids). The transformation of the coarsely granulated cells (spermatogonia and primary spermatocytes) into the cells of the middle layer (pachytene spermatocytes) cannot easily be proven. Rather, it can only be concluded by comparing cross-sections of seminiferous tubules at different stages of development. If one carefully and repeatedly compares the images of seminiferous tubules in the II.-VII. developmental stages, one can see how little by little the 5.2–5.5 μm cells of the Wandschicht gradually increase in size, develop a distinct non-imbibed boundary layer around the nucleus, and, after they have risen above the level of the Keimnetz, become 10, 12, 17 μm (likely preleptotene, leptotene, and zygotene spermatocytes), finally 19–24 μm cells (pachytene spermatocytes), which surround nuclei of 7–9 μm diameter. When the cells have reached these dimensions, they remain unchanged for the time being, only to divide and multiply52 during the development of the next generation spermatozoa (condensing spermatids). While the mentioned cells metamorphose in this way, the Wandschicht (in the gaps of the Keimnetz) shows a replacement of the migrated cells, first in the form of nucleus-like structures (likely referring to spermatogonia), which are very close to the tunica propria and often flattened. In the VII. stage, the Wandschicht already shows complete replacement, and the cellular structures, although all of them ultimately originate from the same source, are now distinctly separated into three layers (spermatogonia and early primary spermatocytes, pachytene spermatocytes, and round spermatids), since there are three generations of cells next to each other.
In order to clarify these somewhat complicated relationships of the cells in the seminiferous tubules, the following table (Table 1) shows in form of a graph the growth over time and the subsequent reduction in size of the cells as a result of the division.
On the ordinate, the sizes of the cells in are plotted micro millimeters (μm), while the abscissa shows the developmental stages.
The dashed line, which begins at 21 μm, represents the size changes of the cells, which rest on the Wandschicht in the first development stage. One can see that they become progressively smaller (as a result of the division), and finally disappear in the VIII. developmental stage or in the I. stage of the following generation of spermatozoa.
The solid line corresponds to the coarsely granulated cells (spermatogonia/early primary spermatocytes) of the Wandschicht, which after their migration gradually transform into the large cells (pachytene spermatocytes), which rest on the Wandschicht at the beginning of the development of the next generation spermatozoa (spermatids). This line crosses with the first approximately in the V. developmental stage, so that around this time the cells of the first and second generation are almost equal in size.
The dotted line corresponds to the third cell generation, which suddenly appears in III. or IV. developmental stage in the Wandschicht.
Now the question becomes, where do the cells of the Wandschicht, which migrate and are replaced again, come from?
Before expanding on this point, I will make a few observations about the appearance and behavior of these cells in the fresh state.
Since rats were not always readily available, I used the testes of mice for observations on fresh preparations, which behave like the testes of rats in all essential respects. The pieces of testes taken from a newly killed animal were brought under the microscope, slightly plucked in the aqueous humor of the mouse and protected by glass splinters from the pressure of the coverslip. This last precautionary measure is necessary because all cellular structures of the testis are extremely fragile. If one intends to observe undamaged pieces of seminiferous tubules, it is indispensable because otherwise the cover glass crushes them infallibly.
The first step is to see the cells of the Wandschicht in situ.
A fresh seminiferous tubule of the mouse is transparent enough that it reveals some details. An optical longitudinal cross-section shows the tunica propria as a double contoured stripe, in which the nuclei are recognizable – only vaguely on a fresh preparation, more clearly with time. This is followed by the Wandschicht, which appears jagged, and apart from the numerous shining granules, the sharply contoured pale nucleoli-containing nuclei (of Sertoli cells) of the Keimnetz are recognizable. There is not much evidence of the coarsely granulated cells (spermatogonia and early primary spermatocytes) of the Wandschicht, which we can see on the stained sections. At most, it seems that here and there roundish shiny masses are delimited. Inwards, a shiny granular mass, in which no nuclei are detected, appears. If acetic acid is added, the whole seminiferous tubule becomes very dark, so that observation becomes difficult. However, it can now be observed in the optical longitudinal cross-sections, that numerous coarse-grained coagula of round shape (nuclei) have appeared in the Wandschicht. Likewise, the same coarse-grained coagula (nuclei) are seen in the larger roundish masses that follow the Wandschicht inward. Since Henle’s investigations it is well-established that there are two types of cells in the seminiferous tubules, those developing coarsely granulated nuclei when adding acid and those containing smooth nuclei. We can now conclude that cells with coarsely granulated nuclei occur both in the Wandschicht (spermatogonia and early primary spermatocytes) and in the subsequent middle layer (later primary spermatocytes).
If we now look at the isolated cells, which swim around in great quantities on each scrape preparation, we can see that the cells, which develop so-called granulated nuclei in acetic acid do not show nuclei in completely fresh preparations. They are either finely or coarsely granulated and have an appearance that strongly reminds of the appearance of white blood cells. It is rare to notice pale vacuole-like spots on them. If they rest for some time, the granular mass collects within them, while the periphery become more pale. When acetic acid is added, a coarse-grained, shrinking, roundish mass forms in the middle, while the peripheral layer swells and turns very pale. The cells show the same differences in size as we find in the cells of the Wandschicht and the inner layers on cross-sections. The cells also show the amoeboid movements first described by La Valette St. George on testis cells53.
Returning to the question of where the coarsely granulated cells of the Wandschicht (spermatogonia and early primary spermatocytes) originate, the assumption that they originate from the blood seems the most plausible to me. The coarsely granulated cells of the Wandschicht (spermatogonia and early primary spermatocytes) do not differ from white blood cells. They are the same size, they stain almost exactly the same in carmine and hematoxylin, they look just like them when fresh, and they react to acetic acid like these. By the way, on scrape preparations it is not possible to decide with certainty, whether one is looking at cells of the Wandschicht or white blood cells. Only the larger cells (pachytene spermatocytes), whose size far exceeds that of white blood cells, and which reach a diameter of 16–24 μm, are certain to belong to the seminiferous tubules. These now show amoeboid movements. The types of movement differ, of course. The testicular cells grow roundish hyaline humps, while the white blood cells usually develop sharp, threadlike extensions. However, as the images of La Valette show, even the testicular cells sometimes show movements akin to that of white blood cells. In short, everything indicates that the cells in question are very similar to the blood cells, and that the smaller ones cannot be distinguished from them. Furthermore, it is now necessary to search for the source of the cells of the Wandschicht outside of the seminiferous tubules. They move inward without any trace of division, and as they migrate out of the wall, a new layer of very similar formations (spermatogonia) appears at the tunica propria. It is possible that the new cells come either from the tunica propria or from the Keimnetz. Both assumptions are highly unlikely, though it cannot be denied that white blood cells from the lymphatic glands containing the seminiferous tubules are able to penetrate through the thin tunica propria into the interior of the seminiferous tubules54.
Tomsa has ascertained that white blood cells also occur in the normal testicular lymph. I do not want to deny that the established hypothesis is very far from being proven, but in order to make any observations about the replacement of the coarse-grained cells of the seminiferous tubules, I must accept it.
We now turn to the cells that appear in the inner layers of the seminiferous tubules. It has already been mentioned that one can clearly observe how the large cells (pachytene spermatocytes) sitting on the Wandschicht show clear signs of division processes at the time when the spermatozoa (condensing spermatid) heads develop their unmistakable form (stage II and stage III).
Narrowed cells with two nuclei, cells connected only by a thin connecting bridge, and cell chains, especially in the more advanced stages of development, are easily recognizable. However, I could never see anything that suggested a nuclear division55, while La Valette also claims to have observed the same.
By comparing the seminiferous tubules at different stages of development one can follow step by step how the cells become smaller and smaller as they divide and multiply, so that finally the 16–24 μm cells of the first stage of development (pachytene spermatocytes) turn into a large number of smaller, 7–9 μm cells with nuclei of 4.3–5.2 μm in diameter (round spermatids). These division processes also bring about a change in the nature of the nuclei, which take on a smooth appearance. Fresh specimens, at least of mice, show little evidence of the nuclei, even in the cells, which later contain smooth nuclei. The nucleus is either suggested by a weak, smooth contour, or is not visible at all. Only after the preparation has rested for some time, the nuclei, often more than one, emerge more clearly. They then appear smooth and delimited by a single contour, and occasionally contain one or two larger grains. As long as the nuclei remain invisible, the smooth-nucleated cells are not always easily distinguishable from the cells with granulated nuclei, although generally it can be said that the cells with smooth nuclei show a finer-grained protoplasm.
When treating these cells with acetic acid, a strong shiny mass collects on the periphery of the nucleus56, so that the nucleus is doubly contoured. Inside the nucleus, some grains usually appear. If these grains occur in great quantities, it suggests the transition to the cells with a coarse-grained nucleus57.
On hardened preparations, the nuclei of the inner cell layers (round spermatids) show consistently sharply contoured and pale imbibing nuclei in the higher stages of development. We can only say with certainty that the spermatozoa do not emerge from these cells, although the latest handbooks still describe them as precursors of spermatozoa58. It seems to me, however, that their significance is partly secretory, partly mechanical59.
One can isolate pale, nucleus-free, homogeneously looking spheres, so-called Eiweisskugeln (or glass-like spheres60), on the various sections of the seminiferous tubules in different stages. Such spheres are also found on cross-sections, where the same imbibe, though weakly. These spheres are most abundant in the lumen of the tubule, and are most easily seen in the first stages of development, when the lumen is not yet filled with spermatozoa (condensing spermatid) tails. However, they are, as it seems, most numerous in the later stages of development, and one can see images, especially in the VII. and VIII. stages, that indicate that the internal cells create these glass-like spheres. It is not uncommon to observe that the cells are spherical and very pale, while their nucleus is severely shrunken and parietal61. Such cell shapes and Eiweisskugeln lie next to and above each other. So it seems that the cells eventually die by helping to form the seminal fluid of the testis62.
However, this does not always happen in the tract of the seminiferous tubule in which the cells have formed. I have observed on the testes of some mice that cells completely resembling the inner cells of seminiferous tubules with almost mature spermatozoa can appear between fully-formed spermatozoa (condensing spermatids) and those in the process of detachment. This occurs in such a way that the cells occupy the center of the tubule and are completely separated from the rest of the cells found in seminiferous tubule by the spermatozoa (condensing spermatids), which are in the process of detachment and very close together. This seems to suggest that the cells detached in the upper part of the seminiferous tubules and then, as they move forward, insert themselves between the spermatozoa (condensing spermatids), which are about to be detached. However, this is a phenomenon that seems to occur only on testes whose seminiferous tubules only partially produce spermatozoa, a topic I will return to.
It was previously mentioned that the free cells of the seminiferous tubules are of importance for the removal of spermatozoa. The thin, splitting extensions of the Keimnetz63 (which also form the rudiments of new spermatoblasts64), to which the spermatozoa sheaves about to detach are connected, are unlikely to be suitable for transporting the spermatozoa to the interior of the seminiferous tubules. However, it is quite easy to imagine that the free cells, which multiply abundantly at the time of detachment of the spermatozoa (condensing spermatids), push the spermatozoa (condensing spermatid) sheaves inwards, so that the extensions of the Keimnetz serve as guides to ensure a regular detachment process65. Once the spermatozoa (condensing spermatids) have detached, the multiplying cells (round spermatids), which are transformed into Eiweisskugeln (residual bodies) and seminal fluid, contribute to the ongoing movement of the spermatozoa. This is achieved through the expansion of the walls of the seminiferous tubules by increasing their content, which, by virtue of their elasticity, causes the contents to be moved on to the nearest less strained part of the tubule.
The reader will have noticed that nothing has been mentioned yet about the so-called polynuclear66 cysts in which Kölliker once had the spermatozoa (condensing spermatids) develop and, at least in part, still does67. Henle has already pointed out that the multinucleated cysts are missing in the testes of humans and are not consistently found even in animals whose seminiferous tubules are filled with mature spermatozoa (condensing spermatids). These observations are consistent with my own. In the rats I examined, whose seminiferous tubules showed developing spermatozoa everywhere, I have never observed so-called multinucleated cysts, at most one can see cells with 3–4 nuclei in the V. and VI. developmental stage, but never cells with 8–12 and more, smooth nuclei. In contrast, I often saw such structures in mice whose seminiferous tubules sometimes produced no spermatozoa. It is, incidentally, not easy to establish the existence of the multinucleated cysts on sections of hardened preparations, and I have only had two cross-sections where I thought to have seen them. The cysts were found among the detached cells, which, as mentioned earlier, sometimes move in between the almost mature spermatozoa (condensing spermatids). However, so far I have not been able to find the origin of these peculiar cells. Perhaps they are only accidental formations, which occur here and there as part of the vigorous cell production in the seminiferous tubules68. It is certain that they, too, have nothing to do with spermatozoa (spermatid) development69.
The previously reported findings about the construction of the seminiferous tubules refer to observations on animals, in which the seminiferous tubules produce spermatozoa everywhere. Only the most recent conclusions about the presence of cells with smooth nuclei between detaching spermatozoa and about the so-called multinucleated cysts refer to the testes of mice, which also show numerous seminiferous tubules lacking any trace of spermatozoa development70.
Let us now turn to the seminiferous tubules of those testes in which spermatozoa cannot be observed. The images of such objects are, of course, less diverse than those previously described, but it would be a mistake to believe that all seminiferous tubules show the same appearance. One could say that the differences are the same as in the various stages of development of spermatozoa-producing seminiferous tubules, apart from the spermatoblasts and their products, but with the difference that a cell-free lumen never occurs.
On cross-sections of the testis of a mouse not in heat71, you can see the following:
The Wandschicht just touching the tunica propria is delimited on the inside by a somewhat uneven contour. As in the previously described seminiferous tubules, we can distinguish coarsely granulated cells and the Keimnetz. The existence of the latter is best confirmed by way of scrape preparations. Its substance is not very granular, and the pale, nucleoli-containing nuclei are almost always round in shape and generally more densely packed than in seminiferous tubules which develop spermatozoa. Sometimes it seems that the Keimnetz consists of separate cells, but I could not verify this assumption.
The Keimnetz, from which there are no extensions into the interior of the seminiferous tubules, behaves in the same way in all seminiferous tubules. This is different with the other cells, for which we can distinguish about four different cross-sectional images, which we list again in the order we think they are genetically connected.
Coarse-granulated small cells (likely early primary spermatocytes) are located in the Wandschicht, followed by one or two layers of large cells (pachytene spermatocytes). The lumen is partly filled by small cells with pale nuclei (round spermatids), separated from the large cells by a sharp border, and arranged in clusters.
The Wandschicht is free of coarsely granulated cells and therefore consists only of the Keimnetz (Sertoli cells). Inwardly follow 2, 3 or more layers of cells of almost equal size, which are arranged like radii. The outermost layer is characterized by the fact that the nuclei are coarsely granulated and imbibe strongly. The center of the tubule usually contains a cluster of small cells sharply separated from the other cells.
Coarse-granulated cells are again visible in the Wandschicht, as in stage A; the radii-like cells we saw in the previous stage now often extend to the center of the tubule, leaving few or no small cell clusters in the middle. The cells decrease in size from outside to inside, while the cell layer following the Wandschicht is characterized, as in the previous stage, by the behavior of the nuclei.
The cell layer following the Wandschicht has undergone a marked enlargement of its elements in comparison with the previous stage, while the cells of the inner layers have become smaller and often no longer exhibit a radial arrangement.
It is unmistakable that the described pictures are very similar to what we have seen with respect to the cells in the seminiferous tubules that produce spermatozoa. Looking back at the developmental stages described earlier, we can relate stage I and stage II to stage A72 (many stages appear to be lumped together here, IX-XIV and I-III in the current rat staging system), stages III and IV to stage B (appear to be stages IV-VI in the current rat staging system), stage V to stage C (appear to be stages VI-VII in the current rat staging system), and stage VI, VII, and VIII to stage D (these are clearly VIII-IX in the current rat staging system)73. It should be noted that the absence of the coarsely granulated cells of the Wandschicht in stage B (likely differentiating spermatogonia) can be observed much more clearly than in the analogous stages of spermatozoa-containing (condensing spermatid-containing) seminiferous tubules, where we often see new cells in the Wandschicht, where the old ones have hardly raised themselves above the Keimnetz.
It should be noted that in all developmental states of the spermatozoa-free (condensing spermatid-free) seminiferous tubules clusters can be found in the center of the cells in which, as is most probable, detached and transported cells which have been led away from their place of production, are present.
It seems to be the rule that at least some of the cells do not die where they originated, but move through the seminiferous tubules for a certain distance. Indeed, these cells can even reach the epididymis and the vas deferens. On cross-sections of the epididymis tubules, one often finds the lumen of the ducts lined with ciliated epithelium filled with the same cell-clusters as we observed in the lumen of the seminiferous tubules.
On sections of the epididymis of a rat, whose testes produce spermatozoa everywhere, one observes no morphotic elements in the lumen of the tubules other than countless spermatozoa.
These observations also lead to a better understanding of the fact that clusters of detached cells may get between the spermatozoa (condensing spermatids) in a testis whose seminiferous tubules only partially produce spermatozoa, as previously stated.
It is very likely that the different behavior of cells in seminiferous tubules that produce spermatozoa versus those containing only cells is indirectly related to the development of spermatozoa.
One might consider that the cells, which rapidly disappear in the former case, serve as nutritional material for the developing spermatoblasts.
So far, I have only discussed the construction of seminiferous tubules in rats and mice, and I must confess, that the findings about the testes, which produces spermatozoa only partially or not at all, remain somewhat fragmentary. I hope to fill this gap once I have more rat testes at my disposal than is currently the case.
I will now turn to the discussion of the construction of the seminiferous tubules of some other animals and humans, noting in advance that my studies of these objects had the sole purpose of determining, whether the essence of what was found in rats and mice was also valid for other animals and humans.
This limitation was partly inevitable, as the study offers many obstacles that are extremely difficult to overcome, as I pointed out earlier. In my study, I examined guinea pigs, rabbits, hedgehogs and dogs.
In all these animals I was able to find the Keimnetz, the spermatoblasts, in which the spermatozoa (condensing spermatids) develop to in groups of 6–12, as well as the adjacent cells with coarse-grained and smooth nuclei (spermatogonia and primary spermatocytes). As far as these cells are concerned, similar differences in cross-sectional images were found in the animals mentioned above, as in rats and mice. However, the Keimnetz and the spermatoblasts behave differently. In guinea pigs the former is still quite similar to that of rats and mice, for it usually forms only a single, relatively thin layer, which rarely pushes arcade-like arcs inward. The spermatoblasts are also very similar to those of the rat, but their stems sitting on the Keimnetz are usually much narrower than the latter, and the heads of the spermatozoa are not close to the wall, but always remain at a considerable distance from it.
In rabbits, the Wandschicht is not clearly delineated from the remaining contents of the seminiferous tubules. This results from a stronger development of the Keimnetz, which contains the cells in its gaps, while they sit directly on the Wandschicht in the rat. Thus we now have a network that has developed in thickness. As a result, the spermatoblasts lie more towards the center of the seminiferous tubules, causing the heads of the spermatozoa to be near the center of the tubule in all developmental stages.
Incidentally, it is easy to see cross-sections of the spermatozoa (condensing spermatid) heads in groups of 8–12. More rarely they can be observed swimming in a coarsely granulated mass near the parts of the Keimnetz touching the tunica propria.
In dogs, the Keimnetz is less developed once again. I have rarely encountered protruding arcs that extend from the Wandschicht to wrap around the cells. The spermatoblasts are very long, have thin stems, and carry 10–12 spermatozoa (condensing spermatids). In hedgehogs, the Keimnetz is also poorly developed and forms only a trace of the tunica propria, as in the rat.
The spermatoblasts are similar to those of the rat, and the heads of the spermatozoa can move very closely to the Wandschicht in the middle developmental stages. In contrast, I found the Keimnetz of the cat, of which I only examined a testis without spermatozoa, very developed. It was formed by a mesh, which occupied more than one third of the radius of the seminiferous tubules in thickness, and showed a particularly strong development of the radial bars.
As far as the construction of the seminiferous tubules of man is concerned, we cannot ignore the tunica propria, as there are very contradictory statements about its structure. In Henle74 they consist of lamellae, which are made up of flat, rhombic, nucleated scales (layer of 3–5 PTMs). According to him, these scales seem to merge in the inner layers and also to lose their nuclei. Kölliker,75 however, claims that the wall of the seminiferous tubules consists of a fibrous skin and a structureless membrane propria lying inward of the same. Leydig, Frey, and Hessling essentially agree with Kölliker, while La Valette confirms Henle’s findings. According to Tommasi76, the sheath of the seminiferous tubules is formed firstly by the epithelial cells lining the lymphatic lacunae, secondly by the membrana propria, which is, thirdly, followed by the inner epithelium consisting of small polyhedral cells with large nuclei.
It was previously mentioned that the tunica propria of the rat consists of flat polyhedral cells. The same seems to be true for humans. One can easily isolate pieces of tunica propria from seminiferous tubules stained with bluewood extract, and then find that it consists of a fine-grained skin containing numerous nuclei. The skin often shows short, irregular streaks, which run in different directions. I believe that they correspond to folds. The nuclei are often oval, often irregular or showing a narrowing in the middle. Here and there the nuclei are close to each other. I was unable to get any results treating the seminiferous tubules of man with a silver nitrate solution.
The streaks, which are visible on the sheath of seminiferous tubules, do not have the same function everywhere. The interstitial connective tissue, which can be visibly fibrillary in humans, comes into contact with the tunica propria of the seminiferous tubules in many places and can occasionally give it a denser covering of fiber and cells. Mostly, the streaky appearance of the wall of the seminiferous tubules seems to be due to an irregular folding of the tunica propria. The fact that the tunica propria forms folds and consequently appears streaky on sections is also observable in dogs and rabbits. That is, in animals whose testes are similar to the testes of humans in terms of the coiling of the seminiferous tubules, the formation of lobes, and the behavior of interstitial connective tissue. I was able to confirm with the help of silvering that the tunica propria of the seminiferous tubules in rabbits is of the same composition as in the rat. I therefore believe that the skin of the seminiferous tubules is not actually fibrous, nor is the membrana propria without structure.
I do not doubt that the seminiferous tubules of humans and animals have a tunica propria consisting of only flat polyhedral cells. It is possible that it is covered with endothelial cells, wherever it is directly adjacent to the lymphatic spaces, though I lack specific knowledge to confirm this assumption.
In terms of the contents of seminiferous tubules, the development of a Keimnetz on the testes of humans is especially noteworthy. It presents a strongly developed network, which can extend to over half of the radius inwards. The radial bars are usually highly developed, while the tangentially or irregularly positioned bridges, often arched between radially extending bars, usually remain narrow. The whole Keimnetz is, as it were, a sponge-like tissue with cells in its gaps. The sharply contoured, nucleoli-containing nuclei of the Keimnetz are numerous and usually oval. I only once saw clearly lobed spermatoblasts with spermatozoa (condensing spermatids) on one subject, the testes of a 60-year-old suddenly deceased worker. They were connected to the radial bars of the Keimnetz and contained 8–10 spermatozoa (condensing spermatids).
The Keimnetz contained numerous yellow-colored grains, which I did not notice on the testes of young people. The cells, which rest on the Wandschicht in the first stages of development of the spermatozoa (condensing spermatids) in the rat, lie in the gaps of the Keimnetz in humans. The small cells, which ultimately emerge from the cells mentioned, have nuclei, which impair the clarity of the cross-sectional images as a result of their size and their imbibing power and may lead to their confusion with the heads of spermatozoa (condensing spermatids).
In addition to coagula and glass-like spheres (Eiweisskugeln, residual bodies), clusters of detached cells are also frequently visible in the lumen of the seminiferous tubules. The testes of people, which I was able to examine, did not contain spermatozoa (condensing spermatids) in all seminiferous tubules.
I will now summarize the most important results of my observations:
We have seen that the seminiferous tubules are composed of two essentially different components, one of which is responsible for the development of spermatozoa, the other for the formation of fluid contained in the seminiferous tubules and perhaps also for the supply of nutritional material to the spermatozoa.
The first component, the so-called Keimnetz, forms a wall cover of the seminiferous tubules consisting of fused cells (PTMs), which appears differently in different animals. However, in general it grows extensions (spermatoblasts) into the interior of the seminiferous tubules, in whose widened lobed ends the spermatozoa are formed in groups of 8–1277.
The formation of spermatozoa (condensing spermatids) occurs endogenously in the lobes of spermatoblasts (clusters of round spermatids) without the involvement of a cell nucleus78. The head and midpiece of the spermatozoa must be regarded as protoplasm densifications of the spermatoblast lobes, while the tail emerges from their upper layers.
The spatial distribution of the stages of development of the spermatozoa (condensing spermatids) is such that only one stage of development is present in one cross-section, whereas the successive stages of development occur relatively closely together in the entire course of a seminiferous tubule. Therefore, the complete series of developmental stages can be found in numerous repetitions over the course of one and the same seminiferous tubule.
During the detachment of the spermatozoa (condensed spermatids), which are always positioned with their heads against the wall of the seminiferous tubules during development, the tails precede the head. The detached spermatozoa always first move into sections of seminiferous tubules containing the most advanced stages of development. This explains why in a testis that produces spermatozoa everywhere, those parts of the seminiferous tubules containing the first stages of development of the spermatozoa may have a lumen that is completely free of morphotic constituents.
The second component in the seminiferous tubules consists of cells, which resemble white blood cells and probably migrate into the seminiferous tubules from the lymphatic chambers, then greatly enlarge, become less granular, and obtain smooth nuclei. Finally, by way of continued cell division, they produce a brood of small cells (round spermatids) that finally dissolve to form glass-like spheres (Eiweisskugeln, residual bodies). These metamorphoses occur simultaneously with the development of the spermatozoa (condensing spermatids) in such a way that in the first stages of development two, in the latter three, cell generations occur parallel to one another and move consecutively from the outside to the inside.
The processes occurring in the spermatozoa (condensing spermatid)-free seminiferous tubules are similar, with the noteworthy difference that the cells resulting from division remain longer, and in some cases arrive unchanged in the vas deferens.
Finally, I will add some remarks about recent investigations into our object:
Putting older literature aside, we shall consider studies published since the investigations of Henle on the structure of the seminiferous tubules and the development of spermatozoa of mammals.
Henle’s investigations first established that the then prevalent ideas about spermatozoa development, especially based on Kölliker’s investigations, are untenable. Kölliker first asserted that the spermatozoa formed endogenously in the nuclei of multinucleated cells and later that they formed endogenously in mononuclear or multinucleated cells. According to him, the sperm heads represented transformed nuclei, and the tails were coiled inside the cells, then grew out of the heads and later broke through the cell, a view which Kölliker largely still holds. Regarding the spatial distribution of the developmental stages, Kölliker is of the opinion that they move from the outside inwards, so that the innermost cell layer would consist partly of so-called sperm cysts, which directly produce the spermatozoa.
Henle, however, drew attention to the fact that there are never any cells on cross-sections of hardened testes that show coiled tails inside. He thus denies the normal occurrence of coiled tails, which occur only as a result of the action of certain solutions, and considers the cell body essential to the formation of tails. Henle also rightly emphasizes that the advanced stages of development of spermatozoa (condensing spermatids) are not always found in the center of the tubules but also frequently between the more peripheral (Sertoli) cells79. He thus concludes that the spermatozoa (condensing spermatids) are developing in one and the same place, without the originating cell moving inward. Regarding the spatial distribution of the developmental stages in the seminiferous tubules Henle only states that the previous investigations have not yielded definitive results.
Henle has also found that there are two types of cells in the seminiferous tubules. According to him, the spermatozoa (condensing spermatids) develop from the cells with smooth nuclei, though he admits that the existing investigations only allow for assumptions.
Significant progress in understanding the structure and development of spermatozoa (spermatids) was substantiated by Schweigger-Seidel’s80 research, which found that what had been referred to as the tail of spermatozoa consisted of two essentially different pieces, the midpiece and the actual tail. This made Kölliker’s claim of the tail growing out of the head of the spermatozoa (spermatid) very unlikely81. Schweigger-Seidel also concluded that the midpiece should be considered as transformed cell protoplasm, while the tail represented the cilium of a ciliated cell. Meanwhile, Kölliker undertook to support his old conception by new observations, and, in the most recent edition of his handbook, draws spermatozoa (condensing spermatids) of bulls showing tails growing out of the head through a tube emanating from it. It is not my place to deny the truth of this claim. I can only say that they do not quite agree with the observations I made on the spermatozoa (condensing spermatids) of rats and mice, unless it is assumed that they were spermatozoa (condensing spermatids), in which the tail and midpiece were already formed.
I believe that I can expand on one point, however. Kölliker says, that on one end the head of the spermatozoa (condensing spermatid) grows into a conical shape (proliferating nucleus contents) within the tube, from which the thread (growing flagellum) protrudes. He suspects that these conical shapes become the midpieces. In earlier developmental stages of the rat’s spermatozoa (condensing spermatid) heads (see Fig. 14) one part of the head is paler and softer, while the other, which becomes a hook, is firmer and shinier. This paler and softer part of the spermatozoa seems to be what Kölliker calls the conical part in bull spermatozoa. In the rat and mouse, however, it corresponds to the part of the spermatozoa head, which later extends beyond the point of insertion of the midpiece and therefore cannot become the midpiece.
La Valette St. George82, who made the interesting discovery that the cells of the seminiferous tubules are capable of amoeboid change of shape, has more recently agreed with Henle’s and Schweigger-Seidel’s views regarding spermatozoa (condensing spermatid) development and denies Kölliker’s statements about the occurrence of tails, which are coiled up in the cells as well as the outgrowth of the tails from the heads83. Owsjannikov, however, claims to have observed spermatozoa with coiled tails in the rat. Like Kölliker and in contrast to Henle and Schweigger-Seidel, La Valette still believes that the spermatozoa (condensing spermatids) are formed, at least in part, from the nuclei of multinucleated cells84. Since this assumption, as Kölliker points out, makes it difficult to understand how the tails develop without the participation of the nuclei, La Valette proposes the somewhat peculiar-sounding explanation that in this case every seminal filament originates from a special cell, only the cell substance of the individual cell is not separated from each other85.
All these observers consider the cells of the seminiferous tubules as precursors of spermatozoa. The Keimnetz was either overlooked or completely misunderstood.
The same was first described by Sertoli in humans, where it is strongly developed. The branched anastomosing cells86, which, according to him, form the outermost cell layer of the seminiferous tubules, are evidently what I call the Keimnetz. Later, Merkel described the same cell network as supporting cells and Boll, who tried, albeit in vain, to find a connection between these cells and the tunica propria, explained it as an analogue of the intra-alveolar connective tissue network that appears in the acinus glands. La Valette has also isolated and drawn parts of the Keimnetz from the testes of bulls and dogs. Thus far, only Letzerich has clearly seen and illustrated spermatoblasts in mammals. He briefly notes that a group of spermatozoa (condensing spermatids) developed endogenously in the swollen end of a cell. Incidentally, many of the structures regarded by the earlier observers as multinucleated cysts might better be interpreted as ruptured spermatoblasts, because the lobing of spermatoblasts often occurs only after the development of spermatozoa (condensing spermatid) heads. On non-hardened preparations, the lobing of these structures is often difficult to see. Pieces of the Keimnetz, it seems, have also been confused with polynuclear cysts.
Letzerich is also the first to assert for mammals that spermatozoa originate independently of cell nuclei as separate formations, a view that Grobe also assumes to be valid. Remak has already ascertained that the spermatozoa of the frog form without the involvement of a nucleus.
Several observers (Zenker, Keferstein, Metzernikow, Balbiani) have indicated that spermatozoa develop independently of cell nuclei for several arthropods and one snail87.
Acknowledgments
There are no conflicts of interest to report. This work was supported by a grant from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (R01-HD090083) to C.B.G.
Work supported by a grant to C.B.G. from the NICHD/NIH (R01-HD090083)
Footnotes
Original title: “Studies of the Structure of the Seminiferous Tubules and the Development of Spermatozoids in Mammals and Humans”; Presented to the school of medicine in pursuit of the Venia Decendi (comparable to the current Associate Professor rank) by Dr. Victor von Ebner, Assistant at the Physiological Institute of Graz, Leipzig Germany. Press of Wilhelm Engelmann, 1871,
Annual Report for 1856, p. 14.
Handbook of Systemic Anatomy. Braunschweig (Germany), 1866. Vol. H. p. 354.
Compare V. LA Valette St. George in Stricker’s to Manual of Human and Comparative Histology. Leipzig, 1870. p. 536.
Likely refers to Russian contemporary scientist Petr Ivanovich Peremeschko (1833–1893)
The testis interstitium lacks proper lymphatic vessels, so this may be a descriptive term common in that era
This seems to be a main issue; Ebner did not apparently realize that the darkly-staining heads of spermatids actually contain nuclei, which would have prevented him from appreciating the significant nuclear changes that spermatids undergo; thus, he sought a different origin for spermatozoa
This is likely a typographical error – should be 5–7 mm
Made up of the peritubular myoid cells surrounding the seminiferous tubules
These may be the same as Sertoli’s “niches,” but we have kept “gap” throughout this document as it more closely expresses the German word “Lücke.”
Also called Prussian blue, a dark blue dye.
These are now termed peritubular myoid cells (PTMs), which are wrapped around rodent seminiferous tubules in a single layer
It is unclear who Ebner is referring to here, might be Boll and Leydig, who are mentioned earlier
We now know that in humans and other animals, there are 3–5 layers of PTMs (instead of 1 in rodents), and this likely caused the confusion
Ebner’s stages I and VIII appear to be the same stage (modern stage IX); he does recognize the continuity of development, that a particular stage precedes (or follows) another
The literal translation is “wall layer,” and it seems to denote the tubule periphery. According to Sertoli, this describes the “peripheries of Sertoli cells” (Sertoli, 1878); This term is meant to describe what is essentially the basal compartment, which contains spermatogonia, early primary spermatocytes (preleptotene), and basal portions of Sertoli cells, and generally their nuclei
Ebner may be referring to the blood testis barrier, which had not been formally described yet in history; it is formed by tight junctions between adjacent Sertoli cells to physically separate the basal and adluminal compartments
This appears to be (based on modern terminology) a stage IX tubule, so these basal cells would be leptotene spermatocytes
These are likely to be the nuclei of what are now termed Sertoli cells, named after von Ebner’s contemporary and scientific rival
This is a first indication that Ebner is missing the developmental continuity of the germline; he is describing elongating spermatids, but describes them in the context of an association with Sertoli cells
This is correct – that the spermatocyte nuclei demarcate them as distinct cells from adjacent Sertoli cells
Literal translation is “germinal network”, and was left in the German original by Sertoli and other authors, so we will follow this lead; according to Sertoli, Ebner used this term to describe the basal aspect of Sertoli cells; however, Ebner also seemingly used this term to include basally-located germ cells (spermatogonia, spermatocytes) present at the periphery of the seminiferous tubules along with the peritubuluar myoid cells around the tubules
These “extensions” are Sertoli cell cytoplasm that indeed extend to the lumina
Likely describing primary spermatocytes in the diplotene stage that are dividing to form 2 secondary spermatocytes
This term describes Sertoli cell cytoplasm encasing a cluster of round spermatids, which are the precursor to elongating and condensing spermatids that will ultimately become testicular sperm; Ebner’s contemporary Sertoli realized this progression, and was convinced (correctly) that sperm arose from sequential development of spermatogonia to spermatocytes to spermatids; Ebner recognized these spermatoblasts were precursors to sperm, but did not identify them as direct products of spermatocytes and thus was unable to realize the progression of spermatogenesis
Wonder if this is a typo, as this figure is also used for Stage III, below.
Convention now is that haploid germ cells are called spermatids while resident in the epithelium; these appear to be elongating spermatids
Referring to segments of the flagellum, with the midpiece being proximal to the spermatid head, and the “tail” perhaps referring to the principal and end pieces
Likely referring to Sertoli cell nuclei
Recognized, although did not realize it at the time, that spermatocytes cross the blood-testis-barrier to transit from the basal to the adluminal compartment
Ebner appears to be describing the first meiotic division, which produces short-lived secondary spermatocytes that undergo the second meiotic division to form haploid round spermatids
These are likely clumps of round spermatids, potentially undergoing apoptosis
Likely referring to discarded spermatid cytoplasm forming the so-called residual bodies that become phagocytosed by Sertoli cells
Unclear why he would make this statement “to the degree that this takes place in the testes at all”?
This refers to the cytoplasmic droplet (CD), which is a small remnant of spermatid cytoplasm retained on and moves down the flagellum as sperm leave the testis and transit the epididymis
This would now be termed the apical ectoplasmic specialization, a Sertoli cell structure that encases the spermatid heads until spermiation
This type of statement makes it clear Ebner did not recognize that spermatoblasts (clusters of round spermatids) became sperm
Ebner was so close to making the proper conclusion regarding continuity of the germline – specifically, that round spermatids undergo elongation and then nuclear condensation. It is unclear to me how he missed that after making these careful observations.
This is because Ebner’s stages I and VIII are the same (modern stage IX)
glass-like spheres, likely referring to residual bodies
It is unclear what Ebner is referring to here
Here Ebner seems to consider the possibility that spermatoblasts transition to elongating spermatids and then eventually into sperm
Based on the Figures presented, Ebner seems to be referring to Sertoli cell nuclei, some of which can appear “pointed”
It appears Ebner was expecting to see a cell division corresponding to appearance of condensing spermatids (which he calls spermatozoa). We now know that the morphogenetic transformation of spermatids during spermiogenesis does not include a cell division
This is a key observation; here Ebner is likely describing what we would call today a very late stage VIII tubule section in which most, but not all, of the condensed spermatids have been released into the lumen at spermiation
Flagellum formation actually begins much earlier, in round spermatids, although Ebner did not observe this; indeed, it is nearly impossible to observe this even using modern light microscopes in 5 μm sections. If Ebner had observed this, he most certainly would have concluded that round spermatids metamorphosed into sperm
“thread,” is the German term, and is used throughout; consistent with the technical term “filament”
Interesting concept, but spermatid flagella do not form valves as described
The PTMs, with their rich F-actin cytoskeleton, are thought to contract in a manner analogous to smooth muscle of the gut and exert a peristalsis-type movement to extrude the non-motile testicular sperm from the seminiferous tubules
Likely referring to primary spermatocytes
Likely describing different stages of spermatids – both round and elongating
Perhaps into secondary spermatocytes, which we now know rapidly divide again to form round spermatids
Unsure what exactly is meant here; there is evidence that undifferentiated spermatogonia are motile, and can move within the testis
This is an interesting, but we now know erroneous supposition, that new generations of spermatogonia and primary spermatocytes arise from white blood cells that extravasate to enter the seminiferous epithelium
This is correct, that spermatids are indeed post-mitotic
Likely referring to the acrosome
This is problematic, as Ebner seems to suggest that spermatids can become spermatogonia
This was one of the contentious points with Enrico Sertoli; Ebner held the erroneous opinion that spermatids did not give rise to spermatozoa; unsure what is referred to as “latest handbooks”, but perhaps these are writings of his contemporaries?
It is unclear what rationale underlies this supposition regarding the role of round spermatids
Residual bodies
Ebner observed normal nuclear morphogenesis during spermiogenesis and concluded that the cells died as a result
Another incorrect conclusion; these “glass-like spheres” are the shed excess cytoplasm (residual bodies) from developing spermatids, which are eventually phagocytosed by Sertoli cells
Here, he appears to be referring to Sertoli cells
Ebner incorrectly states that round spermatids form from Sertoli cells
This is correct; the Sertoli cells have an F-actin rich structure termed the apical ectoplasmic specialization that anchors condensing spermatids prior to their release at spermiation
Used interchangeably with “multinucleated”
After the spermatogonial stem cell stage, male germ cells develop clonally, connected by intercellular bridges; the apparent controversy Ebner describes may arise from some scientists’ abilities to visualize clones based on their choice of fixative and stain and the resolution quality of their microscope; indeed, Sertoli in 1878 correctly described the presence of these clones
This is a prescient observation; as a high throughput system, errors are commonly made, and entire clones of germ cells at various developmental points do get removed by apoptosis
Of course Ebner had no way of knowing in 1871 that 135 years later it would be shown that clone formation is actually essential for the completion of meiosis and spermatid formation (Greenbaum et al., Proc Natl Acad Sci 2006)
It is unclear what Ebner is referring to here; perhaps these were from a naturally occurring mutant male, or perhaps he observed the so-called “straight tubules”, which are a transition zone between the seminiferous tubules and the rete testis, and contain Sertoli cells but not complete germ cell development
It is unclear whether Ebner is referring to a naturally-occurring mutant or a young mouse that does not yet produce spermatozoa
Ebner appears to start his first stage (A) at the current stage IX; completed condensed spermatids were just released from the epithelium at spermiation (now termed testicular spermatozoa), and the remodeling process to produce the next set of condensed spermatids has begun with the formation of elongating spermatids
Although a bit crudely done, Ebner appears to have correctly ordered the stages of the seminiferous epithelium based on spermatid morphogenesis, starting with elongating spermatids and completing with condensed spermatids; it is unclear how he missed that round spermatids make a gradual and readily observable transition to elongating spermatids between stages VIII-IX (stage D in his system); this realization would have likely led him to properly conclude that the spermatids give rise to sperm
Eingeweidelehre, p. 353. Translators note: Jacob Henle’s book Handbuch der Eingeweidelehre des Menschen (literally: Handbook of the Anatomy of Humans, 1866) has not been translated into English. There have been reprints in German as recent as 2017 and the Royal College of Physicians of London has the German e-book available online under the spine title Henle’s Anatomy.
Handbuch der Gewebelehre 1867, p. 524. Translators note: Albert Kölliker’s book with the full title Handbuch der Gewebelehre des Menschen : für Aerzte und Studirende (literally: Handbook of the Anatomy of Humans: for Doctors and Students) also remains untranslated.
Virchov, Archive. Vol. XXVIII, p. 373.
This is the basis of Sertoli’s most important criticism of Ebner’s conclusions; Ebner postulated the origin of sperm (spermatoblasts) arose from Sertoli cells
It is unclear to me what he is basing this on; he may not realize his “glass-like spheres” (which I take to be the superfluous spermatid cytoplasm that are the so-called residual bodies) are leftovers from condensed spermatids and not the origination of a new generation of developing spermatids
Spermatids can and do move up and down as they develop within the crypts between adjacent Sertoli cells
Archive of M. Schulze. Vol. 1, p. 309.
Yet, we know now this is what occurs
In this section of the text, von Ebner uses the name “De La Valette St. George,” whereas in earlier passages he only writes “La Valette St. George” or “La Valette.” The German scientist’s full name was Adolph von La Valette-St. George. The “von” signifies an aristocratic title and we assume that von Ebner used the French “de” to account for that in some parts of the text. However, other texts typically leave the “von” out, so we have done the same here
The latter of which is how flagella form
If by “multinucleated cells” La Valette means clumps of interconnected round spermatids, then he was indeed correct
Again, although Ebner classified this as “peculiar-sounding”, La Valette was correct
Also termed “sustentacular cells” at that time, later became known as Sertoli cells
Seems like an odd way to end such a comprehensive report, with offhand mentions of potential for nucleus-free sperm development, which we of course now know to be incorrect (since the nucleus and its contents are the payload delivered to the egg by the sperm!
This may be a depiction of transition of modern stages XIII-I, during which primary spermatocytes divide to form secondary spermatocytes and then round spermatids; since 4 spermatids are produced from each primary spermatocyte, they do indeed appear to “multiply”; however, this did not appear to line up to his stage D description for III developmental stage, above
Ebner seems to not realize, as Sertoli did, that spermatogenesis proceeds from the periphery to the lumina of tubules; mistakenly calls pachytene spermatocytes “second generation” and round spermatids “first generation”
This is a portion of a fragmented tubule, and thus difficult to interpret
All pictures refer to preparations of the testes of a rat in heat hardened in Müller solution. The sections were stained with bluewood extract and rendered transparent with clove oil.
References
- Geyer CB (2018). A historical perspective on some “new” discoveries on spermatogenesis from the laboratory of Enrico Sertoli in 1878. Biol Reprod, 99(3), 479–481. doi: 10.1093/biolre/iox125 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Koelliker RA (1841). Contributions to the knowledge of the sexual relations and the seminal fluidity of invertebrates. In. Berlin. [Google Scholar]
- Roosen-Runge EC (1977). The process of spermatogenesis in animals. New York City, NY: Cambridge University Press. [Google Scholar]
- Sertoli E (1878). The structure of seminiferous tubules and the development of spermatids in rats In Archivio per le scienze mediche (pp. 1–76). Turin: Vincenzo Bona. [Google Scholar]