HISTORY O F THE DECIDUAL PLASMODIA OR GIANT CELLS O F CITELLUS TOWNSENDI ROBERT H. KREHBIEL Zoulogical Laboratory, State College of Washington FIVE FIQURES IXTRODUCTION During the time of gestation certain changes take place in the uterine mucosa which are distinctly characteristic of the period. Included among the many changes is found one phenomenon which is outstanding in that it is restricted to only a few species of mammals, and is in a true sense of the word an isolated occurrence. This phenomenon, which is found in the decidua of the uterus and with certain reservations in the trophoblastic cells of the embryo, is the formation of giant cells. Due to differences observed in their structure, different names, origins, and functions have been assigned to them, until to-day the accumulated data make it impossible to draw any general conclusions regarding their true nature. The origin and also the structure are subjects of controversy. While there are a few objections, the generally accepted theory of their function is one of phagocytosis. The giant cells have been described in the mouse, rabbit, hedgehog, and man, the most extensively in the mouse. In the different species of the latter animal variations are found as to the structure and origin of these enlarged cells. Duval ('91) states that they are derived from the foetal ectodermal wall, and is supported by Sobotta ( ' 0 3 ) , who finds the giant cells undoubtedly arising from foetal tissue. Jenkinson ( '02) attributes the origin to foetal tissue, saying that they are enlarged trophoblastic cells which migrate into the decidua. 275 THE ANATOXICAL RECORD, VOL. 5 0 , NO. 3 276 ROBERT H. KREHBIEL I n absolute contradiction to this, Kolster (’03) and Disse (’06) find that the growth of decidual cells gives rise to the giant cells. There is a marked resemblance in the structure of the cells described by Jenkinson and by Disse. I n the rabbit, Chipman (’03) finds the giant cells to be transformed epithelial cells of the glands of the uterus, whereas Sansom (’27) attributes the origin to the trophoblastic cells. I n the hedgehog, Hubrecht (’89) attributes the origin to maternal tissue. I n man, Bryce and Teacher (’08) state that the decidua is the probable source of the giant cells. I n every case it is to be noted that a giant cell is regarded as the resultant of the enlargement of a single cell. This described variation of origin would seem to suggest a possible difference in structure and function. On the other hand, variations in structure other than size are very slightin fact, so slight that they do not hinder the investigators in the acceptance of a general theory as to their function. Thus it may be said that the ‘megalokarocytes’ of Jenkinson, the ‘Riesenzellen’ of Sobotta and Disse, and the ‘giant cells’ of Hubrecht, Chipman, Sansom, and Bryce and Teacher are all f o r the purpose of bringing about changes in the uterine wall, which will facilitate implantation and placentation. Allow me to repeat that in all cases the investigators have come to the conclusion that they are phagocytes. Having done so, they apparently have interpreted the decidual and blood cells contained in the cytoplasm of the giant cell in such a manner as t o substantiate their theory of the function of these structures. Should such a conception be accepted, it would necessitate the isolation of these cells as distinct in nature from all other known cells. I n all of the cellular transformations which are found in any organism it is impossible to find a comparable enlargement or similar origin of phagocytes. They exceed in size any other single cell of metazoan bodies known; the only cells which may be compared in this respect are the large phagocytic cells of the blood stream. Also, it would be without precedent to assign hemopoietic activity to the trophoblast or to an unmodified mucosa. GIANT C,ELLS O F CITELLUS TOWNSENDI 277 Evidence that contradicts the above accounts has been found in the Townsend ground squirrel (Citellus townsendi). From a purely structural viewpoint I have found that these structures are not single hypertrophied cells, but are the result of the fusion of a group of cells, being purely a passive product, rather than an active structure as previous investigators have concluded. They resemble a plasmodia1 structure greatly, and due t o the fact that this is contradictory to previous findings, I shall use the term plasmodium rather than put forth a definite name or use the existing term of ‘giant cell’ in this paper. Should these bodies found in Citellus townsendi be of an entirely different nature than those found in other species, and should the other structures be true phagocytes, it seems highly important to assign a name to them which will distinguish them from the giant cells found in the blood stream. The knowledge concerning these ‘cells’ is not complete enough at the present time to give them a definite name. However, it seems that a term such as decidual gialzt cells or decidual plasmodia would serve the purpose until further research has provided a basis for different nomenclature. MATERIAL AND METHODS The uteri were taken from tlie animals immediately after they were shot and fixed in Zenker’s bichromate solution. The enlarged areas, the sites of implantation, were sectioned in series at 6 and 7 v. The sections were stained very lightly in Delafield’s haematoxylin before being stained in Giemsa ’s modification of azur-eosin stain (Grubler ’s prepared solution). Mallory’s triple stain of acid fuchsin, anilin blue, and orange G, and iron haematoxylin were also used. ORIGIN AND F A T E O F DECIDUAL PLASMODIA The plasmodia originate in close approximation to the epithelium and spend their life in the same region. They are generally confined to the antimesometrial end of the lumen. Here they occupy the greater part of the decidual region, 278 ROBERT H. HREHBIEL being from two to five rows in thickness. Decidual cells may be seen to be distributed among them as well a s connective tissue. I n general, the epithelium is intact until a general .. . .' . .- .r @ Q' , P' I Fig. 1 Drawing showing tho distribution of the plasmodia. a, mesometriuni; a', antimesometrial end of lumen; bl, blood in lumen; p , plasmodia; p', region X 38. Fig. 2 Section of plasmodia1 region. A, formative stage; B, nuclear stage; C, late nucleolar stage. n, wall of nuclear structure; n', remains of nucleolar formation ; T , reticulum formed by disintegration of individual nuclei ; o, vacuolar remains of a single deteriorated cell; ue, uterine epithelium. X 650. of placentation. sloughing off of the plasmodia takes place. At the mesometrial end of the lumen the decidua is destroyed in such a manner as t o allow the blood t o flow from the opened capillaries into the lumen. This is distinctly a different type of GIANT CELLS O F CITELLUS TOWNSENDI 279 necrosis from that found in the antimesometrial end (fig. 1). As soon as the egg has come to rest in the lumen at the site of implantation, the below-mentioned changes start. The period through which the plasmodia1 structures are found persist up to that time when the placental formation is well under way. A few may be found at the stage in which the chorionic vesicle has completely filled the lumen and started to stretch the uterine wall. The metamorphoses which take place in the plasmodia necessitate a description of the different stages of their life history. It seems advisable to divide the span of their existence into four stages, namely: 1) formative, 2) nuclear, 3) nucleolar, and, 4) vacuolar stages. It is to be understood, however, that at no time is there an abrupt change which would allow one to divide the life history into four distinct stages. I n fact, the different cells, which go to make up a single plasmodium, may be in very different stages of deterioration. 1. Formative stage I n the very early periods of gestation the mucosa of the uterus, throughout most of its extent, is apparently normal and is lined on the lumen surface by a normal epithelium. I n a few places, however, marked changes can be seen in the cells that are located in close approximation to the epithelial cells of either the lumen or the glands. This alteration consists of the loss of the cytoplasmic contour of the individual cells. The breaking of the cell wall brings about the liberation of the cytoplasm and allows the several cells to fuse. At first, definite areas of cytoplasm can be seen to inclose several nuclei, but as the formation of the plasmodium continues, it can be seen that there is a marked tendency of the cytoplasm to become a homogeneous mass (fig. 2, A). Not only is there a general fusion of the cytoplasm, but also of the nuclei. When the nuclei are liberated by the breaking of the cytoplasmic boundaries, they assume a central position. At first, this action can only be detected by the presence of several nuclei in contact with each other. This 280 ROBERT H. KREHBIEL nuclear clumping continues until there is a large nuclear aggregate in the center of each plasmodium. Occasionally several such aggregates will be formed, and, in some cases, there is practically no clumping (fig. 2, B). The deterioration of the cells in a plasmodia1 region is not regular. There may be found, in the same region of formation, cells which have lost every trace of cytoplasmic integrity. At the edge of these regions individual nuclei can be seen Fig. 3 Plasmodium in nuclear stage. a, nuclei being taken into the plasmodia1 structure ; d, decidual cells ; i, individual nucleus with its chromatic material clumped in the center of the vacuole formation; TL, nuclear formation; n', resultant nucleolar formation of a few aggregated disintegrated nuclei; 2), vacuole with the limiting membrane gone; v', vacuole still retaining a portion of the membrane. X 650. which are apparently entering the decidual formation (fig. 3, a ) by the flowing of the cytoplasm out around them. The nuclei show a similar variation of deterioration. The chromatin material is generally in the form of small granules, the majority of them being located at the periphery of the nucleus. Even in the very early stages there is a lack of oxychromatic material in all the nuclei, and a few of them consist of nothing more than a deeply staining broken wall and a single, central body of chromatin (fig. 2, r ) . GIANT CELLS O F CITELLUS TOWNSENDI 281 2. Nuclear stage The nuclei, which have been in different stages of deterioration up to this time, now reach the same point of disintegration in which they consist of nothing more than small vacuoles with chromatic central masses (fig. 3, i). As the nuclei become contiguous their adjacent walls disintegrate, so that a simple fusion of material forms the large nucleus. The outer part of the membrane of the nuclei at the periphery of the mass remains and becomes the wall of the fusion nucleus. Extending from this wall can be seen parts of the nuclear wall of the individual nuclei which have gone t o make up the plasmodium. The remainder of the walls of the nuclei go to form a broken reticulum of the late nuclear and the early nucleolar stage. The chromatic material, upon the disintegration of the walls of the individual nuclei, may become scattered throughout the nuclear structure or aggregated into one or more large masses. The latter is more frequently found. 3. Nucleolar stage The assumption of a central position by the chromatic material causes the plasmodium to have the appearance of a typical, large, single cell (fig. 4). The large nucleolar structure found in this stage, judging from staining quality, undoubtedly arises as a result of a fusion of the chromatic material of the individual nuclei of the first two stages. At times, much in the same manner as the nuclei were aggregated to form the nucleus, the chromatic material forms more than one nucleolus. The nucleus still retains its wall, which becomes somewhat more even in outline than it was previously. Vestiges of an incomplete reticulum, formed by the broken walls of the individual nuclei, can still be seen. Up until the very late periods of this stage a few independent nuclei may be detected, as shown at a in figure 4. 282 ROBERT H. IiREHBIEL 4. Vacuolar stage The destruction of nuclear material leads to the disappearance of the nucleolus and the broken reticulum, leaving the irregular nuclear wall to surround the vacuole thus formed. First the nucleolar structure disintegrates, then the reticulum surrounding it, and finally the nuclear wall itself. There is no dispersion of material in the process of its final disintegration, but merely a very gradual fading from view of all the nuclear constituents. The nuclear wall gone, a vacuole is Fig. 4 Plasmodium in nucleolar stage. a , individual nucleus contained in nuclear structure; n, nuclear wall; n', nucleolar formation ; ue, uterine epithelium. X 650. the only indication, in the very late stages of the plasmodium, of a previous clumping of nuclei (fig. 5, v). The cytoplasm, which was assuming a homogeneity in the formative stage, attains that condition in the nuclear stage. It remains as such until the nuclear structures have disappeared. However, it contains within it vacuoles of varying size which were undoubtedly formed by independent small groups of nuclei going through the process of disintegration. A very fine fibrillation can be detected in the cytoplasm throughout most of its existence. This characteristic does GIANT CELLS OF CITELLUS TOWNSENDI 283 not carry over into the final stages, in which the cytoplasm assumes an extremely coarse fibrous character. The boundaries of the cytoplasm are not distinct. It can be seen to pass into the surrounding tissue in many places, thus engulfing nuclei (fig. 3, u ) . Connective tissue is found between the plasmodia, and often there are only a few strands separating two adjacent ones. After the nuclear structure has become merely a vacuole, the cytoplasms of the contiguous plasmodia fuse, and thus Fig. 5 Plasmodium in vacuolar stage. d, decidual cells; ue, uterine epithelium; w, vacuole without membrane ; w', vacuole still retaining membrane. X 650. in the very last stages one finds a wall of cytoplasm, containing connective tissue and vacuoles, lining the lumen of the uterus. Upon reaching this condition, which has required the destruction of the decidual cells, and leaving only connective tissue to replace it, the cytoplasm, in turn, deteriorates. The deterioration of the latter consists of a loss of its compactness (fig. 5) and ends in it being passed off into the lumen in small portions of much the same size as the original individual plasmodia, there to become converted into a finely fibrous structure, and finally disappear. 284 ROBERT H. K R E H B I E L The destruction of the mucosa by the formation and disintegration of the plasmodia quite naturally leaves the uterine wall much thinner at the site of implantation than it is elsewhere, since little more than the muscular wall remains. DISCUSSION From the foregoing description there can be no doubt that the plasmodia are of decidual origin, and that they arise as a result of the fusion of a large number of mucosal cells. If the conclusions drawn by previous investigators are right, the plasmodia herein described must be regarded as structures which have no likeness to the giant cells they described. However, when a structural comparison is made, such does not seem to be the case. The interpretation of past investigators has accounted for the decidual and blood cells contained in the large ‘cells’ on a basis of supposed phagocytic action of the giant cell. On the other hand, the present account, it seems to me, makes that interpretation wholly untenable. The ‘megalokarocytes’ of Jenkinson ( ’02) and the ‘Riesenzellen’ of Disse (’06) show an inclosure of decidual cells which are undoubtedly deteriorating, but to ascribe the function of phagocytosis on this basis alone, without accurately determining their origin, seems rather unjustifiable. In the ‘megalokarocytes one finds a structure which resembles the plasmodia in every phase other than in their supposed origin and disappearance. I n his illustrations Jenkinson has shown individual nuclei inclosed in the cytoplasm of the ‘megalokarocytes’ and also groups of several nuclei clumped as a result of their deterioration. Disse ( ’06), on the other hand, did not account for the inactivity of the ‘Riesenzellen’ as they pass through the capillaries from the base of the decidua, the site of the origin, to the implantation cavity, where they supposedly resume phagocytic activity. I am inclined t o believe that a restudy of Jenkinson’s and Disse’s material would lead to a classification of these giant cells as plasmodia of the same type as those found in Citellus townsendi. GIANT CELLS OF CITELLUS TOWNSENDI 285 The striking resemblance of the ‘cells’ found in the mouse to the plasmodia is overshadowed by the resemblance of the latter to the ‘cells’ found in the rabbit, as described by Sansom ( ’27). This admirably illustrated paper shows giant cells located under the uterine epithelium in the same manner as in CItellus townsendi. It is very hard to associate the structures which he has shown as giant cells with phagocytic activity. The improbability of such structures coming from foetal tissue seems great at the present time. Had Sansom not failed to correlate his findings with those of the other investigators working with the same animal, he would have noted resemblances. It seems highly probable that the structures shown in his figure 12, plate 23, are, for the most part, plasmodia in the later nucleolar stage. Also, it seems rather inconceivable to attribute the power of division and growth to such a protoplasmic formation. However, it must be kept in mind that his giant cells appear at a much later date than the plasmodia of Citellus townsendi. Sobotta’s ‘Riesenzellen’ are decidedly smaller than the other giant cells. They are distinctly part of the chorion and are merely enlarged cells, the enlargement not being great enough to set them aside as distinct structures. I n man the giant cells appear at a much later date and are clearly a part of placentation; I believe that it may be said with safety that these cells are, as Sobotta’s cells, entirely of a different nature. With the close resemblance of the plasmodia to several of the previously described structures the question arises : What should our interpretation of the significance of these structures be if they are not to be regarded as phagocytes? It would seem that they serve no purpose other than to bring about a thinning of the uterine wall in the site of implantation, and, as such, may correspond very closely t o the vacuolization found in the higher forms of implantation, in so far as function is concerned. An explanation of the formation of the plasmodia upon a purely microscopical basis is comparatively simple and com- 286 ROBERT H. XREHBIEL plete; an ascertation of the cause of this phenomenon is by no means so easy. Since the nature and source of the factors which bring about this change are unknown, an explanation of this alteration must be purely hypothetical. I n view of the present theory, which attributes the changes that are brought about in the decidual area of the uterus during implantation and early placentation to the action of the blastodermic vesicle, the first thought would be, of course, that the blastodermic vesicle is the instigator of the formation of the plasmodia. The distribution of the plasmodia, and the different type of necrosis found in the mesometrial end of the lumen, however, make this conception appear unsound. On the other hand, there is no positive evidence to show that it is the maternal system which brings about the change, although there is less to refute such a supposition. It can readily be seen that this question is but a part of the general problem of the mechanics of implantation and the activity of the blastodermic vesicle and maternal system in effecting it, of which a comprehensive discussion at the present time is unwarranted. The formation of the plasmodia suggests some very interesting problems in the field of physical chemistry. It would seem that the process is perhaps somewhat related to fertilization, in its physical aspects. A solution of these problems is absolutely essential to a complete understanding of the phenomena associated with implantation, at least in Citellus townsendi. An explanation at this time cannot be advanced didactically, but rather as a suggestive interpretation of the physical actions involved, which may be supported or rejected by future research. The intrinsic value of such a hypothesis is, naturally, slight, but it may be of great value by way of suggesting possible lines of research designed to isolate and elucidate the factors of implantation. The first assumption necessary in an explanation of the phenomena is that the destructive force is acting from two or more localized points, with the areas of activity located in a central position in a group of cells. The force acting GIANT CELLS O F CITELLUS TOWNSENDI 287 from these points attacks the surrounding cells, and brings about a lowering of surface tension. Thus, the portion of the cell next to the point of action will be attacked first and there will be, upon the lowering of the surface tension, a flow of the cytoplasm of the cell in the direction of the attack, due to the fact that each molecule of the cytoplasm has a tendency to disrupt the interfacial membrane of the cell. Once the restrictive forces of like molecules are withdrawn, they can migrate out into clear space. With this happening, the area of activity will become filled with cytoplasm and, due to the greater flow in this direction, the nuclei of that area will be drawn closer together. As the destructive force continues to spread out from the point of origin it gradually brings about a lowering of the surface tension in the entire region. However, the surface of an area which lies in close approximation to another such area will have a lower tension than the rest of the surface of that area, due to the fact that the force from another area is also acting there. Thus there will be a general flow of the cytoplasm in that direction and the nuclear groups in turn will be carried closer to each other. The like attraction of nuclear material will, in turn, draw the same into the form of a large nuclear structure. I n the same manner the chromatic material will assume the form of a nucleolar formation after the walls of the individual nuclei break. This theoretical explanation of the physical factors involved in the formation of decidual plasmodia in Citellus townsendi is more in harmony with our knowledge of the behavior of protoplasmic masses than any other which I can devise. Now, if we go one step further and try to see the ultimate source of the destructive force, it would be very difficult to conceive of the trophoblastic vesicle producing the destructive force in the localized manner assumed. Thus, I am led to suggest that we should seek for the factors of implantation within the maternal system. 288 ROBERT H. KREHBIEL SUMMARY 1. Previous investigators have regarded the ‘giant cells ’ as merely large phagocytic cells originating from single hypertrophied trophoblastic or mucosal cells. 2. I n Citellus townsendi the decidual plasmodia originate from the decidual cells, by a process of fusion of deteriorating tissue. The nuclei of the contributory cells fuse, causing the plasmodium to resemble a very large single cell, All of the chromatin becomes collected in masses which have the appearance of giant nucleoli. 3. Complete disintegration of nuclear material leaves only the cytoplasm with its vacuole. These vacuolar structures line the lumen of the uterus. 4. There is a striking resemblance between the plasmodia and the cells described by Jenkinson, Disse, and Sansom. 5. Contrary to previous interpretations, the decidual plasmodia are regarded as passive products of uterine necrosis whose formation and dissolution serve no purpose other than to contribute to the process of enlarging the lumen of the uterus. 6. It is suggested that the factors underlying this process should be sought within the maternal system. I wish t o acknowledge the criticism received from Dr. J. McA. Kater during the course of this work. BIBLIOGRAPHY BRYCE,T. H., AND J. H. TEACHER1908 Contributions to the study of early development and embedding of the human ovum. GIasgow. DISSE, J. 1906 Die Vergrosserung der Eikammer bei der Feldmaus. Arch. f . mikr. Anat., Bd. 68. DCVAL,M. 1891 Le placenta des rongeurs. Jour. de I’Anat. et de la Phys., T. 27. HUBRECHT, A. A. W. 1889 The placentation of Eriiiaeeus europaeus. Quar. Jour. Micr. Science, vol. 30. KOLSTER,R. 1903 Zur Kenntiiis der Embryotrophe. Anatomische Hefte, Bd. 68. sax so^, G. S. 1927 Giant cells in placenta of rabbit. Proe. of Roy. Soe. of London, vol. 101B. SOBOTTA, J. 1903 Die Eiitwicklutig des Eies der Maus vom Sclilusse der Furchungsperiode bis zum Auf treten der Amniosfalten. Arch. f . mikr. Anat., Bd. 61.