- Email: [email protected]
In vitro aggregation of syncytia and cells of a hexactinellida sponge
DEVELOPMENTAL AND COMPARATIVE IMMUNOLOGY, Vol. 6, pp. 15-22, 0145-305X/82/010015-08503.00/0 Printed in the USA. Copyright (c) 1982 Pergamon Press Ltd. All rights reserved.
1982.
IN VITRO AGGREGATION OF SYNCYTIA AND CELLS OF A HEXACTINELLIDA SPONGE
Max Pavans de Ceccatty Laboratoire d'Histologie.LA CNRS 244 Universit~ Claude Bernard - Lyon I 43 bd du Ii Novembre 69622 Villeurbanne cedex France
ABSTRACT. Membrane recognition has been analyzed by studying the reaggregation of syncytia and cells dissociated from the tissues of the sponge Rhabdocalyptus dawsoni. Four obvious features are recognized. I) There is true morphogenesis,led by the syncytia, which gives rise to compact,often very large (0.5cm) aggregation bodies. Each body is a giant macrophagic syncytium surrounding groups of discrete undifferentiated cells. 2) Special couplin~ structures, the cytoplasmic perforate septa, establish communication between all body compartments, both syncytial and cellular. The whole aggregate can,therefore, be compared to one vast multinucleated cell.3) During in vitro aggregation, membrane recognition processes permit different syncytia and cells to fuse, and internal compartmentalization is then achieved by the cytoplasmic septa. 4) Fragments of the flagellated choanosome, which is characteristic of sponges, are internalized by the syncytia. Following the collapse of the aggregate they degenerate. The survival of the body appears to depend on skeleton formation.
INTRODUCTION Recognition is related to phenomena as diverse as cell and tissue integrity and the basis for immune processes. Recent data have confirmed earlier disputed reports of Ijima (i) and of Schulze (2) which state that some of the most primitive multicellular animals,the sponges Hexactinellida, have a predominantly syncytial organization. This confirmation (3,4) is all the more important since the Hexactinellida date back to the Precambrian and since sponges of other classes are truly cellular and not syncytial. Hexactinellida thus provide exceptional models for the study of developmental processes involving recognition or rejection of like or unlike tissular components. 15
16
AGGREGATION
IN SPONGES
Vol.
6, No.
Preliminary investigations of Reiswig, in 1979 (3), prompted Mackie and his coworkers (4) to examine in detail,at the light and electron microscope levels, the histology of Rhabdocalyptus dawsoni Lambe 1893. According to both the earlier and recent reports, the sponge body consists of cellular and syncytial portions. The syncytia form a thin dermal membrane at the surface of the sack-shaped individual, as well as general internal trabecular network with specialized regions such as the reticulated layer of the flagellated choanosome.Clearly cellular are : the multipotent archeocytes and cells containing various inclusions. Three modes of interaction between these tissular components are represented : i) total fusion into a portion of syncytium without septa: 2) fusion with partial separation by means of cytoplasmic bridges provided with perforate septa : 3) distinct segregation into totally independent cells. Due to the fact that the ontogenesis of these sponges is still practically unknown, one of the best approaches to the problem is to examine the reaggregation of these components following experimental dissociation of adult tissues. The aim of the present study is to investigate the process of membrane recognition and the potential for morphogenesis in syncytia and cells of Rhabdocalyptus. Although this work is not strictly immunological, it does relate specifically to the general problem of r e c o g n i t i o n which is necessary for immune phenomena to be initiated.
MATERIAL
AND
METHODS
Sponge samples were collected by scuba diving at a depth of 30 meters near the Bamfield Marine Station in Barkley Sound (B.C.Canada). They were kept in tanks of running sea water at 13°C. A total of 12 animals were used for our experiments. Each sponge was cut into small pieces and mechanically dissociated by squeezing through a silk cloth or nylon net with a mesh of 150~. A dense suspension flowed out and was filtered repeatedly through a 100p mesh silk or nylon netting. After being allowed to settle for 15 min. the last filtrate was decanted and gently centrifuged at less than i000 rpm for i minute . The resulting supernatant was evenly divided among 15 to 20 Petri dishes, 5cm in diameter, which were maintained at constant temperature (13°C) by a film of sea water running on the culture table. Microscopic examination of the supernatant showed numerous single cells, mainly archeocytes, plus variously-sized fragments of general trabecular syncytia and of reticulated syncytial choanosome.The proportions of these components put into each Petri dish could not be rigorously controlled. For E.M. observations, adult tissues, reunition bodies and aggregates were systematically fixed and embedded using a variety of procedures in order to circumvent the numerous failures reported by Reiswig(3). All fixatives, buffers and rinsing solutions were adjusted to an osmotic pressure of 900-950 mOsm, that is slightly higher than Baker Sound sea water osmolarity : 850 mOsm. The specimens were never allowed to emerge from the solution. All changes of bathing solutions were applied by perfusion flow which renewed progressively the medium in the same container. In the case of aggregation bodies, the best results were obtained by simple osmium fixation:l% for lh. at 13°C. and by embedding in mixtures of expoxy resins. For normal
1
Vol.
6, No.
1
AGGREGATION
IN SPONGES
sponge tissues used as controls, the method of Mackie and Singla(4) was adopted : that is, simultaneous double fixation in 3% glutaraldehyde and i% osmium. It gave good ultrastructural preservation even after the specimens had been desilicified for I hour in 2% hydrofluoric acid. Thin sections were d o u b l e - s t a i n e d with uranyl acetate and lead citrate. They were viewed with the Philips 300 and Hitachi 12 electron microscopes. The aggregation process was studied in 12 experimental series (one series for one individual), each one comprising 15 to 20 Petri dishes.
RESULTS
In each series,the events which accompany reaggregation were found to be similar in nearly all (90%) containers.At the beginning, syncytia and discrete cells do not display obvious amoeboid behavior nor do they adhere very firmly to the substrate. The flagella of the choanosome fragments create local turbulences in the water causing the suspended elements to collide. These collisions give rise to minute agglomerates which grow, fuse and then, sometimes, spread on the bottom. Consequently the first reunition bodies have various shapes, sizes and compositions. An essential event arises at the end of an initial 24 h. period. Time lapse cinematography (i frame per one or two minutes) reveals gradual fusion of adjoining small agglomerates leading to general retraction of the area occupied by the culture. Thus, new compact aggregates are formed in the Petri dishes. Sometimes, there may be just a single very large body in the form of a circular plate, I cm in diameter. The total duration of the process may last nearly 12 hours. As these phenomena continue, the edges of the aggregate are raised transforming the shape into that of a bowl or a cup...a sponge shape in a word (Fig.l A and B). After 48 hours each dish contains either several polymorphic middle-sized units, or one large spongelike body accompanied by small agglomerates. The most stricking feature of the cytology of the aggregates is the formation of a sort of giant m ul t i n u c l e a t e d macrophage which encloses groups of individual cells, in internal extra-syncytial compartments (Fig.2 A,B). It is made up of components of the general trabecular and epithelial syncytia found in the intact sponge. Many identical nuclei are surrounded by complementary cytoplasm exhibiting various modalities of organelles assembly,just as in normal sponge. In any case, phagocytic activities in the syncytium of the aggregation bodies are considerably enhanced above that which is usually observed in the syncytia of normal sponges. Numerous phagosomes enclose all sorts of debris : fragments of degenerated cells, pieces of spicules and an occasional alga or bacteria.
17
18
AGGREGATION
Pictures
IN SPONGES
FIG. I from time lapse cinematography
Vol.
6, No.
recordings.
Same magnification for both frames : scale 0.6 cm. A) Reunition bodies in a 24h.-old culture, before the process of general retraction. B) One of the compact cup-shaped aggregation bodies , 36 h.-old after general retraction.
During aggregation, all cells in the culture dish are gradually engulfed and internalized by the syncytial unit. The multipotent archeocytes are internalized and do not participate intensively in phagocytosis which is handled almost entirely by the general syncytium, nor they undergo division. Aggregation bodies are the result of the simple reorganization of syncytia and individual cells without nuclear division or archeocyte differentiation. After 24 hours,most of the choanosome fragments, along with discrete cells, are also internalized. The flagella extend into cavities which are completely close and have no relation to the external milieu. Subsequently, collar microvilli and flagella degenerate. The cavities shrink at the same time as the general fusion and retraction of the aggregates. In spite of the drastic simplification achieved during the aggregation process, the basic components of the intact sponges are preserved. These are perforate septa (4), partitioning off cytoplasmic bridges between either two regions of syncytium, or a syncytial compartment on one side and a discrete cell on the other, or between two discrete cells (Fig.2 C,D). Finally, aggregation bodies have no spicules. As a result,instead of having a three dimensional net of soft tissue superimposed on the crystalline lattice of the skeleton, aggregates are compact masses of tissue enclosing some blind cavities.These bodies are not functional. The phase of aggregate involution and disorganization is accompanied by the disruption of the septa. Discrete cells become isolated from each other and from the syncytium. Cytoplasmic organelles of cells and syncytia degenerate and the pycnosis of nuclei starts.
1
Vol. 6, No. 1
AGGREGATION IN SPONGES
FIG.2 Ultrastructure of aggregates A and B : five day-old aggregate. Same magnification for both pictures: scale bar 4pm. A) At the body surface,the macrophagic syncytium encloses normal or degenerating archeocytes. B) Deep in the body, the syncytium is more reticulated and makes contacts with discrete cells by means of dense septa (arrow). C and D : two day-old aggregate. C) Dense perforate septa (arrows) between archeocytes. A continuous channel of smooth reticulum connects the septa. Scale bar 2 ~ . D) Three septa (arrows) couple different syncytium parts. One is crossed by a cisterna of smooth reticulum. Scale bar O.4~m. Abbreviations:(a) archeocytes, (n)syncytium nuclei : (s) syncytium cytoplasm.
19
20
AGGREGATION
IN SPONGES
Vol.
6, No.
DISCUSSION
The in vitro reaggregration of Rhabdocalyptus tissue constituents obtained by experimental dissociation is characterized at first by the gathering of heterogeneous components. These components either fuse into a single syncytium or, in the case of discrete cells, become enclosed in an internal but still extracellular compartment of that syncytium. Therefore the morphogenetic process is led by the syncytium and not dependent on division, recognition and positioning of multipotent archeocytes, as might be expected (6-12). Moreover, communicating bridges provided with perforate septa, present in normal adult tissues (3,4) are also maintained or reestablished by the cytoplasm of aggregation body constituents. These observations point out that cell and tissue recognition mechanisms have sufficient plasticity to allow plasma membrane fusions to occur in a system composed of differentiated syncytial areas and diverse categories of individual cells. A giant, multinucleated body is created through these fusions but the process does not lead to a general uniformity of the cytoplasm or of the membrane themselves. On the contrary, cytoplasm and membranes can continue to differentiate and produce various local structures, at least in normal sponge. The Hexactinellida have thus found an original way of establishing compatibility between the possession, on the one hand, of membranes which are sufficiently similar to be capable of making contact and fusing, and on the other hand, of distinctive perinuclear cytoplasmic compartments having diverse tissular functions. Contrary to the expectation, the key to this compatibility does not lie in specialized membrane partitions, such as gap junctions, but rather in the existence of unique cytoplasmic structure, the perforate septa. Mackie and coworkers (4) describe this unique structure and suggest that septa act as sieves able t o control the passage of molecules and organelles, and also to provide pathways of ionic signal conduction in non-nervous tissue (5). Syncytium-to-syncytium,syncytiumto-cell, and cell-to-cell communication is established in the same manner. Therefore, a simple cell and syncytium membrane recognition process suffices to build a multifunctional unit in which coarse and fine-grained areas of differentiation are then set up through the density and distribution of cytoplasmic partitio n septa. It is possible, however, that a certain level or type of differentiation is not attainable within this unit and some cell states or some cell categories require individualization. The observed capacity of Rhabdocalyptus to reorganize following dissociation though a necessary first step, is seemingly not sufficient to give rise to a viable organism. In all our experiments, aggregates became abortive after 2 or 3 weeks. The completion of reorganization appears to be skeleton dependent. Assuming that collagen and spicule secretion are late activities in aggregation bodies (7,12) as in growing sponges, their failure to take place in vitro may be due to lack of an essential requirement in our culture conditions. Several authors, after Brien (13), have reported that large-sized aggregates never reconstitute a functional sponge. In our experiments, even the small-sized aggregates sometimes accompanying larger ones failed to survive.
1
Vol.
6, No.
1
AGGREGATION
IN SPONGES
Finally, the aggregation bodies of Rhabdocalyptus merit further investigation for their own sake. This preliminary study shows that the organization of Hexactinellida,inadequately understood until now (14), strongly differs from that of other sponges which feature more often in phylogenetic consideration of developmental immunology (15-19). The principal characteristic of Rhabdocalyptus is that a multifunctional syncytium (partitioned by septa),rather than multipotent discrete cells, appears to be the basic unit of the recognition and patterning process.
Acknowledgements This work was prompted by Prof. G.O. Mackie and entirely supported by grants from his laboratory, Dpt of Biology, University of Victoria, B.C. Canada. I am gratefully indebted to him and to Dr. C.L. Singla, and also to Drs R.E. Foreman and I. Lawn for their welcome in the Bamfield Marine Station. E.M. examinations were done partly in the University of Victoria, partly in the Universit4 de Lyon (CMEABG) , France, with the technical assistance of Liliane Pavans de Ceccatty. I acknowledge A.S.G. Curtis and G. Van de Vyver for their advice and E. Wurdak for her assistance in revising the manuscript.
REFERENCES
I. IJIMA,I. Studies on the Hexactinellida. I. J. Coll. Sci. Imp. Univ. Tokyo, 15, 1-299, 1901. 2. SCHULZE, F.E. Zur Histologie der Hexactinelliden. Sb. Akad. Wiss. Berlin, 14, 198-209, 1899. 3. REISWIG, H.M.. Histology of Hexactinellida (Porifera). In Sponge Biology, Intern. Coll. CNRS n°291.L4vi C.and BouryEsnault N. eds, CNRS Paris, 173-183, 1979. 4. MACKIE,G.O. and SINGLA, C.L. Histology and ultrastructure of Rhabdocalyptus dawsoni. In preparation. 5. LAWN,I.D., MACKIE,G.O. and SILVER,G . Conduction system in a sponge. Science , 211, 1169-1171, 1981. 6. BUSCEMA, M. et VAN DE VYVER,G. Etude ultrastructurale de l'aggregation des cellules dissoci~es de l'4ponge Ephydatia fluviatilis. ~2 Sponge Biology, Intern. Coll. CNRS n°291. L@vi C. and Boury-Esnault N. eds, CNRS Paris, 225-232,1979. 7. BUSCEMA,M., DE SUTTER, D., and VAN DE VYVER,G.Ultrastructural study of differentiation processes during aggregation of p u r i f i e d sponge archeocytes. W i l h e l m Roux's Arch. Dvptal Biol., 188 , 45-53, 1980. 8. BOROJEVIC, R. et LEVI,C. Etude au microscope @lectronique des cellules de l'@ponge Ophlitaspongia seriata au cours de la r4organisation aprQs dissociation. Z. Zellforsch.,64,708-725,1964. 9. BOROJEVIC, R. et LEVI, C. Morphog4n@se exp4rimentale d'une 4ponge & partir de cellules de la larve nageante dissoci~es. Z. Zellforsch., 68, 57-69, 1965. i0. BRIEN, P. Les Demosponges. Morphologie et reproduction. In Les Spongiaires. Trait~ de Zoologie: Vol III, Grass4 P.P. ed., 133-461, 1973.
21
22
AGGREGATION IN SPONGES
Vol. 6, No. 1
Ii. KOROTKOVA, G.P. Etude morphologique compar~e du d4veloppement des ~ponges ~ partir de cellules dissoci~es. Cah. Biol. Mar., 11, 325-354, 1970. 12. LEVI, C. Reconstitution du squelette de l'~Don~e O p h l i t a s p o n g i a seriata & partir de suspensions cellulaires. Cah. Biol. Mar., i, 353-358, 1960. 13. BRIEN, P. La r4organisation de l'@ponge apr~s dissociation par filtration et ph4nom~nes d'involution chez E p h y d a t i a f l u v i a t i l i s . Arch. Bioi.,48, 185-268, 1937. 14. TUZET, O. Hexactinellides ou Hyalosponges. In Les Spongiaires, Trait~ de Zoologie, Vol. III, Grass~ P.P. ed.,634-690,1973. 15. CURTIS, A.S.G. Cell-cell recognition : positioning and patterning systems. In Cell-cell recognition, Curtis A.S.G. ed., Soc. Exp. Biol., Cambridge Univ. Press, G.B., 51-82, 1978. 16. VAN DE VYVER, G. Cellular mechanisms of recognition and rejection among sponges. In Sponge Biology, Intern. Coll. CNRS n°291, L~vi C; and Boury-Esnault N. eds, CNRS Paris, 195-204, 1979. 17. VAITH, P., MULLER, W.E.G. and UHLENBRUCK~ G. On the role of D-Glucuronic acid in the aggregation of cells from the marine sponge Geodia cydonium. Dev. Comp. Immunol.,3,259-275,1979. 18. VAITH, P., UHLENBRUCK, G., MULLER, W.E.G. and HOLZ, G. Sponge aggregation factor and sponge hemagglutinin : possible relationships between two different molecules. Dev. Comp. Immunol. 3, 399-416, 1979. 19. VAN DE VYVER, G. Second-set allograft rejection in two sponge species and the problem of an alloimmune memory. In Phylogeny of Immunological memory. Manning M.J. ed. ,Elsevier NorthHolland, 15-26, 1980.
Received December 1980 Accepted November 1981
1982.
IN VITRO AGGREGATION OF SYNCYTIA AND CELLS OF A HEXACTINELLIDA SPONGE
Max Pavans de Ceccatty Laboratoire d'Histologie.LA CNRS 244 Universit~ Claude Bernard - Lyon I 43 bd du Ii Novembre 69622 Villeurbanne cedex France
ABSTRACT. Membrane recognition has been analyzed by studying the reaggregation of syncytia and cells dissociated from the tissues of the sponge Rhabdocalyptus dawsoni. Four obvious features are recognized. I) There is true morphogenesis,led by the syncytia, which gives rise to compact,often very large (0.5cm) aggregation bodies. Each body is a giant macrophagic syncytium surrounding groups of discrete undifferentiated cells. 2) Special couplin~ structures, the cytoplasmic perforate septa, establish communication between all body compartments, both syncytial and cellular. The whole aggregate can,therefore, be compared to one vast multinucleated cell.3) During in vitro aggregation, membrane recognition processes permit different syncytia and cells to fuse, and internal compartmentalization is then achieved by the cytoplasmic septa. 4) Fragments of the flagellated choanosome, which is characteristic of sponges, are internalized by the syncytia. Following the collapse of the aggregate they degenerate. The survival of the body appears to depend on skeleton formation.
INTRODUCTION Recognition is related to phenomena as diverse as cell and tissue integrity and the basis for immune processes. Recent data have confirmed earlier disputed reports of Ijima (i) and of Schulze (2) which state that some of the most primitive multicellular animals,the sponges Hexactinellida, have a predominantly syncytial organization. This confirmation (3,4) is all the more important since the Hexactinellida date back to the Precambrian and since sponges of other classes are truly cellular and not syncytial. Hexactinellida thus provide exceptional models for the study of developmental processes involving recognition or rejection of like or unlike tissular components. 15
16
AGGREGATION
IN SPONGES
Vol.
6, No.
Preliminary investigations of Reiswig, in 1979 (3), prompted Mackie and his coworkers (4) to examine in detail,at the light and electron microscope levels, the histology of Rhabdocalyptus dawsoni Lambe 1893. According to both the earlier and recent reports, the sponge body consists of cellular and syncytial portions. The syncytia form a thin dermal membrane at the surface of the sack-shaped individual, as well as general internal trabecular network with specialized regions such as the reticulated layer of the flagellated choanosome.Clearly cellular are : the multipotent archeocytes and cells containing various inclusions. Three modes of interaction between these tissular components are represented : i) total fusion into a portion of syncytium without septa: 2) fusion with partial separation by means of cytoplasmic bridges provided with perforate septa : 3) distinct segregation into totally independent cells. Due to the fact that the ontogenesis of these sponges is still practically unknown, one of the best approaches to the problem is to examine the reaggregation of these components following experimental dissociation of adult tissues. The aim of the present study is to investigate the process of membrane recognition and the potential for morphogenesis in syncytia and cells of Rhabdocalyptus. Although this work is not strictly immunological, it does relate specifically to the general problem of r e c o g n i t i o n which is necessary for immune phenomena to be initiated.
MATERIAL
AND
METHODS
Sponge samples were collected by scuba diving at a depth of 30 meters near the Bamfield Marine Station in Barkley Sound (B.C.Canada). They were kept in tanks of running sea water at 13°C. A total of 12 animals were used for our experiments. Each sponge was cut into small pieces and mechanically dissociated by squeezing through a silk cloth or nylon net with a mesh of 150~. A dense suspension flowed out and was filtered repeatedly through a 100p mesh silk or nylon netting. After being allowed to settle for 15 min. the last filtrate was decanted and gently centrifuged at less than i000 rpm for i minute . The resulting supernatant was evenly divided among 15 to 20 Petri dishes, 5cm in diameter, which were maintained at constant temperature (13°C) by a film of sea water running on the culture table. Microscopic examination of the supernatant showed numerous single cells, mainly archeocytes, plus variously-sized fragments of general trabecular syncytia and of reticulated syncytial choanosome.The proportions of these components put into each Petri dish could not be rigorously controlled. For E.M. observations, adult tissues, reunition bodies and aggregates were systematically fixed and embedded using a variety of procedures in order to circumvent the numerous failures reported by Reiswig(3). All fixatives, buffers and rinsing solutions were adjusted to an osmotic pressure of 900-950 mOsm, that is slightly higher than Baker Sound sea water osmolarity : 850 mOsm. The specimens were never allowed to emerge from the solution. All changes of bathing solutions were applied by perfusion flow which renewed progressively the medium in the same container. In the case of aggregation bodies, the best results were obtained by simple osmium fixation:l% for lh. at 13°C. and by embedding in mixtures of expoxy resins. For normal
1
Vol.
6, No.
1
AGGREGATION
IN SPONGES
sponge tissues used as controls, the method of Mackie and Singla(4) was adopted : that is, simultaneous double fixation in 3% glutaraldehyde and i% osmium. It gave good ultrastructural preservation even after the specimens had been desilicified for I hour in 2% hydrofluoric acid. Thin sections were d o u b l e - s t a i n e d with uranyl acetate and lead citrate. They were viewed with the Philips 300 and Hitachi 12 electron microscopes. The aggregation process was studied in 12 experimental series (one series for one individual), each one comprising 15 to 20 Petri dishes.
RESULTS
In each series,the events which accompany reaggregation were found to be similar in nearly all (90%) containers.At the beginning, syncytia and discrete cells do not display obvious amoeboid behavior nor do they adhere very firmly to the substrate. The flagella of the choanosome fragments create local turbulences in the water causing the suspended elements to collide. These collisions give rise to minute agglomerates which grow, fuse and then, sometimes, spread on the bottom. Consequently the first reunition bodies have various shapes, sizes and compositions. An essential event arises at the end of an initial 24 h. period. Time lapse cinematography (i frame per one or two minutes) reveals gradual fusion of adjoining small agglomerates leading to general retraction of the area occupied by the culture. Thus, new compact aggregates are formed in the Petri dishes. Sometimes, there may be just a single very large body in the form of a circular plate, I cm in diameter. The total duration of the process may last nearly 12 hours. As these phenomena continue, the edges of the aggregate are raised transforming the shape into that of a bowl or a cup...a sponge shape in a word (Fig.l A and B). After 48 hours each dish contains either several polymorphic middle-sized units, or one large spongelike body accompanied by small agglomerates. The most stricking feature of the cytology of the aggregates is the formation of a sort of giant m ul t i n u c l e a t e d macrophage which encloses groups of individual cells, in internal extra-syncytial compartments (Fig.2 A,B). It is made up of components of the general trabecular and epithelial syncytia found in the intact sponge. Many identical nuclei are surrounded by complementary cytoplasm exhibiting various modalities of organelles assembly,just as in normal sponge. In any case, phagocytic activities in the syncytium of the aggregation bodies are considerably enhanced above that which is usually observed in the syncytia of normal sponges. Numerous phagosomes enclose all sorts of debris : fragments of degenerated cells, pieces of spicules and an occasional alga or bacteria.
17
18
AGGREGATION
Pictures
IN SPONGES
FIG. I from time lapse cinematography
Vol.
6, No.
recordings.
Same magnification for both frames : scale 0.6 cm. A) Reunition bodies in a 24h.-old culture, before the process of general retraction. B) One of the compact cup-shaped aggregation bodies , 36 h.-old after general retraction.
During aggregation, all cells in the culture dish are gradually engulfed and internalized by the syncytial unit. The multipotent archeocytes are internalized and do not participate intensively in phagocytosis which is handled almost entirely by the general syncytium, nor they undergo division. Aggregation bodies are the result of the simple reorganization of syncytia and individual cells without nuclear division or archeocyte differentiation. After 24 hours,most of the choanosome fragments, along with discrete cells, are also internalized. The flagella extend into cavities which are completely close and have no relation to the external milieu. Subsequently, collar microvilli and flagella degenerate. The cavities shrink at the same time as the general fusion and retraction of the aggregates. In spite of the drastic simplification achieved during the aggregation process, the basic components of the intact sponges are preserved. These are perforate septa (4), partitioning off cytoplasmic bridges between either two regions of syncytium, or a syncytial compartment on one side and a discrete cell on the other, or between two discrete cells (Fig.2 C,D). Finally, aggregation bodies have no spicules. As a result,instead of having a three dimensional net of soft tissue superimposed on the crystalline lattice of the skeleton, aggregates are compact masses of tissue enclosing some blind cavities.These bodies are not functional. The phase of aggregate involution and disorganization is accompanied by the disruption of the septa. Discrete cells become isolated from each other and from the syncytium. Cytoplasmic organelles of cells and syncytia degenerate and the pycnosis of nuclei starts.
1
Vol. 6, No. 1
AGGREGATION IN SPONGES
FIG.2 Ultrastructure of aggregates A and B : five day-old aggregate. Same magnification for both pictures: scale bar 4pm. A) At the body surface,the macrophagic syncytium encloses normal or degenerating archeocytes. B) Deep in the body, the syncytium is more reticulated and makes contacts with discrete cells by means of dense septa (arrow). C and D : two day-old aggregate. C) Dense perforate septa (arrows) between archeocytes. A continuous channel of smooth reticulum connects the septa. Scale bar 2 ~ . D) Three septa (arrows) couple different syncytium parts. One is crossed by a cisterna of smooth reticulum. Scale bar O.4~m. Abbreviations:(a) archeocytes, (n)syncytium nuclei : (s) syncytium cytoplasm.
19
20
AGGREGATION
IN SPONGES
Vol.
6, No.
DISCUSSION
The in vitro reaggregration of Rhabdocalyptus tissue constituents obtained by experimental dissociation is characterized at first by the gathering of heterogeneous components. These components either fuse into a single syncytium or, in the case of discrete cells, become enclosed in an internal but still extracellular compartment of that syncytium. Therefore the morphogenetic process is led by the syncytium and not dependent on division, recognition and positioning of multipotent archeocytes, as might be expected (6-12). Moreover, communicating bridges provided with perforate septa, present in normal adult tissues (3,4) are also maintained or reestablished by the cytoplasm of aggregation body constituents. These observations point out that cell and tissue recognition mechanisms have sufficient plasticity to allow plasma membrane fusions to occur in a system composed of differentiated syncytial areas and diverse categories of individual cells. A giant, multinucleated body is created through these fusions but the process does not lead to a general uniformity of the cytoplasm or of the membrane themselves. On the contrary, cytoplasm and membranes can continue to differentiate and produce various local structures, at least in normal sponge. The Hexactinellida have thus found an original way of establishing compatibility between the possession, on the one hand, of membranes which are sufficiently similar to be capable of making contact and fusing, and on the other hand, of distinctive perinuclear cytoplasmic compartments having diverse tissular functions. Contrary to the expectation, the key to this compatibility does not lie in specialized membrane partitions, such as gap junctions, but rather in the existence of unique cytoplasmic structure, the perforate septa. Mackie and coworkers (4) describe this unique structure and suggest that septa act as sieves able t o control the passage of molecules and organelles, and also to provide pathways of ionic signal conduction in non-nervous tissue (5). Syncytium-to-syncytium,syncytiumto-cell, and cell-to-cell communication is established in the same manner. Therefore, a simple cell and syncytium membrane recognition process suffices to build a multifunctional unit in which coarse and fine-grained areas of differentiation are then set up through the density and distribution of cytoplasmic partitio n septa. It is possible, however, that a certain level or type of differentiation is not attainable within this unit and some cell states or some cell categories require individualization. The observed capacity of Rhabdocalyptus to reorganize following dissociation though a necessary first step, is seemingly not sufficient to give rise to a viable organism. In all our experiments, aggregates became abortive after 2 or 3 weeks. The completion of reorganization appears to be skeleton dependent. Assuming that collagen and spicule secretion are late activities in aggregation bodies (7,12) as in growing sponges, their failure to take place in vitro may be due to lack of an essential requirement in our culture conditions. Several authors, after Brien (13), have reported that large-sized aggregates never reconstitute a functional sponge. In our experiments, even the small-sized aggregates sometimes accompanying larger ones failed to survive.
1
Vol.
6, No.
1
AGGREGATION
IN SPONGES
Finally, the aggregation bodies of Rhabdocalyptus merit further investigation for their own sake. This preliminary study shows that the organization of Hexactinellida,inadequately understood until now (14), strongly differs from that of other sponges which feature more often in phylogenetic consideration of developmental immunology (15-19). The principal characteristic of Rhabdocalyptus is that a multifunctional syncytium (partitioned by septa),rather than multipotent discrete cells, appears to be the basic unit of the recognition and patterning process.
Acknowledgements This work was prompted by Prof. G.O. Mackie and entirely supported by grants from his laboratory, Dpt of Biology, University of Victoria, B.C. Canada. I am gratefully indebted to him and to Dr. C.L. Singla, and also to Drs R.E. Foreman and I. Lawn for their welcome in the Bamfield Marine Station. E.M. examinations were done partly in the University of Victoria, partly in the Universit4 de Lyon (CMEABG) , France, with the technical assistance of Liliane Pavans de Ceccatty. I acknowledge A.S.G. Curtis and G. Van de Vyver for their advice and E. Wurdak for her assistance in revising the manuscript.
REFERENCES
I. IJIMA,I. Studies on the Hexactinellida. I. J. Coll. Sci. Imp. Univ. Tokyo, 15, 1-299, 1901. 2. SCHULZE, F.E. Zur Histologie der Hexactinelliden. Sb. Akad. Wiss. Berlin, 14, 198-209, 1899. 3. REISWIG, H.M.. Histology of Hexactinellida (Porifera). In Sponge Biology, Intern. Coll. CNRS n°291.L4vi C.and BouryEsnault N. eds, CNRS Paris, 173-183, 1979. 4. MACKIE,G.O. and SINGLA, C.L. Histology and ultrastructure of Rhabdocalyptus dawsoni. In preparation. 5. LAWN,I.D., MACKIE,G.O. and SILVER,G . Conduction system in a sponge. Science , 211, 1169-1171, 1981. 6. BUSCEMA, M. et VAN DE VYVER,G. Etude ultrastructurale de l'aggregation des cellules dissoci~es de l'4ponge Ephydatia fluviatilis. ~2 Sponge Biology, Intern. Coll. CNRS n°291. L@vi C. and Boury-Esnault N. eds, CNRS Paris, 225-232,1979. 7. BUSCEMA,M., DE SUTTER, D., and VAN DE VYVER,G.Ultrastructural study of differentiation processes during aggregation of p u r i f i e d sponge archeocytes. W i l h e l m Roux's Arch. Dvptal Biol., 188 , 45-53, 1980. 8. BOROJEVIC, R. et LEVI,C. Etude au microscope @lectronique des cellules de l'@ponge Ophlitaspongia seriata au cours de la r4organisation aprQs dissociation. Z. Zellforsch.,64,708-725,1964. 9. BOROJEVIC, R. et LEVI, C. Morphog4n@se exp4rimentale d'une 4ponge & partir de cellules de la larve nageante dissoci~es. Z. Zellforsch., 68, 57-69, 1965. i0. BRIEN, P. Les Demosponges. Morphologie et reproduction. In Les Spongiaires. Trait~ de Zoologie: Vol III, Grass4 P.P. ed., 133-461, 1973.
21
22
AGGREGATION IN SPONGES
Vol. 6, No. 1
Ii. KOROTKOVA, G.P. Etude morphologique compar~e du d4veloppement des ~ponges ~ partir de cellules dissoci~es. Cah. Biol. Mar., 11, 325-354, 1970. 12. LEVI, C. Reconstitution du squelette de l'~Don~e O p h l i t a s p o n g i a seriata & partir de suspensions cellulaires. Cah. Biol. Mar., i, 353-358, 1960. 13. BRIEN, P. La r4organisation de l'@ponge apr~s dissociation par filtration et ph4nom~nes d'involution chez E p h y d a t i a f l u v i a t i l i s . Arch. Bioi.,48, 185-268, 1937. 14. TUZET, O. Hexactinellides ou Hyalosponges. In Les Spongiaires, Trait~ de Zoologie, Vol. III, Grass~ P.P. ed.,634-690,1973. 15. CURTIS, A.S.G. Cell-cell recognition : positioning and patterning systems. In Cell-cell recognition, Curtis A.S.G. ed., Soc. Exp. Biol., Cambridge Univ. Press, G.B., 51-82, 1978. 16. VAN DE VYVER, G. Cellular mechanisms of recognition and rejection among sponges. In Sponge Biology, Intern. Coll. CNRS n°291, L~vi C; and Boury-Esnault N. eds, CNRS Paris, 195-204, 1979. 17. VAITH, P., MULLER, W.E.G. and UHLENBRUCK~ G. On the role of D-Glucuronic acid in the aggregation of cells from the marine sponge Geodia cydonium. Dev. Comp. Immunol.,3,259-275,1979. 18. VAITH, P., UHLENBRUCK, G., MULLER, W.E.G. and HOLZ, G. Sponge aggregation factor and sponge hemagglutinin : possible relationships between two different molecules. Dev. Comp. Immunol. 3, 399-416, 1979. 19. VAN DE VYVER, G. Second-set allograft rejection in two sponge species and the problem of an alloimmune memory. In Phylogeny of Immunological memory. Manning M.J. ed. ,Elsevier NorthHolland, 15-26, 1980.
Received December 1980 Accepted November 1981