Transdifferentiation of larval flagellated cells to choanocytes in the metamorphosis of the demosponge Haliclona permollis.
Most sponges, which are sessile as adults, have free-swimming larvae in their life cycles. The larvae use a surface layer of flagellated cells to swim around for a while, then settle on a substratum. Soon after the onset of sessile life, the locomotive organ disappears from the larval surface. Such metamorphs without flagella are chaotic in organization; they consist mainly of several types of amoeboid cells (Brien and Meewis, 1938). Sometime later, a canal system and choanocyte chambers are formed in the metamorphs. The fate of larval cells during metamorphosis and the origin of the choanocytes of juveniles have been debated for many years among sponge biologists (Fell, 1974; Simpson, 1984; Kaye and Reiswig, 1991; Woollacott, 1993).
In calcareous sponges, larval flagellated cells are transformed into the choanocytes of a juvenile during metamorphosis (Duboscq and Tuzet, 1937; Amano and Hori, 1992, 1993). In demosponges, however, two views have been presented: one is that the flagellated cells are transformed into choanocytes, as in calcareous sponges (Levi, 1956; Borojevic and Levi, 1965; Boury-Esnault, 1976); the other is that during metamorphosis the flagellated cells are lost by exfoliation or phagocytosis and not involved in the formation of juveniles (Brien and Meewis, 1938; Harrison and Cowden, 1975; Bergquist and Glasgow, 1986; Misevic et al., 1990; Kaye and Reiswig, 1991). In the latter view, choanocytes are thought to derive from archeocytes.
Because larval flagellated cells change radically in morphology in the early stages of metamorphosis, it is almost impossible to follow their developmental fates unless the cells have some identification markers. We took advantage of two natural markers: the flagellar axoneme retracted in the cell body and the minute ellipsoid granules characteristic of the larval flagellated cells of Haliclona permollis. Because both of these markers are very small and difficult to visualize, a large number of good-quality electron micrographs of metamorphosing larvae were necessary to reveal the transformation process of the larval flagellated cells. An improved fixation method (Amano and Hori, 1992, 1994) made it possible to obtain electron microscopic images of sponge metamorphs.
In this report, we show that larval flagellated cells transform into choanocytes during metamorphosis. This process of cellular transformation is thought to be another example of transdifferentiation (Schmid, 1992).
Materials and Methods
Sponges and larvae
In June 1993 and 1994, colonies of H. permollis were collected from rafts in Mutsu Bay in Aomori prefecture in northern Japan. They were immediately placed in containers with seawater, brought to the laboratory of Asamushi Marine Biological Station within 1 h, and kept in running seawater.
The exact taxonomy of our specimens is uncertain. Not only is H. permollis thought to be synonymous with H. cinerea (de Weerdt, 1986), but Japanese H. permollis is probably different from European H. permollis (or H. cinerea), and may be a new species. For the time being, the sponge used in this study should be referred to as H. permollis sensu Tanita. The sponge specimens of H. permollis sensu Tanita are now kept by us and are accessible on request. For brevity, however, we refer here to this Japanese sponge simply as H. permollis.
Haliclona permollis releases larvae in early summer in Japan. In the laboratory, only about one-fifth of the sponges released larvae, and most of these released larvae every day for 5 days or more. To collect larvae, we placed adults in still seawater at about 0530. Under natural illumination, larval release began soon after dawn and ceased at about 1100. Larvae began swimming as soon as they were released.
Free-swimming larvae could be picked up on a platinum loop and placed into fixative. Metamorphosing larvae were easily broken during fixation and dehydration and required different handling. Larvae that settled beneath the air-water interface were very adhesive and were easily attached to a small piece (about 1.0 X 2.0 mm) of membrane filter (Millipore, 0.45-[[micro]meter] pore size) that sank swiftly in media, making centrifugation unnecessary and reducing the mechanical stress on the larvae. After fixation and dehydration, the membrane filter was dissolved in propylene oxide, a clearing agent, so that the metamorphs could be embedded without the piece of filter.
Larvae and metamorphs attached to the small piece of filter were fixed in ice-cold 0.1 M cacodylate buffer (pH 7.4) containing 3.0% glutaraldehyde and 1.0% paraformaldehyde. For each 100 ml of the fixative, 14 g sucrose and 0.05 g anhydrous calcium chloride were added. After 2 h, samples were rinsed twice in 0.1 M cacodylate buffer and postfixed in 1.0% osmium tetroxide in 0.1 M cacodylate buffer (pH 7.4) for 1 h. Samples were dehydrated through a graded ethanol series, cleared in propylene oxide, and embedded in Spurr epoxy resin (Spurr, 1969). Semithin sections were stained with toluidine blue for light microscopy. Ultrathin sections were stained with uranyl acetate and lead citrate, then examined and photographed using a Hitachi H-500 electron microscope.
Flagellated cells of free-swimming larvae
The free-swimming parenchymella larvae of Haliclona permollis are about 250 [[micro]meter] in length and 180 [[micro]meter] in width. Flagellated cells make up the entire larval surface, except at the posterior end where many reddish brown pigment granules occur in the cytoplasm.
Parenchymella larvae were fixed for electron microscopy on the morning of the day of their release. The larval surface region consists of a pseudostratified columnar epithelium of elongate flagellated cells; however, a basal lamina is absent [ILLUSTRATION FOR FIGURE 1 OMITTED]. At the free surface of these flagellated cells is a flagellar socket composed of a pit about 2.4 [[micro]meter] deep and 0.8 [[micro]meter] wide; a single flagellum emerges from the bottom of the socket. Flagella possess typical axonemes with the 9 + 2 arrangement of microtubules. The proximal end of the flagellar axoneme is somewhat electron-dense, terminating in a cylindrical basal body [ILLUSTRATION FOR FIGURE 2 OMITTED]. The long flagellar rootlet is seen as a tuft of filamentous matter without striations. Elongate mitochondria (about 1.2 X 0.3 [[micro]meter]) with well-developed cristae occur regularly around the rootlet.
Flagellated cells have the smallest nucleus (about 3.2 [[micro]meter] in length and 1.8 [[micro]meter] in width) among larval cell types. The nucleus is always located in the basal region of the cytoplasm, and the slender intermediate region of the cell connects the nuclear region with the apical region of the cell [ILLUSTRATION FOR FIGURE 1 OMITTED]. Thus the deeper the nucleus occurs, the longer the flagellated cell. We often found nucleolated flagellated cells, but the ratio of anucleolated to nucleolated is unknown. Flagellated cells contain a Golgi apparatus on the free-surface side of a nucleus, elliptical mitochondria, a few phagosomes, various vesicles, and free ribosomes.
We found minute electron-dense granules concentrated in the apical cytoplasm near the free surface of larval flagellated cells [ILLUSTRATION FOR FIGURES 1, 2 OMITTED!. These ellipsoid granules are always encapsulated by a limiting membrane and are usually smaller than 0.5 [[micro]meter] in length and 0.2 [[micro]meter] in width. Sometimes in section they appear square or cylindrical, but highly magnified images of these granules reveal a crystalloid substructure [ILLUSTRATION FOR FIGURE 3 OMITTED]. No minute ellipsoid granules are found in any larval cells except the flagellated cells.
Residual flagellar axoneme in amoeboid cells
After a free-swimming period of 12-36 h, larvae settled on the substratum (a glass surface or an air-water interface) and quickly transformed from an ellipsoid form to a convex disk. For a while after settlement, a gathering of reddish brown pigment granules originating from the posterior end of the larva was visible in the central region of the convex disk. These pigment granules disappeared within 12 h after settlement.
To observe the initial stages of metamorphosis, we fixed the settled larvae that still possessed concentrates of pigment granules (that is, were within 12 h after settlement) for electron microscopy. Figure 4 shows that flagellated cells are losing their characteristic morphology and the pseudostratified columnar epithelium is in the course of disorganization at the surface of such metamorphosing larva. Figure 5 shows a flagellar stub, a remnant of a flagellum on the cell body, at high magnification, and it is apparent that its flagellum was severed at the transitional region between the basal body and the proximal end of the axoneme. The excised flagellar axonemes were promptly retracted into the cell body [ILLUSTRATION FOR FIGURES 4, 5 OMITTED!. Only the axoneme without a flagellar membrane is withdrawn into the cytoplasm. As many as four sections of the axoneme were found per cell, so it must be coiled or cut into a number of pieces in the cytoplasm.
Amoeboid cells derived from flagellated cells migrate to the inner part of the metamorphosing larva; these amoeboid cells have residual flagellar axonemes [ILLUSTRATION FOR FIGURES 6, 7 OMITTED]. Some axonemes are in the process of rapid decomposition, but others still maintain the typical 9 + 2 microtubule construction [ILLUSTRATION FOR FIGURE 7 OMITTED].
Figure 8 shows the distribution of diameter of the nuclei of nucleolated amoeboid cells and archeocytes in the metamorphs. As both of these cell types are amoeboid cells with pseudopodia and it is difficult to estimate their cell size precisely, the diameter of their nuclei was measured instead. Two peaks are evident in the size distribution of the nuclei, one at about 3.4 [[micro]meter], and the other at 2.3 [[micro]meter]. By comparison with amoeboid cells in the larval interior, we conclude that the nucleolated cells whose nuclei are larger than 3.0 [[micro]meter] are archeocytes. These archeocytes are large cells with numerous phagosomes [ILLUSTRATION FOR FIGURE 9 OMITTED], and residual axonemes never occur in such cells. In contrast, residual flagellar axonemes are observed in nucleolated cells whose nuclei are smaller than 3.0 [[micro]meter] (solid parts of the columns in Fig. 8). The proportion of cells with residual axonemes (7/41) among nucleolated amoeboid cells is thought to be an underestimate because the residual axoneme is small and easily overlooked. We conclude that two types of nucleolated cells occur in the metamorphosing larva: one is the archeocytes and the other is the amoeboid cells derived from the larval flagellated cells.
Minute ellipsoid granules in amoeboid cells and choanocytes
We have shown that minute ellipsoid granules are characteristic of the larval flagellated cells of H. permollis [ILLUSTRATION FOR FIGURES 1, 2 OMITTED!. In the initial stages of metamorphosis, those granules are found in amoeboid cells that have residual flagellar axonemes [ILLUSTRATION FOR FIGURE 9 OMITTED]. The nuclei of these cells are small ([less than] 3.0 [[micro]meter]) and have heterochromatin masses. The large archeocyte in Figure 9 has, however, neither minute ellipsoid granules nor residual axonemes. Figure 10 shows a nucleolated amoeboid cell with two sections of residual axoneme and a minute ellipsoid granule. The minute ellipsoid granule together with the residual axoneme evidently reveal the origin of this amoeboid cell.
Twenty-four hours after settlement, the metamorphosing larvae have a convex disk shape and appear chaotic in organization. The concentration of reddish brown pigment granules is no longer evident and spongin fiber formation has begun. Flagellar axonemes have already decomposed, but the minute ellipsoid granules characteristic of larval flagellated cells still occur in the larval cells at this stage of metamorphosis. In Figure 11, a nucleolated cell from a 24-h metamorph can be seen to contain the minute ellipsoid granules. This cell has a large nucleolus and phagosomes. Figure 12 shows similar amoeboid cells with minute ellipsoid granules, but no nucleus is visible in these cells. Phagosomes are more numerous in amoeboid cells with minute ellipsoid granules from 24-h metamorphs than from 12-h metamorphs.
About 36 h after settlement, flagellar chamber formation began in the metamorphosing larvae. Figure 13 shows a flagellar chamber in the course of development and one choanocyte with a minute ellipsoid granule. Figure 14 shows a granule in such a choanocyte at high magnification. We found fewer ellipsoid granules in 36-h metamorphs than in 24-h metamorphs. Figure 15 shows a more developed flagellar chamber in a metamorph 48 h after settlement. The flagellum of a choanocyte is encircled by a collar of microvilli which are swollen in places. The flagellar basal bodies are not associated with rootlets. The choanocytes usually contain a moderate number of phagosomes. Minute ellipsoid granules are not found in such choanocytes.
Two natural markers
Because the flagellated cells of sponge larvae change their anatomy very quickly and markedly during metamorphosis, it is difficult to elucidate the steps of this transformation. In this study, therefore, we employed two natural markers. One of these, the flagellar axoneme retracted in the cell body, was rapidly decomposed and all of them disappeared from the cell between 12 and 24 h after larval settlement. The other marker, the presence of minute ellipsoid granules, is characteristic of the larval flagellated cells of Haliclona permollis, but similar granules were not found in the larvae of another Haliclona sp. (Amano and Hori, 1994). The minute ellipsoid granules are composed of a crystalloid substance, but their chemical nature is unknown. These granules disappeared from flagellated cells between 36 and 48 h after settlement. Therefore, in future studies, the fate of the larval flagellated cells should be traced for longer durations, and also in other sponges.
Loss of flagella
Soon after settlement, sponge larvae lose the flagella in the surface layer of cells. In H. permollis, the flagella are severed at the transitional region between the basal body and the proximal end of the axoneme, as in other animals (Blum, 1971; Sanders and Salisbury, 1989). In the amphiblastula larvae of calcareous sponges, flagellated cells shed and discard their excised flagella (Amano and Hori, 1993). In the parenchymella larvae of H. permollis, however, the flagellated cells withdrew the excised flagellar axonemes into the cell body. This latter course may be common among demosponges because similar internalized axonemes were observed in the metamorphosing larvae of other demosponges (Levi, 1964; Boury-Esnault, 1976). Thus, the residual flagellar axonemes are found in the amoeboid cells of metamorphosing parenchymellae, but not in those of the metamorphosing amphiblastulae.
Amoeboid cells with minute ellipsoid granules and axonemes in the cell body
De-flagellated larval epithelia do not invaginate in a cell sheet, but are disarranged into individual amoeboid cells that each migrate to the inner region of a metamorph by amoeboid movement. Two natural markers are found in the amoeboid cells of metamorphs within 12 h after settlement. These cells retain a basal body and a rootlet in the initial stage of metamorphosis, but these intracellular components of flagella disappear rapidly from the cytoplasm. The minute ellipsoid granules that were concentrated in the apical region of larval flagellated cells are dispersed in the cytoplasm of the amoeboid cells. It is obvious that the amoeboid cells with these markers derive from larval flagellated cells. Like archeocytes, some of the amoeboid cells have a nucleolus, but these cells are smaller and have fewer phagosomes than the archeocytes. It is apparent that the archeocytes have no residual axonemes or minute ellipsoid granules.
The flagellar axoneme observed in the amoeboid cells is evidently the one that was retracted in the cell body of the flagellated cells. The alternative possibility - that it may represent a flagellated cell phagocytosed by other larval cells or archeocytes - is excluded for the following reasons. (1) The flagellar axoneme, excised at the transitional region, is found in cells with a residual basal body and a rootlet. (2) In the cytoplasm of the amoeboid cells, the flagellar axoneme invariably occurs naked, without a covering membrane, and is never found in phagosomes. This is because only the axoneme, not the flagellar membrane, is severed and taken into the cell body. Axonemes without covering membranes are not found in the intercellular space of metamorphosing larvae. (3) In the early stages of metamorphosis, the flagellar axonemes are found only in amoeboid cells whose nuclei are smaller than 3.0 [[micro]meter]. Archeocytes, whose nuclei are larger than 3.0 [[micro]meter] are phagocytes with many large phagosomes, but no flagellar axonemes occur within these cells or especially within their phagosomes.
Origin of choanocytes in H. permollis
Although residual axonemes have already decomposed 24 h after settlement, minute ellipsoid granules still occur in amoeboid cells that have migrated into the inner region of metamorphs. Because these amoeboid cells usually have more phagosomes than those in earlier stages, they are probably growing in the metamorphs. We suggest that such amoeboid cells are precursor cells of choanocytes.
Minute ellipsoid granules were found in choanocytes that were still in the course of differentiation about 36 h after settlement. It is likely that these choanocytes derive from amoeboid cells originated from flagellated cells, but the possibility that some of the choanocytes originate from archeocytes cannot be excluded. Because large numbers of choanocytes are necessary for the flagellar chamber formation of juveniles, multiple progenitor cell types cannot be ruled out. About the time of flagellar chamber formation, the number of minute ellipsoid granules decreases rapidly, and these granules are not found in fully developed choanocytes 48 h after onset of metamorphosis. On the basis of our findings, we suggest that larval flagellated cells are transformed into the choanocytes of the juvenile of H. permollis by the process illustrated in Figure 16.
Origin of choanocytes in other sponges
In calcareous sponges, the consensus is that the choanocytes of juveniles derive from larval flagellated cells (Duboscq and Tuzet, 1937; Amano and Hori, 1993). Similarly, in the demosponge Polymastia robusta, the choanocytes probably derive from larval flagellated cells, because the blastula larva of this sponge consists only of a layer of flagellated cells (Borojevic, 1967). Brien and Meewis (1938) studied the process of metamorphosis in Ephydatia fluviatilis, a freshwater sponge, and concluded that all larval flagellated cells are phagocytosed by archeocytes and play no role in the formation of juvenile sponges. It is likely, however, that light microscopy lacked sufficient resolution to show whether the amoeboid cells derived from the flagellated cells were positioned within or close to the archeocytes. In Microciona prolifera, Misevic et al. (1990) labeled the flagellated cells of free-swimming larvae with 125I. The label was found in the phagosomes of archeocytes in the metamorphs, but not in the choanocytes of juveniles, leading the authors to conclude that all of the flagellated cells were phagocytosed by archeocytes during metamorphosis and did not give rise to choanocytes. The origin of the choanocytes was, however, not shown because the label was not found in the phagosomes of choanocytes, which are thought to derive from archeocytes.
Our conclusions in the present paper are not contrary to the concept that the archeocytes are multipotential and capable of differentiating into choanocytes. In cases where a supply of larval flagellated cells is not available, choanocytes can derive from archeocytes. For example, within the free-swimming larvae of Ephydatia fluviatilis, choanocytes should be differentiated from the archeocytes precociously before larval settlement (Brien and Meewis, 1938). Also, in juvenile sponges reconstituted from the excised central part of parenchymella larvae (Borojevic, 1966) or from cell aggregates of purified archeocytes (Buscema et al., 1980), choanocytes evidently derive from archeocytes. In contrast, when sufficient numbers of choanocytes are supplied by larval flagellated cells, the archeocyte differentiation to choanocytes is probably unnecessary. In H. permollis, some surplus larval flagellated cells seem to be degenerated and phagocytosed by other larval cells during metamorphosis, because we found only a few picnotic cells in metamorphs. Borojevic (1966) argued that the archeocytes are reserved embryonal cells that regulate the equilibrium between the different categories of cells. The archeocytes should eliminate excessive cells by phagocytosis and, on the other hand, differentiate into deficient cell types to fill the shortage.
Transdifferentiation is the process by which cells that have already expressed specific differentiated traits change into another cell type distinguished from the original cells by a set of phenotypic characters (Okada, 1991; Schmid, 1992). It has been demonstrated experimentally that fully specialized somatic cells can undergo a change in their cellular commitment and gene expression. Under certain culture conditions, the retinal pigment epithelium of chick embryos can transdifferentiate to lentoid cells that express typical lens crystallins (Itoh and Eguchi, 1986). In a medusa, Schmid and Alder (1984) showed that striated muscle cells can undergo pluripotent transdifferentiation and form a complex regenerate under some experimental conditions.
In the early stages of the metamorphosis of H. permollis, the highly differentiated flagellated cells of larvae dedifferentiate to amoeboid cells that subsequently differentiate to choanocytes in juveniles. The process of this transformation is not simply the regeneration of an excised flagellum (Sanders and Salisbury, 1989), because the flagellar basal body and rootlet have been completely lost in the amoeboid cells. Thus the results of the present study demonstrate that the flagellated cells of H. permollis larvae transdifferentiate into the choanocytes of juveniles by way of an intermediate amoeboid cell stage.
We express our gratitude to Dr. T. Numakunai and staff of the Asamushi Marine Biological Station of Tohoku University for their hospitality and help during the stay of S. A. We thank also Mr. S. Tamura and Mr. M. Washio who helped in collecting the sponges and Dr. Y. Watanabe of Ochanomizu University, who identified the sponges.
Amano, S., and I. Hori. 1992. Metamorphosis of calcareous sponges. I. Ultrastructure of free-swimming larvae. Invertebr. Reprod. Dev. 21: 81-90.
Amano, S., and I. Hori. 1993. Metamorphosis of calcareous sponges. II. Cell rearrangement and differentiation in metamorphosis. Invertebr. Reprod. Dev. 24: 13-26.
Amano, S., and I. Hori. 1994. Metamorphosis of a demosponge. I. Cells and structure of swimming larva. Invertebr. Reprod. Dev. 25: 193-204.
Bergquist, P. R., and K. Glasgow. 1986. Developmental potential of ciliated cells of ceractinomorph sponge larvae. Exp. Biol. 45: 111-122.
Blum, J. J. 1971. Existence of a braking point in cilia and flagella. J. Theor. Biol. 33: 257-263.
Borojevic, R. 1966. Etude experimentale de la differenciation des cellules de l'eponge au cours de son developpement. Dev. Biol. 14: 130-154.
Borojevic, R. 1967. La ponte et le developpement de Polymastia robusta (Demosponges). Cah. Biol. Mar. 8: 1-6.
Borojevic, R., and C. Levi. 1965. Morphogenese experimentale d'une eponge a partir de cellules de la larve nageante dissociee. Z. Zellforsch. Mikrosk. Anat. 68: 57-69.
Boury-Esnault, N. 1976. Ultrastructure de la larve parenchymella d'Hamigera hamigera (Schmidt) (Demosponge, Poecilosclerida). Origine des cellules grises. Cah. Biol. Mar. 17: 9-20.
Brien, P., and H. Meewis. 1938. Contribution a l'etude l'embryogenese des Spongillidae. Arch. Biol. 49: 177-250.
Buscema, M., D. De Sutter, and G. Van de Vyver. 1980. Ultrastructural study of differentiation process during aggregation of purified sponge archeocytes. Roux's Arch. Dev. Biol. 188: 45-53.
de Weerdt, W. H. 1986. A systematic revision of the north-eastern Atlantic shallow-water Haplosclerida (Porifera, Demospongiae), Part II: Chalinidae. Beaufortia 28: 81-165.
Duboscq, O., and O. Tuzet. 1937. L'ovogenese, la fecondation et les premiers stades du developpement des eponges calcaires. Arch. Zool. Exp. Gen. 79: 157-316.
Fell, P. E. 1974. Porifera. Pp. 51-132 in Reproduction of Marine Invertebrates. Vol. 1. Acoelomate and Pseudocoelomate Metazoans, A. C. Giese and J. S. Pearse, eds., Academic Press, New York.
Harrison, F. W., and R. R. Cowden. 1975. Cytochemical observations of larval development in Eunapius fragilis (Leidy): Porifera; Spongillidae. J. Morph. 145: 125-142.
Itoh, Y., and G. Eguchi. 1986. In vitro analysis of cellular metaplasia from pigment epithelial cells to lens phenotypes: a unique model system for studying cellular and molecular mechanisms of "trans-differentiation." Dev. Biol. 115: 353-362.
Kaye, H. R. and H. M. Reiswig. 1991. Sexual reproduction in four Caribbean commercial sponges. III. Larval behaviour, settlement and metamorphosis. Invertebr. Reprod. Dev. 19: 25-35.
Levi, C. 1956. Etude des Halisarca de Roscoff. Embryologie et systematique des demosponges. Arch. Zool. Exp. Gen. 93: 1-181.
Levi, C. 1964. Ultrastructure de la larve parenchymella de demosponge. 1. Mycale contarenii (Martens). Cab. Biol. Mar. 5: 97-104.
Misevic, G. N., V. Schlup, and M. M. Burger. 1990. Larval metamorphosis of Microciona prolifera: evidence against the reversal of layers. Pp. 182-187 in New Perspectives in Sponge Biology, K. Rutzler, ed., Smithonian Inst. Press, Washington, DC.
Okada, T. S. 1991. Transdifferentiation. Clarendon Press, Oxford.
Sanders, M. A., and J. L. Salisbury. 1989. Centrin-mediated micro-tubule severing during flagellar excision in Chlamydomonas reinhardtii. J. Cell Biol. 108: 1751-1760.
Schmid, V. 1992. Transdifferentiation in medusae. Int. Rev. Cytol. 142: 213-261.
Schmid, V., and H. Alder. 1984. Isolated, mononucleated, striated muscle can undergo pluripotent transdifferentiation and form a complex regenerate. Cell 38: 801-809.
Simpson, T. L. 1984. The Cell Biology of Sponges. Springer-Verlag, New York.
Spurr, A. R. 1969. A low-viscosity epoxy resin embedding medium for electron microscopy. J. Ultrastruct. Res. 26: 31-43.
Woollacott, R. M. 1993. Structure and swimming behaviour of the larva of Haliclona tubifera (Porifera: Demospongiae). J. Morph. 218: 301-321.
|Printer friendly Cite/link Email Feedback|
|Author:||Amano, Shigetoyo; Hori, Isao|
|Publication:||The Biological Bulletin|
|Date:||Apr 1, 1996|
|Previous Article:||Stimulants of feeding behavior in fish: analyses of tissues of diverse marine organisms.|
|Next Article:||Endogenous beta-galactosidase activity in the larval, pupal, and adult stages of the fruit fly, Drosophila melanogaster, indicates need for caution...|