The developmental role of the extracellular matrix suggests a monophyletic origin of the kingdom Animalia.
Key words. - Animalia, collagen, development, fibronectin, gastrulation, homology, integrin, Porifera.
The kingdom Animalia is considered, by many modern workers, to be a polyphyletic group. This concept of the kingdom is distinctly different from the view held during the last century. Most nineteenth century workers considered the kingdom to be a coherent group of related organisms. Over the last hundred years numerous workers have suggested one, two, three, and more independent origins of animals from unicellular or colonial protists. While it has been variously suggested that deuterostomes (Nursall, 1962), protostomes (Inglis, 1985), and cnidarians (Hanson, 1977) independently evolved multicellularity, the group most commonly considered to have been derived from a separate protist lineage is the sponges, phylum Porifera (as evidenced by Barnes, 1980, p. 74). Salvini-Plawen (1978) has argued for a monophyletic origin of animals, but the available ultrastructural and morphological evidence remains unconvincing (e.g., Willmer, 1990, p. 195). On the basis of this evidence, it is difficult to clearly distinguish between single or multiple origins of animals from flagellate protists. The relationship between sponges and the other animals is the central question for the validity of the kingdom Animalia as a monophyletic group of multicellular organisms.
Near the end of the nineteenth century, two anatomists separated sponges from other animals. Sollas (1884) considered sponges to be an independent group because they have a simple organization and sponge choanocytes bear a strong resemblance to choanoflagellate protozoa. Delage (1892) noted dissimilarities in the details of early embryology of sponges and other animals. Their observations underlie modern arguments over the position of sponges. During this century, three major hypotheses for the relationship between sponges and other animals have been discussed. The Animalia has been considered: polyphyletic, but derived from a common choanoflagellate ancestor; monophyletic; and polyphyletic with origins amongst both ciliates and flagellates. While a great many workers have discussed this issue, Hyman (1940), Beklemishev (1969), and Hanson (1977) have variously provided lucid statements of these positions. These authors differ in their interpretations of gastrulation and in their allegiance to adaptive scenarios. Two such scenarios have been widely discussed. Haeckel's (1874) colonial theory posits that animals evolved through unicellular, blastea, and gastrea stages. Syncytial theories, following Hadzi (1953, 1963), derive some animals from multinucleated ciliates through a process of cellularization, similar to that observed in insect embryos.
The taxonomic distribution of "collagen" has been used to support the hypothesis of a monophyletic origin of animals (Garrone, 1978). Unfortunately, three distinct, possibly convergent, families of collagens exist within vertebrates alone (Olsen and Nimni, 1989). Similarity in a single molecule, such as collagen, is a weak basis for phylogenetic hypotheses, especially when evidence for convergence exists. The presence of collagen alone is not a good homology.
Homologies have classically been identified by precise complex correspondences between anatomical features. They are easier to demonstrate in complex anatomical structures than in simple ones. Complex molecular systems may likewise be more reliably compared than single molecules, since functional or structural constraints upon individual molecules may mask convergence (Morris and CoBabe, 1991). I argue that the developmental role of the extracellular matrix (ECM), a comilex set of molecules including collagens, is a primitive feature of multicellularity in animals, and that it provides evidence that animals are monophyletic and derived from a common multicellular ancestor (Fig. 1).
Interactions between cells and extracellular molecules play a fundamental role in the development of the "higher" animals. Two basic types of cells exist in a developing embryo: sheet-forming epithelial cells and motile mesenchymal cells. The properties of these cells and transitions between cell types are mediated by cell-matrix interactions (Hay, 1981, 1984; Greenburg and Hay, 1988). Interactions of cells with collagens, proteoglycans, and adhesive glycoproteins are the mechanism behind fundamental developmental events (e.g., gastrulation, neural crest migration, and epithelial gland formation). While the temporal and spatial distribution of these events is controlled at other levels, the ECM provides the machinery to carry them out. (Other important aspects of animal multicellularity, such as cell-cell interactions, recognition of self and nonself, and control of developmental timing, will not be considered here.)
The same complex set of molecules found in vertebrate extracellular matrices also appears to be involved in cell motility and development in sponges. I argue that this complex system evolved once in a multicellular common ancestor of both sponges and other animals. This major adaptive complex (sensu Frazzetta, 1975) is considered to be a fundamental mechanism for the production of multicellularity in animals and a homology that unites the kingdom Animalia into a monophyletic group of multicellular organisms.
Three features of animals are at the center of the debate over the position of the sponges. These are: the presence of choanocytes in sponges, the perceived simplicity of sponges relative to other animals, and comparisons of early embryological events (other characters, e.g., Bardele, 1983; Garrone and Lethias, 1990; Warrior and Gall, 1985; Salvini-Plawen, 1978; Franzen, 1956, will not be considered here). Choanocytcs have been discredited as evidence separating sponges and other animals because similar, and presumably homologous, cells are found in many animal phyla, including echinoderms (Bergquist, 1985; Leadbeater, 1983; Willmer, 1990, p. 139). These data support the view that all animals are derived from choanoflagellates, but cannot differentiate between single or multiple origins of multicellularity.
Anatomically, sponges are simple. They do not produce complex patterned structures (i.e., organ systems). However, an examination of any good treatment of sponges will demonstrate histological complexity (e.g., Vos et al., 1991). Simpson (1984) lists some 34 differentiated cell types (excluding reproductive and embryonic cells), most of which are specialized motile cells that roam the mesohyl. While this is an order of magnitude fewer than the 200 cell types in a human, it is surely comparable to a flatworm or to the 16 major cell types in Caenorhabditis elegans (Davidson, 1986, Fig. 4.1).
The strongest traditional lines of evidence used to examine the relationships between animal phyla are embryological characters. A variety of unusual embryos exist in sponges (e.g., parenchymella, amphiblastula; Fig. 2G, H, J, K). As noted by Delage (1892), these are not precisely comparable with the gastrulae of other animals (Fig. 2A-F), and cannot be considered homologous. [This it the case even if germ layer inversion does not occur (Misevic et al., 1990) as proposed by Delage.] However, as noted by Beklemishev (1969, p. 24), the underlying pattern of early embryology does seem similar in all animals. A few sponge embryos invaginate and ingress in a manner similar to other animals (Fig. 2I). Later in their ontogeny, sponges clearly exhibit the same dichotomy of epithelial cell sheets and motile mesenchymal cells observed in the embryology of other animals (Fig. 2L).
Gastrulation in all animals does seem to be represented by some combination of ingression and invagination of cells. On an anatomical level, this perceived similarity is difficult to test for homology. The very simplicity of these early embryological events makes comparisons difficult. Characters of greater complexity than the presence of some aspect of gastrulation are needed to evaluate the relationship between sponges and other animals.
While molecular sequence similarities offer great potential for the testing of phylogenetic hypotheses, they have not clearly placed sponges. A monophyletic kingdom Animalia is supported by 5S rRNA sequences (Hori and Osawa, 1987) that place both cnidarians and sponges within the kingdom (however, doubt has been cast upon the utility of 5S sequences for phylogenetic analysis, Halanych, 1991). In contrast, 18S ribosomal RNA sequences (Field et al., 1988) initially produced a tree in which cnidarians appeared to have an independent origin amongst the protists (no sponge was included in this data set). However, reanalyses of these 18S data suggest that the most parsimonious tree places cnidarians within the metazoa (Lake, 1990; Patterson, 1989). Erwin (1989) has also criticized applications of molecular sequence data to the origin of animal phyla. If the major phyla of animals appeared rapidly about 600 million years ago, then these conflicting results may be showing us variation in molecular clock rates, not information about relationships.
Without a clear phylogeny of the major groups of protists, and without inclusion of a suitable sister taxon (choanoflagellates, chytrids) for the multicellular animals, these sequence data are unable to falsify the hypothesis of multiple independent origins of multicellularity among the Animalia. Even an unambiguous tree that, for example, placed a protist lineage as a sister taxon to animals, and sponges as a sister taxon to other animals would be incapable of ascertaining whether the common ancestor of sponges and other animals was multicellular.
While molecular sequence data are in widespread use for the testing of phylogenetic hypotheses, comparative biochemistry is in disrepute. Occurrences of a single molecule, such as chitin or creatine phosphate, across a few taxa of invertebrates are a dangerous basis for a phylogeny (Willmer, 1990). Chemically similar, or even identical, molecules can be independently derived. Such convergences are not readily identified in surveys of a few "representative" taxa. However, molecules may be effectively used to test phylogenetic hypotheses, if the same care is taken in their analysis as has traditionally been taken in the assessment of morphological homologies (Morris and CoBabe, 1991). Homology of anatomical similarities is more certain if numerous complex correspondences exist. Likewise, a complex biochemical system is less likely than a single molecule to have evolved more than once.
The Developmental Role
of the ECM
The molecules of the extracellular matrix (ECM) comprise the connective tissues of animals and form basal laminae underlying sheets of epithelial cells. The ECM is the fabric between the cells of animals, in both embryos and adults. The structure and function of these molecules are best known in vertebrates. In this paper, I will briefly describe the major types of vertebrate ECM molecules. I will then discuss the developmental role of these molecules in vertebrates. This will demonstrate how the ECM functions as a system fundamental to cell motions and morphogenesis during vertebrate development. These well known cases will then provide a basis for comparison with similar molecules and developmental events in invertebrate taxa, including sponges.
There are three classes of molecules in the ECM of vertebrates; collagens, proteoglycans, and adhesive glycoproteins. Collagens and proteoglycans are mechanically important molecules. Collagen fibers are ordered arrays of rigid rod-shaped proteins. Individual molecules are composed of three interwound subunits, forming a rigid triple helix. A collagen-specific enzyme, lysyl oxidase, crosslinks these trimers into fibrils or meshworks. It forms strong covalent bonds between lysine residues at highly conserved sites (Eyre, 1987). These covalent crosslinks give collagen fibers a high tensile strength, as, for example, in tendons (Piez, 1984; Mechanic et al., 1987; Miller, 1985). Disulfide bonds also crosslink a few collagen types.
Proteoglycans are huge branching molecules (a cartilage proteoglycan may fill several cubic microns). They are aggregates of glycosaminoglycans (GAGs), charged amino sugar chains, bound onto a protein core (Fig. 3, PG). The amino sugars of the GAGs bear many charged groups, and are strongly hydrophilic. When hydrated, these chains repel each other. Compressive stress placed on the structure expels water and deforms the molecule. When stress is released, the molecule hydrates and expands. This confers a resistance to compressive stresses, making cartilage ideal for coating joint surfaces. A wide variety of large and small proteoglycans are found in the ECM throughout the body (Hunziker and Schenk, 1987; Hascall and Hascall, 1981; Wainwright et al., 1976).
The adhesive glycoproteins include a wide variety of structurally distinct extracellular proteins. They are related by sharing similar roles in binding cells to the ECM. They bind to other matrix components, and are recognized by receptors on the surface of cells. The amino acid sequence RGD (Arg-Gly-Asp) is common to many of the sites on these molecules recognized by cells. These proteins are glycolated, in a manner typical of glycoproteins, by N-linked oligosaccharides. The best characterized of these molecules are fibronectin and laminin (Hakomori et al., 1984). Other functionally similar molecules include tenascin, syndecan, entactin, chondronectin, and vitronectin (e.g., Durkin et al., 1988; Hewitt et al., 1980; Mackie et al., 1988).
Collagens are triple helical molecules with repeating Gly-X-Y and Gly-Pro-Y amino acid sequences. Nearly one third of the 3000 residues in a fibrillar collagen molecule are sterically constrained to be glycine. Proline is thermodynamically favored in many other sites (Kivirikko and Myllylla, 1984). These structural and functional constraints produce considerable amino acid sequence similarity among collagens. In addition to these similarities, collagens are modified by three unusual post-translational pathways. Some of their lysine and proline residues may be hydroxylated. Lysyl oxidase cross-links are almost entirely restricted to collagens. Simple sugars may be bound to collagens at hydroxylysine residues by unique O-linkages (galactosyl-hydroxylysine and glucolylgalactosyl-hydroxylysine; Piez, 1984).
Some fifteen vertebrate collagens are known. Functionally, they may be divided into four groups; fibrillar, FACIT (Fibril associated), basement membrane (Type IV), and small collagens. Fibrillar collagens are the familiar rod shaped molecules that aggregate into ordered arrays and form banded fibers in tendon and cartilage. The various non-fibrillar collagens have both globular domains and short flexible interrupts of their triple helix. FACIT collagens bind to the surface of collagen fibers and may bind these fibers to proteoglycans. Type IV collagens underlie epithelia where they form an open meshwork in the basal lamina. Some small collagens can form open lattices, others may function in development (Olsen and Nimni, 1989; Piez, 1984; Ninomiya et al., 1990).
The exon structure of collagen genes indicates that collagens belong to three distinct families of proteins. Fibrillar collagens and FACIT collagens are coded for by many 54-bp exons. Several of the small collagens have triple helical domains coded for by a single large exon. Type IV collagen genes are divided into many exons of various lengths (Olsen and Nimni, 1989; Ninomiya et al., 1990; Soinien et al., 1989). Either three long divergent lineages of collagens exist in vertebrates, or collagens are unrelated, convergent, triple helical molecules.
In the last decade, much has been learned about the involvement of the ECM in development. The ECM plays a critical role in morphogenesis. Cell behavior can be regulated by the surrounding ECM. A linkage exists between the actin cytoskeleton, cell surface receptors, and extracellular molecules (Burridge et al., 1988; Horwitz et al., 1986). These molecules include adhesive glycoproteins specifically recognized by several families of cell surface receptors. A model for the connection between the actin cytoskeleton of a fibroblast and the surrounding collagenous matrix is illustrated in Figure 3. In this example, the fibronectin receptor and fibronectin specifically bind to each other to form this linkage.
Elizabeth Hay and others (Greenburg and Hay, 1986, 1988; Zuk et al., 1989) have described stunning examples of the regulation of cellular phenotype by extracellular matrices. Differentiated epithelial cells (including corneal, thyroid, and kidney epithelia, notochord, and adult skin) cultured on collagen plates grow into polarized epithelia, secrete Type IV collagen, and form a basal lamina. The same cells, suspended in a collagen gel, take on the appearance of fibroblasts, become motile, and secrete Type I collagen (a fibrillar collagen).
Two fundamental cell types exist in developing embryos, sheet-forming epithelial cells and motile mesenchymal cells (Hay, 1981). Epithelial cells link together to form sheets. They transcribe cytokeratins, Type IV collagen, and the adhesive glycoprotein laminin. Epithelia frequently have a top-bottom polarity with apical specializations such as cilia or microvilli, an apical Golgi complex, lateral cell junctions, a basal nucleus, and an underlying basal lamina. In contrast, mesenchymal cells migrate individually through loose collagen matrices. They have a front-back polarity and express vimentin, Type I collagen, and the adhesive glycoprotein fibronectin (Fig. 4). The organization of the actin cytoskeletons of these two cell types is regulated by interactions with extracellular molecules (Hay, 1981; Greenburg and Hay, 1986; Rodriguez-Boulan and Nelson, 1989; Kupfer et al., 1987). Several families of cell surface receptors including integrins are bound to actin filaments within the cell. Integrins also bind to adhesive glycoproteins in a specific manner (e.g., one integrin binds to fibronectin, another is a specific receptor for laminin). Concentrations of integrins are visible as adhesion plaques (Greenburg and Hay, 1986; Hynes, 1987).
Gastrulation, in several deuterostome taxa, has been studied in sufficient detail to provide a good example of the roles of ECM components in a fundamental developmental system. In echinoids, gastrulation involves both invagination and ingression. Ingression is a process of epithelial-mesenchymal transformation. During ingression, the motion of cells from the wall of the blastula into the blastocoel is accompanied by changes in the expression of cell surface receptors. Ingressing primary mesenchymal cells lose their affinity for each other and for hyalin (in the hyaline coating of the blastula). They then express cell surface receptors that bind to fibronectin (in the blastocoel), lose their epithelial character, migrate into the blastocoel, and begin secreting and organizing collagen fibers. Ingression of the primary mesenchyme is dependent upon cell-matrix interactions (McClay and Chambers, 1978; Fink and McClay, 1985; Alliegro et al., 1988). Similar epithelial-mesenchymal transitions are responsible for the migration of neural crest cells, while mesenchymal-epithelial transitions produce blood vessels (Hay, 1981).
Epithelia can also interact with the ECM. During the folding of epithelial glands, mesenchymal cells at the tips of budding glands destabilize the ECM, probably by the enzymatic degradation of proteoglycans. Cell division rates are greatest in the epithelial cells closest to these regions, causing the gland to fold out into the destabilized matrix. At the sides of the folds, collagen is secreted at high rates, stabilizing the matrix and the budding gland (Hay, 1981; Wessells, 1977). Cell-matrix interactions have also been implicated, for example, in neural crest cell migration (Mackie et al., 1988), differentiation of mammary gland epithelia (Lee et al., 1984), tumor metastasis (Chiquet-Ehirsmann et al., 1986), neuromuscular junction formation, chondrogenesis, osteogenesis, tendon formation, kidney development, and wound healing (Trelstad, 1984; Bard, 1990).
Phylogenetic Distribution of
The same collagen-proteoglycan-adhesive glycoprotein ECM is found in both protostomes and deuterostomes. However, ECM components are known in detail in only a few species. Fibrillar collagens have been studied for a long time and are well characterized. Integrins have been examined in only a few taxa. Most matrix components are poorly known in all but the few vertebrates, echinoids, insects, and nematodes used as systems for molecular research.
Three levels of complexity are available for assessing homology in these molecules. First are lines of direct evidence for the similarity of peptides, such as 54-bp exons in collagen genes. Second are similarities of post-translational pathways, such as lysyl oxidase crosslinking. This is the weakest type of evidence, as an enzyme that primitively modified one protein may be co-opted to modify another unrelated protein. Finally, the functional role of these molecules in development may provide evidence of common origin.
The appearance of collagen fibers is similar in most invertebrate taxa. The few invertebrate collagens that have been sequenced are similar to vertebrate collagens in function, exon structure, and domain structure. Four of 50 genes for small collagens in Caenorhabditis elegans have been sequenced. Like the small collagens of vertebrates, they possess terminal globular domains, helical interrupts, and a triple helical domain coded for by a single large exon. These collagens crosslink in the cuticle. Mutations in their genes have profound effects upon the development of the nematode (Mende et al., 1988). A gene for a Type IV collagen in Drosophila is similar to vertebrate basement membrane collagens in domain structure, division into multiple exons of varying lengths, locations of cystine residues, interruptions of the triple helix, and restriction to basement membranes (Lundstrum et al., 1988; Blumberg et al., 1988).
A molecule homologous to vertebrate integrins has been identified in Drosophila. It is associated with adhesion plaques, is involved in the binding of cells to fibronectin bearing ECMs and in differentiation of mesenchymal cells into muscle. It and a set of related antigens are also involved in the differentiation of imaginal disks (Brower et al., 1984; Bogaert et al., 1987). Sequence similarities have identified molecules homologous to vertebrate adhesive glycoproteins in invertebrates. Laminin has been identified in sea urchins and in Drosophila, where it functions as an adhesion molecule (Montell and Goodman, 1988). A molecule similar to entactin occurs in Drosophila (Bogaert et al., 1987). Proteoglycans are also known from invertebrates (Fessler et al., 1984; Dietrich et al., 1983).
Two observations may be made about the ECM of sponges. First, strong similarities exist in the kinds of developmental events that occur in sponges and other animals (Fig. 2). There are striking differences in the details of early embryology; the formation of an amphiblastula larva by inversion, or the proliferation of blastomeres into a follicle enclosed bag of nurse cells in a viviparous sponge embryo are radically different from typical animal gastrulation. While the embryological details clearly differ, there are similarities between the kinds of cellular events that occur in sponges and other animals. Development in all animals is characterized by the folding of epithelial sheets and transitions between mesenchymal and epithelial cell types. While a parenchymella of a sponge is not directly comparable in structure to the gastrula of an echinoderm, the migration of cells into the core of the parenchymella in the sphinctozooid sponge Vaceletia (=Neocoelia, Simpson, 1984, p. 400) is, on a cellular level, directly comparable to the ingression of primary mesenchyme in an echinoid gastrula. That is, sheet forming cells transform into motile cells and migrate into the ECM. The same early embryological dichotomy of epithelia and mesenchyme is observed in sponges and other animals.
Second, members of the three groups of extracellular molecules that mediate these processes of cell adhesion and motility in deuterostomes have been identified in sponges. It appears that the molecules responsible for the mechanics of epithelialmesenchymal transformations are present in all animals. The molecules responsible for the interaction of cytoskeleton, cell surface receptors, and extracellular matrix appear to be the same in all animals.
Many fibrillar forms of collagen are found in sponges. Their post-translational processing is the same as in vertebrates. Hydroxylysine and hydroxyproline residues are present and both may be glycolated with simple sugars bound to hydroxylysine (Simpson, 1984). Sponge collagens are crosslinked with dihydroxylysinonorleucine (DHLN) and hydroxylysinonorleucine (HLN) crosslinks (Olsen and Nimni, 1989; Eyre and Glimcher, 1971), products of lysyl oxidase activity. These three unusual post-translational modifications are almost entirely restricted to collagens in other animals. It has been speculated that the beaded fibers of sponges are composed of GAG bearing collagens related to FACIT collagens (Olsen et al., 1989).
In addition to these chemical similarities, there is direct evidence that molecules homologous to vertebrate fibrillar collagens exist in sponges. A sponge collagen gene that has 54-bp exons has been described by Exposito and Garrone (1990). This is strong evidence for the presence of a fibrillar collagen in the nearest common ancestor of sponges, echinoderms, vertebrates, and insects. A second family of sponge collagens with variable length exons has also been identified (Exposito et al., 1990). This gene may be related to the Type IV collagens identified in Drosophila, echinoids, and vertebrates. Sponge collagens coded for by single large exons, equivalent to small collagens of vertebrates, or some of the nematode cuticle collagens, have not yet been identified. At least one type of sponge collagen is homologous to collagens found in other animals.
Glycosaminoglycans, like those in vertebrates, have been identified in sponges both as matrix components and as cell surface molecules (Simpson, 1984). A proteoglycan-like molecule is involved in cellular aggregation in sponges (Misevic and Burger, 1990). In this proteoglycan (MAF), as in vertebrates, GAG chains are bound to a core protein. Unlike other proteoglycans, MAF has a large globular core protein (Misevic and Burger, 1988; Muller et al., 1988), rather than an unfolded core protein.
Antibodies against fibronectin recognize a molecule in sponges. It is associated with the junctions between epithelial cells, the surfaces of other cells, and is disseminated in the ECM. Interference with this molecule prevents reaggregation of sponge cells. A sponge protein with the same weight as a fibronectin monomer binds to antifibronectin antiserum. These data suggest (but by no means prove) that a molecule homologous to fibronectin exists in sponges (Labat-Robert et al., 1981; Akiyama and Johnson, 1983). Two sponge glycoproteins are known that bind to the proteoglycan-like aggregation factor and to the surfaces of cells. These proteins are disseminated through the sponge ECM and function in the same manner as the adhesive glycoproteins of vertebrates (Varner et al., 1988).
Indirect evidence exists for the presence of integrins in sponges. Features similar to the adhesion plaques of vertebrates (sites of integrin-talin-actin complexes) have been found in sponges. Pavans de Ceccatty (1981) described attachment plaques in pinacocytes of a sponge. He observed actin filaments of the cytoskeleton linked to plaques on the basal cell membrane bound, in turn, to collagen fibers in the ECM. These structures indicate that some complex of molecules is linking the actin cytoskeleton to the ECM. Ultrastructural evidence also exists for cell-collagen binding in sponges (Simpson, 1984).
Differences do exist between sponges and other animals. Sponges show no evidence for a basement membrane beneath their epithelial cells (Simpson, 1984), though some acoel flatworms also appear to lack a basal lamina (Hyman, 1940). A large group of sponges, the hexactinellids, have syncytial epithelia (Mackie and Singla, 1983). Syncytia, while uncommon among vertebrates, do occur in a variety of animals including flatworms, mesozoans, and insects (Barnes, 1980).
Taken individually, most of the similarities between sponges and other animals are unconvincing. While a fibrillar collagen is clearly shared by both sponges and other animals, on its own it does not clearly indicate the presence of multicellularity in the nearest common ancestor of these taxa. The putative sponge fibronectin has only been identified immunologically, a potentially unreliable technique. Until its gene has been sequenced the possibility of convergence remains. The core proteins of sponge proteoglycan-like aggregation factors are unlike the core proteins of other proteoglycans. The phylogenetic distribution of chitin suggests that the synthesis pathways of complex sugars could convergently produce molecules such as proteoglycans (Willmer, 1990). The evidence for integrins in sponges is purely circumstantial. However, taken together, there are too many precise similarities in the composition, structure, and function of extracellular matrices of sponges and other animals to infer that this major adaptive complex has evolved more than once.
Sponges appear to share with other animals a collagen, proteoglycan, adhesive glycoprotein based ECM that plays a fundamental role in their development. If the ECM is indeed a homology for the kingdom Animalia then the system should not exist in any other taxa. However, the presence of some components outside of the kingdom would not falsify this hypothesis. All the molecules used in the system did not arise de novo with the first multicellular animal. For example, the actin cytoskeleton is widely used by protists to control cell shape, contraction, and motility. Animals added molecules that control the actin cytoskeleton by interacting with matrix proteins.
Collagens are not known in any taxa other than animals (Garrone, 1978; Labat-Robert et al., 1981; Akiyama and Johnson, 1983). A collagen-like triple helical motif has been reported from a bacterial capsomer protein (Bamford and Bamford, 1990). This is undoubtedly a convergent appearance of the highly constrained collagenous triple helix (Morris and CoBabe, 1991).
Glycosaminoglycans may exist in protists. Similar amino sugars are common, for example, the subunit of chitin, N-acetylglucosamine, is also a subunit of the GAGs keratan sulfate and hyaluronate (Zubay, 1983, pp. 443-446). It is difficult to assess the phylogenetic significance of these molecules. Enzymatically assembled chemicals, especially those built of readily available subunits (e.g., sugars), may be as convergent as the wings of bats and birds. Knowledge of synthesis pathways and enzyme sequence similarities are needed to assess homology in non-peptide molecules (Morris and CoBabe, 1991).
No homologies between molecules in protists and integrins or adhesive glycoproteins have been reported. However, cell surface receptors that recognize RGD sequences in extracellular proteins are phylogentically widespread. Synthetic RGD peptides interfere with plasma membrane-cell wall binding in plants (Schindler et al., 1989). RGD sequences also bind to cell surface proteins in bacteria (Pierschbacher and Ruoslahti, 1984) and slime molds (Gabius et al., 1985). It has been stated that the RGD sequence is a "general membrane-wall-matrix recognition principle that may transcend species and kingdoms" (Schindler et al., 1989).
Two explanations may be proposed for the widespread distribution of RGD recognition. A single family of cell surface proteins may recognize RGD sequences, or RGD recognition may have independently appeared in bacterial, plant, protist, and animal lineages. Chemically, arginine (R) has a basic side chain, glycine (G), hydrogen, and aspartic acid (D) an acidic side chain. This sequence could be functionally suited for extracellular recognition, and it may be convergently derived. Interestingly, some bacteria may employ RGD recognition by animal cell surface receptors to facilitate adherence and invasion (Leininger et al., 1991).
Cellular slime molds are multicellular protists similar in some respects to animals. In their development, cells migrate through an ECM. Their cell surface receptors for extracellular molecules have stage specific expression. Unlike animals, the primary component of their ECM is cellulose. This is concentrated in a sheath that is synthesized at the leading edge of the slug and is left behind as a trail (Freeze and Loomis, 1977). As in animals, cohesion of cells within the slug is mediated by both cell-cell and cell-matrix binding. An extracellular glycoprotein, discoidin, mediates cell-matrix interactions in a manner analogous to fibronectin (Vardy et al., 1986; Smith and Williams, 1988). Cell-cell binding is mediated by a cell surface glycoprotein. Binding between these molecules (gp80) on adjacent cells is associated with an interdigitation of filopodia unlike any cell junctions in animals (Siu et al., 1988; Gerisch, 1986). While similarities do exist in the development of slime molds and animals, the molecular basis of cell motility is different. They clearly represent convergent approaches to multicellularity.
Chordates and echinoderms possess a collagen, proteoglycan, adhesive glycoprotein extracellular matrix. Interactions between the actin cytoskeleton, cell surface receptors (integrins), and the molecules of the ECM are the fundamental mechanism for cell motility and transitions between epithelial and motile cell types. The ECM is a central mechanism in such events as ingression of primary mesenchyme, neural crest cell migration, vascularization of bone, wound repair, and metastasis that involve cell migrations and transitions llmeen epithelial and mesenchymal phenotypes. Similar events are at the heart of the development of all animals, including sponges. The known distribution of collagens, proteoglycans, and fibronectin suggests that this complex developmental system is a fundamental homology that unites the kingdom Animalia into a monophyletic group descended from a single, multicellular ancestor. This proposal would be further supported by the discovery of homologies between the cell surface receptors of sponges and other animals, and by further similarities between the adhesive glycoproteins of vertebrates and the adhesion molecules of sponges.
Multicellularity in all animals is, on a fundamental level, similar. It is unlike other grades of multicellular organization. Plants and fungi achieve large size by the growth of rigid cellulose cell walls. Slime molds develop through the movement of interconnected cells within a carbohydrate extracellular matrix. Animals are infolded balls, coated by sheets of cells, containing a collagen-proteoglycan extracellular matrix, through which other cells move. This view of animal organization suggests that three great branches of animals exist. The sponges, which are mesenchymal cell specialists, form one branch. The cnidarians, with their bewildering diversity of epithelial cell types, form another, and the other animals, which have grouped cells into tissue and organ systems, form a third branch. Early embryos, in all animals, undergo two kinds of developmental events. The embryo may infold, producing an invagination such as the gut. Alternately cells may detach and migrate into its core. Combinations of these two events, both produced by cell-matrix interactions, produce the variety of gastrulae present among all animals. The parenchymella of a sponge is not homologous with the gastrula of a vertebrate, but the mechanism that produces the cell layers in these embryos is homologous.
This is a specific, testable, phylogenetic hypothesis. It predicts that molecules homologous to integrins, proteoglycan core proteins, adhesive glycoproteins, and the enzymes that form Gal-Xyl-O-Serine and Gla-O-Lysine glycoprotein linkages are present in cnidarians, flatworms, placozoans, and sponges. Furthermore, most of these molecules should be absent from chytrids, choanoflagellates and other protists.
During the last century, the kingdom Animalia was considered a cohesive group of related organisms by most workers. This concept has been challenged by many workers. Today, a common position is that the kingdom is monophyletic, but widespread dissent exists. It is generally accepted that sponges could be independently derived multicellular organisms. The shared presence of a coliagen-proteoglycan-adhesive glycoprotein extracellular matrix that plays a fundamental role in development suggests that this view is incorrect. The ECM is a strong, complex character that unites animals into a monophyletic kingdom.
Akiyama, S. K., and M. D. Johnson. 1983. Fibronectin in evolution: Presence in invertebrates and isolation from Microciona prolifera. Comp. Biochem. Physiol. 76B:687-694. Alberts, B., D. Bray, J. Lewis, M. Raff, K. Roberts, and J. D. Watson. 1989. Molecular Biology of the Cell. Garland, N.Y., USA. Alliegro, M. C., C. A. Ettensohn, C. A. Burdstal, H. P. Erickson, and D. R. McClay. 1988. Echinonectin: A new embryonic substrate adhesion protein. J. Cell Biol. 107:2319-2327. Bamford, D. K., and J. K. H. Bamford. 1990. Collagenous proteins multiply. Nature 334:497. Bard,J. 1990. Morphogenesis: The Cellular and Molecular Processes of Developmental Anatomy. Cambridge University Press, Cambridge, UK. Bardele, C. F. 1983. Comparative freeze-fracture study of the ciliary membrane of protists and invertebrates in relation to phylogeny. J. Submicroscop. Cytol. 15:263-267. Barnes, R. D. 1980. Invertebrate Zoology, 4th ed. Saunders College and Holt, Rinehart and Winston, Philadelphia, PA USA. Beklemishev, W. N. 1969. Principles of Comparative Anatomy of Invertebrates: Vol. 1, Promorphology, University of Chicago Press, Chicago, IL USA. Bergquist, P. R. 1978. Sponges. University of California Press, Berkeley, USA. _____. 1985. Poriferan relationships, pp. 14-27. In S. Conway Morris, J. D. George, R. Gibson, and H. M. Platt (eds.), The Origins and Relationships of the Lower Invertebrates. Clarendon, N.Y., USA. Blumberg, B., A. J. MacKrell, and J. H. Fessler. 1988. Drosophila basement membrane procollagen [alpha] 1 (IV). J. Biol. Chem. 263:1828-1837. Bogaert, T., N. Brown, and M. Wilcox. 1987. The Drosophila [sic] PS2 antigen is an invertebrate integrin that, like the fibronectin receptor, becomes localized to muscle attachments. Cell 51:929-940. Brower, D. L., M. Wilacox, M. Piovant, R. J. Smith, and L. A. Reger. 1984. Related cell-surface antigens expressed with positional specificity in Drosophila imaginal disks. Proc. Natl. Acad. Sci. USA 81:7485-7489. Burridge, K., K. Fath, T. Kelly, G. Nuckolls, and C. Turner. 1988. Focal adhesions: Transmembrane junctions betwecn the extracellular matrix and the cytoskeleton. Ann. Rev. Cell Biol. 4:487-526. Chiquet-Ehirsmann, R., E. J. Mackie, C. A. Pearson, and T. Sakakura. 1986. Tenascin: An extracellular matrix protein involved in tissue interactions during fetal development and oncogenesis. Cell 47: 131-139. Davidson, E. H. 1986. Gene Activity in Early Development, Figure 4. 1. Academic Press, Orlando, FL USA. Delage, Y. 1892. Embryogenie des eponges: Developpement post-larvaire. Arch. Zool. Exp. Gen. 10: 345-498. Dietrich, C. P., V. M. P. Paiva, S. M. B. Jeronimo, T. M. O. C. Ferreira, M. G. L. Medeiros, J. F. Paiva, and H. B. Nader. 1983. Characteristic distribution of heparan sulfates and chondroitin sulfates in the tissues and organs of the Ampularidae Pomacea sp. Comp. Biochem. Physiol. 76B: 695-698. Durkin, M. E., S. Chakravarti, B. B. Bartos, S-H Liu, R. L. Friedman, and A. E. Chung. 1988. Amino acid sequence and domain structure of entactin. J. Cell Biol. 107:2749-2756. Erwin, D.H. 1989. Molecular clocks, molecular phylogenies and the origin of phyla. Lethaia 22:251-257. Eyre. D.R. 1987. Collagen cross-linking amino acids. Methods Enzymol. 144:115-139. Eyre, D. R., and M. J. Glimcher. 1971. Comparative biochemistry of collagen crosslinks: Reducible bonds in invertebrate collagens. Biochim. Biophys. Acta 243:525-529. Exposito, J-Y., and R. Garrone. 1990. Characterization of a fibrillar collagen gene in sponges reveals the early evolutionary appearance of two collagen gene families. Proc. Natl. Acad. Sci. USA 87:6669-6673. Exposito,J-Y.,R.Ouazana, and R.Garrone. 1990. Cloning and sequencing of a Porifera partial cDNA coding for a short chain collagen. Eur. J. Biochem. 190:401-406. Fessler, J. H., G. Lundstrum, K. G. Duncan, A. G. Cambel, R. Sterne, H. P. Bachinger, and L. I. Fessler. 1984. Evolutionary constancy of basement membrane components, pp. 207-219. In R. L. Trelstad (ed.), The Role of Extracellular Matrix in Development. Alan Liss, N.Y., USA. Field, K. G., G. J. Olsen, D. L. Lane, S. J. Giovannoni, M. T. Ghieselin, E. C. Raff, N. R. Pace, and R. A. Raff. 1988. Molecular phylogeny of the animal kingdom. Science 239:748-753. Fink, R. D., and D. R. McClay. 1985. Three cell recognition changes accompany the ingression of sea urchin primary mesenchyme cells. Dev. Biol. 107: 66-74. Franzen, [Angstrom]. 1956. On spermatogenesis, morphology of the spermatozoon, and biology of fertilization among invertebrates. Zool. Bidr. Uppsala 31:355-482. Frazzetta, T. H. 1975. Complex Adaptations in Evolving Populations. Sinauer, Sunderland, MA USA. Freeze, H., and W. F. Loomis. 1977. Isolation and Characterization of a Component of the Surface Sheath of Dictyostelium discoideum. J. Biol. Chem. 252:820-824. Gabius, H-J., W. R. Springer, and S. H. Barondes. 1985. Receptor for the cell binding site of Discoidin I. Cell 42:449-456. Garrone, R. 1978. Phylogenesis of Connective Tissue. S. Karger, Basel, Switzerland. Garrone, R., and C. Lethias. 1990. Freeze-fracture study of sponge cells, pp. 121-128. In K. Rutzler (ed.), New Perspectives in Sponge Biology. Smithsonian Institution Press, Washington, DC USA. Gerisch, G. 1986. Interrelation of cell adhesion and differentiation in Dictyostelium discoideum. J. Cell Sci. Suppl. 4:201-219. Giudice, G. 1986. The Sea Urchin Embryo: A Developmental Biological System. Springer-Verlag, Berlin, Germany. Greenburg, G., and E. D. Hay. 1986. Cytodifferentiation and tissue phenotype change during transformation of embryonic lens epithelium to mesenchyme-like cells in vitro. Dev. Biol. 115:363-379. _____. 1988. Cytoskeleton and thyroglobulin expression change during transformation of thyroid epithelium to mesenchymal-like cells. Development 102:605-622. Hadzi, J. 1953. An attempt to reconstruct the system of animal classification. Syst. Zool. 2:145-154. _____. 1963. The Evolution of thc Metazoa. Macmillian, N.Y., USA. Haeckel, E. 1874. Die Gastraea-Theorie, die phylogenetische Classification des Thierreichs und die Homologie der Keimblatter. Jenaische Zeitschrift fur Naturalwissenshaften 8:1-56. _____. 1891. Anthropogenie, Erster Theil: Keimesgeschichte des Menchen. W. Engelmann, Leipzig, Germany. Hakomori, S., M. Fukuda, K. Sekiguchi, and W. G. Carter. 1984. Fibronectin, laminin, and other extracellular glycoproteins, pp. 230-275. In K. A. Piez and A. K. Reddi (eds.), Extracellular Matrix Biochemistry. Elsevier, N.Y., USA. Halanych, K. M. 1991. 5S Ribosomal RNA sequence inappropriate for phylogenetic reconstruction. Mol. Biol. Evol. 8:249-253. Hanson, E. D. 1977. Origin and Early Evolution of Animals. Wesleyan University Press, Middletown, CT USA. Hascall, V. C., and G. K. Hascall. 1981. Proteoglycans, pp. 39-64. In E. D. Hay (ed.), Cell Biology of the Extracellular Matrix. Plenum, N.Y., USA. Hay, E. D. 1981. Collagen and embryonic development, pp. 379-409. In E. D. Hay (ed.), Cell Biology of the Extracellular Matrix. Plenum, N.Y., USA. _____. 1984. Cell-matrix interaction in the embryo: Cell shape, cell surface, cell skeletons, and their role in differentiation, pp. 1-3 1. In R. L. Trelstad (ed.), The Role of Extracellular Matrix in Development. Alan Liss, N.Y., USA. Hewitt, A. T., H. K. Kleinman, J. P. Pennypacker, and G. R, Martin. 1980. Identification of an adhesion factor for chondrocytes. Proc. Natl. Acad. Sci. USA 77: 385-388. Hori, H., and S. Osawa. 1987. Origin and evolution of organisms as deduced from 5S ribosomal RNA sequences. Mol. Biol. Evol. 4:445-472. Horwitz, A., K. Duggam, C. Buck, M. C. Beckerle, and K. Burridge. 1986. Interaction of plasma membrane fibronectin receptor with talin - A transmembrane linkage. Nature 320:531-533. Hunziker, E. B., and R. K. Schenk. 1987. Structural organization of proteoglycans in cartilage, pp. 155-186. In T. N. Wight and R. P. Mecham (eds.), Biology of Proteoglycans. Academic Press, Orlando, FL USA, Hyman, L. H. 1940. The Invertebrates: Protozoa Through Ctenophora, Vol. 1. McGraw Hill, N.Y., USA. _____. 1951. The Invertebrates: Platyhelminthes and Rhynchocoela, the Acoelomate Bilateria, Vol. 2. McGraw Hill, N.Y., USA. Hynes, R. O. 1987. Integrins: A family of cell surface receptors. Cell 48:549-554. Inglis, W. G. 1985. Evolutionary waves; patterns in the origins of animal phyla. Aust. J. Zool. 33:153-178. Kivirikko, K. I., and R. Myllyla. 1984. Biosynthesis of the collagens, pp. 83-118. In K. A. Piez and A. K. Reddi (eds.), Extracellular Matrix Biochemistry. Elsevier, N.Y., USA. Kupfer, A., P. J. Kronebusch, J. K. Rose, and S. J. Singer. 1987. A critical role for the polarization of membrane recycling in cell motility. Cell Motil. Cytoskeleton 8:182-189. Labat-Robert, J., L. Robert, C. Auger, C. Lethias, and R. Garrone. 1981. Fibronectin-like protein in Porifera: Its role in cell aggregation. Proc. Natl. Acad. Sci. USA 78:6261-6265. Lake, J. A. 1990. Origin of the metazoa. Proc. Natl. Acad. Sci. USA 87:763-766. Leadbeater, B. S. C. 1983. Distribution and chemistry of microfilaments in choanoflagellates, with special reference to the collar and other tentacle systems. Protistologica 19:157-166. Lee, E. Y-H., G. Parry, and M. J. Bissell. 1984. Modulation of secreted proteins of mouse mammary epithelial cells by the collagenous substrata. J. Cell Biol. 98:146-158. Leininger, E., M. Roberts, J. G. Kenimer, I. G. Charles, N. Fairweather, P. Novotny, and M. J. Brennan. 1991. Pertactin, an Arg-Gly-Asp-containing Bordetella pertussis surface protein that promotes adherence of mammalian cells. Proc. Natl. Acad. Sci. USA 88:345-349. Lipscomb, D. L. 1989. Relationships among the eukaryotes, pp. 161-178. In B. Fernholm, K. Brenner, and H. Jornvall (eds.), The Hierarchy of Life. Elsevier, Amsterdam, The Netherlands. Lundstrum, G. P., H-P. Bachinger, L. I. Fessler, K. G. Duncan, R. E. Nelson, and J. H. Fessler. 1988. Drosophila basement membrane procollagen IV. J. Biol. Chem. 263:18318-18327. Mackie, E. J., R. P. Tucker, W. Halfter, R. Chiquet-Ehirsmann, and H. H. Epperlein. 1988. The distribution of tenascin coincides with pathways of neural crest cell migration. Development 102:237-250. Mackie, G. O., and C. L. Singla. 1983. Studies on hexactinellid sponges. I. Histology of Rhabdocalyptus dawsoni. Philos. Trans. R. Soc. London 301:365-400. McClay, D. R., and A. F. Chambers. 1978. Identification of four classes of cell surface antigens appearing at gastrulation in sea urchin embryos. Dev. Biol. 63:179-186. Mechanic, G. L., E. P. Katz, M. Henmi, C. Noyes, and M. Yamauchi). 1987. Locus of a histidine-based, stable trifunctional, helix to helix collagen cross-link: stereospecific collagen structure of Type I skin fibrils. Biochemistry 26:3500-3509. Mende, N. von, D. McK. Bird, P. S. Albert, and D. L. Riddle. 1988. dpy-13: A nematode collagen gene that affects body shape. Cell 55:567-576. Miller, E. J. 1985. The structure of fiber forming collagens. Ann. N.Y. Acad. Sci. 406:1-32. Misevic, G. N., and M. M. Burger. 1988. Multiple low affinity carbohydrates as the basis for cell recognition in the sponge Microciona prolifera, pp. 134-152. In G. P. Chapman, C. C. Ainsworth, and C. J. Chatham (eds.), Eukaryote Cell Recognition. Cambridge University Press, Cambridge, UK. _____. 1990. Multiple low-affinity carbohydrates as the basis of cell-cell recognition in Microciona prolifera, pp. 81-90. In K. Rutzler (ed.), New Perspectives in Sponge Biology. Smithsonian Institution Press, Washington, DC USA. 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 K. Rutzler (ed.), New Perspectives in Sponge Biology. Smithsonian Institution Press, Washington, DC USA. Montell, D. J., and C. S. Goodman. 1988. Drosophila [sic] substrate adhesion molecule: Sequence of laminin B1 chain reveals domains of homology with mouse. Cell 53: 463-473. Morris, P. J., and E. A. CoBabe. 1991. Cuvier meets Watson and Crick: The utility of molecules as classical homologies. Biol, J. Linn. Soc. 44: 307-324. Muller, W. E. G., B. Diehl-Seifert, M. Gramzow, U. Friese, K. Renneisen, and H. C. Schroder. 1988. Interrelation between extracellular adhesion proteins and extracellular matrix in reaggregation of dissociated sponge cells. Int. Rev. Cytol. 111:211-229. Ninomiya, Y., P. Castagnola, D. Gerecke, M. Gordon, O. Jacenko, P. LuValle, M. McCarthy, Y. Muragaki, I. Nishimura, S. Oh, N. Rosenblum, N. Sato, S. Sugrue, R. Taylor, G. Vasios, N. Yamaguchi, and B. R. Olsen. 1990. The molecular biology of collagens with short triple-helical domains. In L. J. Sandell and C. D. Boyd (eds.), Extracellular Matrix Genes (Biology of Extracellular Matrix: Vol. 6). Academic Press, San Diego, CA USA. Nursall, J. R. 1962. On the origins of the major groups of animals. Evolution 16:118-123. Olsen, B. R., Am M. E. Nimni (eds.) 1989. Collagen: Volume IV, Molecular Biology. Chapters 1-4. CRC Press, Boca Raton, FL USA. Olsen, B. R., Y. Ninomyia, D. Gerecke, M. Gordon, G. Green, T. Kimura, Y. Muragaki, I. Nishimura, and S. Sugrue. 1989. A new dimension in the extracellular matrix, pp, 2-19. In B. R. Olsen and M. E. Nimni (eds.), Collagen: Vol. IV, Molecular Biology. CRC Press, Boca Raton, FL USA. Patterson, C. 1989. Phylogenetic relations of major groups: Conclusions and prospects, pp. 273-278. In B. Fernholm, K. Bremer, and H. Jornvall (eds.), The Hierarchy of Life, Proceedings from Nobel Symposium 70. Elsevier, Amsterdam, The Netherlands. Pavans de Ceccatty, M. 1981. Demonstration of actin filaments in sponge cells. Cell Biol. Int. Rep. 5:945-952. Pierschbacher, M. D., and E. Ruoslahti. 1984. Variants of the cell recognitions site of fibronectin that retain attachment promoting activity. Proc. Natl. Acad. Sci. USA 81:5985-5988. Piez, K. A. 1984. Molecular and aggregate structures of the collagens, pp. 1-40. In K. A. Piez and A. H. Reddi (eds.), Extracellular Matrix Biochemistry. Elsevier, N.Y., USA. Rodriguez-Boulan, E., and W. J. Nelson. 1989. Morphogenesis of the polarized epithelial cell phenotype. Science 245:718-725. Salvini-Plawen, L. Von. 1978. On the origin and evolution of the lower Metazoa. Z. Zool. Syst. Evolutionsforsch. 16:40-88. Schindler, M., S. Meiners, and D. Cheresh. 1989. RGD-dependent linkage between plant cell wall and plasma membrane: Consequences for growth. J. Cell Biol. 108:1955-1965. Simpson, T. L. 1984. The Cell Biology of Sponges. Springer-Verlag, N.Y., USA. Siu, C., L. M. Wong, A. Choi, and A. Cho. 1988. Mechanisms of cell-cell recognition and cell cohesion in Dictyostelium discoideum cells, pp. 119-133. In G. P. Chapman, C. C. Ainsworth, and C. J. Chatham (eds.), Eukaryote Cell Recognition. Cambridge University Press, Cambridge, UK. Smith, E., and K. L. Williams. 1988. Cell type specific glycoproteins involved in cell differentiation and movement in Dictyostelium discoideum slugs, pp. 107-118. In G. P. Chapman, C. C. Ainsworth, and C. J. Chatham (eds.), Eukaryote Cell Recognition. Cambridge University Press, Cambridge, UK. Soinien, R., M. Huotari, A. G. Anguly, D. J. Prockop, and K. Tryggvason. 1989. Structural organization of the gene for the [alpha] chain of human Type IV collagen. J. Biol. Chem. 264: 13565-13571. Sollas, W. J. 1884. On the development of Halisarca lobularis [sic] (O. Schmidt). Quart. J. Microsc. Sci. 24:603-621. Trelstad, R. L. (ed.) 1984. The Role of Extracellular Matrix in Development. Alan R. Liss, N.Y., USA. Vardy, P. H., L. R. Fisher, E. Smith, and K. L. Williams. 1986. Traction proteins in the extra cellular matrix of Dictyostelium discoideum slugs. Nature 320:526-529. Varner, J. A., M. M. Burger, and J. F. Kaufman. 1988. Two cell surface proteins bind the sponge Microciona prolifera aggregation factor. J. Biol. Chem. 263:8498-8508. Vos, L. De, K. Rutzler, N. Boury-Esnault, C. Donadey, and J. Vacelet. 1991. Atlas of Sponge Morphology. Smithsonian Institution Press, Washington, DC USA. Wainwright, S. A., W. D. Biggs, J. D. Currey, and J. M. Gosline. 1976. Mechanical Design in Organisms. Princeton University Press, Princeton, NJ USA. Warrior, R., and J. Gall, 1985. The mitochondrial DNA of Hydra attenuata and Hydra littoralis consists of two linear molecules. Archiv Sci. (Geneva) 38:439-445. Wessells, N. K. 1977. Tissue Interactions and Development. Benjamin, N.Y., USA. Willmer, P. 1990. Invertebrate Relationships. Cambridge University Press, Cambridge, UK. Zubay, G. 1983. Biochemistry. Addison-Wesley, Reading, MA USA. Zuk, A., K. S. Matlin, and E. D. Hay. 1989. Type I collagen gel induces Madin-Darby canine kidney cells to become fusiform in shape and lose apicalbasal polarity. J. Cell Biol. 108:903-920.
|Printer friendly Cite/link Email Feedback|
|Author:||Morris, Paul J.|
|Date:||Feb 1, 1993|
|Previous Article:||Origins of genotypic variation in North American dandelions inferred from ribosomal DNA and chloroplast DNA restriction enzyme analysis.|
|Next Article:||A quantitative genetic analysis of oviposition preference and larval performance on two hosts in the bruchid beetle, Callosobruchus maculatus.|