Isolation of Ef silicatein and Ef lectin as molecular markers for sclerocytes and cells involved in innate immunity in the freshwater sponge Ephydatia fluviatilis

Sponges (phylum Porifera) have remarkable regenerative and reconstitutive abilities and represent evolutionarily the oldest metazoans. To investigate sponge stem cell differentiation, we have focused on the asexual reproductive system in the freshwater sponge Ephydatia fluviatilis. During germination, thousands of stem cells proliferate and differentiate to form a fully functional sponge. As an initial step of our investigation of stem cell (archeocyte) differentiation, we isolated molecular markers for two differentiated cell types: spicule-making sclerocyte cells, and cells involved in innate immunity. Sclerocyte lineage-specific Ef silicatein shares 45% to 62% identity with other sponge silicateins. As in situ hybridization of Ef silicatein specifically detects archeocytes possibly committed to sclerocytes, as well as sclerocytes with an immature or mature spicule, therefore covering all the developmental stages, we conclude that Ef silicatein is a suitable sclerocyte lineage marker. Ef lectin, a marker for the cell type involved in innate immunity, shares 59% to 65% identity with the marine sponge Suberites domuncula galactose-binding protein (Sd GBP) and horseshoe crab Tachypleus tridentatus tachylectin1/lectinL6. Since Sd GBP and tachylectin1 are known to bind to bacterial lipopolysaccharides and inhibit the growth of bacteria, Ef lectin may have a similar function and be expressed in a specialized type of cell involved in defense against invading bacteria. Ef lectin mRNA and protein are not expressed in early stages of development, but are detected in late stages. Therefore, Ef lectin may be specifically expressed in differentiating and/or differentiated cells. We suggest Ef lectin as a marker for cells that assume innate immunity in freshwater sponges.     

Hydroxyurea: An Inhibitor of the Differentiation of Choanocytes in Fresh-Water Sponges and a Possible Agent for the Isolation of Embryonic Cells

During the development of a fresh-water sponge from its gemmules, most cell types originate from the undifferentiated archaeocytes through a few divisions, whereas each choanocyte chamber, composed of several tens of choanocytes, arises from a single archaeocyte through repeated mitoses.
This process was studied on gemmules incubated in various concentrations of hydroxyurea.
A concentration of 100 μg/ml postponed the hatching by about two days, and blocked the differentiation of the choanocytes and the morphogenesis of the aquiferous system. The resulting organism was a hollow dome of pinacoderm, stretched on spicules, the bottom of which was strewn with embryonic archaeocytes. After washing and incubation in mineral medium, the sponge differentiated its choanocytes and achieved normal development.
The incorporation of ³H-thymidine into DNA was compared throughout the development of normal and hydroxyurea-treated gemmules. Hydroxyurea delayed the first peaks of incorporation and abolished the large peak that normally occurs around 90 h, just before the formation of choanocyte chambers.
When added after 96 h incubation, hydroxyurea did not affect the differentiation of the choanocytes.
These results suggest that the differentiation of the choanocytes and the further morphogenesis of the aquiferous system depend on the repetitive divisions of the archaeocytes that normally occur around 90 h.
Furthermore, hydroxyurea-blocked sponges provide a suitable source for the isolation of pure populations of embryonic archaeocytes.     

Ultrastructure and embryonic development of a syconoid calcareous sponge

Abstract. Recent molecular data suggest that the Porifera is paraphyletic (Calcarea+Silicea) and that the Calcarea is more closely related to the Metazoa than to other sponge groups, thereby implying that a sponge‐like animal gave rise to other metazoans. One ramification of these data is that calcareous sponges could provide clues as to what features are shared among this ancestral metazoan and higher animals. Recent studies describing detailed morphology in the Calcarea are lacking. We have used a combination of microscopy techniques to study the fine structure of Syconcoactum Urban 1905, a cosmopolitan calcareous sponge. The sponge has a distinct polarity, consisting of a single tube with an apically opening osculum. Finger‐like chambers, several hundred micrometers in length, form the sides of the tube. The inner and outer layers of the chamber wall are formed by epithelia characterized by apical–basal polarity and occluding junctions between cells. The outer layer—the pinacoderm—and atrial cavity are lined by plate‐like cells (pinacocytes), and the inner choanoderm is lined by a continuous sheet of choanocytes. Incurrent openings of the sponge are formed by porocytes, tubular cells that join the pinacoderm to the choanoderm. Between these two layers lies a collagenous mesohyl that houses sclerocytes, spicules, amoeboid cells, and a progression of embryonic stages. The morphology of choanocytes and porocytes is plastic. Ostia were closed in sponges that were vigorously shaken and in sponges left in still water for over 30 min. Choanocytes, and in particular collar microvilli, varied in size and shape, depending on their location in the choanocyte chamber. Although some of the odd shapes of choanocytes and their collars can be explained by the development of large embryos first beneath and later on top of the choanocytes, the presence of many fused collar microvilli on choanocytes may reflect peculiarities of the hydrodynamics in large syconoid choanocyte chambers. The unusual formation of a hollow blastula larva and its inversion through the choanocyte epithelium are suggestive of epithelial rather than mesenchymal cell movements. These details illustrate that calcareous sponges have characteristics that allow comparison with other metazoans—one of the reasons they have long been the focus of studies of evolution and development.     

Piwi expression in archeocytes and choanocytes in demosponges: insights into the stem cell system in demosponges

SUMMARY Little is known about the stem cells of organisms early in metazoan evolution. To characterize the stem cell system in demosponges, we identified Piwi homologs of a freshwater sponge, Ephydatia fluviatilis, as candidate stem cell (archeocyte) markers. EfPiwiA mRNA was expressed in cells with archeocyte cell morphological features. We demonstrated that these EfPiwiA‐expressing cells were indeed stem cells by showing their ability to proliferate, as indicated by BrdU‐incorporation, and to differentiate, as indicated by the coexpression of EfPiwiA with cell‐lineage‐specific genes in presumptive committed archeocytes. EfPiwiA mRNA expression was maintained in mature choanocytes forming chambers, in contrast to the transition of gene expression from EfPiwiA to cell‐lineage‐specific markers during archeocyte differentiation into other cell types. Choanocytes are food‐entrapping cells with morphological features similar to those of choanoflagellates (microvillus collar and a flagellum). Their known abilities to transform into archeocytes under specific circumstances and to give rise to gametes (mostly sperm) indicate that even when they are fully differentiated, choanocytes maintain pluripotent stem cell‐like potential. Based on the specific expression of EfPiwiA in archeocytes and choanocytes, combined with previous studies, we propose that both archeocytes and choanocytes are components of the demosponge stem cell system. We discuss the possibility that choanocytes might represent the ancestral stem cells, whereas archeocytes might represent stem cells that further evolved in ancestral multicellular organisms.     

Choanoflagellates, choanocytes, and animal multicellularity

Abstract. It is widely accepted that multicellular animals (metazoans) constitute a monophyletic unit, deriving from ancestral choanoflagellate‐like protists that gave rise to simple choanocyte‐bearing metazoans. However, a re‐assessment of molecular and histological evidence on choanoflagellates, sponge choanocytes, and other metazoan cells reveals that the status of choanocytes as a fundamental cell type in metazoan evolution is unrealistic. Rather, choanocytes are specialized cells that develop from non‐collared ciliated cells during sponge embryogenesis. Although choanocytes of adult sponges have no obvious homologue among metazoans, larval cells transdifferentiating into choanocytes at metamorphosis do have such homologues. The evidence reviewed here also indicates that sponge larvae are architecturally closer than adult sponges to the remaining metazoans. This may mean that the basic multicellular organismal architecture from which diploblasts evolved, that is, the putative planktonic archimetazoan, was more similar to a modern poriferan larva lacking choanocytes than to an adult sponge. Alternatively, it may mean that other metazoans evolved from a neotenous larva of ancient sponges. Indeed, the Porifera possess some features of intriguing evolutionary significance: (1) widespread occurrence of internal fertilization and a notable diversity of gastrulation modes, (2) dispersal through architecturally complex lecithotrophic larvae, in which an ephemeral archenteron (in dispherula larvae) and multiciliated and syncytial cells (in trichimella larvae) occur, (3) acquisition of direct development by some groups, and (4) replacement of choanocyte‐based filter‐feeding by carnivory in some sponges. Together, these features strongly suggest that the Porifera may have a longer and more complicated evolutionary history than traditionally assumed, and also that the simple anatomy of modern adult sponges may have resulted from a secondary simplification. This makes the idea of a neotenous evolution less likely than that of a larva‐like choanocyte‐lacking archimetazoan. From this perspective, the view that choanoflagellates may be simplified sponge‐derived metazoans, rather than protists, emerges as a viable alternative hypothesis. This idea neither conflicts with the available evidence nor can be disproved by it, and must be specifically re‐examined by further approaches combining morphological and molecular information. Interestingly, several microbial lin°Cages lacking choanocyte‐like morphology, such as Corallochytrea, Cristidiscoidea, Ministeriida, and Mesomycetozoea, have recently been placed at the boundary between fungi and animals, becoming a promising source of information in addition to the choanoflagellates in the search for the unicellular origin of animal multicellularity.     

Comparative study of the choanosome of Porifera: 1. The Homoscleromorpha

The choanoderm and pinacoderm of representatives of the two families of Homoscleromorpha sponges, the Oscarellidae and Plakinidae, have been examined by transmission and scanning electron microscopy. Different fixative procedures have shown the dramatic influence of fixation conditions on the morphology of choanocytes. These two families of sponges have the following morphological features in common: flagellated endopinacocytes with short apical microvilli and basal pseudopods; the presence of a very thin and dense sheet of matrix material which limits the mesohyl. There are, however, only minor differences in the flagellar morphology, granule content, and anchoring system of their choanocytes. Two findings are of particular interest: (1) the presence of glycocalyx bridges between the microvilli of the choanocyte collar; and (2) the discovery of a new cell type, the apopylar cell, which has a morphology intermediate between that of pinacocytes and choanocytes. The apopylar cells limit the apopylar opening of the choanocyte chamber and indicate the transition between choanoderm and pinacoderm.     

Feeding in a calcareous sponge: particle uptake by pseudopodia

Sponges are considered to be filter feeders like their nearest protistan relatives, the choanoflagellates. Specialized "sieve" cells (choanocytes) have an apical collar of tightly spaced, rodlike microvilli that surround a long flagellum. The beat of the flagellum is believed to draw water through this collar, but how particles caught on the collar are brought to the cell surface is unknown. We have studied the interactions that occur between choanocytes and introduced particles in the large feeding chambers of a syconoid calcareous sponge. Of all particles, only 0.1-microm latex microspheres adhered to the collar microvilli in large numbers, but these were even more numerous on the choanocyte surface. Few large particles (0.5- and 1.0-microm beads and bacteria) contacted the collar microvilli; most were phagocytosed by lamellipodia at the lateral or apical cell surface, and clumps of particles were engulfed by pseudopodial extensions several micrometers from the cell surface. Although extensions of the choanocyte apical surface up to 16 microm long were found, most were 4 microm long, twice the height of the collar microvilli. These observations offer a different view of particle uptake in sponges, and suggest that, at least in syconoid sponges, uptake of particles is less dependent on the strictly sieving function of the collar microvilli.     

Ultrastructural study of differentiation processes during aggregation of purified sponge archaeocytes

Archaeocytes from the spongeEphydatia fluviatilis were dissociated and then isolated on Ficoll density gradients. Their aggregation and reconstitution processes were studied by transmission electron microscopy to determine their capabilities for differentiation.Archaeocyte aggregates follow a well defined sequence of differentiation to generate the characteristic structures of a sponge. Pinacoderm is the first structure to be regenerated and appears progressively at the surface of the 12 h aggregates. Pinacocytes which have differentiated in archaeocyte aggregates are identical to native ones except that the nucleolus remains in most cells. The choanocytes appear only after 24 h by a two step process. First, small cells (choanoblasts) are formed from archaeocytes by mitosis. These cells then transform into fully differentiated choanocytes possessing collars and flagella. The early choanocyte chambers are small, irregular and randomly dispersed in the aggregates. Finally, collencytes and sclerocytes begin to appear just before the aggregates spread on the substrate.The differentiation of a suspension of pure archaeocytes is a unique model system to study sponge cell differentiation and has allowed us to demonstrate that archaeocytes isolated from developed sponges maintain the capacity to differentiate even though this capacity is not usually expressed.     

Body structure of marine sponges

Summary Each choanocyte chamber of Petrosia ficiformis is formed by a slightly outpocked choanocyte epithelium and by a ring of three or four uniflagellated cone cells surrounding the apopyle. The apopyle opens into a small aphodus, which leads the water flow to larger excurrent canals. Pinacocytes of the incurrent canal system cover the basal surface of the choanocytes and separate them from the incurrent canals and the mesenchyme. The water flows into the chambers by pores in the pinacocyte cover and then through gaps between adjacent choanocytes. To our knowledge this is the first report of a leuconoid canal system in which choanocyte chambers are covered by a pinacocyte epithelium of the incurrent canal system that isolates the chambers from the mesenchyme. A future comprehensive revision of the types of canal systems in sponges seems to be necessary. Permanent affiliation: Department of Biology and Health Sciences, University of Hartford, West Hartford, CT 06117, USA     

Oogenesis and larval development of Scypha ciliata (Porifera,Calcarea)

Summary Scypha ciliata is a syconoid sponge. Its oocytes differentiate from choanocytes located near the apopyle of a flagellated chamber, and initially they remain in that location, in a trophic complex with neighbouring choanocytes. When this first growth phase is completed, the oocyte migrates to the periphery of the sponge. There it undergoes a second growth phase, in which it phagocytizes choanocytes and mesenchyme cells.Fertilization of the mature egg is assisted by a converted choanocyte, the sperm carrier cell. This cell penetrates the oocyte and transfers to it the sperm contained in a carriercell vacuole. No meiotic events have yet been observed.Cleavage is asynchronous, with holoblastic, approximately equal divisions. After the first cleavage steps the blastomeres often contain multiple nuclei. The single-layered blastoderm of the stomoblastula consists of many micromeres with flagella that project into the blastocoel, a few macromeres and four cruciform cells. There is no development of a follicle epithelium.The stomoblastula develops into the amphiblastula by inversion; with the assistance of the maternal choanocyte epithelium, the hollow sphere turns inside out, simultaneously moving out of the mesoderm and into the lumen of the adjacent flagellated chamber. In this process, the blastocoel of the stomoblastula is lost. The flagellated cells that form the wall of the amphiblastula now have their flagella extending outward; the amphiblastula also comprises four cruciform cells, macrogranular and agranular cells. The larval cavity of the amphiblastula is a newly formed structure.Abbreviations AB amphiblastula - AP apopyle - BC blastocoel - aC agranular cell - maC macrogranular cell - miC microgranular cell - CB crystalline body - CC central cavity - Ch choanocyte - fCh flat choanocyte - gCh granulate choanocyte - CM cell membrane - Co collar of choanocyte - CrC cruciform cell - DM dense material - EM electron micrograph - F flagellum - FC flagellated cell - FCm flagellated chamber - FL free larva - FV food vacuole - IR interior region - LC larval cavity - M mesenchyme - Ma macromere - MC mesenchyme cell - Mi micromere - N nucleus - Nu nucleolus - O opening - OC oocyte - P psudopodium - PC pinacocyte - PhM phase-contrast micrograph - Po pore - PP prosopyle - S sperm - SB stomoblastula - SC segmentation cavity - SCC sperm-carrier cell - SV sperm vacuole - lT large trophocyte - sT small trophocyte - V vacuole - VC vesicular cytoplasm - VM vacuole membrane     

Choanocyte ultrastructure in Halisarca dujardini (Demospongiae,Halisarcida)

Understanding poriferan choanocyte ultrastructure is crucial if we are to unravel the steps of a putative evolutionary transition between choanoflagellate protists and early metazoans. Surprisingly, some aspects of choanocyte cytology still remain little investigated. This study of choanocyte ultrastructure in the halisarcid demosponge Halisarca dujardini revealed a combination of minor and major distinctive traits, some of them unknown in Porifera so far. Most significant features were 1) an asymmetrical periflagellar sleeve, 2) a battery of specialized intercellular junctions at the lateral cell surface complemented with an array of lateral interdigitations between adjacent choanocytes that provides a particular sealing system of the choanoderm, and 3) a unique, unexpectedly complex, basal apparatus. The basal apparatus consists of a basal body provided with a small basal foot and an intricate transverse skeleton of microtubules. An accessory centriole, which is not perpendicular to the basal body, is about 45°. In addition, a system of short striated rootlets (periodicity = 50–60 nm) arises from the proximal edge of the basal body and runs longitudinally to contact the nuclear apex. This is the first flagellar rootlet system ever found in a choanocyte. The accessory centriole, the rootlet system, and the nuclear apex are all encircled by a large Golgi apparatus, adding another distinctive feature to the choanocyte cytology. The set of distinct features discovered in the choanocyte of H. dujardini indicates that the ultrastructure of the poriferan choanocyte may vary substantially between sponge groups. It is necessary to improve understanding of such variation, as the cytological features of choanocytes are often coded as characters both for formulation of hypotheses on the origin of animals and inference of phylogenetic relationships at the base of the metazoan tree. J. Morphol., 2009. © 2008 Wiley‐Liss, Inc.     

The spermatogenesis ofHalichondria panicea (Porifera,Demospongiae)

Summary Spermatogenesis of the marine spongeHalichondria panicea begins with the break up of choanocyte chambers, choanocytes constituting the origin of spermatogonia. The transition from choanocytes to spermatogonia is direct, without cell division. Already the spermatogonia are flagellated. The ensuing large aggregates of spermatogonia are enclosed by spermatocyst-building cells. Further development takes place within the spermatocysts, mostly arranged in fields which, however, lack any developmental gradient. Within a single spermatocyst development is mostly synchronous. Spermatogonia transform into first order spermatocytes directly. The transition from spermatid to spermatozoon is characterized by an unusual prolongation of the chromatin, often resulting in a helical form of the chromosome material and a strong enlargement of the mitochondria which align with the nucleus, leading to an irregular shape of the spermatozoon. Another exceptional feature is the virtual absence of a Golgi apparatus during all stages of spermatogenesis. TheH. panicea investigated here contained only male reproductive elements, thus appear to be gonochorists. Some features of the spermatogenesis ofH. panicea, such as dissolving choanocyte chambers, the enclosure of spermatogonia by spermatocyst-building cells and the formation of a synaptonemal complex in first order spermatocytes occur in other sponge species as well; however, the early presence of flagella in spermatogonia, the absence of the Golgi apparatus and the later irregular development of nuclei, mitochondria and the spermatozoa themselves represent features hitherto not observed in sponges.     

Atypical properties displayed by annexin A9, a novel member of the annexin family of Ca(2+) and lipid binding proteins

Annexin A9 is a novel member of the annexin family of Ca(2+) and phospholipid binding proteins which has so far only been identified in EST data bases and whose deduced protein sequence shows mutations in residues considered crucial for Ca(2+) coordination in other annexins. To elucidate whether the annexin A9 protein is expressed as such and to characterize its biochemical properties we probed cell extracts with specific anti-annexin A9 antibodies and developed a recombinant expression system. We show that the protein is found in HepG2 hepatoma cell lysates and that a green fluorescent protein-tagged form is abundantly expressed in the cytosol of HeLa cells. Recombinant expression in bacteria yields a soluble protein that can be enriched by conventional chromatographic procedures. The protein is capable of binding phosphatidylserine containing liposomes albeit only at Ca(2+) concentrations exceeding 2 mM. Moreover and in contrast to other annexins this binding appears to be irreversible as the liposome-bound annexin A9 cannot be released by Ca(2+) chelation. These results indicate that annexin A9 is a unique member of the annexin family whose intracellular activity is not subject to Ca(2+) regulation.     

STRUCTURE AND FORMATION OF THE COLLAR IN CHOANOCYTES OF TETILLA SERICA (LEBWOHL), DEMOSPONGE

There are two types of collar in the choanocytes of adult Tetilla serica : one type is a continuous cytoplasmic tube and the other consists of discontinuous microvilli. The former is found in the small flagellated chamber and is considered to belong to a young choanocyte in the process of differertiation. To confirm this idea, very young choanocytes which are about to differentiate the collar were examined during embryogenesis.
The youngest choanocytes are noticed forming aggregations of small cells in 3-day larvae. Around the flagellum in each choanocyte, there is a depression which will become wider. At first, the collar is observed as a ring of cytoplasm; next this extends outward and becomes thinner, and finally it divides into microvilli. The microvillous collar is formed by the opening of vesicles and fusion of their membranes. These vesicles are considered to be derived from the Golgi complex. The process of collar formation through fusion of vesicles is discussed.     

Continuous cell movements rearrange anatomical structures in intact sponges.

Time-lapse cinemicrography was used to record the active movements of cells in living intact sponges. Each of the three main cell types (pinacocytes, mesohyl cells, and choanocytes) continuously moved and rearranged themselves so that the internal anatomy of the sponge was continuously remodeled. The shape and appearance of the sponges anatomical structures often changed substantially within a few hours. The most motile were the mesohyl cells, with many moving as fast as one cell-length per minute (15 microns/min). Mesohyl cell locomotion was often accompanied by displacements of spicules, canals, and choanocyte chambers; the patterns of these displacements suggested that the mesohyl cells were providing the motive forces for these rearrangements. The locomotion of the pinacocytes varied according to position: those along the outer sponge margins were most active, whereas those in other parts of the surface moved relatively little. Choanocytes were never observed to undergo independent locomotion but were always found grouped together in choanocyte chambers. These choanocyte chambers interacted with pinacocytes and mesohyl cells to form excurrent canals, which continuously moved, fused with, and branched from one another. These observations suggest that the experimental phenomenon of sponge cell-reaggregation and reconstitution, discovered by H. V. Wilson, represents an extreme version of morphogenetic processes that normally go on continuously within intact sponges. The results from the present study also suggest that these cellular rearrangements are controlled by active cell movements and behavioral responses that include but are not limited to selective cell adhesion.     

Fine structures of sponge cell membranes: comparative study with freeze-fracture and conventional thin section methods

Freeze-fracture replicas of sponge cell membranes revealed in general a low density of intramembranous particles, with the exceptions of the membrane (silicalemma) surrounding the siliceous spicules in Ephydatia and the membranes of spherulous cells in Chondrosia. In addition, several types of particle arrangements were observed. A classical necklace is present at the base of the choanocyte flagellum. Rosettes of particles are particularly obvious in the apical membranes of choanocytes, where they are associated with the fuzzy coat covering these cells. Parallel ridges of particles were observed along the microvilli of the choanocyte collar, at sites of insertion of connecting filaments. Rows of particles were observed in the plasma membrane of pinacocytes in Ephydatia where they are located on areas deformed by protruding fibrillar inclusions. Pinacocyte plasma membranes in this species also can contain accumulations of particles which are likely related to desmosomes. Single rows of aligned particles and double rows of staggered particles (sometimes organized in large plates) in addition to rhombic particle arrays were encountered on replicas of marine sponge cell membranes. No classical arrangements corresponding to gap junctions, tight junctions or septate desmosomes were observed. The significance of these data is analysed.     

NK homeobox genes with choanocyte-specific expression in homoscleromorph sponges

Data on nonbilaterian animals (sponges, cnidarians, and ctenophores) have suggested that Antennapedia (ANTP) class homeobox genes played a crucial role in the early diversification of animal body plans. Estimates of ancestral gene diversity within this important class of developmental regulators have been mostly based on recent analyses of the complete genome of a demosponge species, leading to the proposal that all ANTP families found in nonsponges animals (eumetazoans) derived from an ancestral "proto-NK" six-gene cluster. However, a single sponge species cannot reveal ancestral metazoan traits, in particular because lineage-specific gene duplications or losses are likely to have occurred during the long history of the Porifera. We thus looked for ANTP genes by degenerate polymerase chain reaction search in five species belonging to the Homoscleromorpha, a sponge lineage recently phylogenetically classified outside demosponges and characterized by unique histological features. We identified new genes of the ANTP class called HomoNK. Our phylogenetic analyses placed HomoNK (without significant support) close to the NK6 and NK7 families of cnidarian and bilaterian ANTP genes and did not recover the monophyly of the proposed "proto-NK" cluster. Our expression analyses of the HomoNK gene OlobNK in adult Oscarella lobularis showed that this gene is a strict marker of choanocytes, the most typical sponge cell type characterized by an apical flagellum surrounded by a collar of microvilli. These results are discussed in the light of the predominant neurosensory expression of NK6 and NK7 genes in bilaterians and of the recent proposal that choanocytes could be the sponge homologs of sensory cells.     

Comparative study of the choanosome of porifera: II. The keratose sponges

The aquiferous system of representatives of the orders Dictyocer-atida, Dendroceratida, and Verongida has been studied to note its relevance to the systematics of the groups. The volume of the choanocyte chamber, the size and shape of the choanocytes, the number of choanocytes per chamber, the relative development of the mesohyl, and the features of endopinacocytes are estimated from scanning and transmission electron microscopic observations of representatives of most families of the three orders. Although the Dysideidae have a reticulate skeleton and were classified in the order Dictyoceratida, they are actually closer to the Aplysillidae (Dendroceratida) than to dictyoceratids. The anatomy and cytology of the Halisarcidae differ profoundly from those of these three orders and are clearly more closely related to nonkeratose sponges. Some changes in classification lead to a pattern with highly homogeneous orders that clearly differ in their anatomic and cytologic features, which does not support the hypothesis of a common origin of the “keratose” sponges.