Cell lineage analysis in ascidian embryos by intracellular injection of a tracer enzyme. III. Up to the tissue restricted stage

Cell lineages during embryogenesis of the ascidian Halocynthia roretzi were analyzed up until the stage where each blastomere was fated to be only a single tissue type (i.e., the tissue restricted stage) by intracellular injection of horseradish peroxidase using the iontophoretic injection method. Initially, the developmental fates of all blastomeres of the 64-cell stage embryo were examined, and thereafter, only the fates of daughter blastomeres of those blastomeres that were not tissue restricted at the 64-cell stage were traced. The developmental fates of blastomeres were highly invariant except for two candidates for "equivalence groups" (J. Kimble, J. Sulston, and J. White (1979). In "Cell Lineage, Stem Cells and Cell Determination," pp. 59-68. Elsevier, Amsterdam/New York), in which cellular interaction is suggested to be involved in the specification of the fates. The right and left a8.25 cells gave rise to the otolith and ocellus, and the right and left b8.17 cells gave rise to the spinal cord and endodermal strand in a complementary manner. No fixed relationship existed between the position of the blastomere and its derivative. Most restrictions of cell fates occurred early in cleavage. The numbers of blastomeres which generated a single type of tissue were 44 at the 64-cell stage and 94 at the 110-cell stage. Eight pairs of blastomeres had not yet become tissue restricted by the 110-cell stage. Almost complete lineages of epidermis, nervous system, muscle, mesenchyme, notochord, and endodermal tissues were described, and a fate map was constructed for the blastula. For certain tissues, the primordial cells occupied two different regions. Supplementary investigations of the lineage of muscle cells were also performed on embryos of another species, Ciona intestinalis.     

Developmental potential for tissue differentiation of fully dissociated cells of the ascidian embryo

Summary Initially, each tissue-progenitor blastomere of embryos of the ascidian Halocynthia was identified and isolated manually at the 110-cell (late-blastula) stage, the time at which most of the blastomeres have assumed a particular fate, such that each gives rise to a single type of tissue. The isolates were allowed to develop as partial embryos, then tissue differentiation was examined by monitoring the expression of specific molecular markers for differentiation of epidermis, endoderm, muscle and notochord. Essentially, all of the precursor blastomeres of these four kinds of tissue expressed the appropriate features of tissue differentiation in isolation, indicating that determination is already complete in most of the blastomeres by the 110-cell stage. Next, in order to evaluate the absolute capacity of cells for autonomous development, embryos were maintained continuously in a dissociated state from the first cleavage to the 110-cell stage, then the cells were allowed to develop into partial embryos. Tissue differentiation in the partial embryos was examined. The results showed the striking autonomy of the processes of segregation of developmental potential, as well as the autonomy of the processes of expression of differentiated phenotypes, namely those of epidermis and endoderm. Autonomous muscle differentiation was also observed; however, excess formation of muscle partial embryos occurred. The hypothesis that fate determination is mediated by localized maternal information in the egg cytoplasm is supported by the evidence of development of these tissues. By contrast, no evidence of notochord differentiation was observed in the partial embryos.     

Duration of competence and inducing capacity of blastomeres in notochord induction during ascidian embryogenesis

Notochord cells in ascidian embryos are formed by the inducing action of cells of presumptive endoderm, as well as neighboring presumptive notochord, at the 32-cell stage. Studies of the timing of induction using recombinations of isolated blastomeres have suggested that notochord induction must be initiated before the decompaction of blastomeres at the 32-cell stage and is completed by the 64-cell stage. However, it is not yet clear how the duration of notochord induction is strictly limited. In the present paper, the aim was to determine in detail when the presumptive notochord blastomeres lost their competence to respond, and when the presumptive endoderm blastomeres produced inducing signals for the notochord. Presumptive notochord blastomeres and presumptive endoderm blastomeres were isolated from early 32-cell embryos, and were heterochronously recombined at various stages ranging from the early 32-cell stage to the 64-cell stage. Presumptive notochord blastomeres could respond to inductive signals at the early 32-cell stage, and started to lose their responsiveness at the decompaction stage. By contrast, the presumptive endoderm blastomeres persisted in their inducing capacity even at the 64-cell stage. These observations suggest that the loss of competence in presumptive notochord blastomeres limits the duration of notochord induction in intact ascidian embryos.     

Nuclear plasticity and timing mechanisms of the initiation of alkaline phosphatase expression in cytoplasm-transferred blastomeres of ascidians

Egg cytoplasm containing endoderm determinants was transferred to presumptive-muscle or presumptive-epidermis blastomeres isolated from cleavage-stage embryos of the ascidian Halocynthia roretzi. We investigated three aspects of the expression of endoderm-specific alkaline phosphatase (ALP) activity. First, we examined whether ectopic ALP expression, an indication of ectopic endoderm formation, was promoted in cytoplasm-transferred blastomeres isolated at late-cleavage stage. The results showed that the cell fate was converted by the introduced cytoplasm, even in recipient blastomeres in which the cell fate was already restricted to muscle or epidermis, and in those where expression of the muscle- or epidermis-specific genes was already initiated. Next, we examined the formation of endoderm and other tissue in embryos by double staining for ALP and muscle- or epidermis-specific marker. Regions positive for ALP and positive for muscle or epidermis marker were mutually exclusive. These results suggested that muscle- or epidermis-specific genes that were already expressed in the recipient blastomeres were down-regulated in ectopically forming endoderm cells. This is evidence for nuclear plasticity during ascidian embryogenesis. In the last series of experiments, we investigated the timing of the appearance of ALP activity in cytoplasm-transferred embryos. In the partial embryos that were derived from various combination of recipient blastomeres and donor cytoplasm obtained from various staged eggs and embryos, the timing seemed to coincide with the time that starts when cell fusion for cytoplasmic transfer was done. Therefore, the clock that determines the timing of the initiation of ALP expression is likely to start at the moment of cell fusion. Several possible hypotheses for the timing mechanism are discussed.     

Developmental autonomy of presumptive notochord cells in partial embryos of an ascidian

Summary

Ultrastructural features of larval notochord cell differentiation, sheath (membrane leaflets and filaments) and vacuoles of intracellular colloid, were found in some cells of certain partial embryos of the ascidian, Ciona intestinalis. As expected from established lineage fate maps, mature quarter-embryos developing from microsurgically isolated anterior-vegetal blastomeres (A4.1 pair) at the 8-cell stage had some cells with the notochord features. Such cells, however, also occurred in quarter-embryos resulting from the posterior-vegetal blastomere pair (B4.1) and in partial embryos derived from the B5.1 cell pair isolated at the next cleavage of the B4.1 blastomeres. These findings confirm a prediction of additional notochord cell fates from a recent revision of the ascidian lineage map based on cell marking with microinjected horseradish peroxidase. Partial embryos obtained from other lineages of the 8- and 16-cell stages did not develop notochord cells.     

Determinative mechanisms in secondary muscle lineages of ascidian embryos: development of muscle-specific features in isolated muscle progenitor cells

Muscle cells of the ascidian larva originate from three different lines of progenitor cells, the B-line, A-line and b-line. Experiments with 8-cell embryos have indicated that isolated blastomeres of the B-line (primary) muscle lineage show autonomous development of a muscle-specific enzyme, whereas blastomeres of the A-line and b-line (secondary) muscle lineage rarely develop the enzyme in isolation. In order to study the mechanisms by which different lines of progenitors are determined to give rise to muscle, blastomeres were isolated from embryos of Halocynthia roretzi at the later cleavage stages when conspicuous restriction of the developmental fate of blastomeres had already occurred. Partial embryos derived from B-line muscle-lineage cells of the 64-cell embryo (B7.4, B7.5 and B7.8) showed autonomous expression of specific features of muscle cells (acetylcholinesterase, filamentous actin and muscle-specific antigen). In contrast, b-line muscle-lineage cells, even those isolated from the 110-cell embryo (b8.17 and b8.19), did not express any muscle-specific features, even though their developmental fate was mainly restricted to generation of muscle. Isolated A-line cells from the 64-cell embryos (A7.8) did not show any features of muscle differentiation, whereas some isolated A-line cells from the 110-cell embryos (A8.16) developed all three above-mentioned features of muscle cells. This transition was shown to occur during the eighth cell cycle. These results suggest that the mechanism involved in the process of determination of the secondary-lineage muscle cells differs from that of the primary-lineage muscle cells. Interaction with cells of other lineages may be required for the determination of secondary precursors to muscle cells. The presumptive b-line and A-line muscle cells that failed to express muscle-specific features in isolation did not develop into epidermal cells. Thus, although interactions between cells may be required for muscle determination in secondary lineages, the process may represent a permissive type of induction and may differ from the processes of induction of mesoderm in amphibian embryos.     

Muscle cell differentiation in ascidian embryos analysed with a tissue-specific monoclonal antibody

Utilizing a muscle-specific monoclonal antibody (Mu-2) as a probe, we analysed developmental mechanisms involved in muscle cell differentiation in ascidian embryos. The antigen recognized by Mu-2 was a single polypeptide with a relative molecular mass of about 220 X 10(3). It first appeared at the early tailbud stage and continued to be expressed until the swimming larva stage. There were distinct and separate puromycin and actinomycin D sensitivity periods during the occurrence of the antigen, suggesting the new synthesis of the polypeptide by developing muscle cells. Embryos that had been permanently arrested with aphidicolin in the early cleavage stages up to the 32-cell stage did not express the antigen. DNA replications may be required for the antigen expression. Embryos that had been arrested with cytochalasin B in the 8-cell and later stages developed the antigen, and the number and position of the arrested blastomeres exhibiting the differentiation marker almost corresponded to those of the B4.1-line muscle lineage. Furthermore, in quarter embryos developed from each blastomere pair isolated from the 8-cell embryo, all the B4.1 as well as a part of b4.2 partial embryos expressed the antigen, while the a4.2 and A4.1 partial embryos did not show the antigen expression. These results may provide further support for the existence of cytoplasmic determinants for muscle cell differentiation in this mosaic egg.     

Binary specification of nerve cord and notochord cell fates in ascidian embryos

In the ascidian embryo, the nerve cord and notochord of the tail of tadpole larvae originate from the precursor blastomeres for both tissues in the 32-cell-stage embryo. Each fate is separated into two daughter blastomeres at the next cleavage. We have examined mechanisms that are responsible for nerve cord and notochord specification through experiments involving blastomere isolation, cell dissociation, and treatment with basic fibroblast growth factor (bFGF) and inhibitors for the mitogen-activated protein kinase (MAPK) cascade. It has been shown that inductive cell interaction at the 32-cell stage is required for notochord formation. Our results show that the nerve cord fate is determined autonomously without any cell interaction. Presumptive notochord blastomeres also assume a nerve cord fate when they are isolated before induction is completed. By contrast, not only presumptive notochord blastomeres but also presumptive nerve cord blastomeres forsake their default nerve cord fate and choose the notochord fate when they are treated with bFGF. When the FGF-Ras-MAPK signaling cascade is inhibited, both blastomeres choose the default nerve cord pathway, supporting the results of blastomere isolation. Thus, binary choice of alternative fates and asymmetric division are involved in this nerve cord/notochord fate determination system, mediated by FGF signaling.     

Mesoderm formation by isolated and cultivated 8-cell stage blastomeres of the teleost, Leucopsarion ptersii (shiro-uo).

Isolation of cleavage-stage blastomeres and the study of their developmental potential has been used extensively for analyzing the mechanisms of embryogenesis in vertebrates, including amphibians and echinoderms. We devised a method to isolate 8-cell stage blastomeres in the teleost, shiro-uo, by utilizing its unique cleavage pattern of the horizontal 3rd cleavage plane. Removal of all the upper blastomeres at the 8-cell stage allowed almost normal embryogenesis from the remaining lower blastomeres and yolk cell mass. Isolated upper or lower blastomeres formed vesicles and spherical bodies, which later showed morphological changes during cultivation. Mesoderm formation was detected not only in the cultivated lower blastomeres or whole blastomeres but also in the upper blastomeres isolated from the yolk cell mass at the 8-cell stage, although at a lower frequency than the lower blastomeres. These results indicated the presence of very early signaling for mesoderm induction, which is independent from the currently postulated signals from the yolk syncytial layer at later stages. This also indicated non-equivalence or differentiation of the blastomeres from the very early cleavage stage in teleost embryos.     

The pie-1 and mex-1 genes and maternal control of blastomere identity in early C. elegans embryos.

During C. elegans embryogenesis an 8-cell stage blastomere, called MS, undergoes a reproducible cleavage pattern, producing pharyngeal cells, body wall muscles, and cell deaths. We show here that maternal-effect mutations in the pie-1 and mex-1 genes cause additional 8-cell stage blastomeres to adopt a fate very similar to that of the wild-type MS blastomere. In pie-1 mutants one additional posterior blastomere adopts an MS-like fate, and in mex-1 mutants four additional anterior blastomeres adopt an MS-like fate. We propose that maternally provided pie-1(+) and mex-1(+) gene products may function in the early embryo to localize or regulate factors that determine the fate of the MS blastomere.     

Determinative properties of muscle lineages in ascidian embryos

Blastomeres removed from early cleavage stage ascidian embryos and reared to 'maturity' as partial embryos often elaborate tissue-specific features typical of their constituent cell lineages. We used this property to study recent corrections of the ascidian larval muscle lineage and to compare the ways in which different lineages give rise to muscle. Our evaluation of muscle differentiation was based on histochemical localization and quantitative radiometric measurement of a muscle-specific acetylcholinesterase activity, and the development of myofilaments and myofibrils as observed by electron microscopy. Although the posterior-vegetal blastomeres (B4.1 pair) of the 8-cell embryo have long been believed to be the sole precursors of larval muscle, recent studies using horseradish peroxidase to mark cell lineages have shown that small numbers of muscle cells originate from the anterior-vegetal (A4.1) and posterior-animal (b4.2) blastomeres of this stage. Fully differentiated muscle expression in isolated partial embryos of A4.1-derived cells requires an association with cells from other lineages whereas muscle from B4.1 blastomeres develops autonomously. Clear differences also occurred in the time acetylcholinesterase activity was first detected in partial embryos from these two sources. Isolated b4.2 cells failed to show any muscle development even in combination with anterior-animal cells (a4.2) and are presumably even more dependent on normal cell interactions and associations. Others have noted an additional distinction between the different sources of muscle: muscle cells from non-B4.1 lineages occur exclusively in the distal part of the tail, while the B4.1 descendants contribute those cells in the proximal and middle regions. During the course of ascidian larval evolution tail muscle probably had two origins: the primary lineage (B4.1) whose fate was set rigidly at early cleavage stages and secondarily evolved lineages which arose later by recruitment of cells from other tissues resulting in increased tail length. In contrast to the B4.1 lineage, muscle development in the secondary lineages is controlled less rigidly by processes that depend on cell interactions.     

Isolation of an early neural maker gene abundantly expressed in the nervous system of the ascidian, Halocynthia roretzi

Ascidian tadpole larvae possess a primitive nervous system, which is a prospective prototype of the chordate nervous system. It is composed of relatively few cells but sufficient for complex larval behavior. Here we report on HrETR-1, a gene zygotically expressed in a large proportion of the developing neural cells of the ascidian, Halocynthia roretzi. HrETR-1 is an early neural marker which can be used for analyzing neural differentiation. HrETR-1 expression intensified in most neural cells of genes isolated to date, in both central and peripheral nervous systems including palps as early as the 110-cell stage. Using this gene as a probe, we characterized neural cells in the nervous system as well as confirming their origins. Also, we recognized three types of peripheral epidermal neurons which presumably correlate to the larval neurons previously reported for another ascidian. Among these, five bilateral neurons located in the anterior region of the trunk appeared to be derived from a8.26 blastomeres.     

Expression of an antigen specific for trunk lateral cells in quarter embryos of the ascidian, Halocynthia roretzi

Cell-lineage analysis has demonstrated that a pair of the right and left A7.6 cells of a 64-cell embryo of the ascidian Halocynthia roretzi, descendants of A4.1 cells of an 8-cell embryo, give rise to trunk lateral cells (TLCs). In this study, in order to investigate cellular mechanisms involved in the specification of TLCs, we have examined the expression of a TLC-specific antigen in cleavage-arrested embryos and in quarter partial embryos. Although cleavage arrest of embryos by treatment with cytochalasin B at early stages, prior to and including the 16-cell stage, inhibited expression of the TLC-specific antigen, embryos arrested at the 32-cell stage and at later stages developed the antigen. The only blastomeres exhibiting expression of the antigen were the presumptive TLCs, as predicted by cell-lineage assignments. When the developmental potential of quarter embryos that originated from four isolated blastomere-pairs (a4.2, b4.2, A4.1, and B4.1 pairs) of an 8-cell embryo was examined, the A4.1 quarter embryos, which are developmentally fated to give rise to TLCs, rarely showed evidence of expression of the antigen. Expression of the antigen was not observed in a4.2 and b4.2 quarter embryos, which are not associated with the TLC fate. By contrast, expression of the antigen was detected in about a half of the B4.1 quarter embryos which are also not associated with the TLC fate. These results are discussed with reference to the relationship between TLCs and mesenchyme cells.     

The BMP/CHORDIN antagonism controls sensory pigment cell specification and differentiation in the ascidian embryo

We have investigated the role of the bone morphogenetic protein (BMP) pathway during neural tissue formation in the ascidian embryo. The orthologue of the BMP antagonist, chordin, was isolated from the ascidian Halocynthia roretzi. While both the expression pattern and the phenotype observed by overexpressing chordin or BMPb (the dpp-subclass BMP) do not suggest a role for these factors in neural induction, BMP/CHORDIN antagonism was found to affect neural patterning. Overexpression of BMPb induced ectopic sensory pigment cells in the brain lineages that do not normally form pigment cells and suppressed pressure organ formation within the brain. Reciprocally, overexpressing chordin suppressed pigment cell formation and induced ectopic pressure organ. We show that pigment cell formation occurs in three steps. (1) During cleavage stages ectodermal cells are neuralized by a vegetal signal that can be substituted by bFGF. (2) At the early gastrula stage, BMPb secreted from the lateral nerve cord blastomeres induces those neuralized blastomeres in close proximity to adopt a pigment cell fate. (3) At the tailbud stage, among these pigment cell precursors, BMPb induces the differentiation of specifically the anterior type of pigment cell, the otolith; while posteriorly, CHORDIN suppresses BMP activity and allows ocellus differentiation.