m endo 16

kroxl kroxl

Maternal ß-catenin nuclearization TCF/LEF1 System

Writs amplification

veg2 Endomesoderm Specification Genes vegs Endoderm Specification Genes vegi Endoderm Specification Genes

Early micromere signal

Late micromere s ignai

Signal dependant ß-catenin nuclearization veg2 Mesoderm Specification Genes

Notch amplificaron

4th - 6th cleavage

7th - 9th cleavage late blastula


FIGURE 3.3 Conditional and autonomous specification of veg2 endomes-odermal domain in the sea urchin embryo. (A) Expression of the endol6 gene is conditional on signaling from micromeres in the 4th-6th cleavage interval. endol6 activity is visualized by in situ hybridization in 24 h S. purpuratus blastulae, following microsurgical removal of all four micromeres at: (AI) 4th cleavage, (A2) 5th cleavage (about 45 min later), (A3) 6th cleavage, (A4) untreated control. [(A) From Ransick and Davidson (1995) Development 121, 3215-3222 and The Company of Biologists, Ltd.] (B) Expression of SpKroxl also requires signaling from micromeres. Embryos were fixed for in situ hybridization at 24 h. (Bl, B2) Two embryos from which micromeres were removed at 4th cleavage; (B3) control. [(B) From A. Ransick and E. Davidson, unpublished data.] (C) Autonomous accumulation of (3-catenin in vegetal cell nuclei early in development, vegetal pole view. [3-catenin is visualized immunocytologically (Logan et al., 1999). (CI) Lytechinus variegatus embryo, vegetal pole view, at 32-ceII stage; (C2) 60-cell stage. Nuclearized (3-catenin is visible only in veg2 cells, and in large and small micromeres. (D, E) Endomesoderm specification requires (3-catenin nuclearization. (D) Normal pluteus of L. variegatus (control embryo). (E) L. variegatus embryo of same age as (D), developing from egg injected with mRNA encoding the cytoplasmic domain of cadherin, which traps (3-catenin in the cytoplasm and prevents its nuclearization. The embryo consists of a hollow epithelial sphere and contains neither archenteron nor any endomesodermal cell types. [(C-E) From Logan et al. (1999) Development 126, 345-357 and The Company of Biologists, Ltd.] (F) Overexpression of GSK.3 also prevents endomesodermal specification. An embryo of Paracentrotus lividus developing from an egg injected with mRNA encoding GSK3 is shown, at the same age at which controls are fully formed pluteus larvae. Overexpression of GSK3 also blocks (3-catenin nuclearization. [(F) From Emily-Fenouil et al. (1998) Development 125, 2489-2498 and The Company of Biologists, Ltd.] (G) Process diagram for veg2 endomesoderm specification. The boxes demarcated with black lines represent genes encoding transcription factors; no downstream genes are shown. Color coding for the various domains of the embryo are as in Fig. 3.1: veg2 and vegi endoderm, blue; veg2 mesoderm, lavender; skeletogenic micromeres, red; the initial endomesodermal domain is shown in white. Evidence for the role of maternal |3-catenin nuclearization and for the early micromere signal are reviewed in text. The evidence for the Wnt8 loop is that the gene encoding this ligand is expressed throughout veg2 early on (Fig. 3.2F), and that introduction of negatively acting forms of Wnt8 blocks endomesoderm specification (Wikramanayake et al., 2001). The evidence for the late micromere signal and the role of signal mediated P-catenin nuclearization in the late blastular specification of vegi endoderm is cited in text (McClay et al., 2000; Sweet et al., 1999; Davidson et al., 1998). The late micromere signal is likely to be a Notch (N) ligand since mesoderm specification requires N signaling (Sherwood and NcClay, 1997, 1999). The evidence for the N amplification loop is that N internalization occurs in the cells that are to become vegj mesoderm after 7th cleavage; that if N signaling is blocked by introduction of a negatively acting form of this ligand then mesoderm but not endoderm specification is blocked; and conversely, that more mesoderm is


at the next (i.e., 5th) cleavage, enJoiCTexpression is severely reduced (Fig. 3-3A1-2). But by 6th cleavage it is already too late, and micromere removal now has little effect on subsequent endol6 activity (Fig. 3-3A3; compare control Fig. 3-3A4). Conversely, if micromeres are transplanted to the animal pole of a host embryo at the 8- or 16-cell stage, ectopic endol6expression is induced to occur at the site of transplantation (Ransick and Davidson, 1993). This operation in fact results in complete respecification to endomesodermal fate of the adjacent animal pole blastomeres, which normally produce only ectoderm. A second vegetal plate appears at the top of the embryo and this presently gives rise to a complete tripartite archenteron that fuses with the endogenous one at the esophagus (Ransick and Davidson, 1993). These results cap off a long history of experimental embryology (Horstadius, 1939; Khaner and Wilt, 1990) that had already indicated the conditional nature of veg2 specification. Furthermore, the micromere signals directly affect transcription of the SpKroxl gene (Wang et al, 1996), the zygotic endomesoderm regulator of Fig. 3-2A, B. This is illustrated in Fig. 3.3B1-3, in which we see that normal expression of SpKroxl requires direct contact with micromeres just as does normal expression of endol6(..endol6is activated several cleavages later than is SpKroxl and is not a direct target of SpKroxl-, these genes are components of interlocked but initially parallel pathways). The experiment of Fig. 3-3B shows that an early cleavage signaling event is directly required for initiation of a territorial regulatory state in the veg2 blastomeres. Nor is this the only intercellular interaction required: the expression in veg2 cells of the SpWnt8 gene of Fig. 3.2F is also essential, since if the embryo is loaded with mRNA encoding a form of Wnt8 which acts negatively (i.e., it occupies Wnt receptors but fails to cause signaling) then endomesoderm specification is halted (Wikramanayake etal., 2001).

But an autonomous component also contributes critically to endomesoderm specification. This is the nuclearization of maternal (3-catenin, a cofactor for the Tcf/Lefl family of transcriptional regulators. Maternal (3-catenin is present in all the cells of the early embryo, where it is associated with membrane-bound cadherin. In midcleavage it appears in nuclei of all blastomeres that will contribute to the initial endomesodermal and mesodermal territories (Logan et al., 1999). Thus at 5th cleavage, as illustrated in Fig. 3.3C1, an immunological stain reveals nuclearized 3-catenin in nuclei of macromeres, skeletogenic micromeres and small micromeres (cf. Fig. 3.1); after 6th cleavage, when vegi and veg2 blastomere tiers arise from the parental macromeres, it begins to fade from the vegi nuclei; and by 7th cleavage it is strictly localized to veg2 endomesoderm and formed with excess N signaling (Sherwood and McClay, 1997, 1999). The negative transcriptional interaction on endodermal genes within the mesodermal domain follows from the observation that LiCI and many other treatments enlarge the mesodermal domain at the expense of the more peripherally located endodermal domain (while the latter is also enlarged at the expense of the overlying vegi ectoderm; see text).

to the micromere mesodermal domains, as shown in Fig. 3-3C2. Nuclearization of 3-catenin is a perfect marker of the segregation of lineages into the endomeso-dermal territories (blue + violet + lavender in Fig. 3-1), for elsewhere in the embryo it remains associated with the cell membranes. The early (3-catenin nuclearization in these territories is transient, and disappears by the end of cleavage. Nuclear (3-catenin reappears later in the blastula stage in a ring of vegi cells immediately external to the vegetal plate, i.e., in just those veg\ cells that will shortly contribute to hindgut endoderm cf. fate map, Fig. 3-1A). The late nuclearization of 3-catenin thus occurs in these cells at the time they are specified as endoderm, just prior to their invagination.

The early embryonic nuclearization of 3-catenin is an autonomous process, according to two kinds of evidence. First, at 7th cleavage the same number of blastomeres display nuclear [3-catenin as in controls even if the embryos are continuously dissociated, so that from the 2-cell stage onward the blastomeres divide in isolation from one another and sister cells are immediately separated. Second, removal of micromeres (as in the experiments of Fig. 3-3A, B) has no effect on the extent of (3-catenin nuclearization in veg2 cells (Logan et al., 1999).

The nuclearization of (3-catenin is an essential process in endomesoderm specification. It can be prevented by overexpression of the cytoplasmic domain of cadherin, thereby trapping the (3-catenin in the cytoplasm throughout the embryo. The consequence is total failure of endomesoderm specification, that is, absence of expression of endomesodermal markers (e.g., endol6) or of endo-mesodermal cell types (Logan et al., 1999; Wikramanayake et al., 1998). These embryos develop as hollow ciliated spheres, much as do the progeny of isolated animal-half blastomeres. This is a dramatic effect, illustrated in Fig. 3.3E (Fig. 3.3D shows a control embryo of the same species and same age). Furthermore, overexpression of GSK3, which inhibits (3-catenin activation and nuclearization, produces just the same phenotype (Emily-Fenouil et al., 1998; Fig. 3.3F). The system can be pushed back and forth: treatment of early embryos with LiCl, the relevant function of which here is to inhibit GSK3 (Klein and Melton, 1996), has the effect of expanding the endomesodermal domain so that it includes all of vegy and sometimes more (Horstadius, 1973; Cameron and Davidson, 1997). LiCl teatment correspondingly expands the domain of nuclear |3-catenin (Logan et al., 1999). Finally, expression of negatively active forms of Tcf/Lefl also blocks endomesoderm specification, while overexpression of the normal form expands it (Vonica et al., 2000; Huang et al., 2000). The function of (3-catenin nuclearization in endomesoderm specification in fact depends entirely upon Tcf/Lefl.

As it affects endomesoderm specification the (3-catenin system may be considered an autonomous expression of maternal polarization along the A/V axis of the egg (Logan et al., 1999; Davidson et al., 1998; Angerer and Angerer, 2000), while the requirement for a 4th-6th cleavage micromere signal designates the exact lineages in which endomesodermal specification will occur, i.e., in the blastomeres that abut the micromeres. So these are dual inputs, of different informational consequence. Since only advanced echinoids such as our example

5. purpuratus make micromeres, the autonomous system is the basal one for echinoderms. Its remnants can still be seen in the residual endol6 and kroxl gene expression that occurs in the micromere-less embryos of Fig. 3-3A, B. The dual inputs that control veg2 specification in S. purpuratus must ultimately be integrated at the cis-regulatory level (cf. Chapter 2). This job will be performed by gene(s) that respond to both inputs. For example the expression of SpKroxl requires the micromere signal (Fig. 3-3B); and its domain of expression is expanded to include vegi by LiCl treatment, but is obliterated if (3-catenin nuclearization is prevented by excess cytoplasmic cadherin (A. Ransick, C. Livi, and E. Davidson, unpublished data). Whichever genes carry out this integrative function it can be concluded that the lineage-specific transcriptional output that defines veg2 and sets the endomesodermal fate of its progeny is controlled by a combination of spatial inputs. These are the maternal |3-catenin nuclearization system and the zygotic 4th-5th cleavage mesomere signal.

Figure 3.3G shows an outline of the inputs into the stepwise endodermal specification system of the sea urchin embryo (see legend for further details). The diagram illustrates how a combination of maternal and zygotic inputs position endomesoderm fate in veg2 (4th-6th cleavage). Subsequently the late micromere signal (McClay et al., 2000), plus an endomesodermal regulatory output, are responsible for veg2 mesoderm specification (7th-9th cleavage). The remaining veg2 cells become endoderm. Note that both the endomesodermal and the veg2 mesodermal domain seem to be equipped with signal-mediated state amplification devices; these constitute what could be termed a "community effect" (Gurdon et al., 1993). That is, signaling amongst cells within these domains is required in order to maintain and perhaps amplify the respective states of specification.

So we have the flavor of the mechanism that quickly transforms the initial organization of the sea urchin egg into a specified mosaic of blastomeres. These mechanisms turn on regulatory genes in the nuclei of the early blastomeres even as the cleavage process generates the lineage elements that will constitute each territory. It is an interesting historical note that for generations the early sea urchin embryo was known as the quintessential "regulative" system in that (almost) any part could be made to recreate any other part if put in its place. But, as has been clear for a long time now, "regulative" character simply means that cell fate is conditional, neither more nor less; that is, that interblastomere signaling is important for specification (Davidson, 1986). Both conditional and autonomous mechanisms are required for endomesoderm specification, and as we see in the following this is typical.


The embryos of solitary ascidians also solve the problem of embryogenesis by operating Type 1 specification mechanisms. Like sea urchin embryos they carry out direct cell type specification during cleavage, within polyclonal lineage elements that are the territories from which specific larval structures arise after gastrulation. The initial tier of specifications have largely taken place by 5th-7th cleavage, depending on the lineage (for general review of experimental evidence regarding lineage and specification in these embiyos see Satoh, 1994). Though they are both deuterostomes (cf. Fig. 1.6) ascidian embryos differ from those of sea urchins in that their cell lineage is more rigidly invariant further into embryonic development, but this is of no great importance for present considerations except that it provides an experimental leg up. A more significant difference is that ascidians are direct developers, in contrast to most sea urchin species, which develop indirectly. That is, the ascidian embryo gives rise directly to a small (and very simplified) version of an adult chordate body plan, rather than to a single cell-thick feeding larva, which bears no obvious morphological relation to the adult body plan (Peterson etal., 1997, 2000a, b). The postgastrular outcome of embryogenesis in ascidians is therefore anatomically more complex than in sea urchins. It is complex in another sense as well: though the species considered here are motile, their larvae do not feed, but soon settle down and destroy some of their chordate-like structures. They then proceed to generate, in part from undifferentiated larval set-aside cells, a taxon-specific morphology that equips them for life as sessile filter feeders. These are postembryonic features, however, and our interest here is rather with how the initial specification functions of the embryo are controlled.

Territorial Specification

A territorial map of the ascidian embryo is shown in Fig. 3-4, based on a century of cell lineage analyses (reviewed by Satoh, 1994; Davidson, 1986). The 6th cleavage-stage embryo is portrayed in Fig. 3-4A, and an embryo that has partially completed 7th cleavage in Fig. 3-4B; by this point all but a few cells have been finally allocated to one or another of the definitive territories, in a predictable and invariant sequence. The immediate daughters of most (though not all) of these few cells are alternately assigned to one or another territory after the very next cleavage. The structures of the chordate-like larva to which each territory gives rise are indicated by color in Fig. 3-4C. The population of cells denoted "mesenchyme" (MCH) includes the mesodermal set-aside cells that serve as precursors of some adult (i.e., sessile phase) tissues (Satoh, 1994; Hirano and Nishida, 1997), and similarly the trunk lateral cells (TLC) are the precursors of adult hemocytes (Kawaminani and Nishida, 1997).

FIGURE 3.4 Specification and early territorial expression of cell type-specific genes in ascidian embryos. (A-C) Color-coded territorial map based on detailed knowledge of the embryonic cell lineage (Conklin, 1905; Ortolani, 1955; Nishida, 1987). Embryos are viewed from the vegetal pole, anterior up. The dorsal side, not visible, consists of cells fated to become epidermis, except for the anteriormost cells which will contribute to the cerebral vesicle. Gray regions represent yet incompletely specified cells, the progeny of which will contribute more than one cell type; colored regions have been uniquely specified at the stages shown so that all progeny will contribute to one cell type or structure. Specification of the brain, and completion of the anterior/posterior regionalization of the epidermis, occur later (Miya et al., 1996; Wada et al., 1999; review by Satoh, 1994). Territories are: purple, dorsal glial cell column ("spinal cord", SC); orange, notochord (N); blue, trunk lateral cells (TLC; these are precursors of the adult blood cells [Hirano and Nishida, 1997]); yellow, endoderm (En) and endodermal strand (ES); dark green, epidermis (Ep); light green, mesenchyme

FIGURE 3.4 Specification and early territorial expression of cell type-specific genes in ascidian embryos. (A-C) Color-coded territorial map based on detailed knowledge of the embryonic cell lineage (Conklin, 1905; Ortolani, 1955; Nishida, 1987). Embryos are viewed from the vegetal pole, anterior up. The dorsal side, not visible, consists of cells fated to become epidermis, except for the anteriormost cells which will contribute to the cerebral vesicle. Gray regions represent yet incompletely specified cells, the progeny of which will contribute more than one cell type; colored regions have been uniquely specified at the stages shown so that all progeny will contribute to one cell type or structure. Specification of the brain, and completion of the anterior/posterior regionalization of the epidermis, occur later (Miya et al., 1996; Wada et al., 1999; review by Satoh, 1994). Territories are: purple, dorsal glial cell column ("spinal cord", SC); orange, notochord (N); blue, trunk lateral cells (TLC; these are precursors of the adult blood cells [Hirano and Nishida, 1997]); yellow, endoderm (En) and endodermal strand (ES); dark green, epidermis (Ep); light green, mesenchyme

The blastomeres which give rise to endodermal, epidermal, and tail muscle lineages are specified autonomously. Classical and modern evidence has led to this conclusion, which is mainly based on three kinds of result. First, the progeny of isolated blastomeres of these lineages are competent to generate appropriate structures and express cell type-specific markers in the absence of the remainder of the embryo. Second, cleavage arrest with cytochalasin at successive stages (the nuclei continue to divide but no new cell membranes are formed after the drug is added) shows essentially that so long as the nuclei are surrounded with given sectors of maternal cytoplasm they will express the cell type-specific marker genes that would normally be expressed by the blastomere lineages which inherit these same sectors (for review, Davidson 1986, 1990; Satoh, 1994). Third, and most incontrovertible, is the demonstration that fusion of given, localized sectors of egg cytoplasm with certain early blastomeres or nucleated egg fragments is sufficient to endow these with the capacity, in culture, to express muscle, endoderm, or epithelial markers, which they will otherwise not do (Nishida, 1992, 1993, 1994; Marikawa et al., 1994; reviewed by Nishida, 1997).

The remainder of the specification states indicated in the maps of Fig. 3.4A, B are conditional, depending on signaling interactions between specific blastomeres. These interactions take place mainly at 32-64 cell stage, though of course additional interactions follow on at later stages, e.g., within the nervous system. Among the prominent conditional specification events in which 5th-6th cleavage signaling interactions have been identified are the specification of notochord

(Mch); red, tail muscle (Mu); dark blue, brain (B). (A) 64-cell stage; (B) 110-cell stage; (C) Locations of the morphological structures to which these cell types give rise, seen in a midsaggital, and a more lateral saggital view of the completed larva. [(A-C) From Satoh et al. (1996b) Dev. Growth Differ. 38, 325-340.] (D, E) Expression of AsT (brachyury) gene in notochord lineage founder cells. At 64-cell stage this gene is expressed in the A7.3 and A7.7 pairs of blastomeres (D), and at 76-cell stage (E) in the eight progeny of the A7.3 and A7.7 blastomere pairs plus the B8.6 pair, arrows (i.e., the notochord lineages shown in orange in B). [(D, E) From Yasuo and Satoh (1994) Dev. Growth Differ. 36, 9-18.] (F) Expression of AMD (myoD) gene at 76-cell stage, in nuclei of four muscle lineage blastomeres on each side of the midline. [(F) From Satoh et al. (1996a) Proc. Natl. Acad. Sci. USA 93, 9315-9321; copyright National Academy of Sciences, USA; for further details and lineage, see Fig. 3.5A and text.] (G, H) Expression of HrMA4 muscle actin gene. (G) 64-cell stage; (H) 108-cell stage. The gene is expressed in muscle precursor blastomeres (red cells in (A) and (B), respectively. [(G, H) From Satou e£ al. (1995) Dev. Growth Differ. 37, 319-327.] (B-H) are embryos of Halocynthia roretzi. (I) Expression of Cititfl, an endoderm-specific transcription factor, in all endoderm lineage progenitors in Ciona intestinalis at 76-cell stage (i.e., the A7.1, A7.2, A7.5, B7.I, and B7.2 blastomere pairs). [(I) From Ristoratore et al. (1999) Development 126, 5149-5159 and The Company of Biologists Ltd.]

founder cells (Nakatani and Nishida, 1994, 1999); of mesenchyme lineage founder cells (Kim and Nishida, 1999); of trunk lateral cells (Kawaminani and Nishida, 1999); and of anterior and posterior epidermal fates (Wada et al., 1999). Conditional specification of cell fates within the larval nervous system has long been known to occur conditionally (Rose, 1939; Reverberi and Minganti, 1947; Nishida, 1991; Satoh, 1994; Inazawa et al., 1998). The territorial lineages of the ascidian embryo display the essential feature of Type 1 embryonic process: they proceed immediately toward cell differentiation once the founder cells have segregated from other lineage elements and become specified, whether this occurs by autonomous or conditional means.

As in the sea urchin the character of the embryonic mechanisms at work in the ascidian embryo is shown by the early expression of territory-specific genes. This is illustrated in Fig. 3-4D-I. The territories are coincident with given lineage elements, since the relevant portions of the cleavage process are invariant. This makes it possible to define the exact cleavage and the specific cells in which these genes are activated. In Fig. 3-4D we see that a gene encoding a Brachyury transcription factor (AsT) is activated, exclusively in progenitors of the notochord territory (Yasuo and Satoh, 1994). Brachyury is a key regulator of notochord development (Chapter 5). Transcription of this gene begins at the 64-cell stage, when two lineage founder cells for notochord have been specified on either side of the plane of symmetry. After the following cleavage this regulatory gene is expressed in the eight progeny of those cells and in two additional blastomeres which have now also been specified as notochord (Fig. 3-4E). The location of these cells at the respective stages can be seen in the context of the overall territorial maps in Fig. 3-4A and B, respectively. The expression of the brachyury gene is a direct output of the notochord specification mechanism as discussed below. Figure 3-4F-H illustrates expression of genes specific to the territory that gives rise to the tail muscles: at the 76-cell stage (Fig. 3-4F) a gene encoding a MyoD class transcription factor (Satoh et al., 1996a) is expressed in four muscle precursors on each side (muscle specification is discussed in detail below); and Fig. 3.4G, H shows expression of one of the actin genes (HrMA4) in muscle progenitor blastomeres at 64- and 108-cell stages (Satou et al., 1995). At the risk of unnecessary repetition, this is another example of the activation of a gene encoding a differentiation protein in cleavage, in which transcription begins long in advance of the morphogenetic events that formulate the tail muscle structures. Finally, in Fig. 3-41 is illustrated territorial expression during cleavage of a gene encoding a specific homeodomain transcription factor (Cititfl), here seen in five pairs of endoderm precursor cells (Ristoratore et al., 1999). This gene encodes an endoderm-specific factor which continues to be expressed in all endoderm cells into gastrulation, and in postgastrular and larval stages it is utilized in the head endoderm. The colors in the diagrams of Fig. 3-4A and B can be taken to symbolize the institution of territory-specific domains of differential gene expression, all of which have been installed by midcleavage.

Mechanisms and Pathways in Mesoderm Specification

Just how direct can be the "direct cell type specification" of this discussion is evident in the process by which the muscle territory of ascidian embryos is specified. To appreciate this story a glance at the lineage map of the progeny of the 3rd cleavage B4.1 blastomere pair is necessary (Fig. 3-5A; a view of the 3rd cleavage embryo is shown in Fig. 3-5F5). From this pair of 3rd cleavage cells (one on each side) derive all the primary tail muscles and some mesenchyme cells of the larva (red and light green territories in Fig. 3.4A, B). Progeny of the B4.1 blastomeres produce exactly 14 primary tail muscle cells on each side of the larva, in both Ciona intestinalis and Halocynthia roretzi. A few additional "secondary" tail muscle cells positioned at the caudal end of the larva arise from different areas (i.e., ten from the B4.2 pair of blastomeres and four from the A4.1 cell pair in Halocynthia; in C. intestinalis these same two blastomere pairs instead produce a total of eight secondary muscle cells; Nishida, 1987; Meedel et al., 1987). While the secondary muscle cells are conditionally specified, the primary muscle lineages stemming from B4.1 are autonomously specified. The experimental evidence for their autonomous specification includes all of the sorts of demonstrations listed above (for reviews see Satoh, 1994; Nishida, 1997; Davidson 1990). Molecules that activate muscle specific genes are localized in the portion of the cytoplasm inherited by the B4.1 blastomeres. The key observations (in H. roretzi) come from fusing vesicles that contain sectors of maternal cytoplasm to blastomeres that are normally fated to produce only epidermis. This was done by use of an electric field in the presence of polyethylene glycol. Fusion of cytoplasm from B4.1 causes the progeny of epidermal blastomeres to undergo muscle differentiation and express muscle markers, and the same is true if the cytoplasm is obtained from those sectors of unfertilized and fertilized eggs that are later incorporated in the B4.1 blastomeres (Nishida, 1992). Similarly, in Ciona savigny, certain centrifugal fractions of egg cytoplasm confer ability to activate muscle genes if fused with nucleated egg fragments that otherwise produce only epidermis when cultured (Marikawa et al., 1994).

The diagram in Fig. 3-5A shows the origin of the seven tail muscle progenators that arise from the B4.1 lineage on each side, and traces the segregation of cell fates that occurs between 4th and 6th cleavages (the color codes are the same as in the territorial diagrams of Fig. 3.4A-C). Note that in the top half of the lineage there emerges at 5th cleavage an endoderm founder cell, B6.1 (that is, a cell all of whose progeny will become endoderm). A muscle founder cell, B7.4, is segregated at 6th cleavage. The sister cell, B7.3, produces both a mesenchyme and a notochord founder blastomere at the following cleavage. In the bottom half the final segregation of cell fates all occur at 6th cleavage: B7.5 and B7.8 give rise only to muscle; B7.7 to mesenchyme and B7.6 to endodermal strand. Note that two sublineages which have only mesodermal fates separate out at 5th cleavage. These are the B6.2 lineage, which produces muscle, mesenchyme, and notochord cell types; and the B6.4 lineage, which produces mesenchyme and muscle.



Was this article helpful?

0 0

Post a comment