An axis could be specified by:
1) intrinsic organization; (animal-vegetal axis in the frog);
2) environmental cues (Drosophila oocyte and surrounding
nurse cells; mammalian embryo and the uterine wall after
implantation);
3) self-organizing stochastic symmetry-breaking process
(regulative development of the chick blastoderm in vitro).
The initial symmetry is broken in a many different ways in different embryos. Also, morphogenetic movements of gastrulation in vertebrates are quite variable. Nevertheless, the basic body plan at the phylotypic stage is very similar, perhaps due to conservation of signaling processes at the molecular level.
In amphibians, the animal vegetal axis is pre-determined in the oocyte (germinal vesicle-mitochondrial cloud), whereas the dorsoventral axis is specified de novo and depends on the position of sperm entry. Sperm aster initiates asymmetric growth of microtubules, which may be used as railroad tracks for cortico-cytoplasmic rotation (Ancel and Vintemberger, 1948). Cortical rotation appears to be necessary for DV axis specification, because the agents that disrupt it, including UV light, cold, and nocadazole also cause ventralization. Such embryos could be rescued by 15-min tilting (Schultze rotation, 1898; Scharf and Gerhart, 1980; Youn and Malacinski, 1981). These findings were interpreted to mean that the gravity effectively accomplishes what is normally accomplished by the microtubule-mediated movement. There is evidence, that in the teleost fish, such as zebrafish and medaka, microtubules are also involved in dorsoventral axis specification.
The situation is different in Chondrostei (jawed fish), in which a micropyle is located exactly at the animal pole and determines the position of SEP. In sturgeon (Acipenser), the long animal-vegetal axis of the egg is parallel to the substrate. After fertilization the zygote is released from the envelope, becomes round and rotates by 90o. Simultaneously, the cortical cytoplasm is rearranged to form the ‘clear crescent’ corresponding to the future plane of bilateral symmetry. Thus, it is the gravity, rather than SEP, is responsible for symmetry breaking in this case.
Determination of axis in the chick also involves gravity (Kochav and Eyal-Giladi, 1971) . The first axis (animal/vegetal = dorsal/ventral) is likely to be pre-determined in the bird oocyte (similar to amphibians) by the nucleus-centrosome axis, developing as asymmetric deposition of yolk and the formation of the cytoplasmic island around the oocyte nucleus. As a result, cleavage is meroblastic, with cell divisions restricted to the animal pole surface. The second, anteroposterior, axis is biased by gravity. Rotation of the egg in the oviduct (~0.2 rpm) with the pointed end towards the cloaca results in the blastoderm being tilted in the direction of rotation. The posterior marginal zone develops at the uppermost side of the blastoderm and initiates the primitive streak. When the direction of rotation was changed or when the eggs (w/o the shell) were positioned vertically, the posterior end always developed according to the uppermost part of the blastoderm.
In mammals, there is virtually no yolk and eggs do not have a visible animal vegetal axis, although it may exist (Gardner, 1997). Inside/outside differences generate the inner cell mass and the outer trophoblast. ICM becomes eccentrically located to generate the dorsoventral or embryonic/abembryonic axis of the cylindrical blastocyst (64-128 cells). Thus, an axis may be determined in mammals spontaneously, in the absence of environmental cues.
Recent findings established that the animal-vegetal axis (defined by the second polar body extrusion) is always orthogonal to the future dorsoventral axis (dorsal ICM cells face trophoblast and ventral cells face the blastocyst cavity), but not fixed (Gardner, 1997). The AP axis, which is first defined by the appearance of the primitive streak at the future posterior end, may be generated by a special mechanism (as discussed below).
Axis specification may involve inductive interactions and cytoplasmic determinants.
Mosaic development in Xenopus. When Spemann (1901-1902) constricted gastrulating embryos with a hairloop, ventral halves gave rise to ventralized embryos, whereas dorsal halves were somewhat dorsalized. Blastomere removal and isolation experiments (Kageura and Yamana, 1983-84) have shown that two animal, one ventral and one dorsal blastomere are required for normal axial development. One dorsal blastomere of four-cell embryos can form a normal embryo, but the ventral blastomere cannot. These studies suggest the presence of localized determinants.
Regulative development in the chick (Spratt and Haas, 1960). If the blastodisc (60,000 cells) is cut into 4 parts, they will form 4 primitive streaks. If a blastodisc has twice the amount of the posterior marginal zone (PMZ) in four parts, it still forms one axis, not two or four.
A candidate inducer (tissue or factor) must fulfil the following criteria:
1) It has to be present in the right place and at the right time to
do the job.
2) It should be capable of triggering this program de novo(e.g.,
ectopically).
3) Its removal or inactivation should prevent the execution of this
program.
Cytoplasm transfer experiments (Yuge et al., 1990; Fujitsue et al., 1993; Holowacz and Elinson, 1993). Vegetal cytoplasm (VC) triggers secondary axis formation upon ventral microinjection. This suggests that VC contains determinants that are sufficient for dorsal development. Interestingly, this activity is found vegetally before cortico-cytoplasmic rotation, and move to the dorsal margin after cortical rotation. The removal of cytoplasmic determinants by constriction of the vegetal region (Sakai, 1996) resulted in ventralized development. This demonstrates that ‘dorsal determinants’ are necessary for dorsal development. Grey crescent transplantation experiments, and their re-interpretations: (Curtis, 1960; Gerhart et al., 1980; Kageura, 1997). Role of the cytoskeleton in asymmetry generation and axis determination.
The molecular nature of dorsal determinants is unknown. Activation of a gene called Siamois (Lemaire et al., 1995) is a reliable indicator of dorsal determinant activity, because it is also activated vegetally in the absence of rotation. Since Siamois is a target for b -catenin-dependent pathway ( the Wnt pathway), it appears that this pathway is involved.
Evidence: activation of the pathway leads to ectopic axis and enhanced dorsoanterior development, blocking the pathway causes ventralization. For example, dorsalization by LiCl (Kao and Elinson, 1986) is likely to result from the inhibition of GSK3, a negative regulator of the Wnt pathway.
In the mouse, isolated inner cell mass blastomeres are totipotent (ES cells) . This suggests that there are no cytoplasmic determinants in the mouse embryo. Yet the signaling pathways are conserved. For example, Axin, a regulator of Wnt signaling, has a role in axis specification in the mouse. It is not clear whether the same or different mechanisms lead to the activation of the Wnt pathway in different vertebrates.
Advantages of the amphibian system: big size of the embryo, large numbers, defined medium, fast development, volume unchanged, easy to access early stages of development.
A. State of determination and developmental potential.
A major aim of embryology is to know the state of determination of every embryonic cell or group of cells at any given time during development.
Cytoplasmic determinants vs inductive interactions.
State of determination and developmental potential is assessed experimentally by placing cells/tissues in different environments and following their fates by histological analysis and molecular marker analysis (especially at the early stages, when cells are not that different from each other). Positional markers.
Cell transplantation into blastocoel (Heasman et al., 1986). Dissociation-reaggregation of cells (Gurdon's community effect). Single cell culture (Godsave and Slack, 1987). Cell fate determination is a gradual process. Many genes may be first activated, then, maintained via cell-cell interactions (Sokol, 1994).
Evidence for inductive interactions comes from the comparison of a fate map and a specification map.
Critical tests for inductive interactions were developed by J. Holtfreter:
- Recombination experiment ("sufficiency test"). This experiment may show whether one tissue is able to induce or change fate of another tissue (Holtfreter sandwiches, 1930s).
- Isolation experiments assess whether a cell/tissue requires signals from the neighboring tissues to develop towards a specific direction.
B. Germ layer specification. Mesoderm induction.
Germ layers are specified separately from body axes. Dalq and Pasteels (1937) speculated that there are two gradients of activity, each specifying two orthogonal axes. This model is supported by the existence of two independent molecular pathways. b-catenin pathway is responsible for dorso-ventral differences. The other pathway (involves VgT) is critical for generation of three germ layers along the animal-vegetal axis. Partial removal of VegT caused the respecification of endoderm into mesoderm (Zhang et al., 1998).
Ogi (1967) and Nieuwkoop (1969) proposed different explanations for mesoderm formation in recombinants of prospective endoderm and ectoderm. Induction or restoration of the gradient? Proof of induction came from the lineage tracing and transfilter experiments.
Models for mesoderm formation:
Regional induction was revealed in recombinant sandwiches of animal pole ectoderm with dorsovegetal and ventrovegetal cells (Boterenbrood and Nieuwkoop, 1973; Dale and Slack, 1987). These results were interpreted as evidence for existence of distinct dorsal and ventral inducers. Regional response was observed for dorsal and ventral animal cap cells stimulated in a homogeneous solution of the mesoderm inducing factor activin (Sokol and Melton, 1991). It was interpreted as pre-pattern in the responding tissue.
Homologous germ layers in the chick, zebrafish and mouse. Reaggregated hypoblast induces mesoderm in the epiblast in the chick(Mitrani and Eyal-Giladi, 1981). Induction of mesoderm in zebrafish: a role for the yolk syncitial layer (YSL, Mizuno et al., 1996).
C. Role of different germ layers in axis specification.
Nieuwkoop’s experiments on rotation of vegetal hemisphere (1969) led to the conclusion that the D/V axis is determined by the vegetal hemisphere.
Vegetal + Animal hemisphere = Axis determined by vegetal cells
Waddington’s experiments on hypoblast rotation in the chick (1933) also revealed the primary role for hypoblast in axis determination. However, when the chick hypoblast cells are dissociated and reaggregated, the axis forms according to the initial epiblast position, even though such hypoblast is still capable of induction (Mitrani and Eyal-Giladi, 1981).
Epiblast + Hypoblast = Axis is determined by hypoblast
Thus, the processes of mesoderm induction and axis specification can be experimentally separated. Hypoblast may be equivalent to vegetal endoderm in the frog, whereas epiblast has its own polarity and an ability to direct axis formation.
Similarly, polarity can be observed in Xenopus animal pole ectoderm as
differential response to activin (a mesoderm inducing factor, Sokol and Melton,
1991). Thus, regional induction may be a result of mesoderm induction in
differentially responsive tissues.
The organizer experiment (Spemann and Mangold, 1924). The organizer is a special inducing center around the dorsal blastopore lip, which was discovered during transplantation and lineage tracing experiments. Spemann organizer equivalents: the node in mice, shield in zebrafish, Hensen's node in chicks. Does the organizer form as a result of action of a localized determinant(s) or by inductive interactions?
Nieuwkoop center was defined as an early dorsovegetal signaling center
Is Nieuwkoop Center required for axial development? Axial development occurs even when the vegetal tier cells (tier D containing NC) are removed at the 32-cell stage (Kageura, 1995). Thus vegetal blastomeres are sufficient but not necessary to induce the organizer. |
Dorsovegetal blastomeres at the 32-64-cell stage have
the capacity to rescue ventralized embryos by inducing the organizer in the marginal zone
("sufficiency", Gimlich and Gerhart, 1984). Axis-inducing activity
is maximal in dorsovegetal blastomeres at morula stages, whereas is it maximal
in dorsal marginal cells at late blastula stages. This is the evidence for the
existence of the Nieuwkoop Center that is spatially and temporally
distinct from the Spemann organizer.