More than a thousand different species of frogs have been characterized. The genus Xenopus ("strange foot" in Greek) arose more than 120 million years ago and includes 17 species. These aquatic frogs live in ponds in southern (sub-Saharan) Africa. One of these species, Xenopus laevis, has become, together with the mouse and chick, an attractive vertebrate model system for embryologists and developmental biologists. In fact, most of what is known of vertebrate development comes from the study of the amphibian embryo. Already popular in Europe, X. laevis was first introduced in the United States in the early 1950s as a way to test for pregnancy in humans. When injected subcutaneously into female Xenopus, the gonadotropin hormone present in pregnant women’s urine would induce her to lay eggs the next day. Unlike most other frogs, egg laying in Xenopus is not seasonal; these tests could thus be performed all year round.
For embryologists, a major advantage of this frog is the number and size of its eggs. A single female can lay up to several thousand eggs a day. This is particularly important for modern molecular techniques: the source of biological material such as complementary DNA (cDNA), RNA, or protein is virtually unlimited. In addition, a X. laevis egg is approximately 1 mm in diameter, which makes it one of the largest cells in the world, visible to the naked eye, and amenable to microsurgery and microinjection of molecular factors. Moreover, fertilization of the egg is external, so the amphibian embryo can be studied from the beginning, when the embryo is a single cell. Finally, perhaps one of the most important attributes of this system is the amount of knowledge accumulated over more than a century on the development of the amphibian embryo. There is an amazingly rich body of literature, from descriptive to experimental approaches, on the development of many different types of amphibian embryos, including Xenopus. Associated with this literature is a remarkable cast of characters, including Mangold, Spemann, Roux, His, Holtfreter, Hamburger, and Nieuwkoop. Their work and their vision established the foundations of vertebrate experimental embryology, on which modern molecular embryological approaches are built. Although powerful genetic approaches in Caenorhabditis elegans and Drosophila, and more recently zebrafish and mouse, have contributed substantially to our knowledge of early embryonic development, experimental embryology in the amphibian had already established many key features of embryonic development. For example, the concept of embryonic induction, so widely accepted today as a means to establish embryonic cell fate, was first defined in amphibians (1). This article presents (1) a descriptive overview of Xenopus laevis development, including the concepts of fate maps and gene maps; ( 2) a review of molecular approaches currently used in the Xenopus system in oocytes and in embryos; (3) the use of explants and experimentally perturbed embryos in molecular studies of early development; and ( 4) genomics, nuclear transplantation, transgenesis, and maternal knockout strategies.
1. The Xenopus Embryo
1.1. Early Embryonic Development
Large numbers of Xenopus embryos can be obtained year-round, either by natural mating or by in vitro fertilization. For "natural" mating, both the male and female are injected with human chorionic gonadotropin and left in a quiet dark tank overnight to copulate. A large number of embryos at different stages of development can be harvested the following day. For in vitro fertilization, females are injected subcutaneously with the chorionic gonadotropin and left overnight. The next day, a male frog is sacrificed, and fragments of surgically removed testis are used to fertilize eggs gently squeezed from the female. Each squeeze provides a few hundred eggs. Females can be squeezed every hour for several hours. The advantage of in vitro fertilization is that all the eggs for a given squeeze develop synchronously; thus, a large number of embryos of a given stage are available. When laid, the eggs have a single axis of cylindrical symmetry running from the pigmented pole at the top, also called the animal pole, to the yolky bottom or vegetal pole. This axis of symmetry is imposed maternally during the development of the oocyte (2).
The sperm can enter at any point around the circumference. The first cell cycle is unusually long, about 90 minutes at room temperature. About 30 min after sperm entry, a drastic cytoplasmic movement, referred to as cortical rotation, takes place: the inside of the egg rotates about 30° relative to the outside. The consequence of cortical rotation is that the original axis of cylindrical symmetry is broken, and one side of the egg is different from the other (2). With 85% accuracy, the side opposite the sperm entry point will become the dorsal side. A series of quick cell divisions follow, without change in the volume of the embryo; thus, the cells become smaller in size following each division (Fig. 1a). During the first 4 hs of development, all cells of the embryo divide synchronously. The length of each cell cycle, after the first, is about 20-30 min at room temperature; the early blastomeres of the Xenopus embryos have barely enough time to replicate their genome before cytokinesis occurs and the next cleavage furrow becomes apparent. The development of the Xenopus embryo for the first few hours is entirely under the control of maternal determinants deposited in the egg.
Figure 1. Early embryonic stages of Xenopus laevis development. ( a) The first 6 h of Xenopus laevis development. The numbers on top of each panel, or set of panels, represents the developmental stage (45). Each panel is indicated by a number for reference in the text. The first row shows animal pole views; the middle row, lateral views; and the third row, vegetal pole views. The embryo undergoes 13 synchronous cell divisions without an increase in size. (b) Gastrulation. Panel 18 is an animal pole view; 19, a lateral view; and panels 20-23, vegetal posterior views. (c) Neurulation and the formation of a tadpole. Panels 24 and 25 are dorsal views of neural plate and neural groove stage embryos, respectively; panels 26-28 are anterior, head-on views showing the progressive closure of the anterior neural tube; panel 29 is a dorsal view of the neurula with the entire neural tube closed; and panels 30, 31, and 32 are lateral views of late neurula, tailbud, and tadpole, respectively. 
Dramatic changes occur after 13 synchronous cell divisions, when the embryo reaches about 4000 cells (Fig. 1b): (1) the synchronicity of cell division is lost, and embryonic cells divide at different rhythms in different regions of the embryo; (2) zygotic transcription begins (3); and (3) finally, cell and tissue movements begin. Among the first of these movements is the spreading of the animal pole region toward the marginal zone, called epiboly. Also, the cells of the marginal zone begin to move toward the dorsal side by a process called convergence and extension, a process that will continue during gastrulation (4). At this stage of development, the three embryonic germ layers—ectoderm, mesoderm, and endoderm—have been specified. These early cell movements set the stage for gastrulation, which begins about 9 h after fertilization.
It is during gastrulation that both the antero-posterior and right-left axes are determined and that cells of the prospective mesoderm and endoderm move to the interior of the embryo. Cells in a subequatorial region on the dorsal side involute first (Fig. 1, panels 19 and 20). This invagination is recognizable morphologically by the formation of a small arc of pigment in the dorsal side, called the blastopore lip. The emergence of the blastopore lip defines the dorsal midline. A group of cells located within a 30° angle at each side of the dorsal midline have the capacity to direct development of the entire dorsal axis, including axial mesodermal derivatives, such as somite and notochord, and the entire central nervous system (5). These group of cells are called the " organizer" (Fig. 2) because classic experiments demonstrated that removing this group of cells from a donor early gastrula embryo and implanting it in the ventral axes (6). One axis is generated by the host organizer; the other one forms under the influence of the ectopic organizer implant.
Figure 2. Gastrula fate map. Three views of the superficial gastrula, with the fate map adapted from Keller (1975) (D—is a dorsal view, L—a lateral view, V—a ventral view). The ectoderm is derived from the top of the embryo (animal pole, shades of blue). Mesoderm is derived from the equatorial region (marginal zone, shades of red), and endoderm from the bottom (vegetal pole, yellow).
Involution gradually spreads from the dorsal side to the mediolateral and ventral region of the embryo; this is manifested by the expansion of the blastopore lip arc (Fig. 1, panels 20-23). The lip finally becomes a closed circle when cells located in the entire circumference of the embryo have invaginated. The first cells that involute on the dorsal side are the prechordal plate (or head) mesoderm; these cells will contribute to mesodermal components of the craniofacial structures and are therefore anterior in nature. Cells that involute later, on the dorsal side, give rise to somites and notochord (4). As gastrulation proceeds, the blastopore ring becomes smaller and smaller until it closes to engulf the entire vegetal pole (endoderm). The closure of the blastopore demarcates the end of gastrulation and the beginning of neurulation. The blastopore will ultimately become the anus of the tadpole. Thus, at the end of gastrulation the embryo has successfully brought the entire endoderm inside, covered its surface with ectoderm, and placed mesoderm in between.
At the end of gastrulation, the cells of the dorsal ectoderm thicken to form the neural plate (which is two cell layers thick). This is the first morphological sign of the nervous system (Fig. 1c). The neural plate is flanked laterally by the neural crest and anteriorly by the sensory placodes and an amphibian-specific organ, the cement gland. Within the neural plate itself, mature neurons emerge in three stripes on each side of the dorsal midline (7). The ones closest to the midline will become motor neurons, those in the intermediate strip will contribute to the interneurons and the most lateral stripe will give rise to sensory neurons. During neurulation, morphogenetic movements elevate the lateral edges of the neural plate to give rise to a neural groove (8). The neural groove will ultimately close to generate the neural tube, with a brain at the anterior and the spinal chord at the posterior end. In Xenopus laevis, neural crest migration begins at the neural groove stages and continues following neural tube closure. The eye buds and optic vesicles, as well as the cement gland, become morphologically distinct. Neurulation ends with the closure of the neural tube and the appearance of a tailbud at the posterior end of the embryo. The cells of the heart primordia that flank the organizer on both sides meet in the ventral midline and fuse to form a tube that will later fold to form a functional heart. The progenitors of other organs have begun their differentiation as well. These include the embryonic kidney (pronephros) from lateral mesoderm, as well as the blood islands in the ventral mesoderm that will give rise to cells of the entire hematopoeitic pathway.
By the tailbud and tadpole stages, the embryo has a functional nervous system (Fig. 1, panels 31 and 32). Most neuronal axons have reached their targets, and the embryo is motile. A few hours later in the tadpole, the folded heart pumps blood derived from the blood islands into the newly formed arteries and veins. The development of internal organs, such as pancreas, gut, and intestine, is complete. The maternal yolk, which sustained embryonic life up to this point, is exhausted, and the tadpole begins feeding. Concomitant with this is the emergence of specific behavior.
The tadpole continue to grow in size until metamorphosis occurs. During this dramatic stage of the frog’s life, major remodeling of almost the entire body changes the morphology and the physiology of the tadpole, to generate a frog. There has been tremendous progress in the understanding of the molecular basis of metamorphosis in amphibians, and excellent reviews are available on this topic (9).
1.2. Fate Maps and Developmental Commitment
Fate maps, a fundamental concept in embryology, tell us what regions of the early embryo will contribute to later structures. Of course, the fate map does not tell us about how those structures are formed or their state of commitment, only about their spatial origin. In Xenopus laevis, several detailed fate maps of different stages of embryogenesis, as early as the 16- and 32-cell stage, are available (10). The pioneering work of Keller has provided a very high resolution fate map of gastrula and neurula embryos (11). In addition to fate maps of the whole embryo, there are fate maps of the neural plate and sensory placodes (12). At the gastrula stage, the fate maps described in Xenopus laevis are very similar to those described in other amphibians, including X. tropicalis, X. borealis, Rana pipiens, and Triturus (newt). Fate maps in the amphibian are usually constructed by marking cells with vital dye or by injecting them with other lineage tracers, followed by a statistical assessment of where the labeled cells end up later in the tadpole. Knowledge of the amphibian fate maps has been key in the understanding the molecular processes involved in early development (see text below).
To determine the state of commitment of a region of the embryo, approaches such as specification and determination assays can be used (13). In a specification assay, different regions of the early embryo are removed by microsurgery and cultured in a simple saline solution. Because a source of food is contained within each cell, these cultures do not need extrinsic factors to develop. By comparing the behavior of the explants cultured in isolation and the behavior of this region in the context of the embryo, it is possible to assess the state of commitment of that tissue. If the explant in culture gives rise to the same structure as in the embryo, then the explant is specified. Additionally, if the explants are derived at different times, it is possible to assess the state of commitment of a tissue at a given time toward a given fate. In determination assays, a cell or group of cells are transferred from a donor embryo into a different environment of a host embryo of the same or different developmental stage, and their development is compared to when they are cultured alone (as in specification assays). A region of an embryo is determined if it can give rise to the same structures regardless of its environment (in culture alone or placed in different region of the embryo). Slack (1984) suggested that analysis of the state of commitment can be expanded to include not only tissue and single cells, but nuclei also. Molecular analyses of these commitment states are used in molecular embryology to assess the molecular pathways involved in cell commitment and cell fate determination (see section on use of Xenopus embryos and embryonic explants in molecular biology, below).
1.3. Gene Maps
While early embryonic cells can appear morphologically identical, they do not necessarily express the same repertoire of genes (Fig. 3). This observation has allowed the unveiling of what Kirschner and Gerhart call the "second anatomy" of the embryo, based on the region-specific distribution of gene products (14). Sometimes the gene map correlates very well with fate maps, and sometimes it does not. When a given gene demarcates the same territory as the fate map for a given time in development, the gene can be considered to be a molecular marker for those cell types (Fig. 3). In molecular embryology, the expression of these genes is used as a testimony for the formation of the cell type in which they are expressed. It is very important to remember that, because some genes have an extremely dynamic pattern of expression, a molecular marker for a given cell type will only attest to the presence or absence of that cell at the appropriate time of expression. For example, the gene brachyury (Xbra in Ref. 15) demarcates the entire marginal zone of the early gastrula in Xenopus; fate mapping studies have established that the entire mesoderm will derive from this region. A few hours later, at the end of gastrulation and the beginning of neurulation, Xbra is expressed in the notochord and as a ring around the blastopore. Thus, Xbra is a panmesodermal marker when its expression is measured at gastrula stages; however, at neurula stages Xbra is a marker of notochord and blastoporal ring. Other markers are less dynamic in their expression; for example, the neural cell adhesion molecule (NCAM) (16) is expressed in the neural plate from the onset of its formation and remains entirely neural-specific until at least very late tadpole stages.
