Rotational cleavage of the C. elegans egg

The C. elegans zygote exhibits rotational holoblastic cleavage (Figure 8.42C). During early cleavage, each asymmetrical division produces one founder cell (denoted AB, MS, E, C, and D), which produces differentiated descendants, and one stem cell (the P1-P4 lineage). In the first cell division, the cleavage furrow is located asymmetrically along the anterior-posterior axis of the egg, closer to what will be the posterior pole. It forms a founder cell (AB) and a stem cell (P1). During the second division, the anterior founder cell (AB) divides equatorially (longitudinally; 90° to the anterior-posterior axis), while the P1 cell divides meridionally (transversely) to produce another founder cell (EMS) and a posterior stem cell (P2). The stem cell lineage always undergoes meridional division to produce (1) an anterior founder cell and (2) a posterior cell that will continue the stem cell lineage.

The descendants of each founder cell divide at specific times in ways that are nearly identical from individual to individual. In this way, the exactly 558 cells of the newly hatched larva are generated. The descendants of the founder cells can be observed through the transparent cuticle and are named according to their positions relative to their sister cells. For instance, ABal is the “left-hand” daughter cell of the Aba cell, and ABa is the “anterior” daughter cell of the AB cell.

Anterior-posterior axis formation

The elongated axis of the C. elegans egg defines the future anterior-posterior axis of the nematode's body. The decision as to which end will become the anterior and which the posterior seems to reside with the position of sperm pronucleus. When it enters the oocyte cytoplasm, the centriole associated with the sperm pronucleus initiates cytoplasmic movements that push the male pronucleus to the nearest end of the oblong oocyte. This end becomes the posterior pole (Goldstein and Hird 1996).

A second anterior-posterior asymmetry seen shortly after fertilization is the migration of the P-granules. P-granules are ribonucleoprotein complexes that probably function in specifying the germ cells. Using fluorescent antibodies to a component of the P-granules, Strome and Wood (1983) discovered that shortly after fertilization, the randomly scattered P-granules move toward the posterior end of the zygote, so that they enter only the blastomere (P1) formed from the posterior cytoplasm (Figure 8.43). The P-granules of the P1 cell remain in the posterior of the P1 cell and are thereby passed to the P2 cell when P1 divides. During the division of P2 and P3, however, the P-granules become associated with the nucleus that enters the P3 cytoplasm. Eventually, the P-granules will reside in the P4 cell, whose progeny become the sperm and eggs of the adult. The localization of the P-granules requires microfilaments, but can occur in the absence of microtubules. Treating the zygote with cytochalasin D (a microfilament inhibitor) prevents the segregation of these granules to the posterior of the cell, whereas demecolcine (a colchicine-like microtubule inhibitor) fails to stop this movement (Strome and Wood 1983). The partitioning of the P-granules and the orientation of the mitotic spindles are both deficient in those embryos whose mothers were deficient in any of the par (partition-defective) genes. The proteins encoded by these genes are found in the cortex of the embryo and appear to interact with the actin cytoskeleton (Kemphues et al. 1988; Kirby et al. 1990; Bowerman 1999).

Figure 8.43

Segregation of the P-granules into the germ line lineage of the C. elegans embryo. The left column shows the cell nuclei (the DNA is stained blue by Hoescht dye), while the right column shows the same embryos stained for P-granules. At each successive (more...)

Formation of the dorsal-ventral and right-left axes

The dorsal-ventral axis of the nematode is seen in the division of the AB cell. As the AB cell divides, it becomes longer than the eggshell is wide. This causes the cells to slide, resulting in one AB daughter cell being anterior and one being posterior (hence their names, ABa and ABp, respectively). This squeezing also causes the ABp cell to take a position above the EMS cell that results from the division of the P1 blastomere. The ABp cell defines the future dorsal side of the embryo, while the EMS cell, the precursor of the muscle and gut cells, marks to future ventral surface of the embryo. The left-right axis is specified later, at the 12-cell stage, when the MS blastomere (from the division of the EMS cell) contacts half the “granddaughters” of the ABa cell, distinguishing the right side of the body from the left side (Evans et al. 1994).

Control of blastomere identity

C. elegans demonstrates both the conditional and autonomous modes of cell specification. Both modes can be seen if the first two blastomeres are experimentally separated (Priess and Thomson 1987). The P1 cell develops autonomously without the presence of AB. It makes all the cells that it would normally make, and the result is the posterior half of an embryo. However, the AB cell, in isolation, makes only a fraction of the cell types that it would normally make. For instance, the resulting ABa blastomere fails to make the anterior pharyngeal muscles that it would have made in an intact embryo. Therefore, the specification of the AB blastomere is conditional, and it needs the descendants of the P1 cell to interact with it.

Autonomous specification

The determination of the P1 lineages appears to be autonomous, with the cell fates determined by internal cytoplasmic factors rather than by interactions with neighboring cells. It is thought that protein factors might determine cell fate by entering the nuclei of the appropriate blastomeres and activating or repressing specific fate-determining genes. Have any transcription factors been found in autonomously determined cell lineages? While the P-granules of C. elegans are localized in a way consistent with a role as a morphogenetic determinant, they do not enter the nucleus, and their role in development is still unknown. However, the SKN-1, PAL-1, and PIE-1 proteins are thought to encode transcription factors that act intrinsically to determine the fates of cells derived from the four P1-derived somatic founder cells, MS, E, C, and D.

The SKN-1 protein is a maternally expressed polypeptide that may control the fate of the EMS blastomere, the cell that generates the posterior pharynx. After first cleavage, only the posterior blastomere, P1, has the ability to autonomously produce pharyngeal cells when isolated. After P1 divides, only EMS is able to generate pharyngeal muscle cells in isolation (Priess and Thomson 1987). Similarly, when the EMS cell divides, only one of its progeny, MS, has the intrinsic ability to generate pharyngeal tissue. These findings suggest that pharyngeal cell fate may be determined autonomously by maternal factors residing in the cytoplasm that are parceled out to these particular cells. Bowerman and his co-workers (1992a,b, 1993) found maternal effect mutants lacking pharyngeal cells, and isolated a mutation in the skn-1 gene. Embryos from homozygous skn-1-deficient mothers lack both pharyngeal mesoderm and endoderm derivatives of EMS (Figure 8.44). Instead of making the normal intestinal and pharyngeal structures, these embryos seem to make extra hypodermal (skin) and body wall tissue where their intestine and pharynx should be. In other words, EMS appears to be respecified as C. Only those cells that are destined to form pharynx or intestine are affected by this mutation. Moreover, the protein encoded by the skn-1 gene has a DNA-binding site motif similar to that seen in the bZip family of transcription factors (Blackwell et al. 1994).

Figure 8.44

Deficiencies of intestine and pharynx in skn-1 mutants of C. elegans. Embryos derived from wild-type females (A, C) and females homozygous for mutant skn-1 (B, D) were tested for the presence of pharyngeal muscles (A, B) and gut-specific granules (C, (more...)

A second possible transcription factor, PAL-1, is also required for the differentiation of the P1 lineage. PAL-1 activity is needed for the normal development of the somatic descendants of the P2 blastomere. Thus, embryos lacking PAL-1 have no somatic cell types derived from the C and D stem cells (Hunter and Kenyon 1996). PAL-1 is regulated by the MEX-3 protein, an RNA-binding protein that appears to inhibit the translation of the pal-1 mRNA. Wherever MEX-3 is expressed, PAL-1 is absent. Thus, in mex-3-deficient mutants, PAL-1 is seen in every blastomere. SKN-1 also inhibits PAL-1 (thereby preventing it from becoming active in the EMS cell).

A third putative transcription factor, PIE-1, is necessary for germ line fate. PIE-1 appears to inhibit both SKN-1 and PAL-1 function in the P2 and subsequent germ line cells (Hunter and Kenyon 1996). Mutations of the maternal pie-1 gene result in germ line blastomeres adopting somatic fates, with the P2 cell behaving similarly to a wild-type EMS blastomere. The localization and the genetic properties of PIE-1 suggest that it represses the establishment of somatic cell fate and preserves the totipotency of the germ cell lineage (Mello et al. 1996; Seydoux et al. 1996).

Conditional specification

As we saw above, the C. elegans embryo uses both autonomous and conditional modes of specification. Conditional specification can be seen in the development of the endoderm cell lineage. At the 4-cell stage, the EMS cell requires a signal from its neighbor (and sister), the P2 blastomere. Usually, the EMS cell divides into an MS cell (which produces mesodermal muscles) and an E cell (which produces the intestinal endoderm). If the P2 cell is removed at the early 4-cell stage, the EMS cell will divide into two MS cells, and endoderm will not be produced. If the EMS cell is recombined with the P2 blastomere, however, it will form endoderm; it will not do so, however, when combined with ABa, ABp, or both AB derivatives (Figure 8.45; Goldstein 1992).

Figure 8.45

Results of isolation and recombination experiments, showing that cellular interactions are required for the EMS cell to form intestinal lineage determinants. (A) When isolated shortly after its formation, the EMS blastomere cannot produce gut-specific (more...)

The P2 cell produces a signal that interacts with the EMS cell and instructs the EMS daughter that is next to it to become the E cell. This message is transmitted through the Wnt signaling cascade (Figure 8.46; Rocheleau et al. 1997; Thorpe et al. 1997). The P2 cell produces the C. elegans homologue of a Wnt protein, the MOM-2 peptide. The MOM-2 peptide is received in the EMS cell by the MOM-5 protein, the C. elegans version of the Wnt receptor protein, Frizzled. The result of this signaling cascade is to down-regulate the expression of the pop-1 gene in the EMS daughter destined to become the E cell. In pop-1-deficient embryos, both EMS daughter cells become E cells (Lin et al. 1995).

Figure 8.46

Cell-cell signaling in the 4-cell embryo of C. elegans. The P2 cell produces two signals: (1) the juxtacrine protein APX-1 (Delta), which is bound by GLP-1 (Notch) on the ABp cell, and (2) the paracrine protein MOM-2 (Wnt), which is bound by the MOM-5 (more...)

The P2 cell is also critical in giving the signal that distinguishes ABp from its sister, ABa (Figure 8.48). ABa gives rise to neurons, hypodermis, and the anterior pharynx cells, while ABp makes only neurons and hypodermal cells. However, if one experimentally reverses their positions, their fates are similarly reversed, and a normal embryo is formed. In other words, ABa and ABp are equivalent cells whose fate is determined by their positions within the embryo (Priess and Thomson 1987). Transplantation and genetic studies have shown that ABp becomes different from ABa through its interaction with the P2 cell. In an unperturbed embryo, both ABa and ABp contact the EMS blastomere, but only ABp contacts the P2 cell. If the P2 cell is killed at the early 4-cell stage, the ABp cell does not generate its normal complement of cells (Bowerman et al. 1992a,b). Contact between ABp and P2 is essential for the specification of ABp cell fates, and the ABa cell can be made into an ABp-type cell if it is forced into contact with P2 (Hutter and Schnabel 1994; Mello et al. 1994).

Moreover, these studies show that this interaction is mediated by the GLP-1 protein on the ABp cell and the APX-1 (anterior pharynx excess) protein on the P2 blastomere. In embryos whose mothers have mutant glp-1, ABp is transformed into an ABa cell (Hutter and Schnabel 1994; Mello et al. 1994). The GLP-1 protein is a member of a widely conserved family called the Notch proteins, which serve as cell membrane receptors in many cell-cell interactions, and it is seen on both the ABa and ABp cells (Evans et al. 1994).^* As mentioned in Chapter 5, one of the most important ligands for Notch proteins such as GLP-1 is another cell surface protein called Delta. In C. elegans, the Delta-like protein is APX-1, and it is found on the P2 cell (Mango et al. 1994; Mello et al. 1994). This APX-1 signal breaks the symmetry between ABa and ABp, since it stimulates the GLP-1 protein solely on the AB descendant that it touches, namely, the ABp blastomere. In doing this, the P2 cell initiates the dorsal-ventral axis of C. elegans, and it confers on the ABp blastomere a fate different from that of its sister cell.