BIOLOGICAL OVERVIEW

Cbl (pronounced "cable") was initially identified as a mammalian oncogene, a gene which, when altered, causes cancer. Cancer-causing Cbl, which stands for Casitas B-lineage lymphoma, arose in mice by a recombination between the Cas-Br-M virus and c-cbl sequences, the cellular DNA sequences that code for Cbl. The transformation product is a fusion protein (v-cbl) containing viral sequences and additional sequences from cellular Cbl. Only the 355 N-terminal amino acids of the total of 913 amino acids of cellular Cbl are present in v-Cbl.

Cbl is a multidomain protein. The human Cbl consists of a long C-terminal proline rich region that is not present in either Drosophila or C. elegans Cbl proteins. Conserved in the Cbl proteins of humans, flies and C. elegans is a central ring finger motif (C-terminal in worms and flies); adjacent to this ring finger is another motif (70Z/3) that when deleted, renders human Cbl oncogenic. To date, the functions of the ring finger motif and the 70Z/3 motif are unknown, but ring fingers are implicated as protein interaction motifs and deletion of the 70Z/3 sequence renders Cbl constitutively active (oncogenic).

Cbl is of interest to developmental biology because Cbl acts in both C. elegans and Drosophila as a negative regulator of receptor tyrosine kinase signaling. Cbl acts to diminish signaling from the Epidermal growth factor receptor in the development of the R7 photoreceptor in Drosophila eyes (Meisner, 1997) and in the induction of vulval differentiation in C. elegans (Yoon, 1995).

Cbl is active in a number of pathways. It is thought that mammalian Cbl is recruited to the mammalian EGF receptor via interaction of Cbl with Grb2 (Meisner, 1995a). Human Cbl interacts via its proline-rich C-terminus with Grb2, a component of the ras pathway that directly interacts with receptors. Cbl also recruits (binds) phosphatidylinositol 3'-kinase (Meisner, 1995b) and an adaptor protein Crk (Buday, 1996). Cbl interacts with and is phosphorylated by a number of kinases: Fyn, Lyk (Donovan, 1994) and Abl (see Drosophila Abl) (Andoniou, 1994). Finally, Cbl possesses a nuclear localization signal, suggesting that Cbl might have a direct role in the regulation of transcription.

Even though Drosophila Cbl lacks the C-terminal proline-rich sequences of mammalian Cbl and fails to bind Grb2, Drosophila Cbl interacts with the Epidermal growth factor receptor in response to Epidermal growth factor. EGF also causes tyrosine phosphorylation of Cbl but no association of phosphatidylinositol 3-kinase is detected. Since an N-terminally deleted form of Drosophila Cbl fails to associate with Egf-receptor, it seems likely that the most highly conserved region of Cbl, amino acids 205 to 330, may be sufficient for binding directly to the EGF receptor. Thus the N-terminal region of Cbl proteins appears to provide a second means (other than binding to Grb2) for EGF receptor association (Meisner, 1997).

Mutations in Drosophila Cbl have not yet been identified. Despite this lack of information, studies have been made of possible developmental roles for Cbl. In the developing eye, the differentiation of R7 photoreceptor neurons depends on signaling through the Sevenless and EGF-receptor tyrosine kinases. Transformant flies were constructed that carry Drosophila Cbl cDNA under the transcriptional control of a promotor that causes the expression of Cbl in all cells that express Sevenless, including the R7 photoreceptor precursor. To assess the role of D-Cbl in R7 development, these transgenic flies were tested in a sensitized genetic assay. In this assay, signaling through Sevenless is compromised using a partially disabled Sevenless kinase. In such flies the dosage of genes participating in receptor tyrosine kinase signaling, and the fraction of ommatidia developing R7 cells provide an accurate readout of the strength of the transduction signal. A copy of Cbl transgene was introduced into the sensitized background. In such flies, the development of R7 cells is essentially eliminated. This result strongly argues that in the Drosophila eye, Cbl acts as a negative regulator of one or more receptor tyrosine kinase pathways that are essential for photoreceptor differention (Meisner, 1997).

In C. elegans, mutations in the Cbl homolog, sli-1, interact in four of the five developmental phenomena known to involve the Epidermal growth factor receptor, LET-23. The sli-1 (suppressor of lineage defect) locus was defined by extragenic suppressor of let-23 reduction-of-function mutations. The let-23(rf) mutation causes at least five phenotypes: defects in (1) viability, (2) hermaphrodite fertility, (3) male spicule development, (4) posterior epidermal development, and (5) vulval differentiation. The sli-1 (rf) mutations suppress all known defects of let-23 (rf) with the exception of sterility (Yoon, 1995).

Colocalization of Drosophila and mammalian Cbl proteins with mammalian EGF receptor can be detected in cultured cells as early as a few minutes after addition of EGF; by 20 minutes, an intense punctate pattern of Cbl and the EGF receptor is evident. It is thought that Cbl-EGF receptor complexes are internalized into endosomes through a coated pit pathway. In C. elegans, sli-1 gene has been shown to interact genetically with unc-101 in the regulation of the EGF receptor pathway leading to vulval differentiation (Lee, 1994). unc-101 encodes a clathrin-associated protein homologous to mouse AP47, which constitutes one of the main components of coated pits and vesicles. It is tempting to speculate that Cbl function in downregulating EGF-R signaling could involve the degradation of receptor-Cbl complexes during EGF-R trafficking within intracellular membranes (Meisner, 1997).

During Drosophila oogenesis, asymmetrically localized Gurken activates the EGF receptor (Egfr) and determines dorsal follicle cell fates. Using a mosaic follicle cell system a mutation has been identified in the Cbl gene that causes hyperactivation of the Egfr pathway. Cbl is required in ventral follicle cells to ensure that ventral patterning occurs correctly in the embryo. Cbl proteins are known to downregulate activated receptors. The abnormal Egfr activation is ligand dependent. These results show that the precise regulation of Egfr activity necessary to establish different follicle cell fates requires two levels of control. The localized ligand Gurken activates Egfr to different levels in different follicle cells. In addition, Egfr activity has to be repressed through the activity of Cbl to ensure the absence of signaling in the ventral most follicle cells (Pai, 2000).

To examine the Cbl expression pattern in the ovary, in situ hybridization experiments were performed. Cbl mRNA is detected both in nurse cells and in follicle cells that are associated with the oocyte. The high levels of expression in nurse cells may indicate a maternal contribution of Cbl to the embryo: this is consistent with its presence at the blastoderm stage. To test the requirement for Cbl in embryonic development, germline clones for Cbl were generated using heat shock Flipase and the ovo^D1 system. A distinct head defect is observed in embryos lacking both maternally and zygotically contributed Cbl. Zygotic Cbl is able to rescue the maternal lack of Cbl. This results in normal embryos that hatch. A small percentage of embryos had dorsalization phenotypes, which is likely due to the simultaneous generation of follicle cell clones induced by the heat shock Flipase. Egfr signaling is also required for ventral ectoderm development during embryogenesis. However, there were no other cuticle phenotypes detected in Cbl germline clone embryos, which suggests that Egfr signaling in cuticle development is not severely affected by loss of Cbl. In particular, no obvious segmentation defects, that would have suggested hyperactivity of the torso pathway, were detected. Taken together, these results demonstrate that in some signaling pathways involving receptor tyrosine kinases, loss of Cbl has no visible phenotypes, but in processes that are very sensitive to levels of receptor activity, such as in the follicle cells, a mutation of Cbl has dramatic effects (Pai, 2000).

To confirm that the dorsalized embryonic phenotype is caused by follicle cell clones homozygous for Cbl, and to determine in which follicle cells Cbl is required, the Cbl mutant cells were marked with a defective chorion 1 (dec) mutation. dec mutant follicle cells produce an abnormal egg shell resulting in an almost transparent appearance of the follicle cell imprints on the egg shell, in contrast to the opaque appearance of wild-type follicle cell imprints. Dorsalized embryos were observed only within mosaic egg shells. Furthermore, by correlating the position of the mutant clones on the egg shell with the region of dorsalization in the embryo, the spatial requirement for F165 (a Cbl mutation) activity in the establishment of ventral cell fates in the embryo could be determined. Surprisingly, only clones at ventral positions within the follicle cell epithelium result in a dorsalized embryo phenotype. Clones that were confined to the dorsal half of the egg shell do not produce a visible embryonic mutant phenotype. From these observations, it is concluded that Cbl is required in ventral follicle cells to ensure that ventral patterning occurs correctly in the embryo (Pai, 2000).

Although Cbl mutant dorsal follicle cell clones do not alter the pattern of the embryo, an effect of these clones was observed on the pattern of the egg shell when the clones occupied the dorsoanterior regions adjacent to the dorsal appendages. When dorsolateral cells are mutant for Cbl, dorsal appendages are shifted to more lateral positions. This shift of dorsal appendage could be due to an expansion of midline cell fates caused by higher Egfr activity at the dorsal-lateral position. This possibility was tested by staining for the expression of argos, a dorsal midline, cell-specific gene. Argos expression was expanded in several mosaic egg chambers with dorsal Cbl mutant clones. When cells situated more laterally, adjacent to the dorsal appendages, are mutant for Cbl, extra or wider dorsal appendages are produced. These egg shell phenotypes suggest that Egfr signaling is elevated in these dorsal Cbl mutant follicle cell populations (Pai, 2000).

To directly analyze the effects of the Cbl mutation on the Egfr pathway, the expression of a primary downstream target gene, kekkon 1 (kek), was examined in mutant clones. kek expression is induced in follicle cells in response to Egfr activation triggered by the germline specific ligand, Gurken. In grk mutants, kek expression is lost; conversely, expression of a constitutively active form of Egfr in all follicle cells results in kek expression in all follicle cells. In order to detect kek expression, a kek enhancer trap line, BB142, was used in which LacZ is expressed under the control of the kek promoter, and the ovaries were stained with anti-ß-galactosidase antibody. To visualize the Cbl mutant cells, Cbl was placed in trans to a c-myc-tagged chromosome. Homozygous mutant clones were detected by the absence of c-myc expression, and the wild-type cells were detected by staining positively with anti-c-myc antibody. In wild-type stage 7 and 8 egg chambers, kek is highly expressed in the posterior follicle cells that overlie the oocyte, and in a small number of the most anterior follicle cells. When Cbl mutant clones are induced in the anterior half of the egg chamber, these mutant cells ectopically express kek. In wild-type egg chambers at stages 9 to 10, after the oocyte nucleus has migrated to the dorsal anterior corner of the oocyte, kek expression becomes restricted to the dorsal cells that are exposed to Gurken signal. Ectopic kek expression is detected in mutant clones localized at the ventral side of stage 10 egg chambers. At both stages, this ectopic expression of kek is cell autonomous: all the mutant cells express kek, even at the clone boundary. The induction of kek expression in mutant clones demonstrates that the Cbl phenotype reflects activation of the Egfr pathway and indicates that this activation occurs in a cell autonomous manner. These results are consistent with the molecular nature of Cbl, given that Cbl downregulates receptor activity in signal receiving cells (Pai, 2000).

The ectopic expression of kek indicates that in the F165 mutant clones, the Egfr pathway is active in ventral follicle cells. This raises the question whether the effect of F165 on Egfr activity requires Gurken. To answer this question, F165 mutant clones were generated in a gurken mutant background and examined for kek expression. Surprisingly, no kek expression was detected in ventral Cbl mutant clones in gurken mutant egg chambers. The fact that ectopic kek expression was lost from the mutant clones in the absence of Gurken strongly suggests that ectopic activation of the Egfr pathway in Cbl mutant cells requires the presence of ligand to activate the receptor (Pai, 2000).

Ventral follicle cell clones mutant for Cbl produce a dorsalized embryo phenotype. Since restricted pipe expression in the follicle cells is of crucial importance for embryonic patterning, pipe expression was examined in mutant clones by RNA in situ hybridization on ovaries. pipe expression is detected in ventral follicle cells in wild-type stage 9 egg chambers and persists until stage 10B. When egg chambers were examined that contained follicle cell clones mutant for Cbl, pipe expression was found to be abolished in ventral mutant clones. Interestingly, there is a sharp boundary between cells with and without pipe expression. Since Egfr activity is elevated in Cbl mutant cells in a cell autonomous manner, the effects of Egfr signaling on pipe expression could be directly examined. The wild-type cells were marked with a c-myc antibody and simultaneously pipe in situ hybridization was carried out. There was an exact correspondence between the absence of pipe expression and the absence of c-myc staining on the ventral side of stage 9 or 10A egg chambers. This demonstrates that the ectopic Egfr activation in the mutant clones suppresses pipe expression in a cell autonomous fashion. Since pipe expression in ventral follicle cells is required for ventral cell fate determination in the embryo, these results also suggest that the elimination of pipe expression in mutant clones can account for the dorsalization phenotype of the resulting embryo. The cell by cell correspondence of mutant clones and pipe expression further implies that the Egfr pathway can downregulate pipe expression directly, rather than acting via a secondary, diffusible signal that would repress pipe at a distance (Pai, 2000).

Along the dorsoventral axis, there are four types of follicle cells that can be distinguished by their gene expression patterns as well as their imprints on the egg shell: the dorsal midline cells, which express argos and are located between cells that will give rise to two dorsal appendages; the dorsolateral cells, which express Broad-Complex and secrete the appendages; the ventral cells that express pipe; and the lateral cells located between ventral and dorsolateral cells. The different embryonic and egg shell phenotypes that were generated by follicle cell clones mutant for Cbl have demonstrated that all of the thresholds that define these cell populations are affected by Cbl. In follicle cells lacking Cbl, Egfr signaling is elevated. In ventral follicle cells, this level of elevation of Egfr signaling is sufficient to suppress pipe expression, and results in dorsalized embryos. However, this ectopic activation of Egfr in ventral mutant cells is not sufficient to induce dorsal appendage formation. A higher level of activity is required to reach the threshold necessary to produce dorsal appendages. Only Cbl mutant follicle cells in lateral regions close to dorsal appendages can produce ectopic dorsal appendages. Furthermore, dorsal midline cell fates, which initially have the highest level of Egfr activity, can only be induced in cells next to the dorsal midline, since the expansion of argos expression was observed in mutant cells adjacent to the dorsal midline cells, but not in ventral follicle cells. However, it is known that high levels of uniform Egfr activity can transform all follicle cells into dorsal midline or dorsolateral cell fates. Therefore, in the absence of Cbl, an underlying Egfr activity gradient exists in follicle cells along the dorsoventral axis, which needs to be regulated by the activity of Cbl (Pai, 2000).

A pivotal question of dorsoventral patterning in the egg chamber is how Gurken/Egfr signaling on the dorsal side of the follicular epithelium determines ventral follicle cell fates, which in turn establish the ventral cell fates in the embryo. The highly asymmetric distribution of Gurken along the dorsoventral axis had led to the model that the Egfr is only activated in a relatively small population of dorsal follicle cells along the dorsal midline. Genetic mosaic studies of the activity of pipe and wind in follicle cell epithelium have demonstrated that there is a ventral region in the follicle cell epithelium, comprising about one third of the circumference required for the establishment of the embryonic ventral cell fates. It had appeared therefore that there might be a considerable distance between the dorsal Egfr signaling and the ventral zone of pipe activity, suggesting the possibility that a secondary signal induced by Egfr signaling might be involved. This study of Cbl in axis formation has, however, shown that Egfr signaling itself appears to directly determine the size of the ventral pipe expressing region. From these results it appears that all follicle cells have some Gurken-dependent Egfr activity, but that this occurs at different levels along the dorsoventral axis. In this system, the essential role of Cbl is to control the level of Egfr signaling by destroying activated Egfr in order to keep the Egfr activity below certain thresholds, thus ensuring the full range of follicle cell fates. This conclusion is supported by the observation that ectopic kek expression depends on the presence of Gurken. In addition, the fact that in the Cbl mutant clones, pipe expression is repressed in a cell autonomous manner, shows that Egfr activity affects ventral patterning directly, and not at a distance. In the absence of Cbl, the low levels of Egfr signaling triggered by Gurken in ventral follicle cells are sufficient to repress pipe expression, which leads to the loss of embryonic ventral cell fates (Pai, 2000).

Two possible models have been suggested for the establishment of a gradient of Egfr activity by Gurken. One is that asymmetrically localized Gurken activates Rhomboid, which then processes and activates Spitz, another Egfr ligand. Activated Spitz or other unidentified ligands might diffuse away from the dorsal source and establish a gradient of Egfr activity. The alternative model is that Gurken distribution is broader than it appears, judging from antibody staining, and that some Gurken actually reaches the ventral side of the egg chamber and represses pipe expression directly. This could involve either diffusion of processed Gurken in the space between the follicle cells and the oocyte membrane, or it could involve diffusion of membrane tethered Gurken protein within the oocyte membrane. Mutant follicle cell clones for spitz cause only relatively minor defects in egg shell morphology, but not in embryonic development. This argues against Spitz acting as the global cell fate determinant. Therefore, the favored model is that Gurken is present in a graded distribution in the egg chamber and activates Egfr in ventral follicle cells. Similar to the results from the analysis of spitz and argos clones, defects in dorsal follicle cell patterning have no effects on embryonic patterning, since embryos develop normally and hatch out of eggs containing dorsal Cbl mutant clones. This further supports the model that the ventral region responsible for the ventral cell fate establishment in the embryo is directly defined by the Gurken signal, and that the second phase of Egfr signaling involving amplification through Rho, Spitz, and Argos only affects egg shell morphology, but not the dorsoventral axis of the embryo. Cbl activity is required in all follicle cells along the dorsoventral axis, and affects both patterning events. Its activity essentially acts as a sink that ensures that the full range of the Egfr activity gradient is reliably achieved and/or maintained in the follicle cell epithelium (Pai, 2000).

GENE STRUCTURE

Bases in 5' UTR - 313

Bases in 3' UTR - 1317

PROTEIN STRUCTURE

Amino Acids - 448

Structural Domains

A comparision of the amino acid sequences of human-Cbl, Drosophila-Cbl and C. elegans Cbl (Sli-1) shows no homology over the first 40 amino acids or beyond residue 426 of Drosophila Cbl. However, within the conserved region of approximately 380 amino acids, the three species show 63% similarity between amino acids 46 and 205 and 93% similarity between amino acids 205 and 330. Within the N-terminal conserved region, several tyrosine residues found in human Cbl and not present in Drosophila Cbl. A high degree of similarity is also observed between amino acids 354 and 425. This region includes a ring finger domain and a sequence of 17 amino acids (called 70Z/3) which, when deleted, renders c-Cbl transforming. Interestingly, Drosophila Cbl is only 93 residues longer than v-Cbl, but these amino acids include the important 70Z/3 sequence and the ring finger. Finally, unlike the protein found in C. elegans or mammals, Drosophila Cbl contains no proline-rich motifs (Meisner, 1997).

date revised: 24 November 2000 

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.