c-SRC is a common oncogenic partner of the epidermal growth-factor receptor (EGFR), as both c-SRC and EGFR are overexpressed in human malignancies such as breast cancer. However, the mode of oncogenic cooperation between these two proteins has been unclear. Now, Yosef Yarden and colleagues report that c-CBL — a ubiquitin ligase that is known to be a regulator of EGFR endocytosis — is the 'middle leg' in the relay of oncogenic signalling between c-SRC and EGFR.

To investigate whether c-SRC affects the expression of EGFR, the receptor was co-expressed with different forms of c-SRC in receptor-negative Chinese hamster ovary (CHO) cells. Both wild-type c-SRC and an active mutant form resulted in increased levels of expression of EGFR, in contrast to a kinase-defective mutant, which led to decreased EGFR expression levels. As the active form of c-SRC was shown to have no effect on receptor synthesis (mRNA levels) or maturation, the authors propose that c-SRC stabilizes the mature, cell-surface form of EGFR.

c-CBL ubiquitylates EGFR, which results in its endocytosis and degradation and, therefore, receptor desensitization in response to growth-factor signalling. Could effects on c-CBL be responsible for the change in expression of EGFR in response to c-SRC? Ectopic expression of c-CBL increased the removal of EGFR from the surface of SYF cells (which lack SRC, YES and FYN), but this receptor endocytosis was inhibited by the co-expression of an active-mutant c-SRC. The authors found that c-SRC prevents the interaction between c-CBL and EGFR (which is required for receptor ubiquitylation), specifically downregulates the expression of c-CBL and recruits c-CBL to vesicles.

So, how does c-SRC decrease the level of expression of c-CBL? Expression of active mutant c-SRC leads to polyubiquitylation of c-CBL, which targets this protein for degradation by the proteasome. This process does not occur for a RING-finger mutant c-CBL (lacking ubiquitin-ligase activity), which indicates that c-SRC activates a self-ubiquitylating function of c-CBL. However, the RING-finger mutant is still susceptible to the degradative effects of c-SRC, so it is probable that alternative, ubiquitin-independent mechanisms for the degradation of c-CBL are also used.

These data indicate the following mechanism of oncogenic cooperation. Increased expression of c-SRC or expression of a constitutively active form leads to the destruction of c-CBL (through self-ubiquitylation or other means). In turn, reduced levels of c-CBL mean that EGFR ubiquitylation and endocytosis (receptor desensitization) are inhibited and growth-factor signalling is increased, thereby potentiating EGF-induced mitogenesis.