EGFR signaling attenuates Groucho-dependent repression to antagonize Notch transcriptional output

Abstract

Crosstalk between signaling pathways is crucial for the generation of complex and varied transcriptional networks. Antagonism between the EGF-receptor (EGFR) and Notch pathways in particular is well documented, although the underlying mechanism is poorly understood. The global corepressor Groucho (Gro) and its transducin-like Enhancer-of-split (TLE) mammalian homologs mediate repression by a myriad of repressors, including effectors of the Notch, Wnt (Wg) and TGF-beta (Dpp) signaling cascades. Given that there are genetic interactions between gro and components of the EGFR pathway (ref. 9 and P.H. et al., unpublished results), we tested whether Gro is at a crossroad between this and other pathways. Here we show that phosphorylation of Gro in response to MAPK activation weakens its repressor capacity, attenuating Gro-dependent transcriptional silencing by the Enhancer-of-split proteins, effectors of the Notch cascade. Thus, Gro is a new junction between signaling pathways, enabling EGFR signaling to antagonize transcriptional output by Notch and potentially other Gro-dependent pathways.

Crosstalk between the EGFR and other signalling pathways at the level of the global transcriptional corepressor Groucho/TLE

“…Importantly, the activation of EGFR, or of downstream components of the pathway such as Ras or MAPK, attenuates Gromediated repression in vivo. Conversely, when either EGFR or Ras are mutated, an opposite effect is observed, that is, Gro-mediated repression is strengthened (Hasson et al, 2005). Consistent with these results, repression is less potent by a derivative of Gro that mimics a constitutively pseudo-phosphorylated form (Gro DD ), in which the phospho-acceptor residues found in the two MAPK sites have been substituted for by negatively charged Aspartate (D) amino acids.…”

“…As mentioned above, Gro/TLE-dependent repressors function not only downstream of N, but also as effectors of the Dpp/TGF-b and Wg/Wnt pathways. The regulation of Gro by RTK signalling is expected to similarly affect the transcriptional output of these cascades ( Figure 2C) (Hasson et al, 2005;our unpublished data).…”

“…Groucho contains two putative, evolutionarily conserved MAPK consensus sites, and is phosphorylated in cell culture and in vivo in response to EGFR signalling (Hasson et al, 2005). Importantly, the activation of EGFR, or of downstream components of the pathway such as Ras or MAPK, attenuates Gromediated repression in vivo.…”

“…Consistent with these results, repression is less potent by a derivative of Gro that mimics a constitutively pseudo-phosphorylated form (Gro DD ), in which the phospho-acceptor residues found in the two MAPK sites have been substituted for by negatively charged Aspartate (D) amino acids. Reciprocally, blocking phosphorylation of the MAPK sites by changing these residues to Alanine (A) (Gro AA ) leads to the overpotentiation of Gro's corepressor capacity (Hasson et al, 2005). Significantly, phosphorylation of Gro in response to MAPK activation hinders the E(spl) proteins -the nuclear effectors of the Notch pathway -from effectively silencing their target genes, as will be described below.…”

“…Several lines of evidence now suggest that EGFR activity promotes vein and bristle formation at least in part by phosphorylating Gro and relieving E(spl)-mediated, Gro-dependent repression, and hence interfering with N signalling ( Figure 1B). First, the misexpression of the pseudophosphorylated Gro-derivative (Gro DD ), the presumed end product of MAPK signalling vis-à-vis Gro, generates similar effects to those seen when the EGFR pathway is overactivated, that is, extra vein material and supernumerary bristles are formed (Hasson et al, 2005). Second, the overexpression of Gro AA , the nonphosphorylatable derivative of Gro, renders the N pathway refractory to antagonistic EGFR activity, and the phenotypes caused by the misexpression of Gro AA are comparable to those observed when the Notch pathway is ectopically activated, that is, loss of vein material and of bristles (de Celis et al, 1997;Culi et al, 2001;Hasson et al, 2005).…”