EGF receptor endocytosis and degradation

The major interest of our current work is to understand the mechanisms involved in negative regulation of EGF receptor signaling pathways (Dikic and Giordano, 2003). Terminal EGFR signal inactivation is accomplished by endocytosis and degradation of activated receptors and associated signalling proteins. These processes are essential to avoid constitutive signalling and tumorigenesis. Ligand-induced ubiquitination of EGF receptors has been linked to their negative regulation by internalization and endocytic targeting to destruction in the lysosome. The Cbl family of ubiquitin ligases plays pivotal roles in these processes (Dikic et al. 2003). Cbl can directly bind to phosphorylated EGF receptors via its tyrosine kinase binding (TKB) domain, while the RING finger domain of Cbl recruits ubiquitin-conjugating enzymes (E2, Ubc) and mediates the transfer of ubiquitin to the receptor.

 

Figure 1. Members of the Cbl protein family. Domain structure of mammalian Cbl, Cbl-b and Cbl-c, Drosophila melanogaster D-Cbl short and long as well as Caenorhabditis elegans SLI-1 are shown.

Recent lines of evidence show that Cbl mediates monoubiquitination of EGF receptors and that ubiquitin carries both internalization and degradation signals that control trafficking and sorting of receptors for degradation in the lysosome (Haglund et al. 2003). In addition, we have recently discovered a novel pathway by which Cbl and Cbl-b regulate internalization of the EGF receptor (Soubeyran et al, 2002; Szymkiewicz et al., 2002). This pathway is functionally separable from the ubiquitin ligase activity of Cbl and it depends on binding of the adaptor protein CIN85 to Cbl and recruitment of endophilins in complexes with activated EGF receptors (Dikic, 2002; Soubeyran et al, 2002; Szymkiewicz et al., 2002). Subsequently, we have shown that CIN85 associates with Cbl/receptor complexes along the endocytic pathway, where Cbl directs monoubiquitination of CIN85 (Haglund et al, 2002). These events seem to be critical for proper routing of EGF receptor complexes for degradation in the lysosomal compartment (Haglund et al, 2002).

 

Figure 2. Intracellular trafficking of activated receptors.
The activation of EGF receptors by EGF induces its phosphorylation and the recruitment of Cbl that mediates its multi-monoubiquitination. This process is rapidly followed by internalization of activated receptors/Cbl complexes and their subsenquent trafficking toward various intracellular compartments leading, in fine, to either the recycling of the receptors to the cell surface or their concomitant degradation in the lysosome. This mechanisms are regulated by divers proteins able to interact with ubiquitin moeties via specific domains (UIM:Ubiquitin Interacting Motif; UBA:UBiquitin Associated; UBC:UBiquitin Conjugating enzyme like) and that can also undergo mono-ubiquitination.

Major questions we are currently addressing deal with the role of monoubiquitin signals in endocytic sorting of cargo (i.e EGFR and associated receptor complexes) as well as functions of ubiquitin binding proteins that serve as ubiquitin receptors along the endosome (see Haglund et al. 2003b). In addition, we are particularly interested to understand the mechanisms which define the specificity in determining mono-, multi- or poly-ubiquitination of distinct Cbl-substrates?

 

 

 

 

Ubiquitin signaling in the NF-kB pathway

NF-kB is a key transcription factor that regulates various pathological and biological conditions, such as immune response, cancer, skin or bone development and so on.
It has been long known that NF-kB is activated by different stimuli, ex. TNF-a, LPS or IL-1b in above conditions. Importantly, a small modifier, ubiquitin, is involved in control of NF-kB activation in different aspects. Firstly, by these stress-inducing stimulations, enzymes of ubiquitylation such as TRAFs (E3 liagse), IAPs (E3 ligase), Ubc13 (E2 ligase) are activated. Secondly, signal mediators, RIP, TRAFs and NEMO (IKK subunit, also known as IKKg) are ubiquitylated with help of these enzymes. So far, RIP and TRAFs are considered to be poly-ubiquitylated via K63-linked chains. On the other hand, NEMO is modified either by single ubiquitin molecule, K63-linked or linear ubiquitin chains. Thirdly, ubiquitylation of these molecules assembles a complex with proteins that contains ubiquitin binding domains, such as TAB2, ABINs and NEMO. The assembly of these networks activates IKK kinases that are the key player in NF-kB signaling. Lastly, IKKs phosphorylate a substrate, IkBa that leads itself K48-linked ubiquitylation.
Polyubiquitylated IkBa is degradated by proteasome leading NF-kB components, p65 and p50 being freed and target transcriptions. A regulatory mechanism of NF-kB has been long studied, however there are still missing links that are in a black box. For example, a modification via K63-linked ubiquitin chains of RIP by Ubc13 is a critical step for NF-kB activation. However, knock down of Ubc13 in mice does not affect the NF-kB signaling. It is not yet clear how this gap is connected. Another example is how TAB2/TAK selectively activate NF-kB and MAPK pathways. On top of that, the major question, which still remains is “how IKK kinases are activated”.

Our group is focusing on how these ubiquitylated molecules are involved in regulation of NF-kB signaling pathway. Modification of proteins by ubiquitin changes; 1) conformation of proteins themselves, 2) interactoms, and 3) stability of proteins themselves, in hence controls many biological functions. Currently, we are trying to elucidate how ubiquitin modification changes protein-protein interactions and structural confirmation in regulation of NF-kB signaling. For example, we recently identified a critical role of novel ubiquitin binding domains in ABINs, NEMO and Opteneurin in NF-kB signaling pathway (Wagner et al, 2008; Bloor et al, 2008). We also clarified an important role of ubiquitin binding of Tax1BP1 Zinc finger in NF-kB regulation both in vivo and in vitro (Iha et al, 2008).

 

Terzic J, et al. Biochem Soc Trans. 2007 Nov;35(Pt 5):942-5. Review.

 

 

 

 

Specificity in Ubiquitin signaling

Ubiquitin is a highly conserved 76-amino-acid polypeptide that is covalently attached to target proteins via an isopeptide bond between the carboxyl-terminal glycine of ubiquitin and the e-amino group of a lysine in substrate proteins. This occurs through a three-step process involving ubiquitin-activating (E1), ubiquitin-conjugating (E2) and ubiquitin-ligating (E3) enzymes (Hofmann and Pickart, 2001). The addition of a single ubiquitin molecule to a substrate is defined as monoubiquitination (Figure 1) (Dunn and Hicke, 2001). Alternatively, several lysine residues in the substrate can be tagged with single ubiquitin molecules, giving rise to multiple monoubiquitination, also referred to as multiubiquitination (Figure 1) (Dikic et al., 2003; Haglund et al., 2003; Mosesson et al., 2003). Moreover, ubiquitin contains seven lysine residues, which can also be targeted by another ubiquitin in an iterative process, known as polyubiquitination, that leads to the formation of a ubiquitin chain attached to a single lysine of a protein substrate (Figure 1) (Hofmann and Pickart, 2001). It is now clear that different types of ubiquitin conjugates are involved in the regulation of different cellular processes (Figure 1). Monoubiquitination is implicated in endocytosis of plasma membrane proteins, sorting of proteins to multivesicular bodies (MVB), budding of retroviruses, DNA repair, histone activity and transcriptional regulation (Dunn and Hicke, 2001). On the other hand, polyubiquitin chains formed via lysine 48 of two ubiquitins have a well characterized role in targeting proteins to degradation by the 26S proteasome, while ubiquitin chains formed through lysine 63 are involved in endocytosis processes and DNA repair (Ettenberg et al., 2001; Hofmann and Pickart, 2001).

The wide variety of proteins known to be modified by Ub in cells points to the existence of a large Ub-linked network participating in numerous cellular processes. The recent identification of Ub-binding domains found in proteins with known function in diverse biological processes has provided some insights in how specific Ub-interactors may regulate Ub functions in cells. Currently known domains able to directly interact with Ub include UIM (ubiquitin-interacting motif), UBA (ubiquitin-associated domain), UBC (ubiquitin-conjugating enzyme-like)/UEV (ubiquitin E2 variant), CUE (Cue1-homologous), PAZ (polyubiquitin-associated zinc finger) and NZF (novel zinc finger) domain (recently reviewed in DiFiore et al 2003 and Schnell and Hicke, 2003).

Ub-binding domains have different structural folds recognizing distinct patches on the Ub surface. They are thus able to bind Ub with different affinities and/or may specifically recognize different Ub modification such as Ub-monomer versus Ub-chains (Di Fiore et al., 2003; Schnell and Hicke, 2003). Several recent studies have suggested that Ub-binding proteins may participate in the control of Ub-dependent physiological and pathological processes, including intracellular transport, DNA repair, gene transcription or silencing, HIV budding, and neurological disorders, like Alzheimer and Parkinson (Di Fiore et al., 2003; Schnell and Hicke, 2003). However, much more work remains to be done to prove that interactions between ubiquitinated substrates and Ub-binding proteins are essential for regulation of these processes in particular in vivo. We are currently analyzing functional roles of several Ub-binding proteins by inducible expression of small interfering RNA in embryonic stem (ES) cells. The siRNA target vectors will be stably integrated into the genome of ES cells that can be differentiated into neuronal and hematopoietic cell types, and following induction of siRNA the effects on processes like apoptosis, proliferation, cell cycle, protein degradation and gene transcription will be examined. Taking advantage of the totipotency of ES cells, we will be able to generate transgenic mice from those cell lines that exhibit the most significant phenotypes.

 

 

Regulation of translesion DNA synthesis by ubiquitin

Replicative DNA polymerases are known for the high fidelity with which they replicate DNA. This is made possible thanks to the structure of their catalytic pockets which cannot accommodate anything but correct base pairs. As a consequence, replication forks get stalled whenever high-fidelity polymerases encounter a lesion. A solution to this problem exists and is called post-replication repair. One of its pathways involves specialized polymerases of the Y-family and is known as translesion synthesis (TLS). In this process, the replicative polymerase, blocked at a damaged site, is exchanged to one of the Y-polymerases: iota, eta, kappa or Rev1. The catalytic pockets of these enzymes are much more open, and thus can accommodate distorted DNA and replicate across it. In this way, the lesion is bypassed and replication can be accomplished. It has been shown that upon treatment with DNA damaging agents PCNA undergoes monoubiquitylation and that this modification is necessary for TLS to occur. However, what monoubiquitylation of PCNA does to promote TLS has been an open question for a long time.
Our group has characterized two novel ubiquitin-binding domains (UBDs) embedded in the structure of all Y-polymerases. They were named Ub-Binding Motif (UBM) and Ub-Binding Zinc-finger (UBZ) and were shown to be necessary for these proteins to localize to replication foci in order for TLS to be effective and efficient. We have demonstrated that it is the interaction between the UBDs of those polymerases and the ubiquitin (Ub) moiety attached to PCNA that allows this process to occur upon replication block (Figure 1). This mechanism is universal for all Y-family polymerases and evolutionary conserved from yeast to mammals. Moreover, these polymerases were proven to undergo monoubiquitylation themselves in an UBD-dependent manner and our preliminary results suggest that this modification regulates their access to replication forks.
We have now identified a mutant of polymerase eta, which cannot undergo monoubiquitylation and are using it as a tool to study the function of this modification. We propose that monoubiquitylation of polymerase eta inactivates this protein due to an intramolecular interaction between its UBZ domain and a ubiquitin moiety covalently attached to the polymerase. This possibly represents yet another level of regulation of TLS and might hold true also for the other three members of the Y family of polymerases.

Figure 1. Upon treatment with agents that introduce DNA lesions that are recognized by the TLS machinery (here UVC as an example), the Rad6/Rad18 complex monoubiquitylates PCNA at replication forks stalled by the damage. This leads to displacement of normal replicative polymerases such as pol epsilon, and recruitment of one of the Y-polymerases. The latter not only recognize PCNA via their PIPbox (PCNA Interacting Peptide) motifs but the appearance of a Ub moiety on PCNA creates an additional surface recognized by their UBD domains.