Cbl ubiquitin ligase: Lord of the RINGs

Not just quite interesting - really interesting!

A cell must be able to degrade proteins when their activity is no longer required. Many eukaryotic proteins are labelled for destruction by addition of a 'flag' called ubiquitin. This 'ubiquitination' directs proteins along a pathway for degradation.

RING E3 ligase transfers ubiquitin from the E2 active site to the substrate.
Defects in such pathways are associated with diseases such as cancer, neurodegenerative disorders and viral infections. Flagging a protein with ubiquitin is done by enzymes called ubiquitin ligases. There are many such enzymes in the cell which flag different sorts of proteins, and one which is quite interesting is Cbl (also called c-Cbl) (view-1). Cbl is a member of the RING E3 class of ubiquitin ligases, so called because it contains a RING (Really Interesting New Gene) domain. Members of the RING E3 family are the final protein in a cascade of three enzymes and act to transfer ubiquitin from a donor protein, E2 ubiquitin-conjugating enzyme (E2), to the protein that is to be flagged for destruction. Cbl specifically transfers ubiquitin to proteins in the receptor tyrosine kinase (RTK) family which is important in cell signalling.

One RING to bind E2

Cbl is a multi-domain protein (view-1). A tyrosine kinase-binding domain (TKBD) which binds RTK substrates is found at the N-terminus. It is connected to the RING domain via a linker-helix region. The RING domain is followed by a proline-rich region and, finally, a ubiquitin-associated (UBA) domain. Only the first three domains, TKBD, linker helix and RING, are needed for activity. The structure of this portion of Cbl has been solved (PDB 1fbv) in complex with its two substrates, E2 and a peptide from the tyrosine kinase. The structure reveals the domain architecture of Cbl and the location of the substrate-binding sites. It turns out that the RING domain binds the ubiquitin-donor protein E2 and the TKBD binds the RTK substrate. However, these two substrates, ubiquitin donor and acceptor, are located on opposite sides of the Cbl molecule, nearly 70 Å apart (view-2)! It would be impossible to transfer the ubiquitin between the two substrates in this conformation.
Cbl acts not only as a E3 ubiquitin ligase but also as an adaptor in RTK signal transduction. In this role, the enzyme must bind its target protein, but crucially should not modify it by addition of ubiquitin, i.e., its enzymatic function should be suppressed. Which of the two functions Cbl performs is controlled by phosphorylation of a single tyrosine residue at position 371, located in the linker helix. Phosphorylation enhances the ligase activity and removing the phosphate drastically reduces it. Defects or mutations in Cbl, particularly those involving Tyr-371, are related to human myeloid disorders. In fact, this residue is frequently mutated in people with certain forms of leukemia. Obviously, it would be really interesting to find out how phosphorylation of Tyr-371 controls the enzyme’s activity.
The Huang laboratory in Glasgow has recently determined the structure of Cbl with and without substrates. They have also determined the structures of Cbl in both the de-phosphorylated and phosphorylated forms, shedding light on how the phosphorylation switch activates Cbl and influences the substrate-binding sites (ref. 1).

Moving the RING

In de-phosphorylated Cbl (PDB 2y1m), the RING domain is packed against the TKBD, with its E2-binding surface buried, thus preventing E2 from binding. This state (view-3) is therefore an auto-inhibited conformation. Binding of a substrate peptide to the opposite face of the TKBD induces a slight movement of the linker helix (PDB 2y1n), establishing a new hydrophobic interaction between the linker helix and the TKBD that shifts the RING domain into a half-open conformation (view-4). In this conformation, E2 is able to bind to the RING, but it competes with the TKBD for binding. Tyr-371 and Tyr-368, residues located in the linker helix, are buried in the TKBD, thereby locking the linker helix to the TKBD (view-5). Therefore, when Cbl is de-phosphorylated and substrate is bound, the RING domain fluctuates between a closed, auto-inhibited conformation and an open conformation. Though this open state may be catalytically competent, the E2 and the substrate are still on opposite sides of Cbl and the enzyme has a very low catalytic rate. This makes the de-phosphorylated Cbl a poor enzyme, but allows it to act as an ideal adaptor.

RING flipping

Upon phosphorylation, Tyr-371 can no longer be accommodated by the hydrophobic pocket on the TKBD, and therefore the linker helix is unlocked from the TKBD. This movement fully exposes the E2-binding face of the RING domain, enabling it to bind E2 without competition. Thus, phosphorylation of Tyr-371 enhances ligase activity by abolishing the auto-inhibition of Cbl.
But phosphorylation doesn't just free up the RING domain, it also induces huge conformational changes that flip the linker helix 180° compared with the de-phosphorylated state (PDB 4a4c), (view-6). As a result, the RING domain and its bound E2 move to the same side of Cbl as the substrate peptide. This shortens the distance between the active site of the E2, where ubiquitin is attached, and the TKBD-substrate-binding site to 28Å. This is still a large distance, but in this structure only a small fragment of the substrate is bound; the real substrate is a protein of 619 amino acids.
The structures show that phophorylation of Tyr-371 enhances Cbl’s enzymatic activity both by preventing the RING domain from adopting its auto-inhibited state and by moving it into a much more favourable position for ubiquitin transfer.
Mutations or deletions in the linker helix and RING domains would affect the activity of Cbl and potentially provide a starting point for the design of new anti-cancer therapeutics.

Acknowledgement

This Quips article was developed in collaboration with Hao Dou and Danny Huang at the Beatson Institute for Cancer Research.