Membrane budding, remodeling and repair events within the cell are often driven by the proteins of the endosomal complex required for transport (ESCRT) machineries (ESCRT-0, I, II, and III). Tumor susceptibility gene 101 (Tsg101) is a member of the ESCRT-I complex, where it functions as both an adaptor to other ESCRT members and cargo sensor. This is accomplished by its modularity, with an N-terminal ubiquitin E2 variant (UEV) domain followed by a proline-rich region, a coiled-coil domain, and a C-terminal steadiness box. The Tsg101 UEV is structurally homologous to E2s, which serve as the intermediate carrier in the ubiquitylation pathway. Unlike E2s, however, Tsg101 UEV lacks the cysteine necessary for covalent attachment of ubiquitin (Ub), instead featuring a unique Ub interface with an extended β-hairpin and a peptide binding pocket that recognizes PT/SAP motifs in host proteins and which is exploited for retroviral egress. Despite this, we have found Tsg101 still interacts with a subset of E3 ligases, as a protective chaperone that prevents ligase degradation. To learn more about how these differences have arisen between E2 enzymes and the Tsg101 UEV, we have solved the structure of the closest homologue to the Tsg101 UEV domain, UEVLD, using solution NMR and pseudocontact shift (PCS), residual dipolar coupling (RDC), and NOE restraints. The resulting structure is nearly identical to Tsg101 UEV yet fails to bind known Tsg101 interactors. We then back-designed mutants to reveal key determinants in Tsg101 UEV that are being kept inactive in UEVLD allowing us to better understand the mechanisms of Ub and PT/SAP binding in Tsg101. The structural similarity of the two proteins without functional concurrence raises questions regarding additional, shared interactions across these proteins, both within the ESCRT context and beyond.
Nyenhuis et al. (Sun,) studied this question.