Examination Date

4-23-2014

Degree

Dissertation

Degree Program

Biochemistry and Molecular Biology

Examination Committee

Arthur Haas, PhD; Shyamal Desai, PhD; Suresh Alahari, PhD, FAAS, FASCB; David Worthylake, PhD; Andrew Hollenbach, PhD

Abstract

Ubiquitin-dependent regulation has remarkably pleiotropic effects on cell metabolism and homeostasis though its ability to modify protein function, to target protein degradation via the 26S proteasome, or to direct endocytic trafficking. The specificity and consequence of these roles are mediated by a large subset of cellular ubiquitin ligases that append monoubiquitin or specifically linked polyubiquitin signals to specific cellular target proteins. Among these enzymes, the relatively small superfamily of Hect (Homologous to E6-associated Carboxyl Terminus) ligases serve critical functions in cell maintenance; however, the mechanism(s) by which these conjugating enzymes function remains poorly understood and controversial. In the present work, we exploited the ability of HECT domain ligases to form 125 I-polyubiquitin chains in the absence of their cognate substrates as a functional readout of enzyme activity in order to probe the mechanism of the related Nedd4.1 and Nedd4.2 HECT domain paralogs. Biochemically defined initial velocity assays under E3-limiting conditions and precisely determined concentrations of various recombinant E2 paralogs showed broad specificity for the E2 carrier proteins but relative selectivity for members of the Ubc4/5 clade. Thus, while most ligases characterized to date show restricted specificity for a single E2 family, neither Nedd4.1- nor Nedd4.2-catalyzed reactions are supported by a single cognate E2. We speculate that the broad E2 specificity is an inherent consequence of the HECT ligase mechanism, which proceeds directly through a high-energy HECT domain~ubiquitin thioester during target protein conjugation. Thus, while other ligase classes must involve direct interaction between E3-bound E2~ubiquitin thioester and target protein during isopeptide bond formation, the E2 is not present during the analogous step for the HECT ligases. Other rate studies demonstrate that both ligases assemble both unanchored (free) and anchored (conjugated) polyubiquitin chains linked through the Lys 63 moiety of ubiquitin. This clarifies the linkage specificity for Nedd4.1, which had previously been suggested to require Lys 48, thereby committing targeted substrates for 26S proteasome-dependent degradation.

Detailed kinetic assays revealed hyperbolic kinetics with respect to E2 concentrations in the nano molar range and substrate inhibition at E2 concentrations above lμM, consistent with an ordered mechanism comprising two functionally-distinct E2~ubiquitin thioester binding sites. Other studies indicated conservation of a two-step mechanism for polyubiquitin chain assembly involving initial formation of the HECT~ubiquitin thioester intermediate (Site 1) followed by polyubiquitin chain elongation (Site 2), consistent with empirical observation of functionally-distinct E2~ubiquitin thioester binding sites. Further studies of Nedd4.2 with a stable Ubc5BC85S-ubiquitin oxyester substrate analog showed competitive inhibition with respect to wild type Ubc5B~ubiquitin thioester, as did parallel studies with a Ubc5BC85A product analog. In an effort to identify the two E2~ubiquitin thioester binding sites, Ubc5BT98A and Ubc5BF62A point mutants were prepared to disrupt binding at the canonical E2 binding site identified first identified for E6AP/UbE3A [Huang et al., Proc. Natl. Acad. Sci. US.A. 286, 1321-1326 (1999)]. Recombinant Ubc5BT98A supported Nedd4.2-catalyzed 125I-polyubiquitin chain formation with kinetics identical to that of wild type Ubc5B, suggesting that the kinetically-derived Km values reflect binding at Site 1 during HECT domain~ubiquitin thioester formation. In contrast, Ubc5BF62A was unable to support 125 I-polyubiquitin chain elongation but formed the obligate HECT domain~125 I-ubiquitin thioester with rates qualitatively identical to wild type Ubc5B, demonstrating that the two steps can be functionally uncoupled. Additionally, Ubc5BF62A~ubiquitin thioester functioned as a non-competitive inhibitor of wild type Ubc5B~ubiquitin thioester. These observations collectively indicate a site that binds the initial E2~ubiquitin thioester for HECT domain~ubiquitin thioester formation (Site 1) and the canonical E2~ubiquitin thioester binding site that is required for chain elongation (Site 2) prior to stochastic transfer of the polyubiquitin chain to the target protein.

Recent studies based on a previously-discounted crystal structure for E6AP/UbE3A proposes that HECT ligases form a trimer in their catalytically active form [Ronchi et al., J Biol. Chem. 289, 1033-1048 (2014)]. Various studies indicate that intercalation of the absolutely conserved Phe 727 of E6AP/UbE3A into a conserved hydrophobic pocket of the adjacent subunit is critical for stabilizing the Hect ligase oligomer. Sequence analysis using Clustal W identified Phe 823 as the paralogous stabilizing residue for Nedd4.2. Recombinant GST-Nedd4.2F823A exhibited ablated rates of chain formation, with a 10-fold decrease in ~at and an increased Km for the Ubc5B~ubiquitin thioester binding that is consistent with the requirement for trimerization for catalytic activity. Trimer formation could also be blocked by the Phe 823 mimic Ac-Phe-NH2 (Ki= 20 mM) which presumably competes with Phe823 for binding to the adjacent hydrophobic pocket. Other mutation studies identified amino acids Glu 646 and Arg 604 as catalytic residues involved in HECT domain ~ubiquitin thioester formation for Nedd4.2.

Finally, a model assay was developed to examine the complete catalytic cycle ofNedd4.2- catalyzed target protein polyubiquitin chain assembly using an extrinsic GST-FLAG-ENaC r subunit target protein. In the presence of the extrinsic substrate, Nedd4.2 showed greater affinity for Ubc5B~ubiquitin thioester than in its absence (Km = 127 ± 49 nM versus Km = 8 ± 2 nM, respectively), indicating a two-substrate random addition mechanism.

The present studies define for the first time, the detailed mechanism of polyubiquitin chain assembly by the Nedd4 family of HECT domain ligases. Quantitative studies confirm conservation in the mechanism of polyubiquitin chain formation. among the HECT ligase family and reconciles the lack of E2 specificity during Nedd4-catalyzed polyubiquitin chain formation. Other studies identify additional conserved amino acid residues critical for catalytic activity and stabilization of the active Nedd4.2 trimer. Finally, development of a well-characterized model substrate provides the platform for further mechanistic characterization of Nedd4.2 and high throughput drug screens.

Download

Share

COinS