Just a Trim, Please: Refining ER Degradation through Deubiquitination

Abstract

ER-associated degradation clears the secretory pathway of misfolded proteins and mediates the regulated degradation of some ER resident proteins. Only a minor increase in the interaction between a protein and a ubiquitin ligase is sufficient to signal substrate degradation. Zhang et al. have identified deubiquitination as a signal amplifier.

The decision to destroy misfolded proteins in the cell is not made lightly, as there is always the hope that proteins having transitional conformations may simply be en route to their native structures. This is especially true in the secretory pathway because soluble misfolded substrates are recognized in the lumen of the endoplasmic reticulum (ER) and then must be exported into the cytoplasm, where they are destroyed via the ubiquitin-proteasome pathway. This process is known as ER-associated degradation (ERAD). The ERAD of integral membrane proteins presents a special challenge, as membrane-spanning domains must be liberated from the lipid bilayer before the protein is threaded into the 26S proteasome. Genetic and in vitro analyses have delineated the varied pathways taken during the degradation of membrane proteins, with the spotlight directed at E3 ligases that append ubiquitin onto a proteasome-targeted substrate. In this issue of Cell, Hegde and colleagues redirect the spotlight toward an opposing reaction, the processive removal of the polyubiquitin chain, which amplifies subtle differences in E3-client interactions to generate a polyubiquitin chain that is sufficient for proteasome-mediated degradation (Zhang et al., 2013).

Mammals encode > 600 E3s, so one might envision that each E3 recognizes a misfolded conformation adopted by a subset of the proteome (Varshavsky, 2012). In turn, each protein might be identified by a select group of E3s. Indeed, functional redundancy among E3-client interactions is frequently observed. However, due to complexities inherent in the folding pathway, a protein displays a range of misfolded conformations. Moreover, previous studies uncovered relatively minor differences in the recognition of an ERAD substrate versus its wild-type counterpart by an E3 ubiquitin ligase (Gardner et al., 2001; Ishikura et al., 2010; Meacham et al., 2001). How are these differences magnified to ensure that folding-competent proteins do not fall victim to the ubiquitin-proteasome system or do so rarely?

To address this question, Zhang et al. co-opted an HIV product, Vpu, which when phosphorylated recruits an E3 ligase (SCF^betaTrCP) to catalyze CD4 degradation in the ER. CD4 is the receptor that HIV uses to enter host cells, and its destruction prevents superinfection and facilitates maturation of a viral coat protein (Nomaguchi et al., 2008). The authors showed first that the transmembrane domain (TMD) and the cytosolic region of CD4 and Vpu associate. When assessing how mutations in the TMD affect CD4’s fate in transfected cells, even a modest (~30%) reduction in interaction efficiency resulted in substantial effects on substrate ubiquitination and degradation, which is in accord with previous studies (see above). To better understand the nature of this phenomenon, a “mini-CD4” containing only the TMD and a defined cytosolic region was constructed. The substrate was then reconstituted into liposomes in the presence or absence of Vpu. When essential purified components (E1 ubiquitin-activating enzyme, E2 ubiquitin-conjugating enzyme, the E3, ubiquitin) and ATP were added, Vpu- and phosphorylation-dependent polyubiquitination of mini-CD4 was evident. Unexpectedly and in contrast to the results obtained in cells, there was little change in substrate ubiquitination when the TMD was mutated to compromise mini-CD4 association with Vpu. Thus, the reconstituted system lacked a factor that is vital for substrate discrimination.

The identity of the missing factor became apparent when the polyubiquitin chain in a microsome-based CD4 expression system was examined. Zhang et al. discovered that the longest polyubiquitinated chains were enriched in these vesicles compared to the reconstituted system, suggesting that the chains had been processively trimmed. One class of factors that copurify with microsomes and are absent from the proteoliposomes is deubiquitinating enzymes (DUBs); a few DUBs are ER associated and/or have been implicated in ERAD (see, for example, Sanyal et al., 2012; Sowa et al., 2009). When the catalytic domain of a DUB was titrated into the reconstituted system that contained Vpu and either wild-type or TMD-mutated mini-CD4, the magnitude of substrate discrimination rose. Because DUB activity decreases the dwell time of a polyubiquitinated substrate on an E3 and assuming that an E3 continues to display even a slight preference for one substrate over another, repeated cycles of ubiquitination and ubiquitin chain trimming may be needed to deliver ERAD substrates to the proteasome yet spare a protein that binds the E3s with somewhat lower affinity (Figure 1). This hypothesis was supported when data from a mathematical model were analyzed.

Figure 1. Substrate Discrimination during the Vpu-Dependent ERAD of CD4.

(A) After phosphorylation by casein kinase 2, Vpu recruits the SCF^betaTrCP ubiquitin ligase (“E3”), which ubiquitinates CD4 on cytoplasm-resident Lys side chains. Based on the strong interaction between Vpu and CD4 (denoted by the three lines within the membrane), repeated cycles of ubiquitin addition and removal occur, and the polyubiquitin chain is maintained. Ultimately, CD4 is degraded by the proteasome. (B) An ~30% reduction in Vpu-CD4 interaction (denoted by two lines within the membrane and experimentally determined using the CD4-M1 mutant) also results in substrate ubiquitination, although the length of the polyubiquitin chain is somewhat shorter. In this case, after cycles of ubiquitin addition and removal, the chain may be depleted, which prevents proteasome-mediated degradation.

The model developed by Hegde and colleagues will prompt a search for substrate-specific, ERAD-requiring DUBs, especially as a relatively small number of these enzymes have been characterized. It will also be critical in future efforts to determine whether DUB activity is sufficient for discrimination or whether the myriad domains that reside in DUBs are needed to augment ERAD substrate binding and discrimination (Reyes-Turcu et al., 2009). DUB activity may even regulate components of the ERAD machinery. In addition, the authors have used a substrate whose degradation is most likely chaperone independent, but the discrimination of most ERAD substrates is aided by molecular chaperones that deliver unfolded proteins to E3s (Vembar and Brodsky, 2008). Furthermore, a future challenge will be to determine whether the DUB-dependent discrimination model holds for other substrates, such as the largest class of ERAD substrates, i.e., those that are misfolded. Finally, the retrotranslocation, ubiquitination, and degradation of soluble substrates are tightly coupled, so repeated cycles of ubiquitin addition and removal may be absent. Nevertheless, the reconstitution of regulated CD4 ubiquitination provides an elegant system to examine the interplay between E3s and DUBs, a pursuit of great significance as many E3s associate with DUBs. Together, the DUBs are clearly important actors in ERAD substrate discrimination, and their time to share the spotlight with E3s has arrived.