Application and Analysis of the GFP^u Family of Ubiquitin‐Proteasome System Reporters

Abstract

The relevance of the ubiquitin proteasome system (UPS) to disease and fundamental cellular processes has generated a demand for methods to monitor the activity of this system in living cells and organisms. Here we describe the GFP^u family of UPS reporters. These reporters are constitutively degraded, ubiquitin‐dependent proteasome substrates that can be used to monitor UPS function in the living cell. The GFP^u reporter family consists of three variants that can report on global, nuclear, and cytoplasmic UPS function. This article focuses on the properties and design of these reporters and highlights appropriate techniques and applications for their use.

Introduction

Three predominant methods are used to assess UPS function in cells, tissues, or whole animals. Fluorogenic peptidase assays are widely used and permit rapid assessment of 20S or 26S proteasome catalytic activity. Because these reporters are cell impermeant, they are suitable only for assessment of proteasome activity in cell lysates or tissue homogenates. Moreover, because cleavage of these fluorogenic peptides does not require ATP‐dependent unfolding, a rate‐limiting step in proteolysis (Thrower et al., 2000), or ubiquitin conjugation, they more closely resemble the products of the UPS than its substrates. Another alternative is to monitor the levels of a short‐lived endogenous substrate protein by immunoblotting or immunofluorescence microscopy. Signal‐to‐noise ratios can plague immunofluorescence analysis, and quantitative immunoblotting is complicated by the altered mobility of ubiquitylated species on SDS‐PAGE gels. These antibody‐based detection methods can be circumvented through the use of reporter protein fusions, but caution must be exercised to avoid dominant negative effects of overexpressed fusion proteins.

The use of luciferase or GFP fusions to a UPS‐specific degron, like GFP^u, overcomes many of the aforementioned limitations (Neefjes and Dantuma, 2004). They do not mimic or participate in any known cellular signaling or metabolic pathways, do not require cell or tissue disruption, and are true polypeptide substrates that require protein unfolding and ubiquitylation (with the exception of ornithine decarboxylase [ODC]–based reporters) before degradation. Exploitation of enzymatic or fluorescent detection helps to reduce the level of expression required to achieve a high signal‐to‐noise ratio.

UPS reporters now use a targeting element of the ubiquitin‐independent substrate ODC (Li et al., 1998), an uncleavable ubiquitin fusion (UFD) (Dantuma et al., 2000), a cleavable ubiquitin fusion that permits the creation of an N‐end rule substrate (Dantuma et al., 2000), and the CL1 degron (Bence 2001, Gilon 1998). The major difference among the non‐ODC degrons is the ubiquitin conjugation pathway traversed en route to the proteasome. These reporters are all constitutively degraded and are widely applicable UPS monitoring reagents.

GFP^u is a UPS reporter that relies on a 16 amino acid degron (CL1) fused to the carboxyl terminus of GFP (Fig. 1A,B). The CL1 degron was first identified in a yeast screen for sequences that destabilize β‐galactosidase in a Ubc6‐ and Ubc7‐dependent manner (Gilon et al., 1998). Structural predictions indicate that the peptide sequence can form an amphipathic helix, perhaps mimicking the endogenous Ubc6/7‐dependent degron of the yeast protein Matα (Gilon et al., 2000). Although GFP^u is degraded in mammalian cells in an ubiquitin‐dependent manner by the proteasome, it is not known whether GFP^u degradation is Ubc6/7 dependent. The failure of a dominant‐negative Ubc6 homolog to influence GFP^u degradation implies at the very least that Ubc6 is not the sole E2 enzyme responsible for GFP^u degradation (Lenk et al., 2002). Interestingly, to promote GFP^u accumulation with pharmacological proteasome inhibitors, the chymotryptic activity of the proteasome must be attenuated in excess of 70%. This is consistent with the observations made for the ubiquitin fusion GFP reporter (Masucci and Dantuma, 2000) and may reflect either a robust cellular reservoir of proteasome activity or the need to also inhibit the tryptic or peptidyl‐glutamyl peptide hydrolytic (PGPH) activities of the proteasome to impair proteolysis (most inhibitors show the highest affinity for the chymotryptic site and only affect the latter sites at high concentrations).

Supporting a role for ubiquitin in the degradation of GFP^u, the reporter coimmunoprecipitates with ubiquitin, and degradation is impaired by the expression of dominant‐negative K48R ubiquitin (unpublished data). The CL1 degron contains an internal lysine that may act as an ubiquitin acceptor, but the mutation of the CL1 lysine to a glutamate (CL1 K‐E) increases the half‐life from 30 min to approximately 1 h but does not alter the eventual proteasomal fate of the reporter (both in yeast and mammalian cells). GFP itself contains a number of lysine residues on the surface of its β‐barrel structure that may act as interchangeable ubiquitin acceptors similar to the multiple lysine residues on the yeast protein Sic1 (Petroski and Deshaies, 2003).

Additional CL1 amino acid changes also affect the properties of the degron. Mutation of the two histidine residues to alanine circumvents the Ubc6/7 enzymes and diverts ubiquitylation to an unidentified E2 pathway in yeast (Gilon et al., 2000). Our unpublished experiments have shown that the corresponding mutations in the GFP^u degron still yield an unstable UPS reporter protein in mammalian cells, although it is currently unknown whether these mutations alter the ubiquitylation pathway as in yeast. As more is understood about the degradation pathway of the CL1 degron and its sequence variants, it should be possible to further tailor reporters to the study of specific ubiquitylation pathways.

Recently, nuclear and cytoplasmic versions of the GFP^u reporters have been generated and characterized (Bennett et al., 2005). Both reporters contain tandem GFP molecules (to exceed the ∼60‐kDa diffusional limit of the nuclear pore complex) with a CL1 degron and either a nuclear localization sequence (NLS) or a nuclear export sequence (NES) (Fig. 1B,C). The nuclear and cytoplasmic GFP^u reporters (NLSGFP^u and NESGFP^u) are useful for the study of localized UPS insults that may affect the nuclear or cytoplasmic UPS pools differently. The nuclear and cytoplasmic GFP^u reporters, in combination with the original GFP^u and its sequence variants, comprise a powerful set of tools for UPS research.