Ubiquitin Fusion Technique and Related Methods

Abstract

The ubiquitin fusion technique, developed in 1986, is still the method of choice for producing a desired N‐terminal residue in a protein of interest in vivo. This technique is also used as a tool for protein expression. Over the past two decades, several otherwise unrelated methods were invented that have in common the use of ubiquitin fusions as a component of design. I describe the original ubiquitin fusion technique, its current applications, and other methods that use the properties of ubiquitin fusions.

Introduction

The ubiquitin (Ub) fusion technique was invented through experiments in which a segment of DNA encoding the 76‐residue Ub was joined, in‐frame, to DNA encoding Escherichia coli β‐galactosidase (βgal) (Bachmair 1986, Varshavsky 1996b, Varshavsky 2000). When the resulting protein fusion was expressed in the yeast Saccharomyces cerevisiae and detected by radiolabeling and immunoprecipitation with anti‐βgal antibody, only the moiety of βgal was observed, even if pulse‐labeling was close to the time (1–2 min) required for translation of the Ub‐βgal's open reading frame (ORF). It was found that the Ub moiety of the fusion was rapidly cleaved off after the last residue of Ub (Fig. 1) (Bachmair et al., 1986). The proteases involved are called deubiquitylating enzymes (DUBs) (Amerik 2004, Baker 1996, Gilchrist 1997, Hemelaar 2004, Pickart 2004, Verma 2002, Wilkinson 1998, Wilkinson 2000). A mammalian genome encodes at least 80 distinct DUBs that are specific for the Ub moiety. The in vivo cleavage of a Ub fusion at the Ub‐polypeptide junction is largely cotranslational (Johnsson 1994b, Turner 2000).

A note on terminology: ubiquitin whose C‐terminal (Gly‐76) carboxyl group is covalently linked to another compound is called the ubiquityl moiety, with derivative terms ubiquitylation and ubiquitylated. The acronym Ub refers to both free ubiquitin and the ubiquityl moiety. This nomenclature (Varshavsky 1997, Webb 1992), which brings Ub‐related terms in line with standard chemical terminology, has been adopted by most Ub researchers. Shorthand for “degradation signal” is “degron” (Dohmen 1994, Gardner 1999, Varshavsky 1991). Through the use of prefixes, subscripts, or superscripts, this acronym can be employed to denote, in a uniform and succinct way, different types of degradation signals. For example, “N‐degron” denotes one class of degradation signals recognized by the N‐end–rule pathway, specifically those in which an essential determinant is a substrate's destabilizing N‐terminal residue (Bachmair 1989, Rao 2001, Suzuki 1999, Varshavsky 1996b).

One physiological function of DUB‐mediated cleavage reactions (Fig. 1) is the excision of Ub from its natural DNA‐encoded precursors, either linear poly‐Ub (Finley et al., 1987) or Ub fusions to specific ribosomal proteins (Finley 1989, Redman 1989). Many DUBs that process linear Ub fusions can also cleave Ub off its branched, posttranslationally formed conjugates, in which Ub is conjugated either to itself, as in a branched poly‐Ub chain, or to other proteins. A Ub–protein conjugate usually is composed of a single poly‐Ub chain covalently linked to an internal Lys residue of a substrate protein. The ubiquitylated substrate is recognized (in part through its poly‐Ub chain) and processively degraded by the 26S proteasome, an ATP‐dependent multisubunit protease (Baumeister 1998, Rechsteiner 2005). For reviews of the Ub system, see Fang 2004, Hershko 2000, Hicke 2003, Petroski 2005, Pickart 2004.

The finding of a rapid in vivo deubiquitylation of Ub fusions (Fig. 1) led to the discovery of N‐end rule, a relation between the in vivo half‐life of a protein and the identity of its N‐terminal residue (Fig. 2) (Bachmair et al., 1986). First, it was shown that the cleavage of a Ub‐X‐polypeptide after the last residue of Ub takes place regardless of the identity of a junctional residue X, proline (Pro) being the single exception. By allowing a bypass of “normal” N‐terminal processing of a newly formed protein, this finding yielded an in vivo method for placing different residues at the N‐termini of otherwise identical proteins. It was found that the in vivo half‐lives of resulting test proteins were strongly dependent on the identities of their N‐terminal residues, a relation referred to as the N‐end rule (Fig. 2) (Bachmair et al., 1986). The underlying, universally present N‐end rule pathway has a variety of functions; their list continues to expand (Du 2002, Kwon 2002, Kwon 2003, Rao 2001, Turner 2000, Varshavsky 1996b, Varshavsky 2003, Yin 2004).

The Ub fusion technique (Fig. 1, Fig. 2) remains the method of choice for producing, in vivo, a desired N‐terminal residue in a protein of interest. The requirement for a “technique” to do so stems from a constraint imposed by the genetic code. All nascent proteins bear N‐terminal Met (formyl‐Met in prokaryotes). The known methionine aminopeptidases (MetAPs) that remove N‐terminal Met would do so if, and only if, a residue at position 2, to be made N‐terminal after cleavage, is small enough (Bradshaw 1998, Varshavsky 1996b). Specifically, MetAPs do not remove N‐terminal Met if it is followed by any of the 12 destabilizing residues in the yeast‐type N‐end rule (Fig. 2). The exception, in metazoans, is Cys, whose side chain is small enough to allow cleavage of the Met‐Cys bond by MetAPs. N‐terminal Cys is a destabilizing residue in mammals and (apparently) other multicellular eukaryotes, but a stabilizing residue in fungi such as S. cerevisiae (Gonda 1989, Kwon 2002). (All destabilizing residues, including Cys, can be made N‐terminal through cleavages by other intracellular proteases, such as separases, caspases, and calpains, which act, in this capacity, as upstream components of the N‐end rule pathway.) The Ub‐specific DUB proteases are free of constraints imposed by the preceding property of MetAPs, except when the residue X of a Ub‐X polypeptide is Pro, in which case the cleavage still takes place but at a much lower rate (Bachmair 1986, Johnson 1992, Johnson 1995). However, there also exists a DUB that can efficiently cleave at the Ub‐Pro junction (Gilchrist et al., 1997).

Ub fusions can be deubiquitylated in vitro as well (Baker 1996, Catanzariti 2004, Gonda 1989). High activity and specificity of DUBs make them reagents of choice for applications that involve, for example, the removal of affinity tags from overexpressed and purified proteins. A particularly efficacious version of the Ub fusion technique for high‐level production and easy purification of recombinant proteins expressed in E. coli was described by R. Baker and colleagues (Fig. 3) (Baker 2005, Catanzariti 2004).

Yet another advantage of the Ub fusion technique stems from the finding that expression of a protein as a Ub fusion can dramatically augment protein's yield (Baker 1994, Butt 1989, Ecker 1989, Mak 1989). The yield‐enhancement effect of Ub was observed with short peptides as well (Pilon 1997, Yoo 1989). This and other applications of Ub fusions are described in the following, with references to original articles and specific constructs.