O-GlcNAc glycosylation: a signal for the nuclear transport of cytosolic proteins?

Abstract

Year 2004 marks the 20th anniversary of the discovery of O-linked N-acetylglucosamine (O-GlcNAc) by Gerald W. Hart. Despite interest for O-GlcNAc, the functions played by this single monosaccharide remain poorly understood, though numerous roles have been suggested, among which is the involvement of O-GlcNAc in the nuclear transport of cytosolic proteins. This idea was first sustained by studies on bovine serum albumin that showed that the protein could be actively carried to the nucleus when it was modified with sugars. In this paper, we will review data on this puzzling problem. We will first describe the well-established nuclear localisation signal (NLS)-dependent nuclear transport by presenting the different factors involved, and then, we will examine where and how O-GlcNAc could be involved in nuclear transport. Whereas it has been suggested that O-GlcNAc could interfere at two levels in the nuclear transport both by modifying proteins to be translocated to the nucleus and by modifying the nucleoporins of the nuclear pore complex, according to us, this second idea seems unlikely. Part of this study will also be dedicated to a relatively new concept in the nuclear transport: the role of the 70-kDa heat shock proteins (HSP70). The action of the chaperone in nuclear translocation was put forward 10 years ago, but new findings suggest that this mechanism could be linked to O-GlcNAc glycosylation.

Introduction

O-Linked N-acetylglucosamine (O-GlcNAc) glycosylation is the modification of proteins by a single residue of N-acetylglucosamine on serine and threonine groups. Usually, this monosaccharide is neither epimerized nor elongated. O-GlcNAc was first evidenced on the surface of intact lymphocytes using bovine milk galactosyltransferase as an impermeant probe (Torres & Hart, 1984). After much investigation, it appeared that the finding of O-GlcNAc constituted a major breakthrough in glycobiology, and that it challenged two dogmas in our understanding of glycosylation: for the first time, a glycosylation type was found to have a high content in the cytosolic and nuclear compartments of eukaryota, whereas, until that date, glycosylation had been confined in the lumen of the endoplasmic reticulum and Golgi apparatus or to membranous proteins and secretory pathways; and secondly, the versatility of O-GlcNAc demonstrated that a glycosylation could not be always static (like N-glycans or classical O-glycans). As was mentioned above, one of the main features of O-GlcNAc is its versatility, since it is often compared to phosphorylation (Hart et al., 1995; Whelan & Hart, 2003). Like phosphorylation–dephosphorylation processes, that are regulated by a set of kinases and phosphatases, the glyco–deglyco process is also controlled by two cytosolic and nuclear distributed enzymes (Iyer & Hart, 2003a), i.e. the O-GlcNAc transferase (OGT) that transfers the monosaccharide from UDP–GlcNAc to the protein (Iyer & Hart, 2003b) and glucosaminidase that hydrolyses the sugar (Wells et al., 2002). Moreover, O-GlcNAc could compete with phosphorylation at the same site or in the vicinity, causing mutual exclusion (Comer & Hart, 2001; Slawson & Hart, 2003). O-GlcNAc is found in the cytosolic and nuclear compartments on a high number of proteins that are part of structural proteins (Cieniewski-Bernard et al., 2004); the transcriptional machinery including transcription factors as Sp1 (Yang et al., 2001) and Stat5 (Gewinner et al., 2004), RNA polymerase II (Comer & Hart, 2001); heat shock proteins (Guinez, Lemoine, Michalski, & Lefebvre, 2004; Walgren, Vincent, Schey, & Buse, 2003).

Despite its abundance, functions played by glycosylation remain unclear even though many roles have been put forward. Among them, O-GlcNAc could have a crucial importance in the transcriptional activity (Comer & Hart, 2001; Yang et al., 2001), in enzymatic activity (Cieniewski-Bernard et al., 2004), in the protection against proteasome (Guinez et al., 2004, Zachara et al., 2004) and in intracellular transport, especially in nuclear transport, since a few years ago, it was potentially presented as an alternate nuclear transport signal to the traditional nuclear localisation signal (NLS) controlled by O-GlcNAc (Duverger, Roche, & Monsigny, 1996).