A role for ubiquitin in the pathogenesis of human diseases was first suggested some two decades ago, from studies that localized the protein to intracellular protein aggregates, which are a feature of the major human neurodegenerative disorders. Although several different mechanisms have been proposed to connect impairment of the UPS (ubiquitin–proteasome system) to the presence of these ‘ubiquitin inclusions’ within diseased neurones, their significance in the disease process remains to be fully clarified. Ubiquitin inclusions also contain ubiquitin-binding proteins, such as the p62 protein [also known as SQSTM1 (sequestosome 1)], which non-covalently interacts with the ubiquitinated protein aggregates and may serve to mediate their autophagic clearance. p62 is a multifunctional protein and, in the context of bone-resorbing osteoclasts, is an important scaffold in the RANK [receptor activator of NF-κB (nuclear factor κB)]–NF-κB signalling pathway. Further, mutations affecting the UBA domain (ubiquitin-associated domain) of p62 are commonly found in patients with the skeletal disorder PDB (Paget's disease of bone). These mutations impair the ability of p62 to bind to ubiquitin and result in disordered osteoclast NF-κB signalling that may underlie the disease aetiology. Recent structural insights into the unusual mechanism of ubiquitin recognition by the p62 UBA domain have helped rationalize the mechanisms by which different PDB mutations exert their negative effects on ubiquitin binding by p62, as well as providing an indication of the ubiquitin-binding selectivity of p62 and, by extension, its normal biological functions.
Ubiquitin and neurodegenerative diseases
It is over 20 years since the first demonstrations of the localization of the small protein ubiquitin to intracellular protein aggregates or ‘inclusions’ (which are a feature of many of the major human neurodegenerative disorders) intimated a connection between ubiquitin and human disease [1,2]. In fact, it is now realized that ubiquitin inclusions characterize not only diseased brains, but also cells outside of the nervous system in a range of human disorders [2,3]. Within these inclusions the principal components appear to be forms of a single disease-related protein which have adopted non-native conformations; for example, the main component of Alzheimer's disease neurofibrillary tangles is (filamentous) deposits of the microtubule-associated protein tau, and of Parkinson's disease Lewy bodies is the pre-synaptic protein α-synuclein (reviewed in ). Ubiquitin can singly (mono-ubiquitination) or multiply (polyubiquitination) modify target proteins, the latter involving multiple copies of ubiquitin linked via isopeptide bonds involving different lysine residues of individual ubiquitins within the polyubiquitin chain. In general, polyubiquitin chains linked via Lys48 serve as signals for proteasomal degradation of the target protein, whereas those linked via Lys63 serve to regulate non-degradative processes, such as intracellular signal transduction pathways . That in vivo ubiquitinated tau includes forms that are modified with Lys48-linked polyubiquitin [6,7] is broadly supportive of the notion that an impairment of the UPS (ubiquitin–proteasome system) may be a feature of human neurodegenerative disorders, be they primary (causal) or secondary (in the pathological state). Certainly, genetic evidence, principally the identification of mutations that affect the activity of the ubiquitin pathway enzymes in familial cases of Parkinson's disease and parkinsonism , supports the notion that altered function of the UPS can directly cause neurodegeneration. Further, several different mechanisms have been proposed which could account for an age-dependent impairment of the UPS that may underlie disease progression in the more common sporadic forms of neurodegeneration, including: direct UPS impairment by protein aggregates ; decreases in proteasomal gene expression and/or neuronal ATP levels ; and dominant inhibition of the 26S proteasome by a frameshift mutant form of ubiquitin which arises via a process known as ‘molecular misreading’ . However, despite considerable efforts over the past two decades, it remains the case that the full significance of the presence of ubiquitin inclusions in diseased tissue, and indeed whether these structures are indicative of cellular dysfunction or in fact a protective cellular response to protein aggregation, remains to be clarified.
A role for the ubiquitin-binding protein p62 in neurodegenerative disorders
Modification of target proteins with ubiquitin is only the first stage of a ubiquitin-mediated process; often the ubiquitination ‘signal’ is then transduced by selective recognition of the ubiquitinated target by ubiquitin-binding proteins, which non-covalently interact with the ubiquitin tag . Perhaps unsurprisingly, inclusions within diseased neurones contain not only ubiquitinated (covalently modified) proteins but also ubiquitin-binding proteins, and one such ubiquitin-binding protein called p62 [also known as SQSTM1 (sequestosome 1)] appears to be a common feature of ubiquitin inclusions . p62 is a multifunctional ubiquitin-binding protein, which has been implicated in a variety of processes including cell signalling, receptor internalization and protein turnover . Quite why p62 accumulates within ubiquitin inclusions is unclear, although a recent role in mediating the autophagic clearance of ubiquitinated protein aggregates is of particular interest . Notably, there is some evidence that p62 may have a binding preference for Lys63-linked polyubiquitin chains in vivo , and the demonstration that Lys63-linked ubiquitination may promote the autophagic clearance of protein aggregates  is consistent with the notion that in some instances p62-positive ubiquitin inclusions could represent attempts by the cell to clear protein aggregates via the autophagic pathway.
p62 mutations and PDB (Paget's disease of bone)
In an entirely different context, p62 has also been shown to play an important role in a different human condition, the skeletal disorder PDB. PDB is a disorder of osteoclasts (bone-resorbing cells) that involves abnormal bone turnover at distinct sites throughout the skeleton . In 2002, mutations affecting the SQSTM1 gene, which encodes the p62 protein, were first identified in PDB patients [19,20] and over 20 different mutations have been described so far . Importantly, all of these mutations cluster within the C-terminus of p62, the region of the protein which contains an UBA domain (ubiquitin-associated domain; residues 387–436), which is the ubiquitin-binding module of p62. Truncating mutations remove most or all of the UBA domain, whereas missense mutations result in amino acid substitutions within or close to the UBA sequence .
Effects of p62 PDB mutations on protein and cellular function
Consistent with the mutations affecting the region of p62 which contains the UBA domain, the molecular defect in PDB with SQSTM1 mutations is an impaired ability of the p62 protein to bind to ubiquitin, i.e. disordered ubiquitin recognition [21,22]. The precise role of p62 within osteoclasts is unclear, although the protein appears to be a key scaffold within the osteoclast RANK [receptor activator of NF-κB (nuclear factor κB)]–NF-κB signalling pathway, which is required for osteoclastogenesis and osteoclast activity . Within this pathway, the ubiquitinated targets of p62 are not fully characterized, although p62 has been suggested to regulate the ubiquitination of other signalling proteins, such as TRAF-6 (tumour-necrosis-factor-receptor-associated factor-6)  and NEMO (NF-κB essential modulator) , to mediate specific protein–protein interactions required for signal transduction. Specifically, mutant forms of p62 (with impaired ubiquitin-binding function) appear to promote NF-κB-dependent cellular responses, perhaps by resulting in a loss of p62's ability to repress the RANK–NF-κB signalling pathway , with the net effect being an increased osteoclast bone-resorbing activity.
Mechanism of ubiquitin recognition by p62 and structural rationalization of the effects of PDB mutants
Studies investigating the mechanism of ubiquitin recognition by the p62 UBA domain have provided insights into the ubiquitin-binding specificity of p62, as well as a structural rationalization of the mechanism(s) by which different mutations exert their negative effects on ubiquitin binding. The structure of the unbound form of the p62 UBA domain was previously determined by protein NMR and found to adopt a three-helix bundle with an overall topology similar to other UBA domains, with a hydrophobic surface patch presumed to represent the ubiquitin-binding interface . However, more recently, we have extended these analyses to show that, unlike other UBA domains, the bound conformation of the p62 UBA domain (in complex with mono-ubiquitin) adopts a different conformation to the unbound, specifically involving repackaging of the three-helix bundle . This occurs via a slow-exchange structural reorganization and appears to optimize both hydrophobic and electrostatic surface complementarity with ubiquitin. Accurate mapping of the binding interface using chemical shift perturbation measurements allows the detrimental effects of some missense mutations on the ubiquitin-binding properties of the p62 UBA to be rationalized, as these affect a contact residue in the UBA domain (e.g. M404V and M404T). Other missense mutations may exert their effects by influencing UBA domain stability, and the intriguing possibility that a subset of mutations (e.g. G425R) may negatively influence ubiquitin binding by actually stabilizing the unbound UBA conformation and antagonizing the conformational rearrangement is also suggested from CD melting curves . Finally, some missense mutations that (i) are peripheral to the binding interface, (ii) do not affect UBA domain stability, and (iii) in in vitro assays affect ubiquitin binding of full-length p62 but not of the minimal UBA domain (e.g. P387L and P392L) [21,22] are more difficult to rationalize, but suggest the involvement of other UBA (and p62) sequences in the subtleties of the conformational rearrangement and ubiquitin recognition. NMR titrations using Lys48-linked di-ubiquitin also provide an indication of the binding specificity of the p62 UBA and, by extension, of ‘normal’ biological function . These experiments reveal that the p62 UBA binds individual ubiquitin moieties in the ubiquitin dimer in an identical manner to mono-ubiquitin, with no evidence for a second UBA-binding surface that would allow simultaneous recognition of both distal and proximal ubiquitin units in the Lys48-linked dimer. Further, the binding affinity of the p62 UBA for Lys48-linked di-ubiquitin was found to be approx. 4-fold lower than for mono-ubiquitin, reflecting competition between binding of individual ubiquitins in the dimer to the UBA and intramolecular self-association of di-ubiquitin to form a ‘closed’ state involving the same binding surfaces. This observation is consistent with the apparent in vivo preference of p62 for Lys63-linked polyubiquitin  (in these chains the individual ubiquitins are presented as ‘beads on a string’ and do not self-associate ) and p62's important roles in regulating cell signalling pathways mediated principally through Lys63-linked polyubiquitination .
Considerable inroads into understanding the function and dysfunction of ubiquitin-mediated processes in human health and disease have been made in recent years, and it is increasingly clear that the ubiquitin protein pervades almost all physiological and pathophysiological systems. An understanding, at the molecular level, of the mechanisms by which disrupted ubiquitin-mediated processes underlie the progression of many of the major human degenerative and chronic disorders is the first stage in the design of rational drug treatments to manipulate the activity of the UPS and other cellular pathways that rely on ubiquitin conjugation under these conditions.
Ubiquitin and Ubiquitin-Like Modification in Health and Disease: Biochemical Society Irish Area Section Annual Meeting held at Royal College of Surgeons in Ireland, Dublin, Ireland, 22–23 November 2007. Organized and Edited by Caroline Jefferies (Royal College of Surgeons in Ireland).
Our work is supported by the BBSRC (Biotechnology and Biological Sciences Research Council) and the National Association for the Relief of Paget's disease of bone.