August has only just begun, but I have already found one paper that deserves featuring in my ‘interesting papers‘ series. I will also talk about a second paper I found today, although I am not yet totally sure if it is really that much interesting.
- As I mentioned in a previous post, I feel attached to the UBA domain, a small but quite abundant ubiquitin recognition domain. This month’s edition of Molecular Cell features a paper from Kalle Gehring’s group, entitled “Structural Basis for Ubiquitin-Mediated Dimerization and Activation of the Ubiquitin Protein Ligase Cbl-b“. This paper describes a new trick of UBA domains, which expands their repertoire of operating modes. The classical view of UBA and other ubiquitin receptor domains is that they bind to ubiquitin, thus bringing their ‘host protein’ into contact with ubiquitin or ubiquinated proteins. A second mode of action was first described for the UIM motif, but might also apply to other ubiquitin binding domains: the prevention of poly-ubiquitin chain formation. A third aspect was brought into play by a paper I discussed in the July edition, the mediation of mono-ubiquitination independent of an E3 enzyme. Here is a new function: mediating protein heterodimerization, dependent on the presence of ubiquitin chains. I have not read the experimental section of the paper carefully enough to really judge if the authors’ interpretation of their findings is fully substantiated. In brief, they looked at the structure of the UBA domain of Cbl-b in contact with ubiquitin. As expected, the authors found ubiquitin bound to the UBA domain, but surprisingly using a non-conventional binding surface. All other UBA:Ubiquitin complexes studied so far show a ubiquitin binding surface formed by the loop between helices 1 and 2 and some residues from helix 3 of the UBA domain. In the present case, a different surface formed by hydrophobic residues of helix 1 is used. This binding mode leaves a second interaction surface formed by helices 2 and 3, which appears to support UBA dimerization. It should be noted that UBA dimerization has been described before, but has always been assumed to be mutually exclusive with ubiquitin binding. By contrast, the authors of this paper go on to show that the binding of the UBA domains to ubiquitin chains enhances the UBA dimerization tendency. Based on their structural work, they suggest a model where K48-linked tetraubiquitin forms a complex with a UBA dimer. According to this model, the 1st and 4th ubiquitin domains bind to the UBA dimer, using binding surfaces that are opposite of the UBA dimerization interface. This model, if true, would explain some of the known properties of Cbl-b and c-Cbl, e.g. the functional dependence on dimerization. It is interesting to speculate if this ubiquitin-dependent dimerization process might also be relevant for other UBA domains.
- The second paper of potential interest comes from Ron Kopito’s group, has been published in this week’s issue of Nature and is entitled “Global changes to the ubiquitin system in Huntington’s disease“. It has been known for a long time that Huntington’s disease and other neurodegenerative polyglutamine expansion diseases go along with the deposition of intranuclear inclusions, which are rich in ubiquitin (probably in the form of ubiquitinated proteins). The inclusion bodies often contain other components of the ubiquitin-proteasome system (UPS). One explanation for this observation is the overloading (or even breakdown) of the ubiquitin-dependent protein degradation pathway. The authors of the current paper use a mass-spec approach to quantify poly-ubiquitin chains from both human patients and animal models of Huntington’s disease. First, they show that the presence of non-degraded polyubiquitin chains correlates well with the functional status of the degradation pathway (poor degradation <=> more chains). They go on to show that in Huntington’s disease and the disease models there is a strong accumulation of K48-linked poly-ubiquitin chains. This was to be expected, as it is known that proteasomal function is impaired in this disease. More surprisingly, the authors also found an accumulation of K63 and K11-linked poly-ubiquitin chains. As these chains do not target proteins for proteasomal degradation, their accumulation cannot easily be explained by proteasomal dysfunction. The authors conclude that in Huntington’s disease, it is not only the proteasome that is impaired; they rather postulate that the ubiquitin system is affected in a more global way. Unfortunately, the authors do not address the level of mono-ubiquitination, which also might be different – they focus exclusively on poly-ubiquitin chains (which they isolate by affinity capture using an isolated UBA domain!)