Type of ubiquitin modification in cell-surface transmembrane proteins
The efficient recognition of
proteins by the proteasome requires polyubiquitin chains at least four subunits long, which in most cases appear to be linked
via the Lys48 of ubiquitin [9]. It was suggested that
ubiquitylated plasma membrane proteins may escape recognition and subsequent
degradation by the proteasome, probably depending on the type of ubiquitin
chain they receive. Indeed, some yeast plasma membrane proteins appear to
display mono-ubiquitylation, whereas others are modified by the addition of
K63-linked di/tri-ubiquitin residues ([1, 2, 21, 50] for review).
The list of membrane proteins which
are post-translationaly modify with ubiquitin is continually increasing, and
currently comprises thirty three proteins, including transporters,
ABC-transporters, (list in [50]), and receptors. Data concerning
the ubiquitin profiles of six of these proteins are now available. Ste2p is
monoubiquitylated on several lysines (multi-monoubiquitin) [80], as are probably the galactose
permease, Gal2p, and the maltose permease [81-83]. Ste3p, Gap1p, the zinc transporter, Ztr1p, and Fur4p have been shown to
be modified by small chains of two to three ubiquitins, each attached to one,
two or more target lysine residues [71, 74, 84, 85 , 86 ]. The ubiquitylation pattern of Fur4p and Gap1p has been analyzed in
cells lacking the Doa4p ubiquitin isopeptidase (which have lower than normal
intracellular free ubiquitin concentrations [87], and which are therefore impaired
in ubiquitylation processes) and overproducing either wild-type or variant
ubiquitins incompetent for the formation of K29-, K48- or K63-linked ubiquitin
chains. Both transporters carry two target Lys residues [71, 72] that can accept up to two or three
ubiquitin residues, linked via the Lys63 residue of ubiquitin [84, 86]. For both transporters, the
addition of one ubiquitin to the two target lysines (multi-monoubiquitylation)
appears to be sufficient for some endocytosis to occur, but the formation of
Lys63-linked short ubiquitin chains is required for efficient internalization. Moreover, although the fusion of ubiquitin in-frame at the
N-terminus of the variant of Fur4p lacking the two target lysines restores
permease internalization, the rate of uptake is five times higher if ubiquitin
is fused in-frame to the wild-type permease also modified by adding a short
chain of di-ubiquitin to each of its target lysines [5]. Although mono-ubiquitylation is sufficient for Ste3p endocytosis, there
is evidence to suggest that multi-ubiquitylation also increases the rate of
internalization [74]. Whether Fur4p and Gap1p, two of the few known ubiquitylated
substrates carrying Lys63-linked ubiquitin residues, are representative of a
larger class of Rsp5p plasma membrane substrates, remains to be determined. A
recent proteomic study of ubiqutylated yeast proteins has shown that Lys63
chains are far more abundant than previously thought [88].
No comments