Ubiquitylation is a post-translational modification in which one or more ubiquitin proteins are covalently attached to substrate proteins. This process is executed by a number of enzymes including ubiquitin-activating enzyme (E1), ubiquitin-conjugating enzyme (E2) and ubiquitin-protein ligase (E3). After modification, the substrate protein is frequently delivered to proteasomes for degradation. Upon release from the substrate protein, ubiquitin molecules may participate in further rounds of ubiquitylation.
In a review article published in the October 2001 issue of Trends in Plant Science (6: 463-470), Frank Eisenhaber and his associates (Research Institute of Molecular Pathology, Vienna, Austria, and University of Edinburgh, UK) describe a number of plant proteins known or predicted to comprise the ubiquitylation machinery and compare them with ubiquitylation components in Saccharomyces cerevisiae and animals.
The authors illustrate the various stages of ubiquitylation. This process begins when the ubiquitin-activating enzyme E1 covalently links the C-terminal Gly of ubiquitin to an E1 Cys residue, forming a thioester (SH) bond. A subsequent thioester bond forms when the activated ubiquitin is transferred to a Cys residue of ubiquitin-conjugating enzyme, E2. Ubiquitin protein ligase, E3, then facilitates transfer of ubiquitin from E2 to a Lys residue of the substrate. This occurs by two different mechanisms for two types of E3 enzymes, HECT-domain E3 enzymes and RING-finger-type E3 ligases. Additional ubiquitin units may be attached to the first, a process that may require an additional factor (E4).
The authors then describe Arabidopsis genes encoding components of the ubiquitin complex, starting with ubiquitin-coding genes themselves. Sixteen known Arabidopsis ubiquitin genes are listed, though the 82 additionalArabidopsis genes encoding ubiquitin-like proteins or proteins with ubiquitin-like domains are not considered. As in other organisms, ubiquitin genes in Arabidopsis encode precursors with multiple repeats of ubiquitin, frequently arranged in head-to-tail orientation. These precursors are processed by ubiquitin-specific proteases to release individual ubiquitin units. Some precursors are comprised of ubiquitin units fused to ribosomal proteins. Two Arabidopsis genes, UBQ7 and UBQ15, encode precursors in which a single ubiquitin unit is fused to the ubiquitin-like protein RUB. These two genes do not appear to have orthologs in the S. cerevisiaegenome.
The authors then consider genes encoding ubiquitin-activating enzyme. This enzyme (E1) does not play a regulatory role in the process. Thus, the two E1 isoforms reported in Arabidopsis are expected to have the same specificity. The domains architecture of the Arabidopsis E1 enzymes is similar to yeast E1 proteins. In animals, phosphorylation may regulate the distribution of E1 to the nucleus or cytoplasm. Ubiquitylation occurs in both compartments in animals.
About half of the Arabidopsis ubiquitin-conjugating enzymes (E2s) have been characterized biochemically and at the molecular level by Richard Vierstra and his colleagues (U. of Wisconsin, Madison). Ten of the 11 yeast E2 genes have a close homolog in Arabidopsis . However, an Arabidopsis ortholog could not be identified for the only E2 known to be essential in yeast, ScUBC3/CDC34. There are three additional families of E2 genes in Arabidopsis that do not correspond to orthologs in yeast, but do correspond to potential orthologs in animals, possibly indicating additional functional roles for ubiquitylation in metazoans that are not present in S. cerevisiae. Differential transcriptional regulation has been observed for some individual members of E2 gene families in Arabidopsis, as might be expected.
Several proteins have been identified in Arabidopsis that contain a ubiquitin conjugating enzyme motif but lack the conserved Cys essential for catalytic activity. Similar proteins are also found in animals and yeast. These proteins are called ubiquitin-enzyme variants (UEVs), and studies in yeast indicate that they are capable of binding ubiquitin when it is attached to certain substrates but not when it is attached to E1. One such protein in yeast, ScMMS2, can bind to a true E2 protein (ScUBC13), and this dimer can promote the linkage between two ubiquitin molecules through Lys63. Another UEV in yeast, ScVPS23, binds to monoubiquitylated membrane proteins during vacuolar protein sorting.
The authors then summarize what is known about ubiquitin-protein ligases (or E3 enzymes) in plants. All known ubiquitin-protein ligases are characterized by having either a HECT (Homologous to E6-AP Carboxyl Terminus) domain or a RING-finger domain. Both types of domains act as a docking site for E2, although there is a mechanistic difference between HECT and RING-finger E3s for transfer of ubiquitin. HECT domain E3 enzymes form a covalent thioester bond with ubiquitin at a Cys residue when ubiquitin is shifted from E2. RING-finger domain E3 enzymes, however, are not themselves ubiquitylated but rather bring the substrate protein and E2 in close proximity. The HECT domain is encoded seven times in the Arabidopsis genome, and there are over 300 regions encoding the RING-finger domain.
Two other types of multisubunit RING-finger domain enzymes are present in fungi, animals and plants. One is the anaphase promoting complex (APC), which has 12 or more subunits and is essential for cell cycle progression. The E3 present in this complex has not been well characterized in plants. The second type is the Skp1, Cullin and F-box (SKF) complex, named for several of the subunits present in the complex. As over 500 F-box encoding genes are present in Arabidopsis, it appears that plants use the SCF type ubiquitin-protein ligase extensively, and a number of mutant phenotypes have been described resulting from disruption of F-box containing genes in plants.
Various types of ubiquitin chain linkages have been described in other organisms, and these types of linkages are likely to be conserved in plants as well. Single ubiquitin linkages usually occur between Gly76 of ubiquitin and an internal Lys of the substrate. This usually targets the substrate for degradation, although some proteins (such as histone H2A or H2B) are actually stabilized by a single ubiquitin conjugation. Some substrates are ubiquitylated at the alpha-amino group of the N-terminus rather than an internal Lys. Multiubiquitylation occurs when additional ubiquitin molecules are attached to the first one. Linkages between the ubiquitin units occur through Gly76 and Lys48. Multiubiquitylated proteins are usually degraded by the proteosome, but at least one such substrate (ribosomal protein L28) is stable. Studies on L28 and the components involved in its ubiquitylation indicate that, in certain circumstances, ubiquitin modification may play a role in DNA repair. A third type of ubiquitin chain involves Gly76-Lys29 linkages. This chain type is also thought to target substrate proteins for degradation by the proteosome.
The authors summarize several proteins that have been shown to be targets for ubiquitylation in plants. The only plant substrate so far detected in ubiquitylated forms in vivo is oat phytochrome A. However, several degradation signals (degrons) of plant proteins have been reported recently and assigned to E3 pathways. Proteins that may be regulated by ubiquitylation include Aux/IAA transcription factors and the light-regulated transcription factor HY5 (which is acted upon by the RING-finger protein COP1). There is evidence that the tobacco mosaic virus (TMV) movement protein and coat proteins are substrates for ubiquitylation. Ubiquitylation of animal viral coat proteins also appears to be common, and it has been suggested that the ubiquityl moiety facilitates viral budding. Degrons-reporter gene fusions have been used to study the process of substrate degradation in yeast, and these tools are being used in plants as well.
Ubiquitylation is a tightly regulated process, and the authors describe mechanisms of regulation in plants, animals and yeast. Because of the number of gene family members present encoding ubiquitin machinery components in plants, the authors suggest that transcriptional regulation may play a role. Post-translational modification of ubiquitylation machinery components has been reported. For example, APC-type E3 ligases are phosphorylated, and SCF-type E3 ligases are modified by the ubiquitin-like protein RUB in animals and plants. These types of modifications are reversible and affect activity of the components. Changes in substrate molecules frequently influence ubiquitylation. In plants, phytochrome A undergoes a conformational change upon irradiation with red light that appears to expose degradation signals. Post-translational modification of substrates that are ubiquitylated at the N-terminus can lead to exposure of the bulky first amino acid. In animals, phosphorylation of the defense regulator I-kappa-B-alpha is a prerequisite for ubiquitylation. The kinase complex that performs this phosphorylation is itself activated by ubiquitylation, but is also a target for modification by the ubiquitin-like molecule SUMO. Which molecule modifies this kinase complex thus ultimately dictates the level of kinase activity.
The process of de-ubiquitylation has also been investigated. De-ubiquitylating enzymes are abundant in yeast, and have been divided into two structural classes: the ubiquitin C-terminal hydrolase (UCH) family and the ubiquitin protease (UBP) family. Functional homologs of both family types may exist in Arabidopsis and are being investigated at the genetic and biochemical levels. De-ubiquitylation enzymes may be important for regeneration of free ubiquitin for further rounds of ubiquitylation, and may possibly have other regulatory roles.
In conclusion, the authors point out that the proportion of plant genes devoted to the process of ubiquitylation outnumbers those present in animals. Furthermore, there is evidence that ubiquitin ligases (E3s) may have more diverse biological roles for plants than in animals. The authors emphasize the importance of further in planta studies to elucidate the biological functions and specificities of ubiquitylation machinery components.