Cellular processes in which sumoylation is involved include cellular trafficking, channel and receptor regulation, regulation of transcription-factor activity, DNA repair and replication, chromosome dynamics, mRNA processing and metabolism, cellular replication, and cross-talk with ubiquitination. The mechanism of SUMO attachment resembles other ubiquitin-like conjugation pathways. Briefly, mature SUMO is first activated by a heterodimeric SUMOactivating enzyme, E1, before passing to the SUMO-conjugating enzyme, E2. Only one E2 appears to exist in most well studied organisms including human, yeast, rat, and mouse. Unlike with ubiquitination, sumoylation may proceed in an E3-independent manner. This notion is based on the observation that binding of the E2 Ubc9 to the consensus sequence Y-K-X-E present in a target protein is sufficient for sumoylation. Furthermore, grafting of this consensus sequence to a protein not normally sumoylated will result in its sumoylation. Given the apparent E3-independent nature of sumoylation, the existence of SUMO E3 ligases was initially challenged, although evidence hinted at their existence. The involvement of E3 ligases in sumoylation has now been demonstrated. However, while an E3 can enhance target sumoylation, its role in substrate specificity and WZ8040 lysine selection remains debated. The crystal structure of SUMO-RanGAP1- Ubc9-Nup358 complex suggests the E3 merely aligns the E2- SUMO pair for optimal E2 binding and SUMO transfer without itself binding the target protein. Interactions between the target protein and E3 appear to augment efficiency, but sumoylation depends solely upon E2 binding. Furthermore, individual genetic knockout of the mammalian SUMO E3 ligases PIAS1, PIASy, and PIASx in mice does not affect global sumoylation patterns. Similarly in yeast, knockout of the E3 Siz2 does not affect global sumoylation, although the knockout of the E3 Siz1 attenuates robustness. Further studies in yeast examining sumoylation of individual proteins confirm this trend in overlapping E3 function. Differences in local concentrations rather than differences in target recognition may be the mechanism whereby E3 specificity is manifested in vivo but is absent in vitro. Importantly, SUMO E3 ligases are not dispensable in the cellular context as the knockout of every E3 is lethal. Furthermore, emerging evidence suggests that the E3 may play a role in target specificity. Several proteins are modified at nonconsensus sequences and an E3 ligase, not an E2, may be responsible for this modification. For example, Siz1 is required for sumoylation of PCNA’s nonconensus K164 site. Several studies have confirmed that the PINIT domain of the E3 is solely responsible for this K164 lysine specificity. Further, E3s tend to bias the particular SUMO isoform that is attached to the target protein. Several groups have reconstituted E3-independent sumoylation cascades in E. coli. These sumo-engineered E. coli systems have several advantages. First, endogenous levels of sumoylated protein in eukaryotic cells tend to be low.