NEDD8 is a small polypeptide closely related to ubiquitin. Like ubiquitin, NEDD8 is conjugated to other proteins through E1 activating and E2 conjugating enzymes. In contrast to ubiquitination, conjugation of NEDD8 to its only known targets, the cullin proteins, does not target them for degredation. Rather, cullins are part of SCF ubiquitin-ligases. Conjugation of NEDD8 to the cullin subunit stabilizes these complexes, which then mediate the ubiquitination of a subset of ubiquitin targets, leading to their degredation. We identified a temperature-sensitive allele of the C. elegans UBA3 subunit of the E1 activating enzyme for NEDD8, which we named
rfl-1 (ruffle-1), based on its early embryonic phenotype. While in other systems the Nedd8 pathway has mainly been implicated in regulating cell cycle progression, C. elegans embryos lacking RFL-1 function show defects in microfilament and microtubule-related processes as well as interphase delays. Most interestingly, the Nedd8 pathway seems to be required to negatively regulate microfilament contractility:
rfl-1(-) embryos initiate ectopic cleavage furrows during cytokinesis and also show extensive cortical contractility during pronuclear migration. Additionally, rotation of the first mitotic spindle is often delayed or absent in
rfl-1(-) embryos and about 50% of the embryos exhibit a late cytokinesis defect. Using immunocytochemistry and tubulin::GFP labeling of the mitotic spindle, we were able to determine that
rfl-1(-) embryos have unstable mitotic spindles: astral microtubules are shorter and the central spindle is displaced towards one pole of the mitotic spindle in about 50% of the embryos. This central spindle mislocalization always co-incites with the cytokinesis defect, indicating that the cytokinesis failure in
rfl-1 mutants is due to microtubule instabilities, rather than reflecting a direct role for the Nedd8 pathway in the completion of cytokinesis. Similar spindle defects have been previously observed in mutant embryos that mislocalize the microtubule severing complex katanin, consisting of the two proteins MEI-1 and MEI-2, to the mitotic spindle. We were able to show that
rfl-1 mutants also ectopically localize MEI-1 to the mitotic spindle, suggesting that the microtubule severing activity of katanin is accountable for the observed spindle defects. Moreover,
mei-1(RNAi) was able to rescue the spindle orientation and cytokinesis defects observed in
rfl-1 mutants, but failed to rescue some aspects of the ectopic cortical activity and the cell cycle delay. Finally, we were able to show by RNAi that all other previously identified pathway components, including the C. elegans cullin
cul-3, exhibit the same embryonic defects than
rfl-1, indicating that the pathway is intact in the early embryo. As inactivation of
mei-1 only partially rescues the phenotype associated with the
rfl-1 mutation, we propose that the Nedd8 pathway serves multiple purposes in the early embryo. Aside from MEI-1/MEI-2, we also expect proteins that are required for negative regulation of cortical contractility and for cell cycle progression to be regulated in a Nedd8 dependent manner.