Generation of asymmetry in the 1-cell embryo of C. elegans establishes the anterior posterior axis and is necessary for the position and orientation of mitotic spindles. Conserved PAR proteins are asymmetrically distributed and are required for these differences. However, how PAR asymmetry is established and how this polarity is translated into spindle position is still an open question. We have identified a series of components involved in these events: the conserved small G protein CDC-42 plays a role in generating PAR asymmetry and heterotrimeric G proteins in conjunction with AGS-3 proteins transduce polarity information to the mitotic spindle. Cdc42 is a highly conserved small G protein that has been shown to be important for cell polarity during mating and budding in yeast, and for establishing and maintaining epithelial polarity in mammalian cells. We found that
cdc-42(RNAi) embryos have a phenotype nearly identical to
par-3,
par-6, or
pkc-3 mutants, and asymmetric localisation of these and other PAR proteins is lost. CDC-42 is required to maintain, but not to establish the asymmetric PAR domains. The anterior actin cap is lost in
cdc-42(RNAi) embryos as in
par-2 and
par-3 mutants, suggesting that CDC-42 may mediate a transient reorganisation of actin at the anterior which is important to localise factors required for polarity and spindle orientation. Surprisingly, reduction of
cdc-42 function supresses the
par-2 mutant phenotype indicating that CDC-42 and PAR-2 act antagonistically. We further showed that CDC-42 binds PAR-6 in a two-hybrid assay. One possibility consistent with our data and related mammalian data is that CDC-42 activates the PAR-3/PAR-6/PKC-3 complex, and activation is required for its proper localisation. How are these early polarity signals translated into correct spindle position and orientation? GPB-1, a G b subunit of heterotrimeric G proteins, was previously shown to be required for orientation of early cell divisions in C. elegans embryos, but its mechanism of action was unknown. We have shown that two G a subunits (GOA-1 and GPA-16) function redundantly in the early embryo with GPB-1, but that G a and G b control distinct microtubule dependent processes. G b is important in regulating centrosome migration around the nucleus and hence in orienting the mitotic spindle. G a is required for asymmetric spindle positioning in the 1-cell embryo. Although loss of G a results in a symmtetric first cleavage, embryonic polarity appears to be normal. The uncoupling of spindle asymmetry from general polarity suggests that G a might be responsible for translating embryonic polarity from the PAR proteins into mitotic spindle positioning. The link between PAR polarity and the heterotrimeric G proteins may involve homologues of AGS3, a mammalian activator of G protein signaling. AGS3 contains GoLoco domains implicated in binding G a . C .elegans contains three AGS3 homologues, called
ags-3.1,
ags-3.2, and
ags-3.3. Embryos simultaneously depleted for
ags-3.2 and
ags-3.3 display a phenotype identical to G a depleted embryos suggesting that they are regulators of this heterotrimeric G protein. Interestingly, a Drosophila homologue of AGS3, PINS (Partner of Inscuteable), is essential for polarity and spindle orientation in the neuroblast and is found in a complex with Inscuteable, Bazooka (the Drosophila PAR-3 homologue) and a G a . We suggest that in C. elegans an equivalent complex exists which contains PAR-3, G a and AGS-3.2 or AGS-3.3 and that this complex translates polarity into spindle position. We are currently investigating this possibility. We propose a model in which CDC-42 controls the localisation and activity of PAR proteins. PAR proteins are in turn responsible for the activatation of the heterotrimeric G protein via the AGS-3 proteins, and thus for controling spindle orientation and position.