Orientation of cell division plays a critical role in the development of both plants and animals. In early C. elegans embryos, two basic patterns of division occur. The AB lineage displays a typical orthogonal pattern of cleavage. In the P lineage, divisions repeatedly occur on the same axis due to a 90 degree rotation of the nuclear-centrosome complex. In the P1 cell, rotation depends on an interaction between astral microtubules and a site on the anterior cortex. We are studying maternal effect lethal mutations that alter spindle orientation in early embryos. In most embryos from animals homozygous for mutations in
ooc-5 (for abnormal oocyte formation), the P1 nucleus migrates to the anterior cortex but fails to rotate. This results in P1 dividing transverse instead of along the anterior-posterior axis. As an initial step in understanding the
ooc-5 phenotype, we have compared the distribution of cytoskeletal elements in wild-type and
ooc-5 mutant embryos. In the P1 cell of wild-type embryos, both actin and CP accumulate in an anterior cortical dot which correlates strongly with nuclear rotation (Waddle et al., 1994, Devel. 120, 2317-2328). The percentage of
ooc-5 embryos with a cortical accumulation of CP in the P1 cell is comparable to what we observe in wild-type embryos. These results suggest that the failure of rotation is not due to a lack of CP accumulation. In addition, the microtubule cytoskeleton appears normal in
ooc-5 embryos as observed using standard immunofluorescence microscopy. Specifically, astral microtubules are seen projecting toward the cell periphery and appear long enough to interact with the cortex, suggesting that failure of rotation is not caused by microtubules being too short. Early C. elegans embryos are highly polarized. To determine whether
ooc-5 mutations disrupt this polarity, we examined
ooc-5 embryos for both P granule and PAR-3 localization. In one and two-cell embryos, P granules are localized to the posterior pole of
ooc-5 embryos, just as in wild type. In the majority of 4-cell
ooc-5 embryos examined however, P granules are seen in two of the four cells. This is presumably due to the failure of nuclear rotation resulting in P1 dividing transverse to the axis of P granule localization. The PAR-3protein is required for embryonic polarity and becomes localized to the anterior periphery of wild-type one-cell embryos (Etemad-Moghadam et al., 1995, Cell 83, 743-752). Our preliminary results show that PAR-3 protein is properly localized to the anterior periphery of one-cell
ooc-5 embryos. Together, these observations suggest that altered spindle orientation in
ooc-5 embryos is not a result of an overall polarity defect. Mutations in
ooc-5 are pleiotropic however, resulting in a germline defect in addition to disrupting spindle orientation in P1. In
ooc-5 mutant animals, the germline contains a double row of oocytes that are smaller than wild-type oocytes. Embryos produced by
ooc-5 mutants are approximately half the size of a wild-type embryo while their nuclei are approximately 80% wild-type size. This observation raises the possibility that failure of nuclear rotation in these embryos is due to the steric effect of having a near normal size nucleus or microtubule array in a small cell. We have examined other strains that produce small embryos to see if reduced embryo size generally correlates with a failure of nuclear rotation. Homozygous
ooc-3 mutants also have a double row of oocytes; the embryos produced are smaller than wild type and nuclear rotation fails to occur in the P1 cell. One interpretation of the
ooc-5 and
ooc-3 phenotypes is that the reduced distance between centrosomes and the anterior cortex prevents nuclear rotation, for example because microtubules from both asters make stabilizing connections with the anterior cortical site. However, small embryos produced by
ceh-18 hermaphrodites can undergo nuclear rotation. In two
ceh-18 embryos examined, the distance from the center of the nucleus to the anterior cortex was shorter than in several
ooc-5 and
ooc-3 embryos. In addition, we have observed several cases in wild-type embryos where the P1 nucleus moves up against the anterior cortex before rotating. These data suggest that lack of rotation in
ooc-5 and
ooc-3 embryos is not a result of having a shorter distance between the centrosomes and the anterior cortical site. However, we can not rule out other steric effects since
ceh-18 and wild-type embryos are larger than
ooc-5 embryos in other dimensions. We are beginning a molecular analysis of the
ooc-5 gene to gain more insight into its role in germ line formation and spindle orientation.