In animals, female meiotic spindles mediate the expulsion of chromosomes into polar bodies to generate a haploid egg. Time-lapse imaging of fluorescent protein fusions has revealed that wild-type C. elegans meiotic spindles undergo discrete length changes during each division. Meiosis I spindles, for example, maintain a constant metaphase length of 7.7 um for several minutes. Meiosis I spindles then shorten until they reach 58% of their starting length. During this initial shortening phase, the average fluorescence intensity of GFP-tubulin doubles (suggesting an inward sliding mechanism) and the spindles remain parallel to the cortex. After this initial shortening phase, spindles rotate perpendicular to the cortex, then initiate anaphase chromosome separation as they proceed through a second phase of spindle shortening. During the second shortening phase, the average fluorescence intensity of GFP tubulin decreases at poles while increasing in the midzone, suggesting a redistribution of microtubules within the spindle. At the end of the second shortening phase, MI spindles are 2.8 um long (38% of starting length). MEI-1 and MEI-2 are the C. elegans orthologs of the 2 subunits of a conserved microtubule-severing ATPase called katanin. MEI-1 and MEI-2 are localized on meiotic spindles and purified complexes of MEI-1 and MEI-2 sever microtubules in vitro, indicating that MEI-1/MEI-2 may regulate microtubule length within the meiotic spindle. We have analyzed a partial loss of function allele of
mei-2 called
ct98.
mei-2(
ct98) meiosis I spindles exhibit 4 defects. First,
mei-2(
ct98) MI spindles initially assembly with a longer than wild-type metaphase length of 9.6 um. Second, they do not rotate. Third, they do not undergo the second phase of spindle shortening although they undergo the first phase of shortening with wild-type velocity. Fourth, microtubules do not redistribute from the poles to the midzone. As a result of these defects,
mei-2(
ct98) spindles reach a final minimum length of 6.2 um (65% of starting length) and induce formation of larger than wild-type polar bodies. These results are consistent with a model in which microtubule severing reduces the average microtubule length during spindle assembly and during the second phase of spindle shortening. We are currently analyzing a collection of
mei-1 misense alleles provided by Paul Mains to establish a relationship between in vitro microtubule-severing activity and in vivo spindle length control in order to address the perplexing question of why null alleles of
mei-1 or
mei-2 prevent assembly of bipolar meiotic spindles rather than causing assembly of extremely long meiotic spindles.