Being and becoming the spindle: Microtubule dynamics and microtubule motors in mitotic spindle morphogenesis

In order better to understand the biochemistry of mitotic spindle structure and function, we developed a cell-free system from Xenopus eggs that reconstitutes spindle behavior in vitro. We shaw that two steps of bipolar spindle assembly in vivo can be dissociated in vitro: half-spindle assembly is driven by the selective stabilization of microtubule dynamics by activities present in mitotic chromatin., while half-spindle fusion involves further stabilization of half-spindle microtubules. We also studied spindle microtubule dynamics in vitro, with a photoactivatable caged-fluorescent tubulin analog. Upon fluorescently marking the spindle and imaging with a computer-controlled low-light fluorescence videomicroscope, we observed a poleward movement of spindle microtubules. This poleward microtubule flux is arrested by AMP-PNP but unaffected by vanadate, suggesting that a microtubule motor protein may be involved.; To investigate a further role for microtubule motor proteins in spindle assembly and poleward flux, we raised antibodies against conserved peptides derived from kinesin heavy chain sequences. Antibodies recognized spindle and/or kinetochore structures in Xenopus and mammalian tissue culture cells and also recognized a number of microtubule-binding proteins in Xenopus egg extracts. One of these proteins was determined to be the product of the Xenopus Eg5 gene. By immunofluorescence Eg5 is localized to spindle microtubules and enriched toward spindle poles. A recombinant Eg5 fusion protein is a slow, plus-end directed microtubule motor in vitro. Anti-Eg5 antibodies disrupt spindle assembly in vitro, especially spindle pole morphology. These results suggest a structural model for the role of Eg5 in mitosis.; The mechanism of flux and role of Eg5 in flux was further investigated by photoactivating spindle subassemblies generated in vitro. Half-spindle assembly intermediates and astral microtubule arrays formed by the addition of DMSO to extracts were used to test the role of chromatin and centrosomes. All structures exhibited poleward flux after photoactivation, suggesting that poleward flux is a fundamental property of organized microtubule arrays in mitosis. We also tested the role of Eg5 in poleward flux in extracts depleted of Eg5 by immunoprecipitation. Immunodepletion of Eg5 had no apparent effect on poleward flux.