The mitotic spindle is often positioned in a characteristic location during development, for example to enable the proper segregation of developmental determinants [1,2]. When epithelial cells divide, the mitotic spindle is often positioned parallel to the plane of the epithelium, so that both daughter cells contribute to the epithelium [3]. The mechanisms by which mitotic spindles are positioned have not been characterized in great detail, but evidence is accumulating that in some systems the dynein-dynactin microtubule motor complex plays a role [4-6]. Dynein has yet not been localized to cortical sites where it could bind to microtubules and exert a force that might orient the mitotic spindle, however [7,8]. Here, we report that in mitotic polarized epithelial cells, the dynein-dynactin complex accumulates, from prometaphase onwards, along astral microtubules and at cortical spots, into which many of the astral microtubules dock. The spots are assembled at the lateral plasma membrane, in the region below the tight junctions. Their formation is inhibited by cytochalasin D, and under these conditions the spindles do not orient properly. This novel localization of the dynein-dynactin complex is consistent with a role for the complex in the positioning of the mitotic spindle. We also show that, during prophase, the motor complex colocalizes with the nuclear envelope, consistent with it having a role in separating the centrosomes that are associated with the nuclear envelope.

In the present work, we investigated, by use of forskolin, whether adenosine 3',5'-cyclic monophosphate (cAMP) level and Ca movements were modulated sequentially or in parallel by the activation of the beta-adrenergic receptor in the rat parotid gland. Forskolin-induced [3H]protein secretion was dependent on external Ca, whereas isoproterenol-induced secretion was not. This effect was not due to a requirement of adenylate cyclase for Ca, since the cAMP level increase induced by forskolin was not Ca dependent. Furthermore isoproterenol induced 45Ca efflux, whereas forskolin did not. 45Ca efflux was correlated neither to cAMP nor to secretion, since when there was a massive augmentation of cAMP there was no change in 45Ca efflux, and forskolin, which induced much secretion, was unable to induce Ca efflux. Carbachol potentiated the secretion induced by forskolin in the absence of Ca, whereas it did not potentiate the isoproterenol-induced response. From these results we suggest that beta-adrenergic receptor activation would lead to two parallel events, cAMP accumulation and Ca movements, which together would lead to maximal secretion.

After in vitro incubation of Xenopus oocytes with vitellogenin (VTG)-gold conjugate, the gold particles are distributed on the whole plasma membrane. Their concentration in coated pits still occurs at 0 degrees C. At +20 degrees C the label quickly (30 sec) appears in multi-vesicular endosomes (MVE) which segregate together with primary endocytic vesicles into distinct clusters below the plasma membrane. From this step up to crystallization of the yolk platelets, the gold particles stay in the same compartment. During 5.5 h the label progressively increases along the MVE membrane, first (1.5 h) by fusion of primary endocytic vesicles with consecutively enlarging endosomes, then (4 h) by decreasing of the MVE membrane. As concerns the yolk platelet formation, concentration of primordial yolk platelets (PYP) occurs at 5.5 h from the incubation onset, the labeling of preexisting yolk platelets starts at 7 h, while crystallization of PYP begins only after 12-13 h. Our results indicate that VTG receptors are not preclustered in coated pits and their lateral translation is not inhibited at 0 degrees C. The yolk protein processing takes place within one compartment only. The VTG condensation begins with a long concentration phase of receptor-VTG complexes still integrated in the endosome membrane. It occurs in MVE by: i) a repeated fusion of primary endocytic vesicles; ii) removing part of the endosome membrane by internal vesiculation. Fusion between endosomes occurs only after VTG has dissociated from its receptors and VTG dissociates only when when the density of the VTG-receptor complexes in the endosome membrane is sufficient. Crystallization begins after a 7-8 h delay. The endosome migration into the oocyte is also controlled by the binding of VTG to its receptors. Our results also demonstrate that binding of VTG colloidal gold modifies neither the vitellogenic pathway nor the duration of the vitellogenin internalization. However when vitellogenin is bound to colloidal gold, dissociation of ligand-receptor complexes is delayed because the amount of ligand in the incubation medium is necessarily low.