Rings of intermediate (100 A) filament bundles in the perinuclear region of vascular endothelial cells. Their mobilization by colcemid and mitosis.

Vascular endothelial cells cultured from guinea pig aorta or portal vein contain naturally occurring bundles of 100 A (diameter) filaments that completely encircle the nucleus. These rings are phase lucent and birefringent when examined with the light microscope. Perinuclear bundles of 100 A filaments were also seen in endothelial cells in vivo, indicating that they are a normal cytoplasmic component. These filaments did not decorate with S-1, and were not disrupted by glyceination. With these cells, experiments were designed to answer the following questions: (a) does Colcemid have an effect on these naturally occuring bundles? And (b) do these filaments remain during cell division? Endothelial cells grown in the presence of Colcemid were followed over 24 h. The perinuclear ring coiled into a juxtanuclear cap that consisted of disorganized arrays of 100 A filaments. This "coiling" effect was not blocked by cycloheximide, an inhibitor of protein synthesis. In another experiment, dividing cells were examined. During division the bundle of filaments is passively pulled in half into the daughter cells. These bundles did not disappear during the mitosis when mitotic spindle microtubules assemble. These studies suggest that Colcemid may exert a direct effect on 100 A filaments, independent of microtubules. Since these filaments do not disappear during mitosis, it is possible that in these cells the 100 A filaments and tubulin do not share a common pool of precursor proteins.