The distribution and polymerization state ofactin in metaphase rat kangaroo cells was studied by fluorescence microscopy. Formaldehyde-fixed, acetone-extracted cells were labeled with either of two types of actin probes. The first, 7-nitrobenz-2-oxa-1,3-diazole-phallacidin, has high affinity for F actin and does not bind monomeric G actin. The second was a conjugate of DNase I labeled with either tetramethylrhodamine or fluorescein. DNase binds with high affinity to G actin and with lesser affinity to F actin. The polymerization state of actin was deduced by comparing the fluorescence distribution of the phallacidin derivative with that of the fluorescent DNase. The results indicate that the pole-to-chromosome region ofthe metaphase spindle contains G actin but little if any conventional F actin. F actin is found concentrated in a diffuse distribution outside the spindle region in metaphase cells and returns to the interzone area between the chromosomes by early telophase. These results exclude spindle models for chromosomal movement that require more than about five F actin filaments per chromosome, support the hypothesis that F actin is involved in force generation for cell cleavage, and are not inconsistent with the possibility that actin outside the spindle may be involved in chromosomal movement.The mechanism responsible for chromosomal movement during mitosis has not been established. Most investigators have attempted to show that either actin or tubulin in the mitotic spindle is involved in the force production required for chromosomal movement (1-5). The evidence for and against each protein includes biochemical, electron microscopic, and fluorescence microscopic results.Strong evidence exists that tubulin is a spindle component, although there remains some uncertainty regarding the amount present (5-7). Electron micrographs show microtubules inserted into chromosomal kinetochores, structures thought to be tubulin-nucleation centers (5). Each chromosome has one such site of attachment. This has led to speculation that tubulin participates directly in generating force for chromosomal movement (for review, see ref. 5).Biochemical evidence shows that actin is a component of nonmuscle cells (8). However, it is difficult to determine the subcellular location of actin by biochemical techniques. Some experiments have indicated the presence of actin in nuclear fractions (9, 10), but others have indicated that actin in these fractions may be a cytoplasmic contaminant. By using isolated nuclei, Comings and Harris found that actin and myosin constitute at most a very small fraction of nonhistone proteins (6).Electron microscopy has provided evidence for the existence ofspindle actin (11)(12)(13)(14). These experiments have identified actin by the appearance of characteristic arrowhead complexes formed from the binding of heavy meromyosin (HMM) or its nuclease S1 subfragment to actin filaments (15). Glycerination has been a standard step in the preparation ofthese specimens, and it has been suggested that c...