FORCED EXPRESSION OF CHIMERIC HUMAN FIBROBLAST TROPOMYOSIN MUTANTS AFFECTS CYTOKINESIS

Human fibroblasts generate at least eight tropomyosin (TM) isoforms (h TM1, hTM2, hTM3, hTM4, hTM5, hTM5a, hTM5b, and hTMsm alpha) from four distinct genes, and we have previously demonstrated that bacterially p roduced chimera hTM5/3 exhibits an unusually high affinity for actin f ilaments and a loss of the salt dependence typical for TM-actin bindin g (Novy, R.E., J. R. Sellers, L.-E Liu, and J. J.-C. Lin, 1993. Cell M otil. & Cytoskeleton. 26: 248-261). To examine the functional conseque nces of expressing this mutant TM isoform in vivo, we have transfected CHO cells with the full-length cDNA for hTM5/3 and compared them to c ells transfected with hTM3 and hTM5. Immunofluorescence microscopy rev eals that stably transfected CHO cells incorporate force-expressed hTM 3 and hTM5 into stress fibers with no significant effect on general ce ll morphology, microfilament organization or cytokinesis. In stable li nes expressing hTM5/3, however, cell division is slow and sometimes in complete. The doubling time and the incidence of multinucleate cells i n the stable hTM5/3 lines roughly parallel expression levels. A closel y related chimeric isoform hTM5/2, which differs only in the internal, alternatively spliced exon also produces defects in cytokinesis, sugg esting that normal TM function may involve coordination between the am ino and carboxy terminal regions. This coordination may be prevented i n the chimeric mutants. As bacterially produced hTM5/3 and hTM5/2 can displace hTM3 and hTM5 from actin filaments in vitro, it is likely tha t CHO-expressed hTM5/3 and hTM5/2 can displace endogenous TMs to act d ominantly in vivo. These results support a role for nonmuscle TM isofo rms in the fine tuning of microfilament organization during cytokinesi s. Additionally, we find that overexpression of TM does not stabilize endogenous microfilaments, rather, the hTM-expressing cells are actual ly more sensitive to cytochalasin B. This suggests that regulation of microfilament integrity in vivo requires stabilizing factors other tha n, or in addition to, TM.