Abstract. As anaphase began, mitotic PtK~ and newt lung epithelial cells were permeabilized with digitonin in permeabilization medium (PM). Permeabilization stopped cytoplasmic activity, chromosome movement, and cytokinesis within about 3 min, presumably due to the loss of endogenous ATE ATE GTP, or ATP-~-S added in the PM 4-7 min later restarted anaphase A while kinetochore fibers shortened. AMPPNP could not restart anaphase A; ATP was ineffective if the spindle was stabilized in PM + DMSO. Cells permeabilized in PM + taxol varied in their response to ATP depending on the stage of anaphase reached: one mid-anaphase cell showed initial movement of chromosomes back to the metaphase plate upon permeabilization but later, anaphase A resumed when ATP was added. Anaphase A was also reactivated by cold PM (,x,16~ or PM containing calcium (1-10 mM). Staining of fixed cells with antitubulin showed that microtubules (MTs) were relatively stable after permeabilization and MT assembly was usually promoted in asters. Astral and kinetochore MTs were sensitive to MT disassembly conditions, and shortening of kinetochore MTs always accompanied reactivation of anaphase A. Interphase and interzonal spindle MTs were relatively stable to cold and calcium until extraction of cells was promoted by longer periods in the PM, or by higher concentrations of detergent.Since we cannot envisage how both cold treatment or relatively high calcium levels can reactivate spindle motility in quiescent, permeabilized, and presumably energy-depleted cells, we conclude that anaphase A is powered by energy stored in the spindle. The nucleotide triphosphates effective in reactivating anaphase A could be necessary for the kinetochore MT disassembly without which anaphase movement cannot proceed.N 'o consensus currently exists regarding the role of ATP during anaphase A (chromosome-to-pole movement) and it is difficult to evaluate the diverse data on this subject. We (Pickett-Heaps and Spurck, 1982; Spurck et al., 1986a, b) have found that the concentration of metabolic inhibitors needed to cause rapid (within 1 min) and completely reversible cessation of cellular activity (presumably by depleting cellular ATP levels) in live cells is critical; 2,4 dinitrophenol (DNP) 1 is effective in both diatoms and mammalian cells only at the seemingly high concentration of 1 mM, and lowering the concentration even to 0.5 mM results in a significantly delayed and often incomplete response. In contrast, far lower concentrations of metabolic inhibitors were used in most previous work (reviewed by Spurck et al., 1986a, b; Hepler and Palevitz, 1986). Furthermore, many who work with animal cell systems use inhibitors of oxidative phosphorylation alone. We (Spurck et al., 1986a) found that glycolysis must also be blocked by using 2-deoxyglucose (DOG) in conjunction with DNP or other metabolic inhibitors. Without DOG, cellular activity may be maintained in DNP, and glycolysis, we believe, keeps cells alive for the long periods of DNP treatment alone (1--4 h or more) repo...