Mitotic cells Iysed into solutions of polymerizable microtubule protein contain a spindle which is similar to the living spindle in two respects: it will lose and gain birefringence when cooled and warmed, and it will move anaphase chromosomes to the opposite ends of the cell. Early anaphase cells lysed into buffers containing high molecular weight polyethylene glycol and nucleotide triphosphates will continue chromosome motion and spindle elongation in the absence of exogenous spindle subunits. These results suggest that while spindle growth requires microtubule polymerization, anaphase motions do not.Considerable effort has gone into the study of cell-free preparations of the mitotic apparatus (MA). Much has been learned about the structure of the isolated MA (1-4), and something is known of its chemistry (5-12), but two interesting questions have remained unanswered: what is the character of the equilibrium between the spindle and its subunits, and what is the nature of the motors that move the chromosomes?Previous studies have isolated the MA by "stabilizing" it, i.e., making it reversibly less labile by lysing mitotic cells into buffers which are poor solvents for protein (9). In such buffers the MA will keep its birefringence and overall form for hours, but it will not shrink and regrow as a function of temperature the way the spindle does in situ (13). Since the work of Hoffman-Berling on chromosome motion in glycerol-extracted cells (14), no one has, to our knowledge, been able to study the physiology of anaphase-like processes in vitro.In this paper we describe an experimental system for studying spindle functions in vitro and present some of our early results. We have taken advantage of the recent discovery of techniques for the reassembly of isolated microtubule protein (tubulin) (15, 16) to design a procedure based upon an equilibrium between assembled and disassembled subunits. Independent investigations based upon the same experimental design have been initiated by two groups at Woods Hole, Massachusetts. Preliminary reports of their findings have recently appeared in abstract form (17, 18). We lyse mitotic Pt K1 cells with a nonionic detergent into buffers containing various concentrations of polymerizable tubulin, and we monitor the magnitude and longevity of spindle birefringence and spindle size with the light microscope. Our results show that the spindle is stable for more than 1 hr after lysis in solutions containing sufficient concentrations of tubulin. The Abbreviations: PIPES, Piperazine-N-N'-bis[2-ethane Sulfonic Acid]; EGTA, ethylene bis(oxyethylene-nitrilo) tetraacetic acid; GTP, guanosine triphosphate; GEP, GTP + EGTA + PIPES; ATP, adenosine triphosphate; Tx, Triton X-100; MA, mitotic apparatus. stabilized spindles will lose birefringence when cooled and regain at least some of it when rewarmed. Cells deprived of spindle birefringence by cold treatment and then lysed into tubulin solutions will regrow birefringence after lysis when the preparation is warmed to 370. Spindles pre...