© 2000 Macmillan Magazines Ltd brief communications NATURE CELL BIOLOGY VOL 2 DECEMBER 2000 http://cellbio.nature.com 948 The rate of poleward chromosome motion is attenuated in Drosophila zw10 and rod mutants Matthew S. Savoian*†, Michael L. Goldberg‡ and Conly L. Rieder*†§ *Division of Molecular Medicine, Wadsworth Center, New York State Department of Health, Albany, New York 12201-0509, USA †Department of Biomedical Sciences, State University of New York, Albany, New York 12222, USA ‡Department of Molecular Biology and Genetics, Cornell University, Ithaca, New York 14853, USA §e-mail: Rieder@Wadsworth.org Here we show that the rate of poleward chromosome motion in zw10-null mutants is greatly attenuated through- out the division process, and that chromosome disjunction at anaphase is highly asynchronous. Our results show that ZW10 protein, together with Rod, is involved in production and/or regulation of the force reponsible for poleward chromosome motion. P oleward chromosome motion is mediated by kinetochores, which can move on the surface of microtubules or on their disassembling plus ends. Although the microtubule motor cytoplasmic dynein is thought to be involved in this motion 1 , attempts to demonstrate this have been unsuccessful because inhi- bition of dynein gives rise to aberrant spindles that fail to enter anaphase 2,3 . Zw10 is required for localizing dynein to kinetochores in many organisms including humans 4,5 , and analysis of fixed zw10 mutants has shown that division is disrupted at anaphase when chromosomes lag 6 . To determine the origin of this lag, we used video-light microscopy (video-LM) and laser microsurgery to study chromosome behaviour in living Drosophila spermatocytes hemizygous for a null allele of zw10. We found that the rate of pole- ward chromosome motion in these mutants is greatly attenuated throughout the division process, and that chromosome disjunction at anaphase is highly asynchronous. Similar behaviours have also been observed in spermatocytes lacking Rod, a protein that is required for the association of Zw10 with kinetochores 5,7 . Many spermatocytes have a checkpoint that inhibits anaphase when non-natural univalents are present or when spindle forma- tion is disturbed 8 . However, univalents do not prevent anaphase in Drosophila spermatocytes 9 , which indicates that these cells may lack this pathway. To evaluate this, we determined the duration of prometaphase, defined as the time from initiation of chromosome motion until anaphase onset, for both y; ry 506 wild-type (35 ± 2 min; n = 16) and zw10-mutant (37 ± 2 min; n = 19) sper- matocytes, and found them to be similar. We then investigated whether taxol, a drug that delays anaphase in cells by stabilizing microtubules 10 , delays anaphase in wild-type or zw10-null sperma- tocytes. We found that treating wild-type cells in late G2 with 10 μM taxol prolonged the subsequent prometaphase roughly twofold (57 ± 3 min; n = 4), but that similar treatment of zw10-null spermatocytes did not prolong prometaphase (33 ± 2 min; n = 7). Thus, Drosophila spermatocytes possess a short-term spindle- assembly checkpoint that forestalls anaphase when microtubule dynamics are perturbed, and in the absence of Zw10 this pathway is non-functional. Drosophila primary spermatocytes contain four bivalent pairs, but with high-resolution optics only one is usually visible within a single focal plane. Consistent with previous findings 11 , we found that in wild-type cells chromosomes often underwent rapid pole- ward movements during the initial stages of kinetochore attach- ment. In 10 cells chosen at random, 12 chromosomes remained in focus during the first 20 min of prometaphase. During this time each chromosome carried out one or more rapid poleward excur- sions with an average maximum velocity of 11.2 ± 1.2 μm min –1 (range 5.8–22.1 μm min –1 ). By comparison, chromosomes in zw10 mutants were visibly far less dynamic during early prometaphase, and the fastest poleward motions observed in 10 cells chosen at random averaged only 2.4 ± 0.4 μm min –1 (n = 10; range 0.7–5.0 μm min –1 ). It is also noteworthy that, relative to the wild type, metaphase bivalents in mutants exhibited little if any stretch- ing between the poles. Thus, in the absence of Zw10 both the force acting on the kinetochores and the rate of poleward chromosome motion are markedly reduced during the early stages of spindle assembly. The alignment of all bivalents near the spindle equator is not a prerequisite for initiation of anaphase in either wild-type 11 or zw10-mutant spermatocytes. In the wild type (Fig. 1), anaphasic separation of bivalents occurred, as in most organisms 12,13 , over a period of 1–2 min. Once disjoined, the chromosomes then moved polewards at 1.9 ± 0.1 μm min –1 (n = 18; range 1.0–2.8 μm min –1 ). Anaphase lasted 8 ± 1 min (n = 16) and ended when the dyads formed karyomeres near the poles before fusing into a nucleus. In contrast to the wild type, the disjunction of bivalents in zw10 mutants was highly asynchronous, requiring 2–4 min in 11 out of the 16 primary spermatocytes analysed (Fig. 2), and in several cells one or more bivalents failed to disjoin and formed karyomeres near the spindle midzone at telophase (data not shown). After the first bivalent disjoined, the dyads decondensed 4–8 min later regardless of their position relative to the poles. Asynchronous disjunction of chromatids was also seen in zw10-null secondary spermatocytes (data not shown). As is common for cells cultured on glass, cytoki- nesis was prolonged in the flatter wild-type and zw10-mutant spec- imens, but it occurred with normal kinetics in the more rounded cells (Figs 1 and 2). After disjoining, most but not all dyads in zw10-null spermato- cytes exhibited some poleward (anaphase A) motion, but the rate and extent was severely depressed relative to that in the wild type (Fig. 2). For the 20 moving chromosomes analysed, the average rate of poleward motion was only 0.5 ± 0.1 μm min –1 (range 0.1–1.3 μm min –1 ). For a particular chromosome the rate fell into one of two distinct velocity categories, both of which were often seen in a single cell (Fig. 2). Half of the dyads moved polewards at 0.6–1.3 μm min –1 (average 0.8 ± 0.1 μm min –1 ; n = 10), or roughly half of the average poleward velocity of wild-type spermatocytes. However, the other half moved at a rate of 0.1–0.4 μm min –1 (aver- age 0.3 ± 0.1 μm min –1 ; n = 10), or roughly one-seventh of the