Brief Communication 211 Movement of nuclei along microtubules in Xenopus egg extracts Sigrid Reinsch and Eric Karsenti Microtubules are implicated in the movement and positioning of nuclei in many cell types. Nuclei can be moved by forces acting on microtubules nucleated at the spindle pole body, as in fungi [1], or microtubules nucleated at the centrosome, as during migration of the male (sperm) pronucleus towards the centre of the zygote after fertilization [2–4]. The dramatic movements of the female pronucleus towards the male pronucleus potentially involve another mechanism: movement along the microtubules of the sperm aster towards their slower growing, attached or ‘minus’ ends [3,5]. Here, we have reconstituted this last type of nuclear movement in vitro. Synthetic nuclei assembled in cytoplasmic extracts made from interphase Xenopus eggs move along microtubules towards their minus ends. We provide strong experimental evidence that cytoplasmic dynein is the motor for nuclear movement in this in vitro system, and discuss our results in terms of current knowledge of motility of the endoplasmic reticulum. Address: Cell Biology Program, EMBL, Meyerhofstrasse 1, Postfach 10.2209, 69012 Heidelberg, Germany. Correspondence: Sigrid Reinsch. E-mail: Reinsch@EMBL-Heidelberg.DE Received: 26 November 1996 Revised: 16 January 1997 Accepted: 16 January 1997 Electronic identifier: 0960-9822-007-00211 Current Biology 1997, 7:211–214 © Current Biology Ltd ISSN 0960-9822 Results and discussion We sought to develop an in vitro assay system that could reconstitute nuclear movement along microtubules, and to examine the molecular basis of this movement. In order to generate a source of well-defined nuclei, we assembled synthetic nuclei on DNA-coupled magnetic beads in cytoplasmic extracts made from Xenopus eggs (we refer to such synthetic nuclei as bead nuclei) [6]. Egg extracts were fractionated into cytosolic and membrane components in order to assemble either complete nuclei or, as a control, interphase chromatin lacking a nuclear envelope. Figure 1 shows electron microscope images of representative bead nuclei. To determine whether bead nuclei move along microtubules, we assayed preassembled nuclei on microtubule asters nucleated using purified centrosomes in Xenopus Figure 1 Electron microscopy of bead nuclei. (a) Low-magnification montage showing nuclei containing variable numbers of magnetic beads. (b) Single bead nucleus containing two magnetic beads. The lower panel shows a high-magnification view of the region boxed in the upper panel (I, inner nuclear membrane; O, outer nuclear membrane; arrowhead, nuclear pore; R, ribosomes). (c) Nuclei often show membrane continuities with annulate lamellae (A) containing nuclear pores (B, magnetic dynabead; C, chromatin). Scale bar = 1 µm. 212 Current Biology, Vol 7 No 3 Figure 2 (b) (c) (d) Distance from centrosome (µm) (a) CS 40 A 30 C B 20 E D 10 F 0 05:50 06:40 07:30 08:20 09:10 10:00 10:50 Time (min:sec) (f) (g) CS (h) 40 Distance from centrosome (µm) (e) A 30 B 20 C 10 D 0 15:00 15:50 16:40 17:30 18:20 19:10 20:00 Time (min:sec) Bead nuclei move to, and accumulate at, centrosomes in Xenopus cytosol. (a–c,e–g) Stills from typical motility assays. Times are shown as h:min:sec. (d,h) The position of the moving nuclei shown in (a–c) and (e–g), respectively, tracked relative to the centrosome. (a–c) One nucleus is already at the centrosome (CS), which nucleates a large aster. A second nucleus containing two beads moves to the centrosome. (d) Velocities: A→B, 0.3 µm sec –1 ; C →D, 0.45 µm sec –1 ; E→F, 0.31 µm sec –1 ; overall rate A→F, 0.19 µm sec –1 . (e–g) Membraneous tubules extend behind the moving nucleus. (h) Velocities: B →C, 0.76 µm sec –1 ; overall rate A→D, 0.2 µm sec –1 . interphase egg cytosol. As shown in Figure 2, nuclei moved along microtubules toward the centrosome in a saltatory fashion, with periods of rapid movement interspersed with pauses occurring when a nucleus encountered a randomly oriented microtubule in its path. In some cases (20 % of moving nuclei), membraneous tubules of ER extended from the nucleus (Fig. 2e–g), usually trailing behind the moving nucleus. Bead nuclei moved towards slower growing, attached or ‘minus’ ends of the microtubules at an average overall rate of 0.36 ± 0.17 mm sec–1 (mean ± standard deviation) and an average maximum speed of 0.62 ± 0.24 mm sec–1 (n = 47), speeds that are comparable to the rates of pronuclear migration in vivo [7–9]. chromatin beads to the substrate. Video microscopy showed that only about 15% of chromatin beads moved (n = 20). Those that did move oscillated near the periphery of the asters and their movements did not appear to be due to direct attachment to, and movement along, microtubules. Nuclei stripped of their nuclear membrane with the detergent Triton X-100 (data not shown), or magnetic beads lacking DNA but incubated in nuclear assembly extracts (mock assembly), did not move to, or accumulate at, the centrosomes (Fig. 3a). We conclude, therefore, that the nuclear membrane is required for movement of nuclei along microtubules. We then asked whether nuclear motility requires the nuclear envelope. We compared the motility of bead nuclei with that of either interphase chromatin beads or bead nuclei from which the membrane had been stripped using detergent. Most bead nuclei (85–95%) accumulated at centrosomes after 20 minutes, whereas only 40% of the interphase chromatin beads did so (Fig. 3a). The association of interphase chromatin beads with centrosomes does not seem to reflect a dynamic interaction with microtubules (in contrast to the movement of mitotic chromatin in mitotic extracts, which we and others have described previously [6]). Rather, it largely reflects random sticking of interphase Because nuclei moved towards microtubule minus ends, we examined whether inhibitors of the minus-end-directed motor, dynein, could block this movement. Both vanadate and an antibody to the intermediate chain of cytoplasmic dynein (70.1), which is known to inhibit dynein activity during spindle assembly [6], significantly inhibited nuclear movements and accumulation at centrosomes (Fig. 3a). As expected, various anti-kinesin reagents which disrupt spindle assembly in mitotic extracts had no effect on the movement of bead nuclei (data not shown). The effect of vanadate was dose-dependent (data not shown), with 50% inhibition at ~15 mM, which is typical for the inhibition of dynein-mediated motility [10]. Nuclei made abortive, short, Brief Communication 213 Figure 3 Proportion of beads at the centrosome (%) (b) (c) 100 80 Bead nuclei Others 60 40 20 0 C B e hr ad om n at uc l in ei be ad 10 s µM M oc v 25 a k µ nad α- M v ate an dy ad n at α- 0 .5 e dy m g n m– 0. l 1 α- 3 m mg yo 1 ml –1 m g m– l 1 (a) Accumulation of nuclei at centrosomes requires a nuclear membrane and cytoplasmic dynein. (a) Accumulation in cytosol was compared for bead nuclei, interphase chromatin beads, and mock-assembled beads (mock); assembly in cytosol containing vanadate or anti-dynein antibody (α-dyn) was examined for bead nuclei; an anti-myosin antibody (α-myo) was used as a control. Values are mean ± standard deviations of the number of bead nuclei or other beads present at centrosomes as a percentage of the total bead nuclei or other beads counted. (b,c) Two stills of different focal planes of the same field from a sample containing anti-dynein. (b) The centrosome (CS) is on the substrate. (c) The nucleus is above this focal plane; arrowheads indicate an ER tubule which extends from the nucleus. Scale bar = 10 µm. minus-end-directed movements along microtubules, but were then consistently observed to move up from the focal plane of the centrosome and to accumulate at the periphery of the microtubule asters (Fig. 3b,c). No plus-end-directed motility of nuclei along microtubules to the aster periphery was observed. Instead, growing microtubules appeared to move the nuclei upwards, away from the substrate, and ER tubules were often seen to extend from the nucleus, away from the centrosome (Fig. 3c, arrowheads). Furthermore, movement to the periphery of the aster occurred at concentrations up to 100 mM vanadate, at which point kinesin is likely to be inhibited [10]. These data strongly suggest that dynein is the motor underlying the minus-end-directed movement in this system. towards the centrosome at an average rate of 1.2 ± 0.49 mm sec–1 (n = 20); they extended and retracted repeatedly over short times (Fig. 4c), while nuclear movements were longer lasting. Some minus-end-directed ER tubule extension was observed during incubation with vanadate or anti-dynein at concentrations which significantly inhibited nuclear accumulation at centrosomes (data not shown), consistent with previous results [12]. There are clear similarities between nuclear motility and ER tubule extension, though the rates of movement and sensitivity to vanadate are different. Our results indicate that nuclear movement can be produced both by a dynein-driven mechanism (towards the centrosome) and by microtubule polymerization. This is similar to the ER motility characterized in Xenopus extracts (Fig. 3 and see also supplementary material; [11,13,14]). During nuclear movement, however, the dynein-driven mechanism predominates and results in persistent movement along microtubules towards the minus ends, and accumulation at the centrosome. It has been proposed that nuclear movements might reflect motility of the ER [8,11]. We therefore examined the characteristics of minus-end-directed extensions of ER tubules from nuclei in our assay. ER tubules extended Figure 4 (a) (b) (c) Distance from origin (µm) Comparison of the movement of nuclei on microtubules with ER tubule extension. (a,b) Minus-end-directed ER tubule extension. The two frames, taken 5 sec apart, show a nucleus paused on the substrate. ER tubules extend from the nucleus (n); during the 5 sec interval, one ER tubule (arrow) extended towards the centrosome (cs) and another (arrowhead) retracted. (c) Typical time course showing a single ER tubule which extended and retracted repeatedly. 20 15 10 5 00 20 40 Time (sec) 60 214 Current Biology, Vol 7 No 3 While the experiments reported here demonstrate that dynein can drive nuclear movement in vitro, it remains to be shown that this is also the case in vivo. Dynein could potentially be involved in migration of both female and male pronuclei. Female pronuclear movement might use dynein targeted to the outer nuclear membrane. Centration of the sperm nucleus could be mediated by dynein, anchored throughout the egg cytoplasm, acting on the microtubules of the sperm aster [3]. Indeed, in fungi and yeast, nuclei are positioned by the activity of dynein on microtubules nucleated at the spindle pole body [1]. It is clearly becoming increasingly important to determine how dynein is targeted to, and regulated at, its site of action. Materials and methods Nuclear assembly and electron microscopy DNA was coupled to streptavidin dynabeads as described [6]. Fractionated extracts for nuclear assembly were prepared as described [15] with the following modifications: sucrose was added to a final concentration of 125 mM to the 10 000 g cytosol before centrifuging at 200 000 g. Cytosolic fractions were diluted with one volume complete acetate (ABC) buffer (100 mM KAc, 3 mM MgAc, 5 mM EGTA, 10 mM HEPES pH 7.4, 150 mM sucrose, 7.5 mM creatine phosphate, 80 mg ml–1 creatine kinase, 1 mM MgATP, 1 mM DTT and 10 mg ml–1 each of aprotinin, pepstatin and leupeptin) and spun at 100 000 g for 15 min at 4 °C (ABC cytosol). DNA beads (1 mg) were resuspended in 100 ml ABC cytosol, 10 ml membranes, 15 mg ml–1 glycogen, 1 mM ATP, 7.5 mM creatine phosphate, 80 mg ml–1 creatine kinase, and incubated at 20 °C for 3 h. Assembly was assessed by a nuclear transport assay as described [6]. Glycerol was added to 10% and nuclei were snap frozen in 10 ml aliquots in liquid nitrogen and stored at –70 °C. Nuclear assembly reactions were diluted in excess ABC buffer and the nuclei retrieved on a magnet. Nuclei were washed once in ABC buffer, pelleted at 2 000 r.p.m. in a microfuge and fixed with 1% glutaraldehyde in 200 mM sodium cacodylate pH 7.4, for 45 min at 4 °C. Pellets were embedded in Epon, sectioned and observed with a Zeiss EM10 microscope. Motility assay Centrosomes were purified as described [16]. Xenopus high-speed supernatants (HSS) were prepared as described [13] from unactivated Xenopus eggs or eggs 20 min after activation with the calcium ionophore A23187. Chambers for differential interference contrast (DIC) microscopy (a volume of 7 ml) were assembled from two pieces of double-sided tape and an 18 × 18 mm coverslip on a microscope slide. Centrosomes (2 × 10 8) diluted in PE (10 mM K-Pipes pH 7.2, 1 mM EDTA) were flowed into the chamber and allowed to attach for 5 min on ice. The chamber was blocked with 30 ml 5 mg ml–1 casein in PE and incubated on ice until use. Ten minutes before the assay, 10 ml HSS was flowed into the chamber and incubated at room temperature to allow nucleation and aster formation. Nuclear assembly reactions (10 ml) were diluted with 150 ml ABC buffer. The nuclei were retrieved, resuspended in 10 ml HSS, then flowed into the chamber which was sealed with VALAP and observed at 22°C on a Zeiss Axiovert 10 microscope equipped for DIC with a 100× Zeiss Achrostigmat objective. For video analysis, shuttered images were collected with a Hamamatsu CCD C3077 camera. Image averaging and background subtraction were performed with an Argus 10 image processor. Images were stored using a Sony LVR 6000 and home-made software. Images were analyzed using NIH image and processed with Adobe Photoshop. Velocity measurements were calculated for nuclei which moved more than 20 mm. Maximum velocities were calculated over 5 sec intervals; average velocities were calculated for the duration of movement including pauses of no more than 10 sec. Assay for accumulation at centrosomes Chromatin beads were assembled in the same way as bead nuclei, except membranes were omitted, and were assayed as described [6]. ‘Mock nuclei’ are uncoupled M80 streptavidin beads incubated in assembly extracts. Anti-dynein intermediate chain 70.1 and control anti-myosin antibodies were prepared as described [6]. Vanadate was prepared as described [17]. Assays were performed as for motility assays, except that 1 ml diluted vanadate, antibodies, or ABC buffer (control) was added to 9 ml HSS to resuspend the nuclei. Assays were incubated for 20 min at room temperature and bead nuclei or other beads present at centrosomes were scored between 20 and 30 min (50–150 bead nuclei or other beads were scored in each focal plane). Assays were repeated three times using different bead and cytosol preparations. Each centrosome containing bead nuclei or other beads was counted as 1, so stated values for accumulation are minimal. Nuclei and centrosomes were titrated so that the number of centrosomes far exceeded the number of nuclei. Supplementary material available Brief video clips and a supplementary figure are also published with this paper on the internet. Acknowledgements W e thank R. Heald for advice and reagents; S. Eaton, P. Gonczy, C. Murphy and H. Yu for reading the manuscript; A. Blocker, J. Burkhardt and T. Hyman for advice on the motility assay; I. Palacios, C. Dingwall and G-J. Arts for advice and reagents; G. Krohne and T. Schroer for antibodies; and A. Habermann, H. W ilhelm and A. Sawyer for technical support. This work was supported by a Human Frontier Science Program Postdoctoral fellowship to S.R. and a Human Frontier Science Program grant and a Human Capital Mobility Research Network grant to E.K. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. Morris NR, Xiang X, Beckwith SM: Nuclear migration advances in fungi. Trends Cell Biol 1995, 5:278–282. W ilson EB: The Cell in Development and Heredity, 3rd edn. New York: Macmillan; 1928. Hamaguchi MS, Hiramoto Y: Analysis of the role of astral rays in pronuclear migration by the colcemid-UV method. Dev Growth Differ 1986, 28:143–156. Chambers EL: The movement of the egg nucleus in relation to the sperm aster in the echinoderm egg. J Exp Biol 1939, 16:409–424. Schatten G, Schatten H: Effects of motility inhibitors during sea urchin fertilization. Exp Cell Res 1981, 135:311–330. Heald R, Tournebize R, Blank T, Sandaltzopoulos R, Becker P, Hyman A, et al.: Self-organization of microtubules into bipolar spindles around artificial chromosomes in Xenopus egg extracts. Nature 1996, 382:420–425. Schatten G: The movements and fusion of the pronuclei at fertilization of the sea urchin Lytechinus variegatus: time-lapse video microscopy. J Morphol 1981, 167:231–247. Rouvière C, Houliston E, Carr D, Chang P, Sardet C: Characteristics of pronuclear migration in Beroe ovata . Cell Motil Cytoskeleton 1994, 29:301–311. Stewart-Savage J, Grey RD: The temporal and spatial relationships between cortical contraction, sperm trail formation, and pronuclear migration in fertilized Xenopus eggs. Roux’s Archives 1982, 191:241–245. McIntosh JR, Porter ME: Enzymes for microtubule-dependent motility. J Biol Chem 1989, 264:6001–6004. Allan V: Role of motor proteins in organizing the endoplasmic reticulum and Golgi apparatus. Semin Cell Dev Biol 1996, 7:335–342. Allan VJ, Vale RD: Cell cycle control of microtubule-based membrane transport and tubule formation in vitro. J Cell Biol 1991, 113:347–359. W aterman-Storer CM, Gregory J, Parsons SF, Salmon ED: Membrane/ microtubule tip attachment complexes (TACs) allow the assembly dynamics of plus ends to push and pull membranes into tubulovesicular networks in interphase Xenopus egg extracts. J Cell Biol 1995, 130:1161–1169. Allan V: Protein phosphatase 1 regulates the cytoplasmic dyneindriven formation of endoplasmic reticulum networks in vitro. J Cell Biol 1995, 128:879–891. Hartl P, Olson E, Dang T, Forbes DJ: Nuclear assembly with l DNA in fractionated Xenopus egg extracts: an unexpected role for glycogen in formation of a higher order chromatin intermediate. J Cell Biol 1994, 124:235–248. Bornens M, Paintrand M, Berges J, Marty MC, Karsenti E: Structural and chemical characterization of isolated centrosomes. Cell Motil Cytoskeleton 1987, 8:238–249. Kypta RM, Hemming A, Courtneidge SA: Identification and characterization of p59fyn (a src-like protein kinase) in normal and polyoma virus transformed cells. EMBO J 1988, 7:3837–3844.