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.