Article
The Role of Mitotic Cell-Substrate Adhesion Re-
modeling in Animal Cell Division
Graphical Abstract
Highlights
d Cells re-model adhesions as they round up upon entry into
mitosis
d These cell-substrate adhesions are essential for division in
non-transformed cells
d Adhesions can guide migration to divide cells with a
compromised actomyosin ring
Authors
Christina L. Dix, Helen K. Matthews,
Marina Uroz, ..., Michael Boutros,
Xavier Trepat, Buzz Baum
Correspondence
b.baum@ucl.ac.uk
In Brief
Dix et al. show that the integrin-positive
adhesive contacts that remain following
mitotic rounding are essential for division
in non-transformed adherent cells in
culture. Further, these adhesion sites
guide polarized daughter cell migration—
a process that is sufficient to drive
abscission in the absence of a visible
contractile actomyosin ring.
Dix et al., 2018, Developmental Cell 45, 132–145
April 9, 2018 ª 2018 Elsevier Inc.
https://doi.org/10.1016/j.devcel.2018.03.009

----!@#$NewPage!@#$----
Developmental Cell
Article
The Role of Mitotic Cell-Substrate Adhesion
Re-modeling in Animal Cell Division
Christina L. Dix,1,8 Helen K. Matthews,1,8 Marina Uroz,3 Susannah McLaren,1 Lucie Wolf,2 Nicholas Heatley,1 Zaw Win,1
Pedro Almada,1 Ricardo Henriques,1 Michael Boutros,2 Xavier Trepat,3,4,5,6 and Buzz Baum1,7,9,*
1MRC - Laboratory for Molecular Cell Biology, University College London, London WC1E 6BT, UK
2Division of Signaling and Functional Genomics, German Cancer Research Center (DKFZ), and Department for Cell and Molecular Biology,
Medical Faculty Mannheim, Heidelberg University, Heidelberg 69120, Germany
3Institute for Bioengineering of Catalonia (IBEC), The Barcelona Institute for Science and Technology (BIST), Barcelona 08028, Spain
4Unitat de Biofisica i Bioenginyeria, Facultat de Medicina, Universitat de Barcelona, Barcelona 08036, Spain
5Institucio´ Catalana de Recerca i Estudis Avanc¸ ats (ICREA), Barcelona 08010, Spain
6Center for Networked Biomedical Research on Bioengineering, Biomaterials and Nanomedicine (CIBER-BBN), Barcelona 08028, Spain
7Institute for the Physics of Living Systems, University College London, London WC1E 6BT, UK
8These authors contributed equally
9Lead Contact
*Correspondence: b.baum@ucl.ac.uk
https://doi.org/10.1016/j.devcel.2018.03.009
SUMMARY
Animal cells undergo a dramatic series of shape
changes as they divide, which depend on re-modeling
of cell-substrate adhesions. Here, we show that while
focal adhesion complexes are disassembled during
mitotic rounding, integrins remain in place. These
integrin-rich contacts connect mitotic cells to the
underlying
substrate
throughout
mitosis,
guide
polarized cell migration following mitotic exit, and
are functionally important, since adherent cells un-
dergo division failure when removed from the sub-
strate. Further, the ability of cells to re-spread along
pre-existing adhesive contacts is essential for divi-
sion in cells compromised in their ability to construct
a RhoGEF-dependent (Ect2) actomyosin ring. As a
result, following Ect2 depletion, cells fail to divide
on small adhesive islands but successfully divide on
larger patterns, as the connection between daughter
cells narrows and severs as they migrate away from
one another. In this way, regulated re-modeling of
cell-substrate adhesions during mitotic rounding
aids division in animal cells.
INTRODUCTION
As animal cells in culture progress through mitosis they undergo
a series of changes in shape (reviewed by Ramkumar and Baum,
2016). Most round up as they enter mitosis. They then elongate
as they exit mitosis and segregate their DNA, forming a central
contractile furrow that narrows around the central spindle to
form a mid-body, which is cut during the process of abscission
to generate two new daughter cells (reviewed by Fededa and
Gerlich, 2012). Two processes are critical for these drastic shape
changes to occur efficiently: cells must dynamically re-model
both their actomyosin cytoskeleton and the adhesions through
which they attach to the substrate.
At the molecular level, the mitotic re-modeling of the actomy-
osin cytoskeleton is regulated by the RhoGEF Ect2 (Matthews
et al., 2012; Miki et al., 1993; Prokopenko et al., 1999; Tatsumoto
et al., 1999). Mitotic rounding begins with the export of RhoGEF
Ect2 from the nuclear compartment in prophase (Matthews et al.,
2012). This activates RhoA (Maddox and Burridge, 2003;
Tatsumoto et al., 1999) at the plasma membrane, leading to
the construction of a relatively isotropic (Rosa et al., 2015) and
mechanically rigid (Kunda et al., 2008; Stewart et al., 2011)
actomyosin cortex. At mitotic exit, Ect2 is then recruited to the
center of the anaphase spindle (Burkard et al., 2009; Somers
and Saint, 2003), to the region of antiparallel microtubule overlap
(Rappaport, 1985), where it activates Rho (Bement et al., 2005;
Kimura et al., 2000; Nishimura and Yonemura, 2006; Tatsumoto
et al., 1999; Y€uce et al., 2005) at the membrane (Kotynova´ et al.,
2016; Wolfe et al., 2009), leading to local actin filament nucle-
ation and myosin II activation (Su et al., 2011). In combination
with signals from anaphase chromatin (Kiyomitsu and Cheese-
man, 2013; Rodrigues et al., 2015), these events polarize the
cell cortex (Wagner and Glotzer, 2016), causing cells to undergo
cytokinesis as they relax at their poles and constrict at their
center (reviewed by Green et al., 2012).
These actomyosin-dependent shape changes can only take
place if the adhesive contacts that couple cells to the underlying
extracellular matrix are re-modeled upon entry into mitosis (Dao
et al., 2009; Lancaster et al., 2013). Although this process is not
well-understood, mitotic rounding has been shown to require
inactivation of the small GTPase Rap1 (Dao et al., 2009). As a
consequence, cells that express a constitutively active form of
Rap1 fail to round up properly when they enter mitosis, leading
to defects in spindle morphogenesis (Lancaster et al., 2013).
Nevertheless, although Rap1 is inactivated at mitotic entry, cells
do not lose all contacts with the substrate. Instead, as they round
up, cells in culture remain in contact with the substrate through
retraction fibers: narrow actin-rich membrane tubes generated
during the process of rounding (Mitchison, 1992). In some
132
Developmental Cell 45, 132–145, April 9, 2018 ª 2018 Elsevier Inc.

----!@#$NewPage!@#$----
NEP
+3
-3
-6
-18
-12
-24
-30
RPE1 Zyxin-GFP
D
-72
-66
-60
-54
-48
-42
-36
-30
-24
-12
NEP
E
-12min
+18
1-Integrin
+18 Live
RPE1 Zyxin-GFP Live Online fixation
Merge
NEP
+6
-6
-12min
+3
RPE1 Zyxin-GFP Live
-9
-3
Fixed
Integrin
Live 
Zyxin
Live 
Zyxin
F
Interphase
Metaphase
A
1-Integrin
Paxillin
Talin
Merge
B
C
100
0
% of adhesions
Cell 5
Cell 4
Cell 3
Cell 2
Cell 1
-60-54-48-42-36-30-24-18-12 -6 0
6
Time from NEP (min)
100
75
25
50
0
% of adhesions
Time from last
adhesion formed (min)
0
3
6
9
12
15
18
(legend on next page)
Developmental Cell 45, 132–145, April 9, 2018
133

----!@#$NewPage!@#$----
instances, these retraction fibers have been shown to support
significant forces (Fink et al., 2011). Moreover, tension in retrac-
tion fibers (Fink et al., 2011) has been shown to guide orientation
of the metaphase spindle relative to the pattern of adhesions
(Petridou and Skourides, 2016; The´ ry et al., 2005). These retrac-
tion fibers then help guide cell re-spreading as cells exit mitosis
and enter G1 (Cramer and Mitchison, 1993).
Here, using hTERT-immortalized RPE1 cells (Bodnar et al.,
1998) as a model system in which to investigate the dynamics
and function of mitotic adhesion re-modeling, we show that,
while most of the components of focal adhesions are lost as
cells enter into mitosis, active integrins remain in place. The
integrin-based contacts that persist throughout mitosis decorate
the portions of the cell that remain attached to the substrate as
cell margins retract during mitotic rounding. As a result, they
are able to assist in the division process by guiding the rapid
re-spreading of cells as they exit mitosis—as demonstrated by
the fact that mitotic RPE1 cells fail to divide when they are
removed from the substrate. This is especially the case for cells
with a compromised actomyosin cortex, where mitotic adhe-
sions are essential to enable daughter cells migrating away
from one another to undergo abscission, as described previously
in Dictyostelium and other systems (Kanada et al., 2005, 2008;
Nagasaki et al., 2009; Neujahr et al., 1997). Taken together, these
data suggest that it is adhesion remodelling, not actomyosin ring
formtion, that plays the dominant role in enabling the division of
adherent non-transformed human cells in culture.
RESULTS
Adhesion Re-modeling during Mitotic Cell Rounding
To explore the dynamics of adhesion re-modeling that accom-
pany changes in cell shape during passage through mitosis
(reviewed by Ramkumar and Baum, 2016), we chose to image
mitotic progression in an adherent, migratory, diploid human
cell line: RPE1-hTERT cells (Bodnar et al., 1998), which are
widely used to study the cell cycle, cell division, and cell
migration in cell culture and, as such, provide an ideal model
for this analysis of mitotic adhesion re-modeling. In these cells,
we found that many components of focal adhesions present in
interphase were lost upon entry into mitosis (Figures 1A and
1B). This included zyxin, which we followed in cells engineered
to stably express zyxin-GFP—a component of the interphase
focal adhesion complex (Kanchanawong et al., 2010). The
exception to this rule was active b1-integrin (Figure 1A), which
remained in punctae at the interface between the cell and the
substrate throughout mitosis (Lock et al., 2017).
When we followed adhesion re-modeling live in zyxin-GFP
RPE1 cells, we saw that, while the cells were highly motile in
interphase, in the minutes prior to entry into mitosis they stopped
moving, stopped extending lamellipodia, and stopped gener-
ating new zyxin-positive puncta (Figures 1B and 1D). The com-
plete set of zyxin-positive focal adhesions (and paxillin puncta
[Figure 1A; Marchesi et al., 2014]) were then rapidly lost as cells
rounded up (Figures 1B and 1D), with a timing that varied be-
tween puncta (Figure 1E) and between cells (relative to the onset
of prometaphase as seen by the influx of GFP into the nucleus
due to nuclear envelope permeabilization) (Figure 1C). However,
despite having lost their full complement of focal adhesion
complexes, these mitotic RPE1 cells remained attached to the
substrate by thin retraction fibers (Cramer and Mitchison,
1995; Fink et al., 2011; The´ ry et al., 2005), and a small number
of relatively thick, long linear attachments, which we call ‘‘tails.’’
Seventy-nine percent of mitotic cells exhibited tails, 52% of
which were bipolar, with tails aligned along the interphase long
cell axis (N = 76 cells from 13 experiments).
To further explore the path of adhesion re-modeling in RPE1
cells, we fixed and stained zyxin-GFP expressing cells that had
previously been imaged live (using an online fixation protocol
[Almada, 2017]) for active b1-integrin. This revealed b1-integ-
rin-rich puncta decorating both the retraction fibers and the
thicker tails that remained following mitotic rounding—at pre-
cisely the same positions as the zyxin-positive focal adhesions
that were lost during prophase (Figures 1F and S1A). This
suggests that, as focal adhesion complexes are disassembled
during entry into mitosis, they leave behind a stable pool of active
b1-integrin, which serves as a molecular memory of the adhesion
pattern and of interphase cell shape (Lock et al., 2017). We note
that adhesion re-modeling was similar in HeLa Kyoto cells, even
though they round without leaving tails (Figures S2A and S2B);
where it depends on Rap1 inactivation (Lancaster et al., 2013)
and is essential for reliable cell division (Figures S2C–S2E).
To assess the function of these substrate attachments in
RPE1 cells, we used the zyxin-GFP line to track cells as they
underwent a complete cycle of rounding and division (Fig-
ure 2A). Strikingly, although RPE1 cells underwent persistent
directional migration in interphase, this directionality was
erased as cells passed through mitosis (Figure S1B), since the
vast majority of daughter cells migrated away from one another
following division. This was most evident for cells cultured on
Figure 1. Adhesion Re-modeling during Mitotic Cell Rounding
(A) Images depict fixed RPE1 cells in interphase and metaphase stained for talin, active b1-integrin (magenta) and paxillin (yellow). Merge also shows DAPI (blue).
One basal z slice. Scale bar, 20 mm.
(B) Graph showing loss of zyxin-positive adhesion sites as cells progress through mitosis. N = 10 cells from seven experiments. Mean ± SD.
(C) Graphs show adhesion loss relative to nuclear envelope permeabilization (NEP) in six sample cells at mitotic entry.
(D) Image depicts adhesion re-modeling in a representative RPE1 cell stably expressing zyxin-GFP rounding up as it enters mitosis. One basal z slice. Time is
shown relative NEP. Scale bar, 20 mm.
(E) Zoom of boxed region in (D) with three neighboring adhesions highlighted as they are formed and eventually lost. Time is shown relative to NEP. Scale
bar, 2 mm.
(F) RPE1 cells stably expressing zyxin-GFP were released from CDK1 inhibition to synchronize entry into mitosis (see the STAR Methods). The left hand image
shows a basal z plane of a representative cell (from a sample of 15 cells from 3 experiments), imaged during rounding, which was fixed in metaphase and stained
on the microscope to visualize actin (magenta) and integrins (yellow) using phalloidin TRITC and b1-integrin antibody. Scale bar, 20 mm (see Figure S1 for another
sample cell which leaves cytoplasmic tails during mitosis). The right hand image shows a montage of the boxed region of the basal plane of the cell, in which
integrin staining following fixation (yellow) has been overlayed with stills fromlive zyxin-GFP imaging (magenta). Scale bar, 20 mm.
134
Developmental Cell 45, 132–145, April 9, 2018

----!@#$NewPage!@#$----
0
5
10
15
****
Without
tails
With
tails
Anaphase to
re-spreading (min)
B
A
C
T ole phase
Metaphase
Interphase
RPE1 V-Integrin-GFP Live
Time
XY
Kymograph
E
D
-3min
+3
Anaphase
+15
+9
+24
RPE1 Zyxin-GFP Live
+6
+12
+18
+21
+18
+21
+24
+27
+30
+48
-9min
-3
NEP
+3
+9
+15
RPE1 Zyxin-GFP Live
+48
-9min
(legend on next page)
Developmental Cell 45, 132–145, April 9, 2018
135

----!@#$NewPage!@#$----
micro-patterned lines (Figures S1B and S1C), but similar results
were seen on non-patterned substrates (data not shown).
Strikingly, in these same movies, many of the focal adhesions
disassembled during rounding were seen reforming at the
same positions during polar re-spreading in G1 (Figures 2A
and 2B)—as described previously for retraction fibers (Cramer
and Mitchison, 1993). In line with the data from fixed cells (Fig-
ures 1F and S1A), many integrin-based adhesions in live cells
labeled with ectopically expressed aV-integrin remained in
place from mitotic entry to exit (Figure 2C). Finally, daughter
cells that adhered well to the substrate throughout mitosis via
extensive tails re-spread sooner than those that did not (Figures
2D–2E). Together, these results demonstrate that cells entering
mitosis re-model their focal adhesion complexes to leave integ-
rin-based attachments, which aid daughter cell re-spreading at
mitotic exit.
The Importance of Cell-Substrate Adhesion for RPE1
Cell Division
Having followed adhesion re-modeling during mitotic entry, we
wanted to determine whether mitotic cell-substrate adhesion is
important for cell division in RPE1 cells. To do so, we followed
cells as they progressed through mitosis in the absence of a
substrate. Trypsin-EDTA was used to remove cells from a tissue
culture dish, which were then plated as single cells into fibro-
nectin-coated control or non-adherent PLL-PEG-coated micro-
wells. Importantly, while the lack of adhesion induced cycle
arrest in the vast majority of the cells in PEG micro-wells (data
not shown), a proportion of cells, presumably those close to
the G2/M boundary, were still able to enter mitosis while sus-
pended. Despite undergoing mitotic exit and furrow formation,
all of these cells failed to complete abscission (Figures 3A and
3B). This compares with a failure rate of 11.5% for cells plated
in
control
fibronectin-coated
micro-wells
(Figure
3B).
By
contrast, HeLa cells were able to divide in suspension culture
under the same conditions (Figures 3C and 3D). Thus, adhesion
is required for a successful division in RPE1 cells, even though
they can exit mitosis and form a cytokinetic furrow without it.
To test whether the division failure observed in RPE1 cells
passing through mitosis in suspension is due to a functional
requirement for adhesion underneath the cytokinetic furrow,
we plated cells on ring-shaped adhesive islands. At mitotic
exit, the polar regions of these cells re-spread over the perimeter
of the adhesive ring, away from the non-adherent division site.
Strikingly, all of these cells (11/11 cells in 3 experiments),
completed division successfully (Figure S3A), suggesting that
cells do not need to establish adhesions under the cytokinetic
ring to complete cell division. In line with this conclusion, the
region under the furrow rose up off the substrate as RPE1 cells
divided on an adherent substrate (FigureS3B). Furthermore,
the rate of furrow closure was similar in cells dividing on an ad-
hesive structure and over a non-adherent hole (furrow closure
on disc 19.37 ± 2.32 nm/s, N = 5 cells; on ring 18.59 ±
1.55 nm/s, N = 3 cells). Thus, while adhesion is required for cell
division in RPE1 cells, this does not appear to reflect a require-
ment for adhesion under the furrow, as had been reported in
some other systems (Pellinen et al., 2008).
Since adhesions are sites at which cells exert traction on the
substrate, it was also important to test the extent to which trac-
tion is required for cell division. When we asked this question by
imaging cells dividing on gels of differential stiffness, we saw
that, while RPE1 cells were able to divide on an ECM-coated
1.5 kPa gel, the majority of cells failed to divide on soft 0.5 kPa
gels (Figures 3E and S3C). Thus, both cell-substrate adhesion
and traction are required for normal RPE1 cell division.
Integrin-Based Protrusions Allow Division in Ect2-
Depleted Cells
We also wanted to determine whether the persistent mitotic
cell-substrate adhesions we observed are sufficient to enable
cells to divide in the absence of a visible actomyosin ring. We
considered this possibility since, while most eukaryotic cells,
including in HeLa cells (Figure 4A), require an actomyosin ring
for division, which is assembled downstream of the RhoGEF
Ect2 and Rho (reviewed by Green et al., 2012), studies using
several types of adherent eukaryotic cells, e.g., Dictyostelium,
NRK and HT1080 fibrosarcoma cells (Kanada et al., 2005,
2008; Nagasaki et al., 2009), have suggested that cells with a
compromised actomyosin ring can divide via an alternative
mechanism as daughter cells migrate away from one another.
To test whether this is also the case for RPE1 cells, we used
Ect2 RNAi to compromise formation of an actomyosin ring in
cells exiting mitosis (Figure 4B). Importantly for this analysis,
RNAi-mediated silencing of Ect2 was sufficient to deplete cells
of the protein (Figures S4A and S4B), so that Ect2 was no longer
visible at the mid-zone of cells exiting mitosis (Figure S4B).
Although this compromised the assembly of the actomyosin
cortex (Figures S4C–S4E) and mitotic rounding (Figure S4F),
as reported previously (Matthews et al., 2012), it did not alter
the timing of adhesion re-modeling (Figures S4G and S4H).
Strikingly, 58% of these Ect2-depleted cells divided (N = 98
cells from 8 experiments) (Figure 4B). Moreover, the chances
of a cell undergoing a successful abscission event without an
actomyosin ring was positively correlated with the presence of
adhesive structures linking the cell to the underlying substrate.
Thus, Ect2 siRNA cells with adhesive contacts with the sub-
strate (tails) can divide, while the same cells without tails tend
to fail in division (Figure 4B).
Figure 2. Integrin-Based Attachments, Left at Mitotic Entry, Aid Daughter Cell Re-spreading at Mitotic Exit
(A) Image depicts an RPE1 stably expressing zyxin-GFP entering and then exiting mitosis. One basal z slice. Time is shown relative to NEP. Scale bar, 20 mm.
(B) Image shows an overlay of the cell from (A) during interphase (–9 min) and during re-spreading (+48 min). Scale bar, 20 mm.
(C) Images show an RPE1 cell transiently expressing aV-integrin-GFP in interphase, metaphase, and telophase. The dotted line in the first panel was used to
generate the kymograph in the final panel. Arrows show a representative adhesion which was present during interphase, was maintained following mitotic
rounding, into telophase, as the cell respread. Scale bar, 20 mm.
(D) Montage shows a representative RPE1 cell stably expressing zyxin-GFP, exiting mitosis with one tail. Time is shown relative to anaphase. One basal z slice.
Scale bar, 20 mm.
(E) Graph showing the time from anaphase to the onset of re-spreading in daughter cells that either do or do not inherit a tail. Mean ± SD. N = 6 experiments.
Statistics used t test. ****p < 0.00001.
136
Developmental Cell 45, 132–145, April 9, 2018

----!@#$NewPage!@#$----
To better understand how RPE1 cells are able to divide
following Ect2 depletion, we used immunofluorescence to deter-
mine whether proteins associated with cytokinesis are recruited
to the cell mid-zone following Ect2 RNAi-mediated silencing. In
these cells, levels of F-actin within the neck connecting daughter
cells were significantly reduced relative to the control. In addi-
tion, neither p-myosin nor anillin could be detected accumulating
at the mid-zone of dividing cells treated with small interfering
RNAs targetting Ect2 (Ect2 siRNA cells) (Figures 4C and 4D).
By contrast, while Aurora B was still recruited to the mid-zone
of Ect2 siRNA cells, it was not localized with the same degree
of precision as it was in wild-type cells prior to abscission
Brightfield
A
Tubulin
Brightfield
B
N cells=     26          25     
Adhesive
Non-
Adhesive
0%
20%
40%
60%
80%
100%
% of cells
+100
-20min
-10
Anaphase
+10
+20
HeLa LifeAct-RFP
% of cells
N cells=       36         134        117 
5kPa
0%
20%
40%
60%
80%
100%
1.5kPa
0.5kPa
E
D
C
Succeed division
Fail division
Succeed division
Fail division
Succeed division
Fail division
RPE1-soft substrates
RPE1
HeLa
Actin
+40
-20min
-10
Anaphase
+10
+20
RPE1 Tubulin-GFP
N cells=    13           24     
Adhesive
Non-
Adhesive
0%
20%
40%
60%
80%
100%
% of cells
ns
****
****
Figure 3. The Importance of Cell-Substrate Adhesion for RPE1 Cell Division
(A) Montage of phase and tubulin-GFP expression in a representative RPE1 cell that fails to divide when imaged in a non-adhesive PLL-PEG-coated well, in
suspension. Time is shown relative to anaphase. Wide-field image. Scale bar, 20 mm.
(B) Graph showing the percentage of RPE1 cells which succeed and fail division in adhesive and non-adhesive wells. N = 2 experiments. Statistics used the chi-
square test. ****p < 0.00001.
(C) Montage showing phase and LifeAct-RFP expression in a representative HeLa cell imaged in a non-adhesive PLL-PEG-coated well as it divides in
suspension. Wide-field image. Scale bar, 20 mm.
(D) Graph showing the percentage of HeLa cells which succeed and fail division in adhesive and non-adhesive wells. N = 1 experiment. Statistics used the
chi-square test.
(E) Graph showing the percentage of RPE1 cells which succeed and fail division on substrates of different stiffness. N = 4 experiments. See Figure S3C for
montages of sample cells. Statistics used the chi-square test. ****p < 0.00001.
Developmental Cell 45, 132–145, April 9, 2018
137

----!@#$NewPage!@#$----
I
H
G
Sample cell
Mean 
0
10
20
30
40
0
1
2
3
4
5
Time from anaphase (min)
Neck closure
Length increase
+30
+9
+6
+3m ni
+12
+15
Anaphase
Control siRNA
Ect2 siRNA
0
10
20
30
40
0
1
2
3
4
5
Time from anaphase (min)
Normalised
distance
Neck closure
Length increase
0
20
40
60
0
2
4
6
8
Time from anaphase (min)
Neck closure
Length increase
Neck closure
Length increase
0
20
40
60
0
2
4
6
8
Neck closure
Length increase
Time from anaphase (min)
Normalised
distance
Normalised
distance
Normalised
distance
RPE1 LifeAct-RFP
C
Actin
Anillin
Merge
pMyosin
Merge
Tubulin
Control siRNA
Ect2 siRNA
Zoom
Control siRNA
Ect2 siRNA
Zoom
Aurora B Tubulin
E
MDA-MB-
231 Ect2i
MCF10A
Control
RPMI
Control
% of cells
Succeed division
Fail division
N cells = 74              68             53             50              71             70              90             90  
MCF10A
Ect2i
0%
50%
100%
MDA-MB-
231 Control
RPMI
Ect2i
HeLa
Control
HeLa
Ect2i
B
A
0
2
4
6
8
10
***
Control
siRNA
Ect2
siRNA
Intensity
0
10
20
30
****
Intensity
Control
siRNA
Ect2
siRNA
0
2
4
6
8
****
Intensity
Control
siRNA
Ect2
siRNA
F
Time from anaphase
to abscission (min)
Control
siRNA
Ect2
siRNA
0
50
150
250
100
200
ns
D
Actin
Anillin
pMyosin
Control
siRNA
Ect2 siRNA
+ tails
% of cells
N cells = 59              55             36
Ect2 siRNA
- tails
0%
50%
100%
Succeed division
Fail division
(legend on next page)
138
Developmental Cell 45, 132–145, April 9, 2018

----!@#$NewPage!@#$----
(Figure 4E). LifeAct-GFP recruitment to the midzone was also
compromised in Ect2 siRNA cells (Figure 4G). Together these
data suggest that, at anaphase, while control siRNA cells accu-
mulate a central band of actomyosin and form a narrow furrow,
which closes to form the neck that separates the two new
daughter cells (Figures 4G–4I), before undergoing abscission
some time later as described previously (Fededa and Gerlich,
2012; Lafaurie-Janvore et al., 2013), division in Ect2 siRNA-
treated cells occurrs in the absence of a visible actomyosin
ring or a furrow. Despite this, however, the anaphase cortex of
these cells remains polarised, since actin is lost from opposing
poles as Ect2 siRNA cells exit mitosis (Figure 4G). Then, as
daughter cells re-spread and move away from each other, the
connection linking them slowly thins—culminating in cell division
(Figures 4G–4I). Similar results were seen in RPE1 cells treated
with a ROCK inhibitor, which compromises actomyosin ring for-
mation (Figures S4I and S4J; Kanada et al., 2008). Strikingly, this
delay in the rate of neck closure in Ect2 siRNA cells did not trans-
late into a delay of abscission timing (Figure 4F), which was
similar in control and Ect2 RNAi cells. Together these data point
to a critical role for mitotic adhesions in the process of division in
Ect2-depleted cells.
Re-spreading of Daughter Cells Is Required for Division
in Ect2-Depleted Cells
In these experiments, the ability of daughter cells to migrate away
from one another as they exited mitosis appeared to determine
whether or not they succeed in dividing. This suggests, in line
with work in Dictyostelium (Neujahr et al., 1997), that adhesion-
dependent migration (Burton and Taylor, 1997) is required to
generate the traction forces that allow the division of cells in
which actomyosin ring formation has been compromised. As a
test of this hypothesis, we constrained daughter cell movement
by plating cells on micro-patterns. We used patterns of different
sizes that were either circular or elliptical in shape (1:2.5 minor:
major axis). Although control siRNA cells successfully completed
division on all patterns tested (Figures 5A and 5B, top row), Ect2
siRNA-treated cells tended to fail during division on small,
less-elongated patterns (Figure 5A, middle row and 5C). In cases
of division failure, the neck separating daughter cells appeared
to narrow as they exited mitosis, before widening again when
opposing poles of the re-spreading daughter cells reached the
pattern edge (Figures 5A and 5B, middle row). As expected, as
the size and length of patterns was increased, the chances of a
successful division began to approach levels seen for cells on
non-patterned substrates (Figures 5A and 5B, bottom row, and
5C). Similarly, for cells growing on a non-patterned substrate,
the ability of cells to re-spread (measured at 21 min after
anaphase) was positively correlated with the likelihood of their
undergoing a successful division (Figure 5D).
When traction forces were measured during RPE1 cell divi-
sions, tensile forces were found dropping to baseline levels as
cells entered mitosis (Figures S5A–S5G), implying that the tails
that persist through mitosis in RPE1 cells grown an adherent
substrate bear almost no tension. At mitotic exit, traction forces
were then re-established at cell poles (Figures S5A–S5G), in
opposite directions (Figures S5E and S5F). By analogy with a
tug-of-war, the force acting at the neck could be readily
computed as the unbalanced traction exerted at the cell-sub-
strate interface by each of the two daughter cells (Figure S5G).
As daughter cells began to migrate away from each other, this
force was found to increase, peaking at �25 nN (Figures S5H–
S5J). Strikingly, this is of the same order of magnitude as the
tugging force between cells undergoing collective migration (La-
bernadie et al., 2017) and was similar in control and Ect2 RNAi
cells (Figure S5J).
Taken together, these data suggest that scission of the phys-
ical connection linking migrating daughter cells to one another
depends on continuous and polarized lamellipodial extension
at opposing daughter cell poles. This generates tension across
the connection between daughter cells, narrowing the bridge,
enabling abscission in the absence of a visible actomyosin
ring. As a test of this idea, we treated cells with a small molecule,
the Arp2/3 inhibitor CK666 (Nolen et al., 2009), which inhibits
lamellipodial formation and adhesion-dependent cell migration.
Although few cells entered mitosis in the presence of the Arp2/
3 inhibitor, those Ect2 RNAi cells that entered and progressed
through mitosis failed to narrow the region between daughter
Figure 4. Integrin-Based Protrusions Allow Division in Ect2-Depleted Cells
(A) Graph depicting the percentage of cells which succeed or fail division in control siRNA cells, and Ect2 siRNA for four cell types: HeLa, MCF10A (N = 2 ex-
periments), RPMI, and MDA-MB-231 (N = 3 experiments).
(B) Graph depicting the percentage of RPE1 cells which succeed or fail division in control siRNA cells, Ect2 siRNA cells without tails, and Ect2 siRNA cells with
tails. N = 8 experiments. See also Figures S4I and S4J for comparable data with a ROCK inhibitor.
(C) Images show RPE1 control siRNA and Ect2 siRNA cells fixed and stained with phalloidin TRITC (magenta), anillin (yellow), tubulin (magenta), and p-myosin
(yellow) antibodies. DAPI is shown in blue in the merge. One medial z stack. Scale bar, 20 mm. See also Figure S2B for Ect2 antibody staining.
(D) Graphs quantifying of the loss of actin (statistics used t test), anillin (statistics used Mann-Whitney test) and p-myosin (statistics used Mann-Whitney test)
proteins from bridge connecting daughter cells, where the neck measures less than 5 mm. For each cell a 10 310 px (1.1 3 1.1 mm) square in the neck is
normalized to the average intensity of two identical sized boxes in each daughter cell cytoplasm. Mean ± SD. N = 1 experiment. ****p < 0.00001.
(E) Images depicting sample cells treated with either control or Ect2 siRNAs stained for Aurora B (yellow) and tubulin (magenta). Scale bar, 20 mm. Zoom of the
boxed regions shows the mid-body. Scale bar, 2 mm.
(F) Graph showing time from anaphase to abscission in control siRNA and Ect2 siRNA cells. Mean ± SD N = 7 experiments. Statistics used Mann-Whitney test.
(G) Montage of control siRNA and Ect2 siRNA RPE1 cells stably expressing LifeAct-RFP as they exit mitosis. The magenta arrow shows the measurement
taken over time of decreasing neck width. The blue arrow shows the measurement taken over time of the increase in length between the polar leading edge of the
two daughter cells as they migrate away from each other. The Ect2 RNAi cell depicted here re-spreads faster than the average. A 5 mm maximum projection of
basal-medial cell. Scale bar, 20 mm. See also Figure S3D for comparable data in cells treated with a ROCK inhibitor.
(H) Graphs depict the rate at which the width of the connection linking daughter cells decreases and the distance between the daughter cell poles increases, as
cells divide. Data have been normalized to the first time point.
(I) Graphs depicting the mean rate at which the width of the connection linking daughter cells decreases and the distance between daughter cell poles increases,
as cells divide. Data have been normalized to the first time point. N = 12 cells from 2 experiments. Error bars show SD.
Developmental Cell 45, 132–145, April 9, 2018
139

----!@#$NewPage!@#$----
B
A
0
20
40
60
0
50
100
150
Time from anaphase (min)
Neck closure
Length increase
Distance ( m)
0
20
40
60
0
50
100
150
Time from anaphase (min)
Distance ( m)
Neck closure
Length increase
0
20
40
60
0
50
100
150
Time from anaphase (min)
Distance ( m)
Neck closure
Length increase
Anaphase
+6min
+12
+18
+24
+30
+36
+42
+48
+54
Ect2 siRNA
2800 m2  ellipse
Control siRNA
700 m
2
  circle
Ect2 siRNA
RPE1 Zyxin-GFP on Micropatterns
C
E
D
700 m2
700 m2 2800 m2
2800 m2
Circle
Ellipse
Succeed division
Fail division
RPE1 Ect2 siRNA
N cells=  11          13           8           19
% of cells
0%
20%
40%
60%
80%
100%
RPE1 Ect2 siRNA
 DMSO
Arp2/3
inhibitor
N cells=  30           7
% of cells
0%
20%
40%
60%
80%
100%
Succeed division
Fail division
Control
siRNA
Ect2 
siRNA
Spread length ( m)
Succeed division
Fail division
0
50
100
150
*
RPE1
**
F
G
N cells=163         105
% of cells
0%
20%
40%
60%
80%
100%
Control
siRNA
Ect2 
siRNA
Succeed division
Fail division
Control
siRNA
Ect2
siRNA
N cells= 42          36           52          52
% of cells
0%
20%
40%
60%
80%
100%
Succeed division
Fail division
PLL
Control
siRNA
Ect2
siRNA
Fibronectin
RPE1
PE1
R
****
(legend on next page)
140
Developmental Cell 45, 132–145, April 9, 2018

----!@#$NewPage!@#$----
cells. Instead, they simply re-spread as large binuclear cells (Fig-
ure 5E); supporting the idea that Arp2/3-dependent migration is
required for division under these conditions.
Finally, to determine whether Ect2-depleted cells require adhe-
sions and retraction fibers laid down during mitotic rounding to
divide, we removed control and Ect2 RNAi cells that had been
arrested in mitosis from the substrate. Cells were then re-plated
on fibronectin and followed as they re-spread and exited mitosis.
Strikingly, while almost all control siRNA cells successfully divided
under these conditions, the vast majority of Ect2 siRNA cells failed
(Figure 5F). This suggests that it is the cell-substrate contacts left
during the adhesion re-modeling process that accompanies
mitotic rounding that guide adhesion-mediated division in cells
compromised in their ability to construct an actomyosin ring. Simi-
larly, when we grew cells on a non-specific adhesive substrate,
PLL, to prevent the formation of tails during mitotic rounding,
the vast majority of Ect2 RNAi cells failed in division (Figure 5G).
DISCUSSION
In this study, we present a detailed analysis of the dynamics
and function of mitotic adhesion re-modeling. This reveals a
role for the adhesions laid down upon entry into mitosis in
cell division - both in cells with a normal actomyosin cortex
and in those that lack a contractile actomyosin ring. Impor-
tantly, our data suggest that, in order to undergo division in
the absence of a visible contractile ring, RPE1 cells must
retain the adhesions generated during mitotic rounding, since
Ect2 RNAi cells fail to divide when they are re-plated from sus-
pension in metaphase onto an adhesive substrate. Adhesion
can aid division under these conditions because, while the
peripheral proteins normally associated with focal adhesion
complexes dissociate from cell-substrate adhesions during
the process of rounding, stable active b1-integrin puncta
remain throughout mitosis, decorating retraction fibers and
tails that connect the cell to the substrate (Figures 1 and
6A). These adhesions may guide spindle orientation (Fink
et al., 2011; The´ ry et al., 2005). They also aid daughter cell
re-spreading (Cramer and Mitchison, 1993) and cell division
(Figures 3A, 3B, and 6B). In addition, in cells with a compro-
mised actomyosin cortex, these integrin-based adhesions
are absolutely essential for cell division, as was shown to be
the case for cell-substrate adhesion in Dictyostelium and
some other systems (Neujahr et al., 1997; Nagasaki et al.,
2002, 2009; Kanada et al., 2005, 2008). In line with a role for
mitotic adhesions in this process, HeLa Kyoto cells, which
fail to generate long adherent tail-like structures when they
round up and enter mitosis (Matthews et al., 2012), fail to
divide following Ect2 siRNA, but can be induced to undergo
adhesion-dependent division through the expression of an
activated form of Rap1 (Figures S5K, 6CI, and 6CII).
Most animal cells appear to use an Ect2-dependent actomy-
osin-based cytokinetic ring to close the plasma membrane
around the spindle, forming a V-shaped neck, which provides
a substrate for mid-body assembly and abscission. By contrast,
when undergoing an adhesion-dependent division, RPE1 cells
do not assemble a visible actomyosin ring, do not accumulate
anillin and p-myosin II at the future division site, and do not
form a V-shaped neck as they divide. Instead, as daughter cells
move away from one another, the bridge connecting RPE1
sisters slowly thins as the result of traction forces generated by
the polar migration of daughter cells away from one another,
culminating in division. A morphologically similar division
through daughter cell migration has been reported for NRK
and HT1080 fibrosarcoma cells (Kanada et al., 2005) treated
with 30 mM blebbistatin, and a similar migration-based division
mechanism has been reported to enable multinucleate cells in
interphase to undergo fission in interphase (Ben-Ze’ev and
Raz, 1981). Thus, adhesion-dependent division is likely to be a
general phenomenon, which relies on the transmission of trac-
tion forces from the leading edge of polarized migrating cells to
the cytokinetic bridge.
In Dictyostelium, the forces driving adhesion-based cell division
depend on the polarized activity of the Arp2/3 activator, SCAR/
WAVE (King et al., 2010). Similarly, in our system, cell division in
the absence of a visible Ect2-based actomyosin ring depends
on Arp2/3. In this, our work parallels recent developments in the
field of cell migration, where it has recently become clear that
there are a variety of mechanisms by which cells can move.
Thus, it has been proposed that cells can use polarized Rho activ-
ity to squeeze themselves forward, or can use polarized Rac1 ac-
tivity and adhesion to pull themselves forward (L€ammermann and
Sixt, 2009; Sanz-Moreno and Marshall, 2010). In a similar manner,
we suggest that dividing cells can use high Rho activity and for-
min-nucleated actomyosin assemblies to squeeze daughter cells
apart (Figure 6A*), or can employ Rac1, SCAR/WAVE, and Arp2/3
activity, together with adhesion, to generate directional traction
forces to pull daughter cells apart (Figure 6C**).
How then does abscission occur in these conditions? First
tension must be transmitted across the dividing cell to thin the
connection linking daughter cells. How this occurs requires
further investigation. One possibility; however, is that cytokinesis
relies on membrane tension-dependent signaling from the lamel-
lipodium to the cell rear, as described previously for migrating
Figure 5. Re-spreading of Daughter Cells Is Required for Division in Ect2-Depleted Cells
(A) Montages showing control siRNA and Ect2 siRNA cells plated on micro-patterned fibronectin discs surrounded by PLL-PEG. The dotted line denotes the
pattern shape. Wide-field image. Scale bar, 20 mm.
(B) Graphs depicting the decreasing neck width and increasing length between the leading edges of daughter cells during division.
(C) Graph showing the percentage of Ect2 siRNA cells that succeed and fail division on each pattern type. N = 7 experiments.
(D) Graph showing the length of daughter cells 21 min after anaphase in control siRNA and Ect2 siRNA cells that succeed or fail division. N = 3 experiments.
Median ± interquartile range. Statistics used t test. *p < 0.01
(E) Graph showing the percentage of Ect2 siRNA cells in the presence or absence of 300 mM Arp2/3 inhibitor that succeed and fail division. N = 4 experiments.
Statistics used chi-square test. **p < 0.001.
(F) Graph showing the percentage of control and Ect2 siRNA cells that succeed and fail division when re-plated from suspension at mitotic exit. N = 4 technical
replicates from 1 experiment. Statistics used chi-square test. ****p < 0.00001.
(G) Graph showing the percentage of control and Ect2 siRNA cells that succeed and fail division when plated on either fibronectin or PLL.
Developmental Cell 45, 132–145, April 9, 2018
141

----!@#$NewPage!@#$----
cells (Diz-Mun˜ oz et al., 2016). Interestingly, in HeLa cells, the
movement of daughter cells away from one another following
mitotic exit has been shown to delay rather than aid cell
separation (Lafaurie-Janvore et al., 2013). This, the authors
suggested, is a result of tensile forces across the bridge con-
necting daughter cells inhibiting ESCRTIII-mediated abscission
(Lafaurie-Janvore et al., 2013); an effect that may be due in
part to the effect of tension on the actomyosin cortex in the
bridge (Figure 2H; Lafaurie-Janvore et al., 2013) since actin
can inhibit abscission if it is not cleared from the mid-body region
in a timely manner (Echard, 2012). These data imply that assem-
bly of the mid-body during the process of ring contraction (Hu
et al., 2012) functions to inhibit abscission until the appropriate
time, when Spastin and the ESCRTIII machinery are recruited
to induce abscission at a nearby site (Connell et al., 2009;
Guizetti et al., 2011). This may be an important feature of many
systems, e.g., epithelia, where connections between daughter
cells should be established prior to abscission in order to prevent
tissue disruption (Herszterg et al., 2013). It may also make the
system subject to checkpoint-mediated control (Norden et al.,
2006; Steigemann et al., 2009). However, as we show in this
paper, efficient abscission does not absolutely require a well-
structured mid-body formed through cortical ring contraction,
since abscission time is not impaired in Ect2-depleted cells
(Figures 3H and S3D). Moreover, as we show, stretching of the
bridge appears to be essential for division in RPE1 cells that
lack an actomyosin ring. These considerations suggest that the
primary goal of cytokinesis is to thin the bridge connecting the
daughter cells so that it is sufficiently narrow to provide a good
substrate for the abscission machinery (Guizetti et al., 2011;
Mierzwa and Gerlich, 2014).
Finally, this study highlights two different types of division sys-
tem in human cells in culture. Transformed cells, such as HeLa
cells, tend to be very good at mitotic rounding (Matthews and
Baum, 2012), likely as a result of having relatively weak cell-sub-
strate adhesion at mitotic entry and a highly contractile cortical
actomyosin network. As a consequence, these cells are unable
to use adhesions as an aid to division. Instead, like the first divi-
sions in early embryos, they rely on cell-autonomous cues to
centrally position Ect2 and assemble a robust contractile acto-
myosin ring in order to divide—making them an ideal system in
which to study actomyosin ring formation and function (Bodnar
Other focal adhesion proteins
e.g. HeLa or RPE1 on 
non-adherent/soft substrate
Little mitotic adhesion
Little traction forces
No contractile ring
Z
X
DNA
Plasma Membrane
Actin
Integrin
Adhesion
Contractility
Adhesion
Contractility
Adhesion
Contractility
X
B
A
C
I
II
**
*
e.g. RPE1
e.g. HeLa
e.g. RPE1 or
HeLa Rap1 Q63E
Spread mitotic adhesion
Polar traction forces
No contractile ring
Figure 6. Model
(A) A cell rounding to leaveintegrin-positive adhesive contacts before undergoing acto-myosin-based cytokinesis using high Rho activity (*).
(B) A cell undergoing mitosis in suspension and either failing or succeeding to complete division, depending on cell type.
(C) A cell treated with Ect2 siRNA and either failing (I) or succeeding (II) to divide in a manner that depends on Rac-Arp2/3-dependent respreading at mitotic
exit (**).
142
Developmental Cell 45, 132–145, April 9, 2018

----!@#$NewPage!@#$----
et al., 1998; Kotynova´ et al., 2016; Matthews et al., 2012). As a
result, HeLa cells divide in a manner that is relatively independent
of their environment, e.g. in suspension (Figures 3C and 3D) and/
or in soft agar (Cox, 1997). In a previous study, V12 H-Ras-trans-
formed fibroblasts successfully divided in suspension, whereas
the corresponding control cells did not (Thullberg et al., 2007),
suggesting that oncogenic signaling may itself help override
the requirement for adhesion, perhaps by increasing cortical ten-
sion. By contrast, as our study makes clear, adherent non-trans-
formed RPE1 cells rely on interactions with the extracellular envi-
ronment to divide. As a result, they are unable to divide in
suspension. However, because RPE1 cells tend to round rela-
tively little when they enter mitosis, they are very good at using
adhesive cues in the environment to guide spindle positioning
(The´ ry et al., 2005). Although more work must be done before
one can conclude that these findings represent a general mech-
anistic difference between the way normal and cancer cells
divide, this may go some way toward explaining the differences
between cells that divide using an autonomous actomyosin ring,
which we previously suggested might be a general feature of
metastatic cancer cell divisions (Matthews and Baum, 2012),
and adhesion-based cell division mechanisms (Kanada et al.,
2005, 2008; Nagasaki et al., 2009; Neujahr et al., 1997) that
rely on intimate contact between a cell and its extracellular envi-
ronment. If so, in future work it will be important to determine
how cancer progression alters the way adhesion and cortical
contractility are remodelled as cells enter mitosis.
STAR+METHODS
Detailed methods are provided in the online version of this paper
and include the following:
d KEY RESOURCES TABLE
d CONTACT FOR REAGENT AND RESOURCE SHARING
d EXPERIMENTAL MODEL AND SUBJECT DETAILS
B Cell Lines and Culture
d METHOD DETAILS
B Plasmid Transfection
B Ect2 siRNA
B Drug Treatments
B Micropatterning and Surface Treatment
B Wells
B Cell Fixation and Immunostaining
B Live Cell Imaging
B Western Blot
B Preparation of Polyacrylamide Gels
d QUANTIFICATION AND STATISTICAL ANALYSIS
B Traction Microscopy
B Statistical Analysis
SUPPLEMENTAL INFORMATION
Supplemental Information includes five figures and can be found with this
article online at https://doi.org/10.1016/j.devcel.2018.03.009.
ACKNOWLEDGMENTS
C.L.D. was supported by PhD funding from the Medical Research Council, UK
(1214605). H.K.M. was supported on a Cancer Research UK (CRUK) pro-
gramme grant (C1529/A17343) and, along with Z.W., by a CRUK/EPSRC
MDPA grant (C1529/A23335). M.U. and X.T. received funding from the Span-
ish Ministry of Economy, Industry and Competitiveness/FEDER (BES-2013-
062633 and BFU2015-65074-P). X.T. was also funded by the Generalitat de
Catalunya (Cerca Program and 2014-SGR-927), the ERC (CoG-616480) and
the European Commission (project 731957). L.W. was funded by a grant
from the Deutsche Forschungsgemeinschaft (DFG) ‘‘Hallmarks of Skin Can-
cer’’ (RTG2099). Research by B.B. was supported by UCL, Cancer Research
UK (C1529/A17343), and the MRC (MC_CF12266). Research by R.H. and P.A.
was supported by grants from the UK BBSRC (BB/M022374/1; BB/P027431/
1; and BB/R000697/1) and UK MRC (MR/K015826/1 and MC_CF12266). The
authors would like to thank Martial Ballard, Matthieu Piel, Nitya Ramkumar,
and Jeremy Carlton for useful comments on the text.
AUTHOR CONTRIBUTIONS
C.L.D. conceived and carried out most of the experiments and wrote the
manuscript, together with B.B.. H.K.M. conceived and carried out the work
testing whether adhesion is required for division. M.U. carried out traction
force microscopy and soft gel experiments. S.M. and L.W. carried out mi-
cro-patterning experiments. N.H. carried out live and fixed imaging in HeLa
with and without Rap1 Q63E. Z.W. carried out the RPE1 migration experi-
ments. P.A. carried out the online fixation experiments with C.L.D. R.H. helped
oversee the work of P.A. X.T. helped oversee the work of M.U. M.B. helped
oversee the work of L.W. B.B. helped to conceive and oversee the entire proj-
ect and the writing.
DECLARATION OF INTERESTS
The authors declare no competing interests.
Received: June 23, 2017
Revised: January 17, 2018
Accepted: March 13, 2018
Published: April 9, 2018
REFERENCES
Almada, P. (2017). Developing Highly Multiplexed Technology for High-
throughput Super-resolution Fluorescence Microscopy (University College
London).
Bement, W.M., Benink, H.A., and Von Dassow, G. (2005). A microtubule-
dependent zone of active RhoA during cleavage plane specification. J. Cell
Biol. 170, 91–101.
Ben-Ze’ev, A., and Raz, A. (1981). Multinucleation and inhibition of cytokinesis
in suspended cells: reversal upon reattachment to a substrate. Cell 26,
107–115.
Bodnar, A.G., Ouellette, M., Frolkis, M., Holt, S.E., Chiu, C.-P., Morin, G.B.,
Harley, C.B., Shay, J.W., Lichtsteiner, S., and Wright, W.E. (1998). Extension
of life-span by introduction of telomerase into normal human cells. Science
279, 349–352.
Burkard, M.E., Maciejowski, J., Rodriguez-Bravo, V., Repka, M., Lowery,
D.M., Clauser, K.R., Zhang, C., Shokat, K.M., Carr, S.A., Yaffe, M.B., et al.
(2009). Plk1 self-organization and priming phosphorylation of HsCYK-4 at
the spindle midzone regulate the onset of division in human cells. PLoS Biol.
7, e1000111.
Burton, K., and Taylor, D.L. (1997). Traction forces of cytokinesis measured
with optically modified elastic substrata. Nature 385, 450–454.
Connell, J.W., Lindon, C., Luzio, J.P., and Reid, E. (2009). Spastin couples
microtubule severing to membrane traffic in completion of cytokinesis and
secretion. Traffic 10, 42–56.
Cox, G.S. (1997). Correlation in HeLa cells of anchorage-independent growth
and synthesis of the glycoprotein hormone a-subunit. Biochem. Biophys. Res.
Commun. 233, 425–431.
Cramer, L.P., and Mitchison, T.J. (1993). Moving and stationary actin filaments
are involved in spreading of postmitotic PtK2 cells. J. Cell Biol. 122, 833–843.
Developmental Cell 45, 132–145, April 9, 2018
143

----!@#$NewPage!@#$----
Cramer, L.P., and Mitchison, T.J. (1995). Myosin is involved in postmitotic cell
spreading. J. Cell Biol. 131, 179–189.
Dao, V.T., Dupuy, A.G., Gavet, O., Caron, E., and de Gunzburg, J. (2009).
Dynamic changes in Rap1 activity are required for cell retraction and spreading
during mitosis. J. Cell Sci. 122, 2996–3004.
Diz-Mun˜ oz, A., Thurley, K., Chintamen, S., Altschuler, S.J., Wu, L.F., Fletcher,
D.A., and Weiner, O.D. (2016). Membrane tension acts through PLD2 and
mTORC2 to limit actin network assembly during neutrophil migration. PLoS
Biol. 14, e1002474.
Dupuy, A.G., L’Hoste, S., Cherfils, J., Camonis, J., Gaudriault, G., and De
Gunzburg, J. (2005). Novel Rap1 dominant-negative mutants interfere selec-
tively with C3G and Epac. Oncogene 24, 4509–4520.
Echard, A. (2012). Connecting membrane traffic to ESCRT and the final cut.
Nat. Cell Biol. 14, 983–985.
Fededa, J.P., and Gerlich, D.W. (2012). Molecular control of animal cell cyto-
kinesis. Nat. Cell Biol. 14, 440–447.
Fink, J., Carpi, N., Betz, T., Be´ tard, A., Chebah, M., Azioune, A., Bornens, M.,
Sykes, C., Fetler, L., Cuvelier, D., et al. (2011). External forces control mitotic
spindle positioning. Nat. Cell Biol. 13, 771–778.
Franco, S.J., Rodgers, M.A., Perrin, B.J., Han, J., Bennin, D.A., Critchley, D.R.,
and Huttenlocher, A. (2004). Calpain-mediated proteolysis of talin regulates
adhesion dynamics. Nat. Cell Biol. 6, 977–983.
Green, R.A., Paluch, E., and Oegema, K. (2012). Cytokinesis in animal cells.
Annu. Rev. Cell Dev. Biol. 28, 29–58.
Guizetti, J., Schermelleh, L., M€antler, J., Maar, S., Poser, I., Leonhardt, H.,
M€uller-Reichert, T., and Gerlich, D.W. (2011). Cortical constriction during
abscission involves helices of ESCRT-III-dependent filaments. Science 331,
1616–1620.
Herszterg, S., Leibfried, A., Bosveld, F., Martin, C., and Bellaiche, Y. (2013).
Interplay between the dividing cell and its neighbors regulates adherens junc-
tion formation during cytokinesis in epithelial tissue. Dev. Cell 24, 256–270.
Hu, C., Coughlin, M., and Mitchison, T.J. (2012). Midbody assembly and its
regulation during cytokinesis. Mol. Biol. Cell 23, 1024–1034.
Hu, C.K., Coughlin, M., Field, C.M., and Mitchison, T.J. (2008). Cell polarization
during monopolar cytokinesis. J. Cell Biol. 181, 195–202.
Hutchins, J.R.A., Toyoda, Y., Hegemann, B., Poser, I., Sykora, M.M.,
Augsburg, M., Hudecz, O., Buschhorn, B.A., Bulkescher, J., Conrad, C.,
et al. (2010). Systematic characterization of human protein complexes iden-
tifies chromosome segregation proteins. Science 328, 593–599.
Kanada, M., Nagasaki, A., and Uyeda, T.Q. (2005). Adhesion-dependent and
contractile ring-independent equatorial furrowing during cytokinesis in
mammalian cells. Mol. Biol. Cell 16, 3865–3872.
Kanada, M., Nagasaki, A., and Uyeda, T.Q. (2008). Novel functions of Ect2 in
polar lamellipodia formation and polarity maintenance during ‘‘contractile
ring-independent’’ cytokinesis in adherent cells. Mol. Biol. Cell 19, 8–16.
Kanchanawong, P., Shtengel, G., Pasapera, A.M., Ramko, E.B., Davidson, W.,
Hess, H.F., and Waterman, C.M. (2010). Nanoscale architecture of integrin-
based cell adhesions. Nature 468, 580–584.
Kimura, K., Tsuji, T., Takada, Y., Miki, T., and Narumiya, S. (2000).
Accumulation of GTP-bound RhoA during cytokinesis and a critical role of
ECT2 in this accumulation. J. Biol. Chem. 275, 17233–17236.
King, J.S., Veltman, D.M., Georgiou, M., Baum, B., and Insall, R.H. (2010).
SCAR/WAVE is activated at mitosis and drives myosin-independent cytoki-
nesis. J. Cell Sci. 123, 2246–2255.
Kiyomitsu, T., and Cheeseman, I.M. (2013). Cortical dynein and asymmetric
membrane elongation coordinately position the spindle in anaphase. Cell
154, 391–402.
Kotynova´ , K., Su, K.C., West, S.C., and Petronczki, M. (2016). Plasma mem-
brane association but not midzone recruitment of RhoGEF ECT2 is essential
for cytokinesis. Cell Rep. 17, 2672–2686.
Kunda, P., Pelling, A.E., Liu, T., and Baum, B. (2008). Moesin controls cortical
rigidity, cell rounding, and spindle morphogenesis during mitosis. Curr. Biol.
18, 91–101.
Labernadie, A., Kato, T., Brugue´ s, A., Serra-Picamal, X., Derzsi, S., Arwert, E.,
Weston, A., Gonza´ lez-Tarrago´ , V., Elosegui-Artola, A., Albertazzi, L., et al.
(2017). A mechanically active heterotypic E-cadherin/N-cadherin adhesion
enables fibroblasts to drive cancer cell invasion. Nat. Cell Biol. 19, 224.
Lafaurie-Janvore, J., Maiuri, P., Wang, I., Pinot, M., Manneville, J.-B., Betz, T.,
Balland, M., and Piel, M. (2013). ESCRT-III assembly and cytokinetic abscis-
sion are induced by tension release in the intercellular bridge. Science 339,
1625–1629.
L€ammermann, T., and Sixt, M. (2009). Mechanical modes of ‘amoeboid’ cell
migration. Curr. Opin. Cell Biol. 21, 636–644.
Lancaster, O.M., Le Berre, M., Dimitracopoulos, A., Bonazzi, D., Zlotek-
Zlotkiewicz, E., Picone, R., Duke, T., Piel, M., and Baum, B. (2013). Mitotic
rounding alters cell geometry to ensure efficient bipolar spindle formation.
Dev. Cell 25, 270–283.
Laukaitis, C.M., Webb, D.J., Donais, K., and Horwitz, A.F. (2001). Differential
dynamics of a5 integrin, paxillin, and a-actinin during formation and disas-
sembly of adhesions in migrating cells. J. Cell Biol. 153, 1427–1440.
Lock, J.G., Jones, M.C., Askari, J.A., Gong, X., Oddone, A., Olofsson, H.,
Go¨ ransson, S., Lakadamyali, M., Humphries, M.J., and Stro¨ mblad, S. (2017).
Reticular adhesions: a new class of adhesion complex that mediates cell-ma-
trix attachment during mitosis. bioRxiv. https://doi.org/10.1101/234237.
Maddox, A.S., and Burridge, K. (2003). RhoA is required for cortical retraction
and rigidity during mitotic cell rounding. J. Cell Biol. 160, 255–265.
Marchesi, S., Montani, F., Deflorian, G., D’Antuono, R., Cuomo, A., Bologna,
S., Mazzoccoli, C., Bonaldi, T., Di Fiore, P.P., and Nicassio, F. (2014).
DEPDC1B coordinates de-adhesion events and cell-cycle progression at
mitosis. Dev. Cell 31, 420–433.
Matthews, H.K., and Baum, B. (2012). The metastatic cancer cell cortex: an
adaptation to enhance robust cell division in novel environments? BioEssays
34, 1017–1020.
Matthews, H.K., Delabre, U., Rohn, J.L., Guck, J., Kunda, P., and Baum, B.
(2012). Changes in Ect2 localization couple actomyosin-dependent cell shape
changes to mitotic progression. Dev. Cell 23, 371–383.
Mierzwa, B., and Gerlich, D.W. (2014). Cytokinetic abscission: molecular
mechanisms and temporal control. Dev. Cell 31, 525–538.
Miki, T., Smith, C.L., Long, J.E., Eva, A., and Fleming, T.P. (1993). Oncogene
ect2 is related to regulators of small GTP-binding proteins. Nature 362,
462–465.
Mitchison, T.J. (1992). Actin based motility on retraction fibers in mitotic PtK2
cells. Cell Motil. Cytoskeleton 22, 135–151.
Nagasaki, A., de Hostos, E.L., and Uyeda, T.Q.P. (2002). Genetic and morpho-
logical evidence for two parallel pathways of cell-cycle-coupled cytokinesis in
Dictyostelium. J. Cell Sci. 115, 2241–2251.
Nagasaki,
A.,
Kanada,
M.,
and
Uyeda,
T.Q.
(2009).
Cell
adhesion
molecules regulate contractile ring-independent cytokinesis in Dictyostelium
discoideum. Cell Res. 19, 236–246.
Neujahr, R., Heizer, C., and Gerisch, G. (1997). Myosin II-independent
processes in mitotic cells of Dictyostelium discoideum: redistribution of the
nuclei, re-arrangement of the actin system and formation of the cleavage
furrow. J. Cell Sci. 110, 123–137.
Nishimura, Y., and Yonemura, S. (2006). Centralspindlin regulates ECT2 and
RhoA accumulation at the equatorial cortex during cytokinesis. J. Cell Sci.
119, 104–114.
Nolen, B.J., Tomasevic, N., Russell, A., Pierce, D.W., Jia, Z., McCormick, C.D.,
Hartman, J., Sakowicz, R., and Pollard, T.D. (2009). Characterization of two
classes of small molecule inhibitors of Arp2/3 complex. Nature 460,
1031–1034.
Norden, C., Mendoza, M., Dobbelaere, J., Kotwaliwale, C.V., Biggins, S., and
Barral, Y. (2006). The NoCut pathway links completion of cytokinesis to spindle
midzone function to prevent chromosome breakage. Cell 125, 85–98.
Pellinen, T., Tuomi, S., Arjonen, A., Wolf, M., Edgren, H., Meyer, H., Grosse, R.,
Kitzing, T., Rantala, J.K., Kallioniemi, O., et al. (2008). Integrin trafficking
regulated by Rab21 is necessary for cytokinesis. Dev. Cell 15, 371–385.
144
Developmental Cell 45, 132–145, April 9, 2018

----!@#$NewPage!@#$----
Petridou, N.I., and Skourides, P.A. (2016). A ligand-independent integrin
b1 mechanosensory complex guides spindle orientation. Nat. Commun.
7, 10899.
Prokopenko, S.N., Brumby, A., O’Keefe, L., Prior, L., He, Y., Saint, R., and
Bellen, H.J. (1999). A putative exchange factor for Rho1 GTPase is required
for initiation of cytokinesis in Drosophila. Genes Dev. 13, 2301–2314.
Ramkumar, N., and Baum, B. (2016). Coupling changes in cell shape to
chromosome segregation. Nat. Rev. Mol. Cell Biol. 17, 511–521.
Rappaport, R. (1985). Repeated furrow formation from a single mitotic
apparatus in cylindrical sand dollar eggs. J. Exp. Zool 234, 167–171.
Rodrigues, N.T.L., Lekomtsev, S., Jananji, S., Kriston-Vizi, J., Hickson, G.R.X.,
and Baum, B. (2015). Kinetochore-localized PP1-Sds22 couples chromosome
segregation to polar relaxation. Nature 524, 489–492.
Rosa, A., Vlassaks, E., Pichaud, F., and Baum, B. (2015). Ect2/Pbl acts via Rho
and polarity proteins to direct the assembly of an isotropic actomyosin cortex
upon mitotic entry. Dev. Cell 32, 604–616.
Sanz-Moreno, V., and Marshall, C.J. (2010). The plasticity of cytoskeletal dy-
namics underlying neoplastic cell migration. Curr. Opin. Cell Biol. 22, 690–696.
Somers, W.G., and Saint, R. (2003). A RhoGEF and Rho family GTPase-acti-
vating protein complex links the contractile ring to cortical microtubules at
the onset of cytokinesis. Dev. Cell 4, 29–39.
Steigemann, P., Wurzenberger, C., Schmitz, M.H.A., Held, M., Guizetti, J.,
Maar, S., and Gerlich, D.W. (2009). Aurora B-mediated abscission checkpoint
protects against tetraploidization. Cell 136, 473–484.
Stewart, M.P., Helenius, J., Toyoda, Y., Ramanathan, S.P., Muller, D.J., and
Hyman, A.A. (2011). Hydrostatic pressure and the actomyosin cortex drive
mitotic cell rounding. Nature 469, 226–230.
Su, K.C., Takaki, T., and Petronczki, M. (2011). Targeting of the RhoGEF Ect2
to the equatorial membrane controls cleavage furrow formation during cytoki-
nesis. Dev. Cell 21, 1104–1115.
Tatsumoto, T., Xie, X., Blumenthal, R., Okamoto, I., and Miki, T. (1999). Human
ECT2 is an exchange factor for Rho GTPases, phosphorylated in G2/M
phases, and involved in cytokinesis. J. Cell Biol. 147, 921–928.
The´ ry, M., Racine, V., Pe´ pin, A., Piel, M., Chen, Y., Sibarita, J.-B., and Bornens,
M. (2005). The extracellular matrix guides the orientation of the cell division
axis. Nat. Cell Biol. 7, 947–953.
Thullberg, M., Gad, A., Le Guyader, S., and Stro¨ mblad, S. (2007). Oncogenic
H-Ras V12 promotes anchorage-independent cytokinesis in human fibro-
blasts. Proc. Natl. Acad. Sci. USA 104, 20338–20343.
Trepat, X., Wasserman, M.R., Angelini, T.E., Millet, E., Weitz, D.A., Butler, J.P.,
and Fredberg, J.J. (2009). Physical forces during collective cell migration. Nat.
Phys. 5, 426–430.
Wagner, E., and Glotzer, M. (2016). Local RhoA activation induces cytokinetic
furrows independent of spindle position and cell cycle stage. J. Cell Biol. 213,
641–649.
Welman, A., Serrels, A., Brunton, V.G., Ditzel, M., and Frame, M.C. (2010).
Two-color photoactivatable probe for selective tracking of proteins and cells.
J. Biol. Chem. 285, 11607–11616.
Wolfe, B.A., Takaki, T., Petronczki, M., and Glotzer, M. (2009). Polo-like kinase
1 directs assembly of the HsCyk-4 RhoGAP/Ect2 RhoGEF complex to initiate
cleavage furrow formation. PLoS Biol. 7, e1000110.
Y€uce, O¨ ., Piekny, A., and Glotzer, M. (2005). An ECT2-centralspindlin complex
regulates the localization and function of RhoA. J. Cell Biol. 170, 571–582.
Developmental Cell 45, 132–145, April 9, 2018
145

----!@#$NewPage!@#$----
STAR+METHODS
KEY RESOURCES TABLE
REAGENT or RESOURCE
SOURCE
IDENTIFIER
Antibodies
Activated Integrin b1
Abcam
Abcam Cat# ab30394, RRID:AB_775726
Paxillin
Abcam
Abcam Cat# ab32084, RRID:AB_779033
Talin
Sigma
Sigma-Aldrich Cat# T3287, RRID:AB_477572
Ect2
Santa Cruz
Santa Cruz Biotechnology Cat# sc-1005, RRID:AB_2246263
Anillin
C. Field (Hu et al., 2008)
N/A
a-Tubulin
Sigma
Sigma-Aldrich Cat# T9026, RRID:AB_477593
p-Myosin LC2 Ser19
Cell signaling technology
Cell Signaling Technology Cat# 3671, RRID:AB_330248
AuroraB
Abcam
Abcam Cat# ab2254, RRID:AB_302923
Phalloidin TRITC
Sigma
Sigma-Aldrich Cat# P1951, RRID:AB_2315148
Dapi
Invitrogen
Thermo Fisher Scientific Cat# D3571, RRID:AB_2307445
Ect2
Santa Cruz
Santa Cruz Cat# sc-1005,
RRID:AB_2246263
Bacterial and Virus Strains
rLVUbi-LifeAct-TagRFP lentiviral vector
Ibidi
60142
Experimental Models: Cell Lines
hTERT-RPE1
Clontech
N/A
RPE1 a-tubulin-EGFP
D. Gerlich
N/A
MDA-MB-231
E. Sahai
N/A
RPMI 7951
American Type Culture
Collection Number: HTB-66
ATCC Number: HTB-66
HeLa Kyoto
MitoCheck (Hutchins et al., 2010)
N/A
HeLa LifeAct Ruby
Made in lab by M. Fedorova
N/A
MCF10A
E. Sahai
N/A
Chemicals, Peptides, and Recombinant Proteins
Y27632
Sigma
Y0503
CK-666
Sigma
SML0006
Ro-3306
Enzolife Sciences
ALX-270-463
STLC
Sigma
164739
PLL-g-PEG-633
SuSoS
PLL(20)-g[3.5]- PEG(2)/Atto633
Fibronectin
Sigma
F1141
Versene 1X
Gibco
15040-066
Recombinant DNA
pArek1-EGFP-Zyxin plasmid
Welman et al., 2010
N/A
Talin-GFP
Franco et al., 2004
N/A
pRK5-Rap1[Q63E]
Dupuy et al., 2005
N/A
aV-Integrin-GFP
R. Horwitz
Addgene plasmid # 15238
chicken Paxillin-GFP
Laukaitis et al., 2001
N/A
Software and Algorithms
Particle imaging velocimetry software
Trepat et al., 2009
N/A
Other
AllStars negative control siRNA
Qiagen
1027280
Hs_ECT2_6 Flexitube siRNA
Qiagen
SI03049249
e1
Developmental Cell 45, 132–145.e1–e3, April 9, 2018

----!@#$NewPage!@#$----
CONTACT FOR REAGENT AND RESOURCE SHARING
Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact,
Buzz Baum (b.baum@ucl.ac.uk).
EXPERIMENTAL MODEL AND SUBJECT DETAILS
Cell Lines and Culture
hTERT-RPE1 (female) cells (Clontech) and MDA-MB-231 (female) cells (gift from E. Sahai) were cultured in DMEM F-12 Glutamax
(Gibco 31331-028), with 10% fetal bovine serum, 3.4% sodium bicarbonate (Gibco 25080-060), 1% Penstrep (Gibco 15070-063).
RPE1 a-tubulin-EGFP were a gift from D. Gerlich. RPMI-7951 (female) cells (purchased from the American Type Culture Collection
(ATCC) Number: HTB-66) and HeLa (female) cells (MitoCheck (Hutchins et al., 2010)) were cultured in DMEM (Gibco 41965-039) with
10% FBS and 1% PenStrep. HeLa LifeAct-Ruby cells were made in lab by M. Fedorova. MCF10A (female) cells (gift from E. Sahai)
were cultured in DMEM F-12 Glutamax, with 5% Horse serum (Invitrogen 16050), 20ng/ml EGF (Peprotech 100-15R), 0.5mg/ml
Hydrocortisone (Sigma H-0888), 100ng/ml Cholera toxin (Sigma C-8052), 10mg/ml Insulin (Sigma I-1882), 1% Penstrep. Leibovitz’s
L-15 CO2 independent media (Gibco 21083-027) +10%FBS was also used during imaging. Cell lines have not been authenticated.
METHOD DETAILS
Plasmid Transfection
rLVUbi-LifeAct-TagRFP lentiviral vector (Ibidi 60142) was used to infect RPE1 cells. Positive cells were then selected with 1ug/ml
puromycin to generate a stable cell line.
Plasmid transfection with the pArek1-EGFP-Zyxin plasmid (Gift from A. Welman (Welman et al., 2010)) was carried out using
Fugene HD (Promega E2311). Positive RPE1 cells were selected with 500mg/ml G418 (Calbiochem 345812) to generate a stable
cell line. Transient transfection of HeLa with H2B-mCherry, Talin-GFP (Franco et al., 2004), chicken Paxillin-GFP (Laukaitis et al.,
2001) or Zyxin-GFP (Welman et al., 2010), pRK5-Rap1[Q63E] (Dupuy et al., 2005) was carried out with the same protocol.
aV-Integrin-GFP was a gift from R. Horwitz (Addgene plasmid # 15238) and was transiently transfected into RPE1 cells using
Liopfectamine LTX and Plus reagent (Invitrogen 15338–100).
Ect2 siRNA
siRNA treatment was carried out using AllStars negative control siRNA (Qiagen 1027280), Hs_ECT2_6 Flexitube siRNA (ATGACG
CATATTAATGAGGAT-Qiagen SI03049249) and Lipofectamine RNAimax (Invitrogen 13778-075) diluted in optimem (Gibco 51985-
026). Cells were used approximately 20 hours post-transfection.
Drug Treatments
Cells were incubated with 50mM of the small molecule Y27632 (Sigma Y0503) which inhibits ROCK activity and imaged the same day.
Cells were incubated with 300mM of the small molecule CK-666 (Sigma SML0006) which inhibits Arp2/3 activity and imaged
immediately.
To synchronise cells prior to mitosis they were incubated with 9mM Ro-3306 (Enzolife Sciences ALX-270-463) to inhibit CDK1
activity for 15-20hours. This was replaced with drug free media immediately before imaging.
To synchronise cells in metaphase they were incubated with 10mM STLC (Sigma 164739) for 4hours. Then a mitotic shake off was
carried out and the collected cells washed twice before being re-suspended in imaging media, re-plated and imaged immediately.
Micropatterning and Surface Treatment
Standard experimental imaging was carried out on glass bottomed dishes incubated with 10mg/ml fibronectin for 1 hour at 37�C.
Poly-L-Lysine (Sigma P4832) was incubated on a glass bottomed dish for 5 min room temperature with rocking, removed and
allowed to air dry for 2 hours before seeding cells.
For micropatterning, HCL cleaned coverslips were passivated by plasma cleaning for 30seconds before incubating for 30min at
room temperature with 0.1mg/ml PLL-g-PEG-633 (SuSoS PLL(20)-g[3.5]- PEG(2)/Atto633). A drop of MilliQ water was used to attach
the coverslips to a quartz mask custom designed with the desired patterns and exposed to deep UV through the mask for 4min. The
illuminated coverslip surface was then incubated for 1hour at room temperature with 25mg/mL of fibronectin (Sigma F1141) solution in
100mM NaHCO3 pH 8.5 (Gibco 25080-060). Versene 1X (Gibco 15040-066) was used to dissociate cells before plating on coverslips
to ensure fast re-spreading on patterns and cells were imaged within 4hours.
Wells
40mm diameter PDMS wells were adhered to a plasma treated glass bottomed dish and baked at 75�C for 1hour before being
incubated in 0.1mg/ml PLL-PEG (SuSoS PLL(20)-g[3.5]- PEG(2)) overnight. Cells were added to the well immediately prior to
imaging.
Developmental Cell 45, 132–145.e1–e3, April 9, 2018
e2

----!@#$NewPage!@#$----
Cell Fixation and Immunostaining
16% warmed PFA was added to cells in media to a final concentration of 4% and incubated at room temperature for 20min.
Alternatively, 10% cold TCA was added and incubated at room temperature for 20min, or ice cold methanol was added and
incubated at -20�C for 5 min. They were then washed 3 times and 0.2% Triton was added for 5min (only to PFA fixed cells, not
TCA or methanol fixed). 5% bovine serum albumin/PBS was used to block for 30min at room temperature before primary antibodies
were added. Activated Integrin b1 (Abcam Cat# ab30394, RRID:AB_775726), Paxillin (Abcam Cat# ab32084, RRID:AB_779033), Talin
(Sigma-Aldrich Cat# T3287, RRID:AB_477572), Ect2 (Santa Cruz Biotechnology Cat# sc-1005, RRID:AB_2246263), Anillin (Gift from
C. Field (Hu et al., 2008)), a-Tubulin (Sigma-Aldrich Cat# T9026, RRID:AB_477593), p-Myosin LC2 Ser19 (Cell Signaling Technology
Cat# 3671, RRID:AB_330248) and AuroraB (Abcam Cat# ab2254, RRID:AB_302923). Phalloidin Tritc (Sigma-Aldrich Cat# P1951,
RRID:AB_2315148) and DAPI (Thermo Fisher Scientific Cat# D3571, RRID:AB_2307445) were added with secondary antibodies
(Invitrogen 647 anti mouse or 448 anti rabbit).
Fixed samples were imaged on a Leica TCS SPE 2 microscope except for the online fixation experiment.
For online fixation RPE1 Zyxin-GFP CDK1 inhibited cells were imaged on a Nikon Eclipse Ti microscope with Andor Neo-Zyla
camera. Using the pump system established by P. Almada (Almada, 2017), the media was exchanged at the microscope to remove
inhibition and allow the cells to enter mitosis and imaging continued. 30min after media exchange when many cells were in meta-
phase, the PFA fixation protocol as above was triggered to fix and stain the cells at the microscope. The same cells were then imaged
post fixation.
Live Cell Imaging
Widefield imaging was carried out on Nikon Ti inverted microscope or a Zeiss Axiovert 200M microscope at 3 or 5 minute timepoints
using a 20x or 40x objective.
Live confocal imaging was carried out on a Nikon TiE inverted stand attached to a Yokogawa CSU-X1 spinning disc scan head,
using the 40X objective and 3 minute timepoints.
Western Blot
Treated cells were lysed using chilled RIPA buffer on ice. The protein concentration of the supernatant was determined using
Bradford reagent and samples were run on 4-12% Tris Bis gel (Invitrogen NW04122). Gels were then blotted and probed with
Ect2 (Santa Cruz sc-1005) and a-tubulin (Sigma T9026) primary anitbodies, and anti-mouse and anti-rabbit HRP-conjugated
secondary antibodies (Dako).
Preparation of Polyacrylamide Gels
Glass-bottom dishes were activated by using a 1:1:14 solution of acetic acid/bind-silane (M6514, Sigma)/ethanol. The dishes were
washed twice with ethanol and air-dried for 5 min. For 5kPa/1.5kPa/0.5kPa gels, a 500ml stock solution containing PBS, 93.3ml/
62.5ml/50ml acrylamide 40% (A4058, SIGMA), 11ml/10ml/7.5ml bisacrylamide 2% (BP1404-250, FisherScientific), 2.5ml 10% APS
diluted in water (Sigma A7460), 0.25ml TEMED (BioRad 161-0800) and 12ml of 200-nm-diameter red fluorescent carboxylate-modified
beads (F8810, ThermoFisher) was prepared. A drop of 18 ml was added to the centre of the glass-bottom dishes and the solution was
covered with 18-mm-diameter glass coverslip. After polymerization, the coverslip was removed and gels were functionalized using
sulfo-sanpah (102568-43-4). Briefly, 80ml drop of sulfo-sanpah was placed on the top of the polyacrylamide gel and activated by UV
light for 3 min. Sulfo-sanpah was diluted in miliQ water to a final concentration of 2mg/ml from an initial dilution 50mg/ml kept at -80�.
Then, gels were washed twice with miliQ water and once with PBS for 5min each. Afterwards, gels were incubated with 200ml of a
fibronectin solution (0.01mg ml�1) overnight at 4�C.
QUANTIFICATION AND STATISTICAL ANALYSIS
Traction Microscopy
Traction forces were computed using Fourier transform based traction microscopy with a finite gel thickness. Gel displacements be-
tween any experimental time point and a reference image obtained after monolayer trypsinization were computed using home-made
particle imaging velocimetry software (Trepat et al., 2009). The transmitted force was computed as the net traction force generated by
each one of the daughter cells after anaphase (Labernadie et al., 2017).
Statistical Analysis
Apart from the traction force microscopy data, all analysis was carried out manually in Fiji. Graphs were produced in Microscoft Excel
and Graphpad Prism. Statistical tests were carried out in Graphpad Prism. Normal data sets comparing distribution of values were
analysed using the unpaired t test, two tailed. Non-normal data sets were analysed using Mann-Whitney two-tailed test. Binary data
sets were analysed using the Chi-square test. *p<0.01 **p<0.001 ***p<0.0001 ****p<0.00001. Details of the statistical tests used,
exact value of n, definition error bars on graphs are all detailed for each figure in the legend.
e3
Developmental Cell 45, 132–145.e1–e3, April 9, 2018