Supplementary Method:
Plasmid DNAs for Rho GTPases, mDia isoforms
and their mutants.
Complementary DNAs for mDia2 and mDia3 were
isolated by polymerase chain reaction–based
cloning from cDNA libraries of mouse testis and
lung with primers based respectively on the
reported mDia2 cDNA sequence1 and on the partial
3' sequence for mDia3 cDNA (Y15910) and
expressed sequence tags (BB576753, AV318338,
AA921166) homologous to the reported sequence
of hDia2 cDNA (ref. 2). The nucleotide sequences
of the cDNAs for each gene isolated from the testis
and lung libraries were identical and were
consistent with the sequences present in the mouse
genomic DNA database. cDNA for the RBDs of
mDia2 (amino acids 33 to 411) and mDia3 (amino
acids 3 to 294) were subcloned into pThioHis or
pVP-16, and cDNA for the H3 region and the FH2
region of mDia3 were subcloned into pBTM. The
constructs for the RBD, the H3, and the FH2
fragments of mDia1 were previously described3.4.
The pBTM plasmids for RhoA, Rac1, and Cdc42
mutants were as described5. The pGEX-RhoA,
pGEX-Cdc42, and pGEX-Rac1 plasmids were also
described previously3. The cDNAs from plasmids
for dominant active and dominant negative mutants
of RhoA, Rac1, and Cdc42 (ref. 6) were subcloned
into pEGFP as described7. A dominant active
mutant of mDia1,N3mDia1, was described
previously3.
For preparation of an active mDia3
mutant, N-mDia3, the AccI/BamHI fragment of
the full length mDia 3 in pCR vector (Invitrogen)
was subcloned in pEGFP vector. Expression of
N-mDia3 in HeLa cells induced the active mDia
phenotype such as cell elongation and actin
filament alignment as did N3-mDia13. Constructs
for GFP-EB1 and dsRed2-histoneH2B-k will be
described elsewhere.
Isolation of CENP-A as a binding protein of the
FH2 region of mDia1.
We previously found that expression of active
forms of mDia1 such as N3-mDia1 induced
elongation of HeLa cells and aligned MTs in
parallel to the long axis of the elongated cells8. The
MT alignment by active mDia1 was suppressed by
co-expression of the shorter mDia1 fragment
spanning the FH2 region, and mutations to alanine
of the lysine residues at 989, 994 and 999 in the
FH2 region abolished the MT aligning activity of
active mDia1, suggesting that the FH2 region is
critically involved in MT alignment8. In order to
identify putative binding partners for the FH2
region, we carried out yeast two hybrid screening
with the FH2 region of mDia1 as a bait9. We also
used the FH2 region fragment with the K->A
mutation, the FH2-KA3, as a bait for negative
control. cDNAs for mDia1-FH2 (amino acids
829-1106) and mDia1-FH2-KA3 (amino acids
829-1106) were produced by PCR using
pGEX-4T-1-mDia1 and pEGFP-N3-mDia1-KA3
(ref. 8) as a template, respectively, with
5’-GGAATTCCATATGGTAAAAGAGCTGAAAG
TGCTG-3’ as the forward primer and
5’-TCCCCCGGGACGAAGTAGTCACCTAGCTC
-3’ as the reverse primer. The resultant PCR
fragments were digested with EcoRI and SmaI and
then subcloned into pGBKT7 (Clontech). These
plasmids were also digested with BamHI and EcoRI
and the fragments were subcloned into pBTM116.
The
yeast
L40
strain
harboring
pGBKT7-mDia1-FH2 was transformed with
pACT2 fused with a human HeLa cDNA library
(Clontech). Initial transformation yielded 1.0×106
transformants. These transformants were then
amplified in culture medium without uracil,
tryptophan and leucine for 16 h. Approximately 2.0
×108 transformants were obtained. Among these
more than 2×103 clones were grown on His (-)
plates, and about 90% of these clones showed β
-galactosidase activity. Among 50 clones picked up,
17 clones were segregated from the bait plasmid.
The pACT2 plasmids were recovered from all these
clones and cDNA inserts were sequenced. Among
these, 15 clones were derived from the same cDNA,
which encoded human centromere protein A
(CENP-A). We selected one clone (clone 5) and
plasmid from clone 5 was used for
co-transformation of the L40 yeast strain with bait
constructs.
Isolation of heterochromatin protein (HP)-1 as a
binding protein of the H3 region of mDia1.
We previously found that mDia1 was associated
with the spindle MTs in mitotic HeLa cells9.
Analysis using various truncation mutants of mDia1
revealed that the minimum fragment that localizes
to the spindle MT is the C-terminal fragment of the
putative FH3 region of mDia1 designated H3
(amino acids 431-603), and that a H3 mutant with a
mutation of Leu455 in this region to Glu
(H3-L455E) showed the markedly attenuated
localization. We sought for a binding partner of
this region of mDia1 using yeast two-hybrid
screening with H3 as a bait. The yeast L40 strain
harboring pBTM-H3 that encodes LexA
DNA-binding protein fused to H3, was transformed
with pVP16 fused with a mouse embryo cDNA
library. Approximately 3.6 x 107 transformants
were obtained and amplified during the 4 h culture
before spreading on histidine-free plates. Among
9.2 x 107 transformants, 382 clones were isolated as
His+ and LacZ+ and cultured in medium without
tryptophan. 226 clones were segregated from the
bait plasmid. Segregated clones were then mated
with a yeast AMR70 strain bearing either the bait
construct, LexA-fused lamin or LexA-fused
H3-L455E. We obtained 31 clones that were
positive for the bait and negative for lamin and
L455E. To confirm these interaction the pVP16
plasmids were recovered from these clones and
retransformed to L40 strain bearing H3, lamin or
L455E. Fourteen pVP16 plasmids reacted with
the bait but not with lamin or L455E (clone 7 was
shown in Supplementary Fig. 5 as representative
data).
DNA sequencing of these plasmids
revealed that 12 pVP16 plasmids possessed HP1,
that one pVP16 plasmids possessed HP1 and that
one pVP16 plasmids possessed heat shock protein
hsc73. cDNA for HP1 was then cloned from
mouse cDNA library, and used for two hybrid assay
with H3, yielding a similar positive signal.
1.
2.
3.
4.
5.
6.
7.
8.
9.
Alberts, A. S., Bouquin, N., Johnston, L. H. &
Treisman, R. Analysis of RhoA-binding
proteins reveals an interaction domain
conserved in heterotrimeric G protein 
subunits and the yeast response regulator
protein Skn7. J. Biol. Chem. 273, 8616–8622
(1998).
Bione, S. et al. A human homologue of the
Drosophila melanogaster diaphanous gene is
disrupted in a patient with premature ovarian
failure: evidence for conserved function in
oogenesis and implications for human sterility.
Am. J. Hum. Genet. 62, 533–541 (1998).
Watanabe, N. et al. Cooperation between
mDia1 and ROCK in Rho-induced actin
reorganization. Nat. Cell Biol., 1, 136-143
(1999).
Kato, T. et al. Localization of a mammalian
homolog of Diaphanous, mDia1, to the mitotic
spindle in HeLa cells. J. Cell Sci., 114,
775-784 (2001).
Reid, T. et al. Rhotekin, a new putative target
for Rho bearing homology to a
serine/threonine kinase, PKN, and rhophilin in
the Rho-binding domain. J. Biol. Chem. 271,
13556–13560 (1996).
Hirose, M. et al. Molecular dissection of the
Rho-associated protein kinase
(p160ROCK)-regulated neurite remodeling in
neuroblastoma N1E-115 cells. J. Cell Biol.
141,1625–1636 (1998).
Tsuji, T. et al. ROCK and mDia1 antagonize
Rho-dependent Rac activation in Swiss 3T3
fibroblasts. J. Cell Biol. 157, 819–830 (2002).
Ishizaki, T., Morishima, Y., Furuyashiki, T.,
Kato, T., Narumiya, S. Coordination of
microtubules and actin cytoskeleton by a Rho
effector, mDia1. Nat. Cell Biol., 3, 8-14
(2001).
Vojtek, A.B. , Hollenberg, S.M. and Cooper,
J.A. Mammalian Ras interacts directly with the
serine/threonine kinase Raf. Cell, 74, 205-214
(1993).
Supplementary Figure 2 Cyclin B degradation
through mitosis in toxin B-treated mitotic HeLa
cells. Mitotic cells were enriched by nocodazole,
and treated either with toxin B or vehicle. Toxin B
was then removed and the cells were either released
by removal of nocodazole or further incubated in
the continued presence of nocodazole. Cell lysates
were prepared at 0, 0.5, 1, 2 and 4 h after removal
of toxin B, and subjected to Western blot analysis
with anti-cyclin B1 antibody (clone CB169, Upstate
biotechonology). While cyclin B1 degradation was
significantly delayed but occurred in the toxin
B-treated cells, the continued presence of
nocodazole completely suppressed this degradation,
and no progression of toxin B-treated cells to
interphase occurred in the presence of nocodazole
(data not shown). Thus, the spindle check-point
mechanism is affected by the toxin B treatment, but
this effect of toxin B is not seen when MT binding
is completely suppressed by nocodazole treatment.
Supplementary Figure 1 Electron microscopy of
toxin B-treated mitotic cells. HeLa cells
synchronized in prometaphase by nocodazole were
treated with or without toxin B for 2 h. The cells
were fixed and subjected to electron microscopy at
60 min after removal of nocodazole. Compared to
control cells (a), chromosomes in the toxin
B-treated cells were misaligned, and MT
attachment was not found at kinetochores of some
chromosomes (b3). Bar, 1 m (a1, b1) and 200 nm
(a2, a3, b2, b3).
Supplementary Figure 3 Structures of the three
mDia isoforms. Residue numbers for the boundaries
of the Rho-binding (RBD), FH3, FH1 and FH2
domains are indicated, as is the percentage
sequence identity (similarity) for each domains of
mDia2 and mDia3 compared with those of mDia1.
H3, and FH2 fragments of mDia1 and mDia3 used
in the two hybrid assay are also shown.
Supplementary Figure 4 Interaction of FH2
fragment of mDia1 and CENP-A. Yeast two
hybrid screening was performed with the FH2
fragment of mDia1 as a bait, and CENP-A was
isolated as a binding protein as described in
Supplementary Method. Interaction of CENP-A
with FH2, FH2-KA3 and lamin was examined by
co-transformation in L40 yeast strain. Note a
significant
β -galactosidase
staining
on
co-transformation with pGBKT7-mDia1-FH2, little
on co-transformation with pBTM encoding
mDia1-FH2-KA3 mutant and no staining on
co-transformation
with
pBTM
encoding
LexA-fused lamin.
Supplementary Figure 5 Interaction of the H3
fragment of mDia1 and heterochromatin protein
(HP)-1. Yeast two hybrid screening was performed
with the H3 fragment of mDia1 as a bait, and HP-1
was isolated as a binding protein as described in
Supplementary Method. Interaction of HP1 with H3,
H3-L455E and lamin was examined in L40 yeast
strain. Note a significant β-galactosidase staining
on HP-1 interaction with H3 but little with
H3-L455E and lamin.
Supplementary Figure 8 Interaction of mDia3
with RhoA, Rac1, and Cdc42. The yeast two-hybrid
assay was performed with RBD of mDia3 as a
VP16 fusion protein and the indicated wild-type
and mutant Rho GTPases as LexA fusion constructs.
mDia3 bound to RhoA, Cdc42, and Rac1, in a
GTP-dependent manner. whereas mDia1 bound
selectively to RhoA3, and mDia2 bound to RhoA
and Rac1 (data not shown).
Supplementary Figure 9 Specificity of anti-mDia3
antibody. Lysates of control HeLa cells (lane 1, left
panel) or cells expressing Flag-mDia3 (lane 2, left
panel), GFP-mDia1 (lane 1, middle and right
panels), GFP-mDia2 (lane 2, middle and right
panel) and GFP-mDia3 (lane 3, middle and right
panel) were subjected to immunoblot with anti
mDia3 antibody (left and middle panels) or
anti-GFP antibody (right panel). Arrows denotes
the position of endogenous mDia3.
Supplementary Figure 6 (above) and 7 (below)
Interaction of mDia3 with HP1 and CENP-A. The
yeast two-hybrid assay was performed with the H3
of mDia3 as a LexA fusion protein and HP1
isoforms (, , ) as VP16 fusion proteins (above)
or with the FH2 region of mDia3 as a LexA fusion
protein and CENP-A or two CENP-A fragments,
mt1 (residues 8-142) and mt2 (residues 40-142) as
ACT2 fusion proteins. As found for mDia1, the
H3 fragment and the FH2 fragment of mDia3
bound directly to all isoforms ( and ) of HP1
and CENP-A, respectively.
Supplementary Figure 10 Co-localization of
mDia3 with HP-1 in the nucleus of interphase HeLa
cells. HeLa cells were transfected with pFL-HP1
(encoding Flag-HP1), fixed and stained with
antibodies to mDia3 and to Flag. DNA was
visualized with TOPRO-3.
Co-staining for
expressed Flag-tagged HP1 and endogenous
mDia3 is shown. Note that most of the nuclear
mDia3
signals
overlapped
with
HP1
immunofluorescence. Bar, 5 m
a
Supplementary Figure 12 Depletion of mDia3 by
RNAi; immunofluorescence. Cells transfected with
siRNA for mDia3 or that with scrambled sequence
(control) were fixed at 24 h, and stained for mDia3
(green) and CENP-A (red). Bar, 5 m
b
Supplementary Figure 11 a, Specificity of
anti-mDia1 antibody. Lysates of HeLa cells
expressing GFP-mDia1 (lane 1), GFP-mDia2 (lane
2) and GFP-mDia3 (lane 3) were subjected to
immunoblot with anti mDia1 antibody. Arrows
denotes the position of endogenous mDia1. Note
that anti-mDia1 antibody specifically detected
endogenous mDia1 and GFP-mDia1 but not either
endogenous or GFP-tagged mDia2 and mDia3
proteins b, Localization of mDia1.Interphase and
mitotic HeLa cells were satined for mDia1 (green)
and CENP-A (red). DNA was stained with
TOPRO-3. Note that mDia1 is localized mostly in
the cytoplasm in interphase cells. In mitosis,
mDia1 is localized at the poles to polar spindle
MTs and found in the cell cortex. Bar, 5 m
Supplementary Figure 13 Effects of active
mutants of mDia isoforms on chromosome
alignment and segregation. Active form of mDia3,
N-mDia3, was constructed by deletion of the
N-terminal Rho-binding domain. Expression
vectors encoding this cDNA or an active form of
mDia1, N3-mDia1, was microinjected into NIH
3T3 cells synchronized in S phase. Cells were
analyzed at 12 h (above) and 16 h (below) after
injection. Note that chromosomal misalignment
similar to those found in cells transfected with
active Cdc42 or treated with toxin B or subjected to
mDia3 RNAi was observed at 12 h in cells
microinjected of the vector encoding either
N-mDia3 or N3-mDia1 but not the vector alone.
Production of multinucleate cells with abnormally
shaped nuclei was apparent at 16 h also in cells
microinjected of the vector encoding either
N-mDia3 or N3-mDia1. The effects of
N3-mDia1 probably reflects the CENP-A binding
activity shared by this isoform. Bar, 5 m
Supplementary Movie-1 (Figure 1a; mitosis of
Supplementary Movie-3 (mitosis of control cells,
control cells)
merged
images
for
GFP-EB1
and
dsRed2-histoneH2B-k)
Supplementary Movie-2 (Figure 1b; mitosis of
toxin B-treated cells)
Supplementary Movie-4 (mitosis of cells treated
with mDia3 siRNA, merged images for GFP-EB1
Fluorescence video microscopy of mitotic HeLa
and dsRed2-histoneH2B-k)
cells transfected with pQBI25-Xbeta-tubulin1 was
performed
essentially
as
described2.
Cells
expressing a fusion construct of green fluorescent
Supplementary Movie-5 (mitosis of cells treated
with mDia3 siRNA, GFP-EB1 images)
protein (GFP) and -tubulin were arrested in
prometaphase by nocodazole treatment and then
incubated for 2 h in the absence (control) or
presence of C. difficile toxin B (10 ng ml-1). After
removal of nocodazole, DNA was counterstained
with Hoechst 33342, and progression through
mitosis was monitored by fluorescence video
microscopy.
Reference
1. Mimori-Kiyosue, Y., Shiina, N. & Tsukita, S.
Adenomatous polyposis coli (APC) protein
moves along microtubules and concentrates at
their growing ends in epithelial cells. J. Cell Biol.
148, 505–518 (2000).
2. Haraguchi, T., Kaneda, T. & Hiraoka, Y.
Dynamics of chromosomes and microtubules
visualized by multiple-wavelength fluorescence
imaging in living mammalian cells: effects of
mitotic inhibitors on cell cycle progression.
Genes Cells 2, 369-380 (1997).
HeLa
cells
expressing
dsRed2-histone H2B-k were
siRNA for mDia3. At 36 h
fluorescence video microscopy
mitotic cells.
GFP-EB1
and
transfected with
after transfection,
was performed on