Inhibition of microtubule assembly by a complex of actin and antitumor macrolide aplyronine a

Article abstract of DOI:10.1021/ja406580w

Aplyronine A (ApA) is a marine natural product that shows potent antitumor activity. While both ApA and ApC, a derivative of ApA that lacks a trimethylserine ester moiety, inhibit actin polymerization in vitro to the same extent, only ApA shows potent cytotoxicity. Therefore, the molecular targets and mechanisms of action of ApA in cells have remained unclear. We report that ApA inhibits tubulin polymerization in a hitherto unprecedented way. ApA forms a 1:1:1 heterotrimeric complex with actin and tubulin, in association with actin synergistically binding to tubulin, and inhibits tubulin polymerization. Tubulin-targeting agents have been widely used in cancer chemotherapy, but there are no previous descriptions of microtubule inhibitors that also bind to actin and affect microtubule assembly. ApA inhibits spindle formation and mitosis in HeLa S3 cells at 100 pM, a much lower concentration than is needed for the disassembly of the actin cytoskeleton. The results of the present study indicate that ApA represents a rare type of natural product, which binds to two different cytoplasmic proteins to exert highly potent biological activities.

Full text of DOI:10.1021/ja406580w

Article
pubs.acs.org/JACS
Inhibition of Microtubule Assembly by a Complex of Actin and
Antitumor Macrolide Aplyronine A
Masaki Kita,*,† Yuichiro Hirayama,† Kozo Yoneda,† Kota Yamagishi,† Takumi Chinen,‡ Takeo Usui,‡
Eriko Sumiya,§ Motonari Uesugi,§,∥ and Hideo Kigoshi*,†
‡
†Graduate School of Pure and Applied Sciences and Graduate School of Life and Environmental Sciences, University of Tsukuba,
1-1-1 Tennodai, Tsukuba 305-8571, Japan
§Institute for Integrated Cell-Material Sciences (WPI-iCeMS) and ∥Institute for Chemical Research, Kyoto University, Gokasho, Uji,
Kyoto 611-0011, Japan
S
* Supporting Information
ABSTRACT: Aplyronine A (ApA) is a marine natural product that shows potent antitumor activity. While both ApA and ApC,
a derivative of ApA that lacks a trimethylserine ester moiety, inhibit actin polymerization in vitro to the same extent, only ApA
shows potent cytotoxicity. Therefore, the molecular targets and mechanisms of action of ApA in cells have remained unclear. We
report that ApA inhibits tubulin polymerization in a hitherto unprecedented way. ApA forms a 1:1:1 heterotrimeric complex with
actin and tubulin, in association with actin synergistically binding to tubulin, and inhibits tubulin polymerization. Tubulin-
targeting agents have been widely used in cancer chemotherapy, but there are no previous descriptions of microtubule inhibitors
that also bind to actin and affect microtubule assembly. ApA inhibits spindle formation and mitosis in HeLa S3 cells at 100 pM, a
much lower concentration than is needed for the disassembly of the actin cytoskeleton. The results of the present study indicate
that ApA represents a rare type of natural product, which binds to two different cytoplasmic proteins to exert highly potent
biological activities.
targeting agents typically disrupt actin cytoskeleton dynamics at
concentrations below 100 nM without the breakdown of stress
fibers, which induces potent cytotoxicity. In addition, various
actin-polymerization-stimulating or -blocking molecules have
been shown to induce apoptosis; examples include cytochalasin
D (100 nM against T cells),9 jasplakinolide (100 nM against
HL-60 cells10 and T cells11), latrunculin A (1−10 μM against
MKN45 or NUGC-4 cells),12 and mycalolide B (100 nM
against HL-60 cells).13 Thus, these agents are not only useful
for investigating actin dynamics in cells but also may be of
therapeutic value as antitumor compounds.
INTRODUCTION
■
Actin is the most abundant protein in the eukaryotic
cytoskeleton and is essential for the regulation of various
cellular functions, such as muscle contraction, cell division, and
the migration of tumor cells. The dynamics of actin assembly
(polymerization/depolymerization) are regulated by numerous
actin-binding proteins.1 For example, thymosin β4 and profilin
bind to monomeric actin and physically inhibit its addition to
growing filaments. CapZ binds to filament ends and inhibits
both monomer addition and dissociation, while the Arp2/3
complex nucleates actin assembly and stabilizes actin filaments
(F-actin) by cross-linking.
Along with these endogenous proteins, various small agents
that target actin have also been discovered and some show
potent cytotoxicity.2 For example, ulapualides,3 mycalolides,4
kabiramides,5 sphinxolides/reidispongiolides,6 swinholides,7
and bistramides8 are all actin-depolymerizing agents of marine
origin that have recently been characterized. These actin-
Among the actin-targeting natural products, the apoptogenic
marine macrolide aplyronine A (ApA, 1) depolymerizes F-actin
and shows antitumor activities in mouse xenograft models (e.g.,
T/C 545%, 0.08 mg kg−1, against P388 leukemia).14 Previous
Received: June 28, 2013
Published: November 14, 2013
© 2013 American Chemical Society
18089
dx.doi.org/10.1021/ja406580w | J. Am. Chem. Soc. 2013, 135, 18089−18095

Journal of the American Chemical Society
Article
studies showed that ApA inhibits polymerization of actin by
forming a 1:1 complex with the globular monomeric molecule
(Kd = 100 nM).15 X-ray analysis of the actin−ApA complex
revealed that ApA intercalates into the hydrophobic cleft
between subdomains (SD) 1 and 3 of actin by using its side
chain.16 Complexes of actin with other agents, such as
kabiramide C,17 sphinxolide B,18 reidispongiolides A and C,18
swinholide A,19 and bistramide A,20 have also been reported.
The contacts between actin and each macrolide are similar to
those observed in the actin−ApA complex.
binds to tubulin in association with actin, inhibits tubulin
polymerization, and prevents spindle formation and mitosis in
tumor cells. Tubulin-targeting natural products and their
synthetic derivatives have been widely used in cancer
chemotherapy.23 To our knowledge, however, there are no
previous descriptions of microtubule inhibitors that also bind to
actin and affect microfilament dynamics.24 Our studies of ApA
provide further insights into the molecular mechanisms of
structurally diverse natural products that regulate cytoskeletal
dynamics.
RESULTS
■
Low Concentrations of ApA, but Not ApC, Induced
Cell Death without Actin Destabilization. As previously
described, ApA is approximately 1000-fold more cytotoxic than
ApC on the basis of the MTT assay (Figure S1, Supporting
Information).14,21,22 We used fluorescence microscopy to
examine the relationships between the cytotoxicity and actin-
destabilizing properties of aplyronines. Both 100 nM ApA and
ApC caused rapid disassembly of the actin cytoskeleton in
HeLa S3, a human cervical carcinoma cell line (Figure 1).
These properties on HeLa cells were similar to those of typical
actin-targeting agents (e.g., 100−1000 nM latrunculin A,
cytochalasin D, and mycalolide B).25 Meanwhile, treatment
with 100-fold more dilute samples (1 nM) of ApA or ApC
induced no detectable alteration of the actin cytoskeleton.
These results suggest that the potent cytotoxicity of ApA may
not be due solely to its F-actin severing properties.
ApA Binds to Tubulin by Interaction with Actin in
Cells and in Vitro. The C7 ester group of ApA, a rare
modification among natural products, protrudes toward the
bulk solvent region of the actin−ApA complex.16 Due to the
basicity and hydrophilic nature of the trialkylamine structure,
we hypothesized that the C7 trimethylserine ester group might
be involved in interactions with biomolecules other than actin,
as with various posttranslational modifications such as the
Although the interaction with actin is mediated by the C24−
C34 side chain of ApA, results of structure−activity studies
indicated that both the macrolactone part (C1−C23) and the
C24−C34 side chain are necessary for cytotoxicity at pM
concentrations.21 For example, aplyronine C (ApC, 2), which
lacks the C7 trimethylserine ester moiety, exhibits 1000-fold
less cytotoxicity, but as much actin-depolymerizing activity, as
ApA. The binding affinity of ApA to actin is comparable to
those of ApC and mycalolide B. However, ApA induces
caspase-dependent apoptosis in HeLa S3 cells at 1 nM, while
ApC and mycalolide B do not, even at 100 nM.22 Therefore,
the potent cytotoxicity and apoptogenic effect of ApA was not
entirely accounted for by its F-actin-severing properties, and its
molecular targets and mechanisms of action remained unclear.
By using fluorescence microscopy observations and photo-
affinity-tag experiments, we now show that ApA synergistically
Figure 1. Confocal fluorescence images of HeLa S3 cells treated with aplyronines for 2 h. Cells were stained with rhodamine−phalloidin (red, F-
actin) and DAPI (blue, nucleus). Scale bars: 20 μm.
18090
dx.doi.org/10.1021/ja406580w | J. Am. Chem. Soc. 2013, 135, 18089−18095

Journal of the American Chemical Society
Article
Figure 2. Actin−ApA complex binding to tubulin in situ and in vitro. (a) Structures of aplyronine photoaffinity derivatives. (b, c) In situ
photolabeling experiments. In (b), HeLa S3 cells were treated with or without ApA−PB and then irradiated with UV (365 nm) for 15 min. Labeled
proteins were affinity-purified with NeutrAvidin agarose, subjected to 10% SDS-PAGE, and detected with silver stain or HRP-conjugated
streptavidin. The arrowheads show the proteins that were identified as α/β-tubulin. Several nonspecific biotinylated proteins (72−74 and 110−120
kDa) were also detected. In (c), HeLa S3 cells were treated with ApA−PB or ApC−PB (2.5 μM) in the absence or presence of excess ApA (25 μM).
Labeled proteins were detected by immunoblotting analysis. The arrowheads show the protein bands (55 and 58 kDa) that covalently bond to ApA−
PB. (d, e), In vitro binding experiments on photoaffinity derivatives (0.67 μM) to purified actin (from rabbit muscle, 0.67 μM) and/or tubulin (from
porcine brain, 0.33 μM for the heterodimer) at 4 °C. After photoreaction, the reaction mixtures were lyophilized and directly analyzed by SDS-
PAGE. Labeled proteins were detected with HRP-conjugated streptavidin. In (d), tubulin and/or actin was photolabeled with ApA−PB. In (e),
tubulin and actin were photolabeled with the aplyronine derivatives 3−5.
phosphorylation of signaling molecules and the sulfation of
peptidoglycans.
ApA−PB (Figure 2c). In contrast, only actin (45 kDa) was
photolabeled to the ApC−PB control. Together, this suggests
that ApA interacts with both actin and tubulin.
To identify additional target proteins of ApA, we prepared
two photoaffinity biotin derivatives, ApA−PB (3) and ApC−PB
(4) (Figure 2a).26 ApA−PB showed potent cytotoxicity against
HeLa S3 cells (IC50 = 1.2 nM), whereas ApC−PB was ∼260-
fold less cytotoxic than ApA−PB. We previously reported that
modifications to the side chain of ApA, such as C34 alcohol or
C25 acetate derivatives, decreased its affinity for actin.21 Apart
from the hydrophobic binding of the side chain moiety (C24−
C33), the terminal N-methyl enamide part of ApA is located in
a hydrophilic environment and interacts with water molecules
inside actin. It has been shown that this moiety can be replaced
by oximes or hydrazones without a significant loss of
activity.22,27 In fact, ApA−PB inhibited actin polymerization
as with ApA, as confirmed by an F-actin sedimentation assay.26
ApA−PB and ApC−PB showed comparable reactivities toward
actin in cell lysate,26 suggesting that they have similar affinities
for actin.
We performed photolabeling experiments with living tumor
cells instead of cell lysate. Actin (43−45 kDa) and several
proteins (52−55, and 58 kDa) were detected, following in situ
photoreaction with ApA−PB in the HeLa S3 cells cooled on ice
and subsequent affinity purification using NeutrAvidin agarose
and silver staining (Figure 2b). Blotting analysis with
streptavidin−HRP conjugate revealed biotinylated bands at
45, 55, and 58 kDa, consistent with the silver staining results.
Mass analysis of tryptic peptide fragments established that the
silver-stained bands at 52−55 and 58 kDa included α- and β-
tubulin (Figure S2, Supporting Information). Since ApA−PB
most effectively photolabeled actin, considerable amounts of
nonlabeled α/β-tubulins might be eluted as adducts of actin−
ApA−PB conjugate, which were observed as a major silver-
stained band (52−55 kDa). Anti-β-tubulin antibody identified
the 58 kDa band as β-tubulin that was covalently bound to
To demonstrate the specificity of ApA−PB binding to
tubulin, we performed a competition experiment (Figure 2c).
Excess ApA inhibited the covalent binding of ApA−PB to both
actin and β-tubulin, and the total amount of β-tubulin was
reduced to control level. The presence of photoaffinity
derivatives had no significant effect on the amount of α-tubulin
detected. Our results suggest that ApA binds selectively to actin
and β-tubulin in cells.
Through the use of ApC−PB, the amounts of affinity-
purified α- and β-tubulin were slightly higher than those in the
control (Figure 2c and Figure S3a (Supporting Information)).
While actin was predominantly photolabeled, several proteins
(i.e., 51 and 60−65 kDa) were also labeled, as with the control
(ApA−PB + excess ApA). It is possible that a tubulin
heterodimer was purified as the adduct of these unspecific
proteins, without direct interaction with the actin−ApC−PB
conjugate.
In vitro photolabeling experiments with purified proteins
were carried out to investigate specific interactions among ApA,
actin, and tubulin. ApA−PB formed a covalent bond with actin
but not with tubulin alone (Figure 2d). In the presence of actin,
however, tubulin was photolabeled with ApA−PB and detected
as two biotin-labeled protein bands (55 and 58 kDa), as in the
in situ experiments. In contrast, ApC−PB did not bind to
tubulin, even in the presence of actin (Figure 2e).
The density ratios of two tubulin bands (55 and 58 kDa) in
Figure 2d,e were less reproducible (Figure S3b, Supporting
Information). The sizes of the original α- and β-tubulin were
52−55 kDa, and we were unable to distinguish whether the
lower biotinylated tubulin bands corresponded to α-tubulin
only or an α/β-tubulin mixture. Photolabeling of tubulin by
ApA−PB might occur at several places following the random
18091
dx.doi.org/10.1021/ja406580w | J. Am. Chem. Soc. 2013, 135, 18089−18095

Journal of the American Chemical Society
Article
Figure 3. ApA and actin synergistically inhibit tubulin polymerization, enhance microtubule depolymerization, and form a ternary complex with the
tubulin heterodimer in vitro. (a) Effects of ApA and actin on tubulin polymerization. Tubulin (2 mg/mL, 20 μM for the heterodimer) was
polymerized with 20% glycerol at 37 °C, as monitored by an increase in DAPI fluorescence (λex/λem 360/450 nm). (b) Synergistic effect of ApA and
actin on microtubule depolymerization. Various agents were added to polymerized tubulin (2 mg/mL) with paclitaxel (10 μM), and their abilities to
attenuate turbidity were monitored at 350 nm. (c) Gel permeation HPLC analysis of the binding of actin−ApA complex to tubulin: buffer, 50 mM
PIPES·K (pH 6.8), 100 mM KCl, 10 mM MgCl2; column, TSKgel SuperSW3000 (ϕ 4.6 × 300 mm); flow rate, 0.2 mL/min; temperature, 8 °C;
detection, UV 280 nm. Binding was monitored by the appearance of a faster mobility peak (∼15.5 min, corresponding to a 1:1:1 complex of actin−
ApA−tubulin heterodimer, green trace) at 18.3 min, excess actin + ApA. Blue trace: tubulin heterodimer (100 kDa). Pink trace: actin−ApA complex
(45 kDa). Eluted proteins (fractions 1−10) were analyzed by SDS-PAGE under reduceing conditions and detected with CBB stain (bottom). Due to
the high concentration of Mg2+ ion, ligand-free actin was polymerized (see also Figure S6 (Supporting Information)) and was not detected as a
single peak under these tubulin-stabilized buffer conditions.
movement of a flexible diazirine moiety on the ternary complex.
It is possible that labeled β-tubulin was detected as split bands.
Although stoichiometric amounts of tubulin were used, the
amount of tubulin photolabeled by ApA−PB in Figure 2d,e was
far less than that of actin. The C24−C34 side chain of ApA
binds to the hydrophobic cleft of SD 1 and 3 of actin,16,28 and
the aryldiazirine group in ApA−PB is believed to be located
inside of actin. Therefore, we prepared an ApA double-PEG-
linked photoaffinity biotin derivative (ApA−DPB, 5) to
enhance the flexibility of the photoreacting group and allow
it to reach toward the C7 trimethylserine ester moiety (Figures
2a and Figure S4 (Supporting Information)). ApA−DPB
showed potent cytotoxicity against HeLa S3 cells (IC50 0.54
nM), similar to the case for ApA−PB and other ApA
derivatives.26,27 As predicted, treatment with ApA−DPB
increased the amounts of both photolabeled tubulins and
actin, in comparison to treatment with ApA−PB (Figure 2e and
Figure S3a (Supporting Information)).
Due to the present poor availability of ApC, ApC−DPB has
not yet been prepared. To clarify the specificity of the
photoreactions of DPB derivatives toward tubulin, we hope
to perform labeling studies with ApC−DPB in future studies.
ApA and Actin Synergistically Inhibit Microtubule
Polymerization. We next examined whether the actin−ApA
complex affects microtubule assembly. While actin and ApA
alone had little effect on in vitro tubulin polymerization, their
1:1 complex delayed nucleation and growth phases and reduced
the final polymer mass of tubulin in a dose-dependent manner
(Figure 3a). In contrast, the actin−ApC complex failed to
attenuate microtubule growth (Figure S5, Supporting Informa-
tion). Actin and ApA synergistically depolymerized micro-
tubules stabilized by paclitaxel (Figure 3b), and this effect
surpassed that of combretastatin A-4 (CA4), a microtubule
inhibitor that binds to the colchicine site of β-tubulin.
Ultracentrifugation experiments similarly showed that actin
and ApA synergistically depolymerized microtubules stabilized
with paclitaxel (Figure S6, Supporting Information).
When the actin−ApA complex and tubulin were coanalyzed
by gel permeation HPLC, a mobility peak appeared that was
faster (green) than both the α/β-tubulin heterodimer (blue)
and actin−ApA complex peaks (pink), and the last two peaks
were reduced (Figure 3c, top). The retention times and SDS-
PAGE analysis revealed that the faster mobility peak
corresponded to the 1:1:1 complex of actin−ApA−tubulin
heterodimer (145 kDa) (Figure 3c, bottom, and Figure S7
(Supporting Information)). A densitometric analysis of the
CBB-stained bands showed that the quantitative ratios of
tubulin/actin in lanes 2 and 3 were 1.6:1 and 2.7:1, respectively
(Figure S8 (Supporting Information)). These values were
consistent with the calculated value for 1:1 ratios of 100 kDa
tubulin heterodimer/43 kDa actin. While we have no evidence
regarding the stoichiometry of ApA in the ternary complex, it is
likely that the complex binds to a tubulin heterodimer on the
basis of the structure of the actin−ApA complex. This ternary
complex was stable in solution for at least 1 h, despite being in
equilibrium with the actin−ApA complex and the liberated
tubulin heterodimer. The ternary complex peaks increased
depending on the amount of actin, and 49% actin contributed
to the formation of a ternary complex when an equivalent
amount of actin was added (Figure S9a (Supporting
Information)). The binding constant of the actin−ApA
complex to tubulin heterodimer was estimated to be 3.0 ×
106 M−1 on the basis of a Scatchard plot analysis (Figure S9b
(Supporting Information)). In contrast, the actin−ApC
complex did not interact with tubulin, and these compounds
were separately eluted at their original retention times (Figure
S5 (Supporting Information)).
18092
dx.doi.org/10.1021/ja406580w | J. Am. Chem. Soc. 2013, 135, 18089−18095

Journal of the American Chemical Society
Article
ApA Inhibits Spindle Formation and Mitosis in Tumor
Cells. Immunostaining experiments showed that HeLa S3 cells
treated with 100 pM ApA had irregular, multipolar spindle
structures with unaligned chromosomes (Figure 4b). The same
concentration that caused complete disassembly of the actin
cytoskeleton (Figure 1). The observed inhibitory effects of ApA
on spindle formation and mitosis might be essential for its
antitumor activity, in accordance with caspase-dependent
apoptosis.29,13
DISCUSSION
■
We previously reported that ApA does not interact with known
molecular targets of antitumor agents, such as DNA, micro-
tubules, and cell cycle regulating proteins.30 Our present results
show that ApA targets tubulin in association with actin, thereby
inhibiting microtubule assembly at a very low concentration. To
the best of our knowledge, there have been no previous reports
that small actin-binding agents disturb microtubule assembly.
Previous pull-down experiments using biotinylated derivatives
of aplyronines revealed that Arp2 and Arp3 (actin-related
proteins) were affinity-purified as binding proteins along with
actin from HeLa S3 cell lysate.27 However, Arp2 and Arp3 did
not covalently bind to ApA−PB or ApC−PB.26 It is noted that
no tubulin in the cell lysate was photolabeled or affinity-purified
with ApA−PB. In contrast, in situ photolabeling experiments
established that the cellular targets of ApA are both actin and
tubulin, consistent with those of in vitro and in vivo effects on
cytoskeleton filaments. These critical differences between in
vitro and in situ photolabeling experiments might be due to the
instability of microtubules under the lysis conditions we used.
Due to the high concentration of intracellular actin and the
high affinity of aplyronines for actin,15 they might be trapped
within cells as an actin complex. It is possible to imagine that
the intracellular concentrations of aplyronines could be higher
than the extracellular concentrations, which might enhance
their cytotoxic effects. However, we previously showed that
fluorescent derivatives of ApA and ApC are similarly
incorporated into HeLa S3 cells and are not easily excluded
from the cytoplasm.22 These results suggest that ApA and ApC
have similar cellular permeabilities and accumulation levels.
Since these two compounds have almost the same effects on
actin assembly in vitro, the significant differences in
cytotoxicity, mitosis inhibition, and apoptogenic effects22 of
ApA and ApC are likely to be due to the tubulin-binding
properties of ApA.
Through the use of in situ photolabeling experiments, we
demonstrated that ApA binds selectively to actin and β-tubulin
in cells. In contrast to β-tubulin, the amount of affinity-purified
α-tubulin with ApA−PB was similar to that of the control
(Figure 2c). This result suggests that the tight association of the
subunits of the tubulin heterodimer need to be dissociated,
which is incompatible with gel permeation HPLC analysis. A
possible explanation is that ApA−PB covalently bound to β-
tubulin might destabilize the association of tubulin subunits.
Alternatively, it is possible that α-tubulin is more labile than β-
isomer under the cell lysis or affinity-purification conditions we
used, which might enhance its dissociation from the resin.
How is the stable ternary complex of actin−ApA−tubulin
formed? To answer this question, we compared the structures
of the actin−latrunculin A31 and actin−kabiramide C17
complexes with that of the actin−ApA complex (Figure S11
(Supporting Information)). The structures of actin in the three
complexes highly overlapped, and the RMSD values of the
main chain atoms were 0.88 and 0.96 Å, respectively. Thus, the
actin bound to ApA does not have a unique conformation. As
described above, the side chain of ApA, which has the same
structure as that of ApC, binds to actin, but the C7
Figure 4. ApA inhibits spindle formation and mitosis at a
concentration that is much less than that which causes actin
depolymerization. (a) Cell cycle analysis. HeLa S3 cells were treated
with various agents for 18 h followed by PI staining. (b, c) Confocal
fluorescence images of HeLa S3 cells treated with various agents for 6
h. Cells were immunostained with anti-α-tubulin (green) and
costained with DAPI (blue). (b) Metaphase cells. (c) Interphase
cells. Scale bars: 10 μm.
treatment inhibited cell-cycle progression in M-phase (Figure
4a). Irregular spindles were observed in 36%, 52%, and 91% of
synchronized mitotic cells by treatment with 0.1, 1, and 10 nM
ApA, respectively, but in only 13% of control cells (Figure S10
(Supporting Information)). Treatment with 1 nM ApA partially
disrupted microtubule structures in interphase cells (Figure 4c)
and induced apoptosis,22 comparable to effects of treatment
with 10 nM vinblastine. In contrast, treatment with ApC had no
detectable effects on spindle formation and inhibited cell-cycle
progression in M-phase only at 100 nM, the same
18093
dx.doi.org/10.1021/ja406580w | J. Am. Chem. Soc. 2013, 135, 18089−18095

Journal of the American Chemical Society
Article
trimethylserine ester does not. Presumably, the positively
charged C7 ester group that protrudes from the actin−ApA
complex might generate a unique tubulin-binding site on the
actin−ApA complex, driving the interaction with tubulin α/β-
heterodimer (probably on the β-isomer). However, it is also
possible that the actin−ApA complex undergoes a structural
rearrangement upon binding to tubulin.
AUTHOR INFORMATION
■
Corresponding Author
mkita@chem.tsukuba.ac.jp; kigoshi@chem.tsukuba.ac.jp
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
Due to the unstable nature of diluted tubulin, an in vitro
tubulin polymerization assay and gel-permeation HPLC
analysis could not be performed at a sub-μM range. In cell-
based assays, however, ApA potently inhibited spindle
formation and mitosis at 100 pM, while actin filament
disassembly occurred only at 100 nM. Within cells, ApA
might first bind to actin and then interact with liberated tubulin
heterodimer to form a ternary complex. A few molecules of the
ternary complex would then bind at the microtubule plus end
or copolymerize into the microtubule lattice, as in treatment
with substoichiometric concentrations of vinblastine, colchicine,
or other antimitotic agents.23 At low concentrations, vinblastine
does not depolymerize spindle microtubules but blocks mitosis,
and cells die by apoptosis.32 Similarly, low concentrations of
paclitaxel block mitosis by kinetically stabilizing spindle
microtubules without changing the mass of polymerized
microtubules.33 It is possible that the actin−ApA complex
might inhibit spindle microtubule dynamics at the lowest
effective concentration by forming a ternary complex.
Microtubule−actin interactions underlie many fundamental
cellular processes, such as cell motility, neuronal pathfinding,
cell division, and cortical flow.34 A variety of proteins mediate
microtubule−actin interactions and regulate their dynamics.
Among those, mammalian diaphanous-related (mDia) formin
proteins not only nucleate and assemble linear actin filaments
but also directly bind to microtubules and regulate their
stabilization.35a A member of the mDia family, mDia3,
associates with the kinetochore and contributes to chromosome
alignment in the M-phase.35b Another F-actin-associated
protein, drebrin, binds to EB3 to coordinate the F-actin−
microtubule interactions responsible for neuritogenesis.36 It is
possible that ApA modulates the coordination between the
microtubules and actin and affects cytoskeleton dynamics by
mimicking such microtubule-targeting, actin-binding proteins.
Only a few natural products have been shown to
simultaneously interact with more than one biomacromolecule,
such as FK506, an immunosuppressive macrolide, which targets
FKBP and calcineurin.37 Thus, ApA represents a relatively rare
type of compound, which binds to two different cytoplasmic
proteins and forms a ternary complex. Rapamycin, a macrolide
structurally related to FK506, also binds to FKBP with high
affinity, but the FKBP−rapamycin complex targets mTOR to
inhibit the activation of lymphocytes in a different manner.38
Actin is an abundant cytoplasmic protein similar to FKBP.
Likewise, it is possible that actin-targeting agents interact with
multiple cellular targets via protein−protein interactions. The
results of these studies have potential in the design and
development of newly classified pharmacological tools and
therapeutic agents.
■
We thank Dr. Elmar Schiebel (Heidelberg University) for
critical reading of the manuscript. This work was supported in
part by JSPS grants (25702047 to M.K., 23310148 to H.K., and
LR018 to M.U.) and by Grants-in-Aid for Scientific Research
on Innovative Areas from MEXT, Japan, “Chemical Biology of
Natural Products.” Support was also provided by the Naito
Foundation, the Uehara Memorial Foundation, and the Takeda
Science Foundation. iCeMS is supported by World Premier
International Research Center Initiative (WPI), MEXT, Japan.
REFERENCES
■
(1) dos Remedios, C. G.; Chhabra, D.; Kekic, M.; Dedova, I. V.;
Tsubakihara, M.; Berry, D. A.; Nosworthy, N J. Physiol. Rev. 2003, 83,
433−473.
(2) (a) Allingham, J. S.; Klenchin, V. A.; Rayment, I. Cell. Mol. Life
Sci. 2006, 63, 2119−2134. (b) Yeung, K. S.; Paterson, I. Angew. Chem.,
Int. Ed. 2002, 41, 4632−4653.
(3) Vincent, E.; Saxton, J.; Baker−Glenn, C.; Moal, I.; Hirst, J. D.;
Pattenden, G.; Shaw, P. E. Cell Mol. Life Sci. 2007, 64, 487−497.
(4) (a) Saito, S.; Watabe, S.; Ozaki, H.; Fusetani, N.; Karaki, H. J.
Biol. Chem. 1994, 269, 29710−29714. (b) Wada, S.; Matsunaga, S.;
Saito, S.; Fusetani, N.; Watabe, S. J. Biochem. 1998, 123, 946−952.
(5) Tanaka, J.; Yan, Y.; Choi, J.; Bai, J.; Klenchin, V. A.; Rayment, I.;
Marriott, G. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 13851−13856.
(6) Zhang, X.; Minale, L.; Zampella, A.; Smith, C. D. Cancer Res.
1997, 57, 3751−3758.
(7) (a) Bubb, M. R.; Spector, I.; Bershadsky, A. D.; Korn, E. D. J. Biol.
Chem. 1995, 270, 3463−3466. (b) Saito, S. Y.; Watabe, S.; Ozaki, H.;
Kobayashi, M.; Suzuki, T.; Kobayashi, H.; Fusetani, N.; Karaki, H. J.
Biochem. (Tokyo) 1998, 123, 571−578.
(8) (a) Statsuk, A. V.; Bai, R.; Baryza, J. L.; Verma, V. A.; Hamel, E.;
Wender, P. A.; Kozmin, S. A. Nat. Chem. Biol. 2005, 1, 383−388.
(b) Rizvi, S. A.; Courson, D. S.; Keller, V. A.; Rock, R. S.; Kozmin, S.
A. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 4088−4092.
(9) Suria, H.; Chau, L. A.; Negrou, E.; Kelvin, D. J.; Madrenas, J. Life
Sci. 1999, 65, 2697−2707.
(10) Rao, J. Y.; Jin, Y. S.; Zheng, Q. L.; Cheng, J.; Tai, J.; Hemstreet,
G. P., III Cell Biochem. 1999, 75, 686−697.
(11) Posey, S. C.; Bierer, B. E. J. Biol. Chem. 1999, 274, 4259−4265.
(12) Konishi, H.; Kikuchi, S.; Ochiai, T.; Ikoma, H.; Kubota, T.;
Ichiwaka, D.; Fujiwara, H.; Okamoto, K.; Sakakura, C.; Sonoyama, T.;
Kokuba, Y.; Sasaki, J.; Matsui, T.; Otsuji, E. Anticancer Res. 2009, 29,
2091−2098.
(13) Ohno, O.; Morita, M.; Kitamura, T.; Teruya, T.; Yoneda, K.;
Kita, M.; Kigoshi, H.; Suenaga, K. Bioorg. Med. Chem. Lett. 2013, 23,
1467−1471.
(14) (a) Yamada, K.; Ojika, M.; Ishigaki, T.; Yoshida, Y.; Ekimoto,
H.; Arakawa, M. J. Am. Chem. Soc. 1993, 115, 11020−11021.
(b) Yamada, K.; Ojika, M.; Kigoshi, H.; Suenaga, K. Nat. Prod. Rep.
2009, 26, 27−43.
(15) Saito, S.; Watabe, S.; Ozaki, H.; Kigoshi, H.; Yamada, K.;
Fusetani, N.; Karaki, H. J. Biochem. (Tokyo) 1996, 120, 552−555.
(16) Hirata, K.; Muraoka, S.; Suenaga, K.; Kuroda, K.; Kato, K.;
Tanaka, H.; Yamamoto, M.; Takata, M.; Yamada, K.; Kigoshi, H. J.
Mol. Biol. 2006, 356, 945−954.
(17) Klenchin, V. A.; Allingham, J. S.; King, R.; Tanaka, J.; Marriott,
G.; Rayment, I. Nat. Struct. Biol. 2003, 10, 1058−1063.
(18) Allingham, J. S.; Zampella, A.; Auria, M. V. D; Rayment, I. Proc.
Natl. Acad. Sci. U.S.A. 2005, 102, 14527−14532.
ASSOCIATED CONTENT
■
S
* Supporting Information
Text, tables, and figures giving all experimental procedures,
characterization data for all compounds, H and 13C NMR
1
spectra, and details of biological evaluations. This material is
available free of charge via the Internet at http://pubs.acs.org.
18094
dx.doi.org/10.1021/ja406580w | J. Am. Chem. Soc. 2013, 135, 18089−18095

Journal of the American Chemical Society
Article
(19) Klenchin, V. A.; King, R.; Tanaka, J.; Marriott, G.; Rayment, I.
Chem. Biol. 2005, 12, 287−291.
(20) Rizvi, S. A.; Tereshko, V.; Kossiakoff, A. A.; Kozmin, S. A. J. Am.
Chem. Soc. 2006, 128, 3882−3883.
(21) (a) Kigoshi, H.; Suenaga, K.; Takagi, M.; Akao, A.; Kanematsu,
K.; Kamei, N.; Okugawa, Y.; Yamada, K. Tetrahedron 2002, 58, 1075−
1102. (b) Kigoshi, H.; Suenaga, K.; Mutou, T.; Ishigaki, T.; Atsumi, T.;
Ishiwata, H.; Sakakura, A.; Ogawa, T.; Ojika, M.; Yamada, K. J. Org.
Chem. 1996, 61, 5326−5351. (c) Suenaga, K.; Kamei, Y.; Okugawa,
M.; Takagi, M.; Akao, A.; Kigoshi, H.; Yamada, K. Bioorg. Med. Chem.
Lett. 1997, 7, 269−274.
(22) Kita, M.; Yoneda, K.; Hirayama, Y.; Yamagishi, K.; Saito, Y.;
Sugiyama, Y.; Miwa, Y.; Ohno, O.; Morita, M.; Suenaga, K.; Kigoshi,
H. ChemBioChem 2012, 13, 1754−1758.
(23) (a) Jordan, M. A.; Wilson, L. Nat. Rev. Cancer 2004, 4, 253−
265. (b) Dumontet, C.; Jordan, M. A. Nat. Rev. Drug Discov. 2010, 9,
790−803.
(24) (a) Parness, J.; Horwitz, S. B. J. Cell Biol. 1981, 91, 479−487.
(b) Himes, R. H.; Kersey, R. N.; Heller−Bettinger, I.; Samson, F. E.
Cancer. Res. 1976, 36, 3798−3802.
(25) (a) Spector, I.; Shochet, N. R.; Kashman, Y.; Groweiss, A.
Science 1983, 219, 493−495. (b) Sumiya, E.; Shimogawa, H.; Sasaki,
H.; Tsutsumi, M.; Yoshita, K.; Ojika, M.; Suenaga, K.; Uesugi, M. ACS
Chem. Biol. 2011, 6, 425−431.
(26) Kita, M.; Hirayama, Y.; Yamagishi, K.; Yoneda, K.; Fujisawa, R.;
Kigoshi, H. J. Am. Chem. Soc. 2012, 134, 20314−20317.
(27) Kita, M.; Hirayama, Y.; Sugiyama, M.; Kigoshi, H. Angew. Chem.,
Int. Ed. 2011, 50, 9871−9874.
(28) Kuroda, T.; Suenaga, K.; Sakakura, A.; Handa, T.; Okamoto, K.;
Kigoshi, H. Bioconjugate Chem. 2006, 17, 524−529.
(29) Jordan, M. A.; Wendell, K.; Gardiner, S.; Derry, W. B.; Copp,
H.; Wilson, L. Cancer Res. 1996, 56, 816−825.
(30) Suenaga, K.; Kigoshi, H. J. Synth. Org. Chem. Jpn. 2006, 64,
1273−1281.
(31) Bubb, M. R.; Govindasamy, L.; Yarmola, E. G.; Vorobiev, S. M.;
Almo, S. C.; Somasundaram, T.; Chapman, M. S.; Agbandje-McKenna,
M.; McKenna, R. J. Biol. Chem. 2002, 277, 20999−21006.
(32) Jordan, M. A.; Thrower, D.; Wilson, L. Cancer Res. 1991, 51,
2212−2222.
(33) (a) Jordan, M. A.; Toso, R. J.; Thrower, D.; Wilson, L. Proc.
Natl. Acad. Sci. U.S.A. 1993, 90, 9552−9556. (b) Jordan, M. A.;
Wendell, K.; Gardiner, S.; Derry, W. B.; Copp, H.; Wilson, L. Cancer
Res. 1996, 56, 816−825.
(34) Rodriguez, O. C.; Schaefer, A. W.; Mandato, C. A.; Forscher, P.;
Bement, W. M.; Waterman−Storer, C. M. Nat. Cell Biol. 2003, 5, 599−
609.
(35) (a) Ishizaki, T.; Morishima, Y.; Furuyashiki, T.; Kato, T.;
Narumiya, S. Nat. Cell Biol. 2001, 3, 8−14. (b) Yasuda, S.; Oceguera−
Yanez, F.; Kato, T.; Okamoto, M.; Yonemura, S.; Terada, Y.; Ishizaki,
T.; Narumiya, S. Nature 2004, 428, 767−771.
(36) Geraldo, S.; Khanzada, U. K.; Parsons, M.; Chilton, J. K.;
Gordon−Weeks, P. R. Nat. Cell Biol. 2008, 10, 1181−1189.
(37) Liu, J.; Farmer, J. D., Jr.; Lane, W. S.; Friedman, J.; Weissman, I.;
Schreiber, S. L. Cell 1991, 66, 807−815.
(38) (a) Kuo, C. J.; Chung, J.; Fiorentino, D. F.; Flanagan, W. M.;
Blenis, J.; Crabtree, G. R. Nature 1992, 358, 70−73. (b) Brown, E. J.;
Albers, M. W.; Shin, T. B.; Ichikawa, K.; Keith, C. T.; Lane, W. S.;
Schreiber, S. L. Nature 1994, 369, 756−758.
18095
dx.doi.org/10.1021/ja406580w | J. Am. Chem. Soc. 2013, 135, 18089−18095