Relocation of aurora B and survivin from centromeres to the central spindle impaired by a kinesin-specific MKLP-2 inhibitor

Article abstract of DOI:10.1002/anie.201003254

Relocation, relocation: Inactivation of the mitotic kinesin MKLP-2 by a specific small-molecule inhibitor (1) leads to failure in the recruitment of the chromosome passenger protein (survivin, red) to the central spindle (microtubules, green) during anaph

Full text of DOI:10.1002/anie.201003254

Communications
DOI: 10.1002/anie.201003254
Chemical Biology
Relocation of Aurora B and Survivin from Centromeres to the Central
Spindle Impaired by a Kinesin-Specific MKLP-2 Inhibitor**
Sergey Tcherniuk, Dimitrios A. Skoufias, Christophe Labriere, Oliver Rath, Franꢀoise Gueritte,
Catherine Guillou, and Frank Kozielski*
Mitotic kinesin-like protein 2 (MKLP-2), a member of the
kinesin-6 family, is essential for cytokinesis. We screened 8900
small molecules for inhibition of the ATPase activity of
MKLP-2 and identified the first inhibitor, (Z)-2-(1H-indol-3-
yl)-3-(pyridin-3-yl)acrylonitrile, which we named paprotrain
(PAssenger PROteins TRAnsport INhibitor; Figure 1b).[1]
Paprotrain is a reversible inhibitor uncompetitive with ATP
(inhibitor constant Ki = (3.4 Æ 0.1) mm) and noncompetitive
with microtubules (MTs; Ki = (1.6 Æ 0.1) mm). It is specific for
MKLP-2 as it does not inhibit other members of the kinesin
superfamily. Paprotrain is cell permeable and incubation with
10 to 50 mm inhibitor results in binucleated cells, a character-
istic phenotype similar to that observed following RNA
interference (RNAi)-mediated depletion of MKLP-2. Addi-
tional mitotic spindle defects were also observed. Notably,
paprotrain perturbs MKLP-2-mediated relocation of the
chromosome passenger proteins Aurora B and survivin from
the centromeres to the central spindle. Lack of relocation of
passenger proteins to the central spindle is associated with
failure of cytokinesis and generation of binucleated cells. In
contrast, paprotrain does not impair kinesin family member 4
(Kif4)-mediated translocation of protein regulating cytokine-
sis 1 (PRC1), which emphasizes its specificity for MKLP-2.
We conclude that paprotrain is a new lead compound
targeting MKLP-2 by a novel mechanism of action.
The motor domain of human MKLP-2 (residues 1–519)
was cloned, expressed, and purified (Figure 1a). Gel filtration
and analytical ultracentrifugation showed that the protein is
monomeric (data not shown) with an estimated molecular
mass of (77 Æ 5) kDa (see Table S1 in the Supporting Infor-
mation).
Figure 1. a) Bar diagram of human MKLP-2 and the protein construct
used for inhibitor screening. MKLP-2 has an N-terminal motor domain
(amino acids (aa) 56 to 505), an a-helical region (aa 519 to 821), and
a C-terminal tail domain (aa 821 to 890). MKLP-2 contains an
approximately 100-residue insertion (aa 189 to 288) in the motor
domain that is unique for members of the kinesin-6 family. b) Chem-
ical formula of paprotrain. Trx=thioredoxin.
The basal ATPase activity was (0.034 Æ 0.012) sÀ1, with a
KM,ATP of 59.4 mm. The activity decreased with increasing NaCl
concentration and inhibitor screening was therefore per-
formed in the absence of salt. The MT-stimulated ATPase
activity was 1.8 sÀ1, which corresponded to an estimated 53-
fold stimulation compared to the basal activity. The K0.5,MT was
1.6 mm (see Figure S1 and Table S2 in the Supporting Infor-
mation).
[*] Dr. O. Rath, Prof. Dr. F. Kozielski
The Beatson Institute for Cancer Research
Switchback Road, Bearden, Glasgow G61 1BW (UK)
Fax: (+44)141-942-6521
We measured the inhibition of the basal and MT-
stimulated ATPase activity in the presence of paprotrain
(Figure 2). Calculated IC50 values were (1.35 Æ 0.2) mm in
basal (Figure 2a) and (0.83 Æ 0.1) mm in MT-stimulated
ATPase assays (Figure 2b). The measured rates of ATP
hydrolysis in the presence of varying concentrations of
paprotrain and MTs demonstrate that the inhibitor acts
through a mixed noncompetitive mechanism with respect to
MTs with an inhibitor constant Ki = (1.6 Æ 0.07) mm (Fig-
ure 2c). We also analyzed the ATP hydrolysis rate at different
ATP and inhibitor concentrations. Paprotrain is an ATP
uncompetitive inhibitor with Ki = (3.36 Æ 0.09) mm (Fig-
ure 2d).
E-mail: f.kozielski@beatson.gla.ac.uk
Dr. S. Tcherniuk, Dr. D. A. Skoufias
IBS (CEA-CNRS-UJF)
41, rue Jules Horowitz, 38027 Grenoble (France)
Dr. C. Labriere, Dr. F. Gueritte, Dr. C. Guillou
Centre de Recherche de Gif, ICSN-CNRS Bt 27
Avenue de la Terasse, 91198 Gif-sur-Yvette (France)
[**] This work was funded by grants from ARC, the CNRS (C.L.), ANR,
and CR-UK. We thank R.-L. Indorato, S. DeBonis, S. Talapatra, E.
Barrꢀ, M. T. Martin, C. Ebel, and D. Grunwald for excellent technical
support. Prof. J. Y. Lallemand, Prof. H. P. Husson, and Dr. C. Thal
are gratefully acknowledged for their interest in our work.
MKLP-2=mitotic kinesin-like protein 2.
The ATPase activity of a panel of 12 other kinesins[2,3] was
not inhibited by paprotrain, thus demonstrating high specif-
icity within the kinesin superfamily (see Figure S2a in the
Supporting Information). Notably, paprotrain does not inhibit
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201003254.
8228
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 8228 –8231

Angewandte
Chemie
Table 1: Inhibitor analogues investigated to study the structure–activity
relationship of paprotrain-related compounds.
Compound
Basal ATPase activity
MT-stimulated
ATPase activity
IC50 [mm]
IC50 [mm]
1
2
1.35Æ0.2
0.83Æ0.1
20.1Æ3.7
9.1Æ2.5
(35%)[a]
(25%)[a]
61.4Æ9.0
5b
n.i.[b]
(55%)[a]
Figure 2. a) Inhibition of the basal MKLP-2 ATPase activity in the presence of
paprotrain. b) Inhibition of the MT-activated MKLP-2 ATPase activity by the
&
*
inhibitor. c) Paprotrain is noncompetitive when mixed with MTs: 0; 0.35;
5c
5d
7a
7b
n.i.[b]
n.i.[b]
~
^
0.7; 1.1 mm MKLP-2 inhibitor. d) The inhibitor is uncompetitive with ATP:
~
*
^
0; & 0.7; 1.5; 2.5 mm inhibitor.
the two closest related kinesins involved in cytokinesis,
MKLP-1 and M-phase phosphoprotein 1 (MPP1).
25.0Æ6.8
52.0Æ5.2
(55%)[a]
(50%)[a]
MKLP-2–MT co-sedimentation assays showed that the
inhibitor did not influence the binding of MKLP-2 to MTs
(Figure S2b). Compared to nocodazole and taxol, paprotrain
has no significant effect on MT polymerization (Figure S2c).
We synthesized and characterized seven additional inhib-
itor analogues (see Schemes S1–S4 in the Supporting Infor-
mation) and assessed the inhibition of MKLP-2 (Scheme 1
and Table 1). The two heterocycles in 1 are in the trans
position to each other. To evaluate the biological effect of the
double bondꢀs stereochemistry, the trans double bond of
compound 1 was isomerized by photochemical irradiation to
afford compound 2 with a cis double bond. This cis
stereochemistry was confirmed by a coupling constant of
9 Hz between the H3 proton and the C1 carbon atom in the
heteronuclear multiple bond correlation (HMBC) spectrum.
Compound 2 inhibits MKLP-2 with considerably higher IC50
values. In 1, the pyridinyl group is connected in the meta
position to the double bond and it is in the trans position with
the indole scaffold. Two additional compounds that have the
n.i.[b]
n.i.[b]
n.i.[b]
n.i.[b]
8
n.i.[b]
n.i.[b]
[a] Percentage of maximal inhibition. [b] n.i.: no inhibition.
pyridinyl group connected in either the ortho (5b) or para
position (5c) do not inhibit MKLP-2. Compound 5d, in which
the pyridine is replaced by a phenyl group, is a less potent
inhibitor of MKLP-2. Equally striking are the results obtained
with two other compounds, in which the nitrogen atom of the
indole ring is substituted in the same position by either sulfur
(7a) or oxygen (7b). Neither compound inhibits MKLP-2.
Finally, the N-methylated indole analogue (8) is also inactive.
Scheme 1. Synthesis of inhibitor analogues.
Angew. Chem. Int. Ed. 2010, 49, 8228 –8231
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 8229

Communications
Paprotrain was cell permeable and effective in HeLa cells.
chromosomes in anaphase, a phenotype, which resembles
Aurora B localization following RNAi-mediated depletion of
survivin.[8] The absence of passenger proteins at the spindle
midzone is also linked to failure of cytokinesis and may
explain the appearance of binucleated cells following papro-
train treatment. The motor proteins, Kif4, MKLP-1, and
MPP1, are also thought to participate in cytokinesis by
interacting with PRC1, a MT-bundling protein necessary for
central spindle formation.[9,10] In the presence of paprotrain,
the normal PRC1 localization pattern in anaphase was
observed (Figure 3b, right images). Therefore, we conclude
that paprotrain does not affect Kif4- or MKLP-1-mediated
transport of PRC1 to the spindle midzone during cytokinesis.
In proliferation assays, paprotrain inhibits growth with
EC50 values between 7.6 and 22 mm (see Table S3 in the
Supporting Information). Paprotrain is not a substrate for the
efflux pump P-glycoprotein (see Table S4 in the Supporting
Information). The inhibitor induces additional spindle defects
that are not observed in cells following MKLP-2 depletion by
RNAi. Metaphase spindles appeared to have an incomplete
chromosome complement aligned to the metaphase plate
(metaphases in Figure 3b). In untreated cells 91.5 Æ 5% of
metaphase spindles contained a normal bipolar spindle
(Figure 3a, upper panels). In con-
Previous studies have shown that RNAi-mediated depletion
of MKLP-2 led to increased levels of binucleated cells.[4,5]
After 8 h of incubation, the number of binucleated cells
(Figure 3a and b, left panels) increased to more than three
times that of untreated cells (11 Æ 1 vs. 3.5 Æ 0.5%, respec-
tively). Exposure of cells to dihydrocytochalasin B[6] (DCB;
10 mm) resulted in 32 Æ 12% binucleated cells (Figure S3a in
the Supporting Information). After 24 h of treatment with
paprotrain (10–50 mm), a significant number of cells were in
apoptosis, confirmed by immunodetection of cleaved poly-
(ADP-ribose) polymerase (PARP) and activated caspase 3
(Figure S3b).
MKLP-2 function is essential for the relocation of the
chromosome passenger protein complex (CPC) from centro-
meres to the spindle midzone during the metaphase-to-
anaphase transition[7] (Figure 3a). The inhibitor did not affect
the inner centromere recruitment of Aurora B in metaphase,
but it inhibited the relocation of Aurora B and survivin to the
spindle midzone in anaphase (Figure 3b). A similar pheno-
type was observed in cells following RNAi depletion of
MKLP-2 (see Figure S3c).[7] In the presence of the inhibitor,
Aurora B and survivin were diffusely associated with the
trast, after treatment with papro-
train the majority of metaphase
spindles (70 Æ 10% of total meta-
phase cells) had misaligned chro-
mosomes (Figure S3d). Higher
concentrations of paprotrain
(50 mm) increased the percentage
of multipolar spindles at the
expense of normal bipolar spin-
dles (14 Æ 3% of total metaphase
cells at 50 mm versus 6 Æ 1.6% at
20 mm; Figure S3e). Phenotypic
dissimilarities have also been
reported for the Kif18A[11] inhib-
itor and might reflect differences
originating from distinct experi-
mental approaches (inhibition
caused by RNAi versus small-
molecule compounds) requiring
further evaluation of the cellular
phenotypes.
To date, kinesin inhibitors
have only been reported for
Eg5,[12]
CENP-E,[13,14]
and
Kif18A.[11] Herein, we have iden-
tified and characterized the first
inhibitor of the mitotic kinesin
MKLP-2. Paprotrain will be a
useful tool for chemical genetics
to study the function of MKLP-2
in cytokinesis and to gain mecha-
nistic insights, for example, to
exploit mechanistic differences
between the various kinesins. In
the case of MKLP-2, it might shed
Figure 3. Paprotrain induces binucleated cells and impedes the relocation of the CPC from the inner
centromeres to the central spindle during anaphase. Immunostaining of a) control untreated cells and
b) cells treated with paprotrain (50 mm) after 8 h. MTs (TUB) were stained with fluorescein isothiocya-
nate (green), Aurora B (AurB), survivin (SRV), and PRC1 with cyanine dye Cy3 (red), and DNA with
4’,6-diamidino-2-phenylindole (blue).
8230 www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 8228 –8231

Angewandte
Chemie
[8] L. Scrittori, D. A. Skoufias, F. Hans, V. Gerson, P. Sassone-Corsi,
light on the question of whether MKLP-2 is functionally
involved in intracellular transport involving the Golgi.[15]
Finally, inhibitors that target cytokinesis-specific kinesins
may help to validate these proteins as potential targets for
drug development in cancer chemotherapy.[16]
S. Dimitrov, R. L. Margolis, Mol. Biol. Cell 2005, 16, 292.
[9] C. Zhu, W. Jiang, Proc. Natl. Acad. Sci. USA 2005, 102, 343.
[10] Y. Kurasawa, W. C. Earnshaw, Y. Mochizuki, N. Dohmae, K.
Todokoro, EMBO J. 2004, 23, 3237.
[11] M. Catarinella, T. Gruner, T. Strittmatter, A. Marx, T. U. Mayer,
Angew. Chem. 2009, 121, 9236; Angew. Chem. Int. Ed. 2009, 48,
9072.
Received: May 28, 2010
Published online: September 20, 2010
[12] G. Bergnes, K. Brejc, L. Belmont, Curr. Top. Med. Chem. 2005, 5,
127.
[13] M. C. Henderson, Y. J. Shaw, H. Wang, H. Han, L. H. Hurley, G.
Flynn, R. T. Dorr, D. D. Von Hoff, Mol. Cancer Ther. 2009, 8, 36.
[14] K. W. Wood, L. Lad, L. Luo, X. Qian, S. D. Knight, N. Nevins, K.
Brejc, D. Sutton, A. G. Gilmartin, P. R. Chua, R. Desai, S. P.
Schauer, D. E. McNulty, R. S. Annan, L. D. Belmont, C. Garcia,
Y. Lee, M. A. Diamond, L. F. Faucette, M. Giardiniere, S. Zhang,
C. M. Sun, J. D. Vidal, S. Lichtsteiner, W. D. Cornwell, J. D.
Greshock, R. F. Wooster, J. T. Finer, R. A. Copeland, P. S.
Huang, D. J. Morgans, Jr., D. Dhanak, G. Bergnes, R. Sakowicz,
J. R. Jackson, Proc. Natl. Acad. Sci. USA 2010, 107, 5839.
[15] A. Echard, F. Jollivet, O. Martinez, J. J. Lacapere, A. Rousselet,
I. Janoueix-Lerosey, B. Goud, Science 1998, 279, 580.
[16] K. Taniuchi, H. Nakagawa, T. Nakamura, H. Eguchi, H.
Ohigashi, O. Ishikawa, T. Katagiri, Y. Nakamura, Cancer Res.
2005, 65, 105.
Keywords: antimitotics · antitumor agents · cytokinesis ·
inhibitors · kinesin
.
[1] C. Guillou, F. Kozielski, C. Labriere, F. Gueritte, D. A. Skoufias,
S. Tcherniuk, C. Thal, H. P. Husson, EP 09163523, 2009.
[2] F. Kozielski, S. DeBonis, D. A. Skoufias, Methods Mol. Med.
2007, 137, 189.
[3] S. DeBonis, D. A. Skoufias, L. Lebeau, R. Lopez, G. Robin, R. L.
Margolis, R. H. Wade, F. Kozielski, Mol. Cancer Ther. 2004, 3,
1079.
[4] E. Hill, M. Clarke, F. A. Barr, EMBO J. 2000, 19, 5711.
[5] R. D. Fontijn, B. Goud, A. Echard, F. Jollivet, J. van Marle, H.
Pannekoek, A. J. Horrevoets, Mol. Cell. Biol. 2001, 21, 2944.
[6] T. E. Schroeder, Front. Biol. 1978, 46, 91.
[7] U. Gruneberg, R. Neef, R. Honda, E. A. Nigg, F. A. Barr, J. Cell
Biol. 2004, 166, 167.
Angew. Chem. Int. Ed. 2010, 49, 8228 –8231
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
8231