Selective elimination of human pluripotent stem cells by a marine natural product derivative

Article abstract of DOI:10.1021/ja501795c

One of the current obstacles to stem cell therapy is the tumorigenic potential of residual undifferentiated stem cells. The present study reports rediscovery of a synthetic derivative of okadaic acid, a marine polyether toxin, as a reagent that selectively induces the death of human pluripotent stem cells. Cell-based screening of 333 cytotoxic compounds identified methyl 27-deoxy-27-oxookadaate (molecule 1) as a substrate of two ATP-binding cassette (ABC) transporters, ABCB1 (MDR1) and ABCG2 (BCRP), whose expression is repressed in human embryonic stem cells and induced pluripotent stem cells. The results demonstrate that selective elimination of human pluripotent stem cells can be achieved by designing cytotoxic small molecules with appropriate ABC-transporter selectivity.

Full text of DOI:10.1021/ja501795c

Communication
pubs.acs.org/JACS
Selective Elimination of Human Pluripotent Stem Cells by a Marine
Natural Product Derivative
Ting-Fang Kuo,§,‡,⊥ Di Mao,§,‡,⊥ Nao Hirata,§,‡ Bilon Khambu,§,‡ Yasuhisa Kimura,† Eihachiro Kawase,∥
Hiroki Shimogawa,‡ Makoto Ojika,# Norio Nakatsuji,§,∥ Kazumitsu Ueda,§,† and Motonari Uesugi*,§,‡
§Institute for Integrated Cell-Material Sciences (WPI-iCeMS), Kyoto University, Kyoto 606-8501, Japan
‡Institute for Chemical Research, Kyoto University, Uji, Kyoto 611-0011, Japan
†Division of Applied Life Sciences, Graduate School of Agriculture, Kyoto University, Kyoto 606-8502, Japan
∥Institute for Frontier Medical Sciences, Kyoto University, Kyoto 606-8507, Japan
#Graduate School of Bioagricultural Sciences, Nagoya University, Chikusa-ku, Nagoya 464-8601, Japan
S
* Supporting Information
distinguishes between human pluripotent stem cells and
differentiated cells.9 The selectivity results primarily from a
ABSTRACT: One of the current obstacles to stem cell
therapy is the tumorigenic potential of residual undiffer-
entiated stem cells. The present study reports rediscovery
of a synthetic derivative of okadaic acid, a marine polyether
toxin, as a reagent that selectively induces the death of
human pluripotent stem cells. Cell-based screening of 333
cytotoxic compounds identified methyl 27-deoxy-27-
oxookadaate (molecule 1) as a substrate of two ATP-
binding cassette (ABC) transporters, ABCB1 (MDR1)
and ABCG2 (BCRP), whose expression is repressed in
human embryonic stem cells and induced pluripotent stem
cells. The results demonstrate that selective elimination of
human pluripotent stem cells can be achieved by designing
cytotoxic small molecules with appropriate ABC-trans-
porter selectivity.
distinct expression pattern of ATP-binding cassette (ABC)
transporters in human pluripotent stem cells and from the
unique transporter selectivity of KP-1. Expression of ABCB1
(MDR1) and ABCG2 (BCRP), both of which cause the efflux
of KP-1, is repressed in human pluripotent stem cells. However,
ABCC1 (MRP1), which is not involved in KP-1 efflux, is
expressed in human pluripotent stem cells (Figure 1A). The
tem cell therapy using human pluripotent stem cells,
Sincluding embryonic stem cells (hESCs) and induced
pluripotent stem cells (hiPSCs), has the potential to treat a
wide variety of diseases. One of the current problems of such
therapy is the tumorigenic risk of residual undifferentiated
cells.1,2 Several strategies for removing undifferentiated cells in
therapeutic transplants have been reported, including cell
sorting and precipitation using an antibody against a stem cell
surface antigen,3,4 and selective killing of pluripotent stem cells
by a cytotoxic antibody.5 Approaches using small molecules to
eliminate pluripotent stem cells have also been examined. In
one method, an inhibitor of stearoyl-CoA desaturase (SCD1)
eliminates hPSCs by blocking the biosynthesis of oleic acid.6 In
a second method, a glucose-depleted culture medium
containing abundant lactate enables selective survival of
cardiomyocytes and eliminates non-cardiomyocytes, including
undifferentiated stem cells.7 A third method uses chemical
inhibitors of survivin, one of the antiapoptotic factors
overexpressed in pluripotent stem cells.8 The discovery of
new mechanisms that could be combined with such elimination
methods would ensure selective and rapid removal of residual
undifferentiated cells.
Figure 1. (A) Distinct expression pattern of ATP-binding cassette
(ABC) transporters in human pluripotent stem cells and differentiated
cells. (B) Chemical structures of methyl 27-deoxy-27-oxookadaate
(molecule 1), okadaic acid (molecule 2), and methyl okadaate
(molecule 3).
By screening a chemical library, we previously discovered
Kyoto Probe 1 (KP-1), a fluorescent chemical probe that
Received: February 25, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja501795c | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
distinct expression pattern of ABC transporters in human
pluripotent stem cells could provide the basis for a new small
molecule-based approach for selective elimination of those cells.
In theory, selective death of human pluripotent stem cells
should be achieved by cytotoxic molecules with ABC
transporter selectivity similar to that of KP-1.
Table 1. IC50 Values of (A) Molecules 1−3 in Model Cells
and (B) Molecule 1 in hiPSCs, Human Primary Cells, and
Cancer Cells, with or without the Addition of CsA, PSC833,
a
or Ko143
An in-house chemical library of cytotoxic small molecules
(333 natural product derivatives and anticancer drugs) was
screened for their effects on the viability of four cell lines: KB3-
1 cells, which express no detectable levels of ABC transporters,
and KB3-1-derived cells that overexpress ABCB1,10 ABCC1,11
or ABCG29 (KB3-1/ABCB1, KB3-1/ABCC1, and KB3-1/
ABCG2). The screening details are described in Table S1. We
selected compounds that had stronger cytotoxic effects on wild-
type KB3-1 and KB3-1/ABCC1 cells than on KB3-1/ABCB1
and KB3-1/ABCG2 cells. Molecule 1 (methyl 27-deoxy-27-
oxookadaate), a synthetic derivative of a well-known marine
natural product, okadaic acid, exhibited the clearest and most
desirable profile.
Molecule 1 was originally synthesized as a derivative of
okadaic acid (molecule 2) by Yamada and co-workers at
Nagoya University about 25 years ago, but was never described
in the literature. We resynthesized molecule 1 and confirmed its
structure and activity. The chemical structure and synthetic
route of molecule 1 are shown in Figure 1B and Scheme S1,
respectively. We then studied the structure−activity relation-
ships of eight okadaic acid derivatives previously synthesized
(Table S2).12,13 Okadaic acid (molecule 2) itself did not act as a
substrate for any of the transporters, and exhibited potent
cytotoxicity to all of the cells tested. Esterification of the
carboxyl group (e.g., molecule 3) enhanced transport by
ABCB1, consistent with the reported preference of ABCB1 for
hydrophobic molecules.14−16 Among the okadaic acid deriva-
tives we tested, molecule 1 was the only derivative that
exhibited clear efflux both by ABCB1 and ABCG2, although the
structural modifications in molecule 1 compromised overall
cytotoxicity, compared to okadaic acid. The reason why the
simple oxidation of a hydroxyl group to a keto group makes
molecule 1 a suitable substrate of ABCG2 is unknown.
To confirm that the ABC transporters were responsible for
the observed cell resistance, we examined the effects of selective
inhibitors for ABCB1 and ABCG2, on IC50 values (Table 1A
and Table S3). Treatment with selective inhibitors of ABCB1,
cyclosporine A (CsA) and its non-immunosuppressive
derivative PSC833,17 greatly reduced the resistance of KB3-1/
ABCB1 cells to molecule 3, whereas Ko143, a potent non-toxic
inhibitor of ABCG2,18 had no detectable effect on its activity to
KB3-1/ABCG2 cells. On the other hand, the cytotoxicity of
molecule 1 in KB3-1/ABCB1 and KB3-1/ABCG2 cells was
restored by the ABCB1 and ABCG2 inhibitors, respectively,
confirming that molecule 1 is a substrate of both ABCB1 and
ABCG2. Neither inhibitor had detectable effects on the activity
of molecule 2.
a
Part A shows the IC50 values of molecules 1, 2, and 3 in KB3−1-
derived cell lines (n = 12). The level of resistance is indicated in blue.
Cyclosporine A (CsA; 10 μM) and PSC833 (1 μM) are selective
inhibitors of ABCB1, and Ko143 (10 μM) is a selective inhibitor of
ABCG2. Panel B shows the IC50 values of molecule 1 in hiPSCs,
human somatic primary cells, and human cancer cells (n = 8). The
level of cytotoxic potency is indicated in red. It is evident that CsA and
Ko143 restored activity of molecule 1 in resistant cells.
model cells. Collectively, the results indicated that further
analysis should focus on molecule 1.
KP-1 has been shown to selectively stain human pluripotent
stem cells and neuronal cells, which have low expression of
ABC transporters, including ABCB1 and ABCG2.9 As molecule
1 has a selectivity profile similar to that of KP-1, molecule 1 was
expected to induce cell death selectively in KP-1-positive cells.
Therefore, we compared the IC50 values of molecule 1 in
hiPSCs, human somatic primary cells, and human cancer cells
(Table 1B). Molecule 1 was highly toxic to hiPSCs, with an
IC50 value of 0.78 μM (Figure S2), even lower than the IC50
value of astrocytes (2.1 μM, Table 1B), a neuronal cell lineage
known to express low levels of ABC transporters.21 In contrast,
cells primarily involved in secretory functions (adrenal gland),
metabolic functions (liver), barrier functions (bronchia), and
reproductive organs (prostate) tend to express higher levels of
ABC transporters21 and exhibited greater resistance to molecule
1 (IC50 > 5 μM). Addition of CsA or Ko143 increased
cytotoxicity of molecule 1 in these resistant somatic cells.
HepG2 cells, a well-differentiated human hepatocellular
carcinoma cell line that expresses high levels of ABC
transporters,22 had IC50 values as high as 14.7 μM. In contrast,
HeLa cells, a human cervical carcinoma cell line that displays
low levels of ABC transporters,22 exhibited IC50 values as low as
that of hiPSCs (1.2 μM). SNL cells, a mouse embryonic
To confirm that molecule 1 is a direct substrate for ABCB1,
we carried out in vitro biochemical assays, in which ATPase
activity of purified, reconstituted ABCB1 was monitored in the
presence of various concentrations of molecules 1−3 (Figure
S1).19 The amounts of ADP produced were measured using
HPLC. Verapamil, a known substrate of ABCB1, drove the
ATPase activity of ABCB1.20 Molecules 1 and 3 activated
ATPase activity to comparable levels, while molecule 2 failed to
do so, consistent with the selectivity of their cytotoxicity to the
B
dx.doi.org/10.1021/ja501795c | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
fibroblast cell line often used as feeder cells for hESC and
hiPSC culture, were resistant to molecule 1, and addition of
CsA or Ko143 increased cytotoxicity of molecule 1 in these
cells. Thus, the selectivity profile of molecule 1 paralleled KP-1
labeling.
To determine the ability of molecule 1 to distinguish hiPSCs
from differentiated cells, we first added molecule 1 to hiPSCs
cultured on SNL feeder cells. As expected, molecule 1 (5 μM)
selectively eliminated hiPSC colonies, leaving the feeder cells
on the plates (Figure 2A, left panel). A 50-fold lower
concentration of okadaic acid (molecule 2) killed both the
hiPSCs and feeder cells. We next added molecule 1 (5 μM) to
partially differentiated hiPSCs, prepared by treating hiPSCs
with 5 μM of all-trans retinoic acid (RA) for 72 h. These
experiments were carried out under both feeder cell (Figure 2A,
right panel) and feeder cell-free (Figure 2B) conditions. Unlike
hiPSCs, which form colonies, the partially differentiated cells
were flat and spread out on the culture plates. While 0.1 μM
molecule 2 killed hiPSCs, feeder cells, and differentiated cells, 5
μM molecule 1 appeared to selectively eliminate hiPSCs,
leaving the differentiated cells. We obtained similar results using
hESCs, in place of hiPSCs (Figure S3).
To confirm the selectivity of molecule 1, we used
fluorescence staining with a stem cell marker, SSEA-4, and a
differentiation marker, SSEA-1. Partially differentiated hiPSCs
displayed a ∼1:1 distribution of SSEA-4-positive cells and
SSEA-1-positive cells. Treatment of the partially differentiated
sample with DMSO alone for 48 h had no effect on the
distribution of these cells (Figure 2C, upper panels). In
contrast, treatment with molecule 1 for 48 h selectively
eliminated the SSEA-4-positive cells (Figure 2C, lower panels).
Fluorescence-activated cell sorting (FACS) analysis showed
that molecule 1 removed SSEA-4-positive (Figure 2D) and
SSEA-1-negative cells (Figure 2E), confirming the ability of
molecule 1 to selectively eliminate undifferentiated cells from
the mixture.
Okadaic acid (molecule 2) exerts its cytotoxic activity by
inhibiting the protein serine/threonine phosphatases, PP1 and
PP2A.23,24 To examine whether molecule 1 also targets PP1
and PP2A, we carried out in vitro biochemical assays, using
recombinant PP1 and PP2 proteins. Okadaic acid (molecule 2)
showed potent inhibition: IC50 was 1 nM for PP2A and 2.5
nM for PP1 (Figure S4A). Molecules 1 and 3 also inhibited
PP1 and PP2A in a dose-dependent manner (Figure S4BC),
but less potently: IC50 ≈ 1 μM for PP2A and IC50 ≈ 5 μM for
PP1, consistent with their cytotoxicity. Thus, it is likely that
molecule 1 exerts its cytotoxicity at least in part by inhibiting
protein phosphatases.
In conclusion, the present study offers a simple small
molecule-based approach to eliminating undifferentiated
pluripotent stem cells from cell mixtures. The selective
cytotoxicity of molecule 1 is based on its unique ABC-
transporter selectivity. The limitation of this new approach is
the inability of molecule 1 to distinguish between pluripotent
stem cells and neuronal cells, similarly to KP-1 staining. Despite
this limitation, molecule 1 or its analogs might serve as useful
tools for both cell biology and stem cell therapy. Combining
molecule 1 or its analogues with other reagents might improve
the purity of differentiated cell samples for transplantation.
ASSOCIATED CONTENT
■
S
* Supporting Information
Supplementary data, experimental procedures, and character-
ization data. This material is available free of charge via the
Internet at http://pubs.acs.org.
Figure 2. (A,B) Cytotoxic effects of molecule 1 on hiPSCs or partially
differentiated hiPSCs. The cells were treated with molecule 2 (0.1
μM) or molecule 1 (5 μM) for 48 h on feeder cells (A) or under
feeder-free conditions (B), n = 4. (C) Double staining of partially
differentiated hiPSCs by a stem cell marker (SSEA-4) or a
differentiation marker (SSEA-1), after treatment with molecule 1 (5
μM) or DMSO alone for 48 h under feeder-free conditions. (D,E)
FACS analysis of partially differentiated hiPSCs immunostained by
SSEA-4 (D) or SSEA-1 (E), after treatment with molecule 2 (0.1 μM)
or molecule 1 (5 μM) for 48 h under feeder-free conditions. Each
percentage was normalized to the corresponding viability, n = 3.
AUTHOR INFORMATION
■
Corresponding Author
uesugi@scl.kyoto-u.ac.jp
Author Contributions
⊥T.-F.K. and D.M. contributed equally to the work.
C
dx.doi.org/10.1021/ja501795c | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
(18) Allen, J. D.; van Loevezijn, A.; Lakhai, J. M.; van der Valk, M.;
van Tellingen, O.; Reid, G.; Schellens, J. H.; Koomen, G. J.; Schinkel,
A. H. Mol. Cancer Ther. 2002, 1, 417.
(19) Kimura, Y.; Shibasaki, S.; Morisato, K.; Ishizuka, N.; Minakuchi,
H.; Nakanishi, K.; Matsuo, M.; Amachi, T.; Ueda, M.; Ueda, K. Anal.
Biochem. 2004, 326, 262.
(20) Kimura, Y.; Matsuo, M.; Takahashi, K.; Saeki, T.; Kioka, N.;
Amachi, T.; Ueda, K. Curr. Drug Metab. 2004, 5, 1.
(21) Langmann, T.; Mauerer, R.; Zahn, A.; Moehle, C.; Probst, M.;
Stremmel, W.; Schmitz, G. Clin. Chem. 2003, 49, 230.
Notes
The authors declare the following competing financial
interest(s): N.N. is a founder and shareholder of ReproCELL,
Inc.
ACKNOWLEDGMENTS
■
We thank Kiyoyuki Yamada (Emeritus Professor at Nagoya
University) for sharing his collection of natural products and
their synthetic derivatives, and Takayuki Kondo and Haruhisa
Inoue at the Center for iPS Cell Research and Application
(CiRA), Kyoto University, for providing partially differentiated
iPS cell colonies. This work was supported in part by JSPS
(LR018 and 26220206 to M.U. and 25221203 to K.U.). iCeMS
is supported by World Premier International Research Center
Initiative (WPI), MEXT, Japan. This work was inspired by the
international and interdisciplinary environments of the iCeMS
and JSPS Asian CORE Program, “Asian Chemical Biology
Initiative”. The upgrade of the confocal microscope was
supported by the New Energy and Industrial Technology
Development Organization (NEDO) of Japan and Yokogawa
Electric Corporation.
(22) Ahlin, G.; Hilgendorf, C.; Karlsson, J.; Szigyarto, C. A.; Uhlen
́
,
M.; Artursson, P. Drug Metab. Dispos. 2009, 37, 2275.
(23) Maynes, J. T.; Bateman, K. S.; Cherney, M. M.; Das, A. K.; Luu,
H. A.; Holmes, C. F.; James, M. N. J. Biol. Chem. 2001, 276, 44078.
(24) Xing, Y.; Xu, Y.; Chen, Y.; Jeffrey, P. D.; Chao, Y.; Lin, Z.; Li, Z.;
Strack, S.; Stock, J. B.; Shi, Y. Cell 2006, 127, 341.
REFERENCES
■
(1) Miura, K.; Okada, Y.; Aoi, T.; Okada, A.; Takahashi, K.; Okita, K.;
Nakagawa, M.; Koyanagi, M.; Tanabe, K.; Ohnuki, M.; Ogawa, D.;
Ikeda, E.; Okano, H.; Yamanaka, S. Nat. Biotechnol. 2009, 27, 743.
(2) Ben-David, U.; Benvenisty, N. Nat. Rev. Cancer 2011, 11, 268.
(3) Tang, C.; Lee, A. S.; Volkmer, J. P.; Sahoo, D.; Nag, D.; Mosley,
A. R.; Inlay, M. A.; Ardehali, R.; Chavez, S. L.; Pera, R. R.; Behr, B.;
Wu, J. C.; Weissman, I. L.; Drukker, M. Nat. Biotechnol. 2011, 29, 829.
(4) Schriebl, K.; Satianegara, G.; Hwang, A.; Tan, H. L.; Fong, W. J.;
Yang, H. H.; Jungbauer, A.; Choo, A. Tissue Eng. Part A 2012, 18, 899.
(5) Choo, A. B.; Tan, H. L.; Ang, S. N.; Fong, W. J.; Chin, A.; Lo, J.;
Zheng, L.; Hentze, H.; Philp, R. J.; Oh, S. K.; Yap, M. Stem Cells 2008,
26, 1454.
(6) Ben-David, U.; Gan, Q. F.; Golan-Lev, T.; Arora, P.; Yanuka, O.;
Oren, Y. S.; Leikin-Frenkel, A.; Graf, M.; Garippa, R.; Boehringer, M.;
Gromo, G.; Benvenisty, N. Cell Stem Cell 2013, 12, 167.
(7) Tohyama, S.; Hattori, F.; Sano, M.; Hishiki, T.; Nagahata, Y.;
Matsuura, T.; Hashimoto, H.; Suzuki, T.; Yamashita, H.; Satoh, Y.;
Egashira, T.; Seki, T.; Muraoka, N.; Yamakawa, H.; Ohgino, Y.;
Tanaka, T.; Yoichi, M.; Yuasa, S.; Murata, M.; Suematsu, M.; Fukuda,
K. Cell Stem Cell 2013, 12, 127.
(8) Lee, M. O.; Moon, S. H.; Jeong, H. C.; Yi, J. Y.; Lee, T. H.; Shim,
S. H.; Rhee, Y. H.; Lee, S. H.; Oh, S. J.; Lee, M. Y.; Han, M. J.; Cho, Y.
S.; Chung, H. M.; Kim, K. S.; Cha, H. J. Proc. Natl. Acad. Sci. U.S.A.
2013, 110, E3281.
(9) Hirata, N.; Nakagawa, M.; Fujibayashi, Y.; Yamauchi, K.; Murata,
A.; Minami, I.; Tomioka, M.; Kondo, T.; Kuo, T.-F.; Endo, H.; Inoue,
H.; Sato, S.-i.; Ando, S.; Kawazoe, Y.; Aiba, K.; Nagata, K.; Kawase, E.;
Chang, Y.-T.; Suemori, H.; Eto, K.; Nakauchi, H.; Yamanaka, S.;
Nakatsuji, N.; Ueda, K.; Uesugi, M. Cell Rep. 2014, 6, 1165.
(10) Taguchi, Y.; Kino, K.; Morishima, M.; Komano, T.; Kane, S. E.;
Ueda, K. Biochemistry 1997, 36, 8883.
(11) Nagata, K.; Nishitani, M.; Matsuo, M.; Kioka, N.; Amachi, T.;
Ueda, K. J. Biol. Chem. 2000, 275, 17626.
(12) Nishiwaki, S.; Fujiki, H.; Suganuma, M.; Ojika, M.; Yamada, K.;
Sugimura, T. Biochem. Biophys. Res. Commun. 1990, 170, 1359.
(13) Yatsunami, J.; Fujiki, H.; Suganuma, M.; Nishiwaki, S.; Ojika,
M.; Yamada, K.; Levine, L. Toxicon 1991, 29, 1409.
(14) Gutmann, D. A.; Ward, A.; Urbatsch, I. L.; Chang, G.; van Veen,
H. W. Trends Biochem. Sci. 2010, 35, 36.
(15) Sharom, F. J. Pharmacogenomics 2008, 9, 105.
(16) Ueda, K.; Taguchi, Y.; Morishima, M. Semin. Cancer Biol. 1997,
8, 151.
(17) Slater, L. M.; Sweet, P.; Stupecky, M.; Gupta, S. J. Clin. Invest.
1986, 77, 1405.
D
dx.doi.org/10.1021/ja501795c | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX