A Synthetic Hybrid Molecule for the Selective Removal of Human Pluripotent Stem Cells from Cell Mixtures

Article abstract of DOI:10.1002/anie.201610284

A major hurdle in stem cell therapy is the tumorigenic risk of residual undifferentiated stem cells. This report describes the design and evaluation of synthetic hybrid molecules that efficiently reduce the number of human induced pluripotent stem cells (hiPSCs) in cell mixtures. The design takes advantage of Kyoto probe 1 (KP-1), a fluorescent chemical probe for hiPSCs, and clinically used anticancer drugs. Among the KP-1–drug conjugates we synthesized, we found an exceptionally selective, chemically tractable molecule that induced the death of hiPSCs. Mechanistic analysis suggested that the high selectivity originates from the synergistic combination of transporter-mediated efflux and the cytotoxicity mode of action. The present study offers a chemical and mechanistic rationale for designing selective, safe, and simple reagents for the preparation of non-tumorigenic clinical samples.

Full text of DOI:10.1002/anie.201610284

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201610284
German Edition: DOI: 10.1002/ange.201610284
Stem Cells
Hot Paper
A Synthetic Hybrid Molecule for the Selective Removal of Human
Pluripotent Stem Cells from Cell Mixtures
Di Mao, Shin Ando, Shin-ichi Sato, Ying Qin, Nao Hirata, Yousuke Katsuda, Eihachiro Kawase,
Ting-Fang Kuo, Itsunari Minami, Yuji Shiba, Kazumitsu Ueda, Norio Nakatsuji, and
Motonari Uesugi*
Abstract: A major hurdle in stem cell therapy is the
tumorigenic risk of residual undifferentiated stem cells. This
report describes the design and evaluation of synthetic hybrid
molecules that efficiently reduce the number of human induced
pluripotent stem cells (hiPSCs) in cell mixtures. The design
takes advantage of Kyoto probe 1 (KP-1), a fluorescent
chemical probe for hiPSCs, and clinically used anticancer
drugs. Among the KP-1–drug conjugates we synthesized, we
found an exceptionally selective, chemically tractable molecule
that induced the death of hiPSCs. Mechanistic analysis
suggested that the high selectivity originates from the syner-
gistic combination of transporter-mediated efflux and the
cytotoxicity mode of action. The present study offers a chemical
and mechanistic rationale for designing selective, safe, and
simple reagents for the preparation of non-tumorigenic clinical
samples.
undifferentiated cells are required for safe stem cell therapy.
Several strategies for their selective removal have been
reported, namely 1) cell sorting and precipitation, using an
antibody against a stem cell surface antigen,[2] 2) using
cytotoxic antibodies specific to human pluripotent stem
cells,[3] 3) modification of cell culture conditions,[4] and
4) using a lectin–toxin fusion protein that is specific to
human pluripotent stem cells.[5] Some chemical strategies
have also been shown to be effective, such as the use of small-
molecule inhibitors of stearoyl-CoA desaturase (SCD1)[6] and
survivin.[7] Despite these considerable and promising inves-
tigations, none of the elimination methods have reached the
stage of clinical application for regeneration therapy. The
development of new chemical tools with distinct elimination
mechanisms would complement other methods to ensure
complete removal of tumor-prone undifferentiated cells.
We previously described Kyoto probe 1(KP-1), a fluores-
cent molecule that selectively labels hiPSCs.[8] A mechanistic
study showed that its selectivity results primarily from
a distinct expression pattern of ATP-binding cassette (ABC)
transporters in human pluripotent stem cells, and from the
transporter selectivity of KP-1. Expression of two ABC
transporters, ABCB1 (MDR1) and ABCG2 (BCRP), both of
which cause the efflux of KP-1, is repressed in human
pluripotent stem cells.
Although KP-1 exhibits excellent selectivity for human
pluripotent stem cells, KP-1 is not capable of eliminating
them owing to its lack of cytotoxicity. In the present study,
KP-1 was covalently conjugated to a cytotoxic anticancer
drug. We envisioned that the structural diversity, various
mechanisms, long clinical history, and cost effectiveness of
cytotoxic anticancer drugs would lead to the discovery of
selective and relatively safe conjugates if they maintain both
the selectivity of KP-1 and the cytotoxicity of the drug.
Initially, we searched for cytotoxic anticancer drugs that
are substrates for both ABCB1 and ABCG2. A wide range of
cytotoxic anticancer drugs have been described as substrates
of ABC transporters.[9] However, previous studies reported
inconsistent results regarding their transporter selectivity.
Therefore, we first compared the transporter selectivity of
twelve cytotoxic anticancer drugs by analyzing their effects on
four cell lines: KB3-1 cells, which express no detectable levels
of ABC transporters, and KB3-1-derived cells (KB/ABCB1,
KB/ABCC1, and KB/ABCG2), which overexpress
ABCB1,[10] ABCC1,[11] and ABCG2,[8] respectively (see the
Supporting Information, Table S1). Most of the drugs exhib-
ited at least one ABC-transporter-mediated resistance.
Among them, etoposide was a substrate for both ABCB1
H
uman induced pluripotent stem cells (hiPSCs) serve as
a highly valuable resource for both basic research and
regeneration therapy. However, there are a number of
limitations for their clinical application, including the tumori-
genic risk of undifferentiated cells during transplantation.[1]
Complete differentiation or selective elimination of the
[*] D. Mao, Dr. S. Ando, Dr. S. Sato, Dr. Y. Qin, Dr. N. Hirata,
Dr. Y. Katsuda, Dr. T.-F. Kuo, Prof. M. Uesugi
Institute for Integrated Cell-Material Sciences (WPI-iCeMS) and
Institute for Chemical Research, Kyoto University
Uji, Kyoto 611-0011 (Japan)
E-mail: uesugi@scl.kyoto-u.ac.jp
Dr. E. Kawase
Institute for Frontier Medical Sciences
Kyoto University, Kyoto 606-8507 (Japan)
Dr. I. Minami
Institute for Integrated Cell-Material Sciences (WPI-iCeMS)
Kyoto University, Kyoto 606-8501 (Japan)
Dr. Y. Shiba
Institute for Biomedical Sciences and Department of Cardiovascular
Medicine, School of Medicine, Shinshu University
Matsumoto 390-8621 (Japan)
Prof. N. Nakatsuji
Institute for Frontier Medical Sciences and Institute for Integrated
Cell-Material Sciences (WPI-iCeMS)
Kyoto University, Kyoto 606-8507 (Japan)
Prof. K. Ueda
Division of Applied Life Sciences, Graduate School of Agriculture and
Institute for Integrated Cell-Material Sciences (WPI-iCeMS)
Kyoto University, Kyoto 606-8502 (Japan)
Supporting information for this article can be found under:
http://dx.doi.org/10.1002/anie.201610284.
Angew. Chem. Int. Ed. 2017, 56, 1 – 7
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

Angewandte
Communications
Chemie
and ABCC1, and SN38 and mitoxantrone were substrates for
both ABCC1 and ABCG2. However, no drug was found to be
a clear substrate for both ABCB1 and ABCG2. These results
are in agreement with our previous screening experience[8]
and a reported result,[12] both of which represented a technical
challenge in discovering a common substrate for both ABCB1
and ABCG2, even though ABCB1 and ABCC1 share
substrates with ABCC1 and ABCG2, respectively. These
results led us to the idea of conjugating KP-1 to cytotoxic
drugs, with the hope that the resulting conjugates would
remain substrates for both ABCB1 and ABCG2.
We carried out a structure–activity relationship study to
identify a potential conjugation site in KP-1. We first
synthesized 15 KP-1 derivatives with different substituents
on the benzene “head” and xanthone “arm” (Scheme S1), and
recorded their fluorescence spectra (lex and lem; Table S2).
We then evaluated their selectivity for ABC transporters by
analyzing the staining patterns of four cell lines (KB3-1, KB/
ABCB1, KB/ABCC1, and KB/ABCG2; Figure 1 and
Table S2). The KP-1 derivatives, except for molecules 9 and
11, exhibited explicit fluorescent staining of parental KB3-
1 cells, and little staining of KB/ABCB1 cells, indicating that
modification of the benzene head and xanthone arm did not
eliminate selectivity for ABCB1. The selectivity for ABCG2
was more sensitive to structural modification than that for
ABCB1, and was particularly affected by substituents on the
benzene head. Molecule 1 and molecule 2 failed to exhibit as
much ABCG2-mediated efflux as KP-1, indicating the
importance of the para substituent for the selectivity for
ABCG2. Replacement of F with Cl (molecule 3), Br (4), H
(5), propyl (6), or hydroxy (7) caused various degrees of loss
of selectivity for ABCG2. Modification of the amine arm had
a smaller effect on selectivity. Similarly to KP-1, molecules 12
and 14 showed good selectivity for both ABCB1 and ABCG2,
but poor selectivity for ABCC1. Molecule 14 was selected for
conjugation with cytotoxic anticancer drugs.
We initially conjugated molecule 14 to combretastatin A4
(CA4), a cytotoxic anticancer drug. This naturally occurring
tubulin inhibitor is effective with cell lines that highly express
ABCB1 and ABCG2, suggesting that CA4 is not a substrate
for either ABCB1 or ABCG2.[13] We envisioned that a hybrid
of molecule 14 and CA4 (conjugate 16) would maintain the
ABC transporter selectivity of KP-1. The structure of
conjugate 16 and its synthetic route are shown in Figure S1
and Scheme S2, respectively. The tubulin polymerization
inhibition of conjugate 16 was confirmed with a tubulin
polymerization assay (Figure S2).
To evaluate the ABC transporter selectivity of conjugate
16, we simultaneously carried out two cell-based assays,
namely a cell viability assay and a fluorescence imaging assay.
As expected, conjugate 16 was significantly more cytotoxic
than molecule 14, indicating that the toxicity of conjugate 16
comes from the CA4 moiety. Unfortunately, conjugate 16
failed to maintain its selectivity for ABCG2, although it
remained a substrate for ABCB1 (Figure S1B). The fluores-
cence images were consistent with the result of the cell
viability assay (Figure S1C). Conjugate 16 stained KB/
ABCG2 cells, but was excluded from KB/ABCB1 cells.
These results collectively indicate that the conjugate of
Figure 1. Fluorescence microscopy images of KB3-1, KB/ABCB1, KB/
ABCC1, and KB/ABCG2 cells treated with KP-1 or its derivatives 1–15
(1 mm). The structural modifications are indicated in red. Scale bar:
200 mm.
molecule 14 with CA4 maintained the selectivity for
ABCB1, but not for ABCG2.
2
www.angewandte.org
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 7
These are not the final page numbers!

Angewandte
Communications
Chemie
ABCG2 is generally thought to have a narrower profile of
scope. Similarly to KP-1, conjugates 17 and 18 fluorescently
efflux substrates than ABCB1.[12] We hypothesized that
replacement of CA4 with an ABCG2-selective cytotoxic
drug might result in a conjugate that is a substrate for both
ABCB1 and ABCG2. Several cytotoxic anticancer drugs have
been reported as ABCG2 substrates.[14] We first selected
SN38, an anticancer drug with long clinical history, and
produced conjugate 17 (Figure 2A and Scheme S2). In cell
labeled undifferentiated cells more strongly than differenti-
ated cells (Figure 3A). Their selectivities were further con-
firmed by flow cytometric analysis with five distinct types of
somatic cells: hepatocytes, prostate epithelial cells, brain
microvascular cells, adrenal microvascular cells, and astro-
cytes (Figure S4). The staining patterns of conjugates 17 and
18 paralleled that of KP-1 in that they labeled hiPSCs and
astrocytes more strongly than other somatic cells.[8] These
results indicate that conjugates 17 and 18 maintain the
transporter-mediated selectivity of KP-1.
Prolonged incubation with each conjugate efficiently
reduced the number of hiPSCs in cell mixtures. Partially
differentiated hiPSCs, or hiPSCs on SNL feeder cells, were
incubated with conjugate 17 or conjugate 18 (5 mm) for 72 h.
Undifferentiated hiPSCs were detected by colorimetric
staining for alkaline phosphatase (ALP) activity, a pluripo-
tency marker.[16] As expected, treatment with the conjugates
removed ALP-positive cells, but had little effect on ALP-
negative cells (Figure 3B). Time-course experiments indi-
cated that although strong cytotoxicities were observed for
the conjugates after 48 h, 72 h incubation was desirable for
efficient removal of the hiPSCs (Figure S5).
To confirm the elimination of undifferentiated cells, we
carried out quantitative polymerase chain reaction (qPCR)
assays to quantify the expression levels of three representa-
tive pluripotency markers[17] in cells treated with each
conjugate or a DMSO control. The relative expression
levels of nanog, sox2, and oct3/4 were normalized to gapdh,
a housekeeping gene. Gene expression levels in hiPSCs,
human astrocytes, and hepatocytes were used as controls. The
cell mixture treated with DMSO alone exhibited the highest
relative expression of nanog (0.7%), oct3/4 (6.7%), and sox2
(0.4%). Incubation with conjugate 17 or 18 decreased the
expression levels to 0.1% for nanog, 0.8% for oct3/4, and
0.03% for sox2 (Figure 3C–E), which is consistent with the
selective removal of pluripotent cells.
To quantitatively evaluate the selectivity, we conducted
flow cytometric analysis with an antibody against SSEA4,
a cell-surface marker for human pluripotent stem cells. The
cell mixture treated with DMSO alone exhibited 17.3%
SSEA4-positive cells. Treatment with conjugate 17 or 18
decreased the number of SSEA4-positive cells to 5.7% or
5.4%, respectively (Figure S6). These results collectively
support the notion that prolonged incubation with conjugate
17 or 18 efficiently reduces the number of hiPSCs in cell
mixtures.
Further studies were focused on conjugate 17, which was
slightly more selective than conjugate 18 (Table S2 and
Figure S4). The effects of conjugate 17 on five human somatic
primary cells and hiPSC-derived cardiomyocytes were com-
pared with the effects on two hiPSC clones, 201B7 and 253G1.
Hepatocytes, prostate epithelial cells, brain microvascular
cells, astrocytes, adrenal microvascular cells, and hiPSC-
derived cardiomyocytes were resistant to conjugate 17 (IC50 >
10 mm) whereas two hiPSC cell lines were highly sensitive
(201B7: IC50 = 0.2 mm; 253G1: IC50 = 0.4 mm; Figure 3F).
Addition of CsA (an ABCB1-selective inhibitor) or Ko143
(an ABCG2-selective inhibitor) increased the cytotoxicity of
Figure 2. Selectivity of conjugate 17. A) Chemical structure of conju-
gate 17. B) IC50 values of SN38 and conjugate 17 for KB3-1, KB/
ABCB1, KB/ABCC1, and KB/ABCG2 cells. The levels of resistance are
highlighted in blue. Data are given as meanÆSD. n=3. C) Fluores-
cence microscopy images of KB3-1, KB/ABCB1, KB/ABCC1, and KB/
ABCG2 cells treated with conjugate 17 (1 mm) in the presence or
absence of CsA (10 mm) or Ko143 (1 mm). Scale bar: 200 mm.
viability assays, SN38 displayed resistance in KB/ABCG2
cells, as previously reported, but potent cytotoxicity in KB/
ABCB1 and KB/ABCC1 cells (Table S1). Although conju-
gate 17 was slightly less cytotoxic than SN38, it showed
resistance in both KB/ABCB1 and KB/ABCG2 cells (Fig-
ure 2B).
As with KP-1, the fluorescence intensity of conjugate 17
was significantly lower in KB/ABCB1 and KB/ABCG2 cells
than in KB3-1 cells. The fluorescence was restored by
treatment with selective inhibitors of ABCB1 (cyclospori-
ne A, 10 mm) or ABCG2 (Ko143, 1 mm;[15] Figure 2C). These
results indicate that conjugate 17 is a good cytotoxic substrate
for both ABCB1 and ABCG2.
To further validate our approach, we conjugated molecule
14 to mitoxantrone, another ABCG2-selective anticancer
drug, to produce conjugate 18 (Figure S3A and Scheme S2).
Cell viability and fluorescence imaging analysis indicated that
mitoxantrone had good selectivity for ABCC1 and ABCG2,
but poor selectivity for ABCB1 (Table S1). In contrast,
conjugate 18 was a substrate for both ABCB1 and ABCG2,
but not for ABCC1 (Figure S3B,C).
We next evaluated the ability of conjugates 17 and 18 to
label and eliminate hiPSCs from cell mixtures. When hiPSCs
are overgrown, the cells form donut-shaped colonies. The
central parts of the colonies represent differentiated cells.[8]
hiPSC colonies were treated with conjugates 17 (1 mm) or 18
(10 mm) and immediately observed under a confocal micro-
Angew. Chem. Int. Ed. 2017, 56, 1 – 7
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

Angewandte
Communications
Chemie
Figure 3. Selective labeling and removal of hiPSCs by conjugate 17 or
18. A) Selective labeling of partially differentiated hiPSC colonies by
KP-1 (1 mm), conjugate 17 (1 mm), or conjugate 18 (10 mm). The higher
concentration of conjugate 18 was used owing to its low fluorescence
intensity. The central parts of the colonies represent differentiated
cells. Scale bar: 200 mm. B) Effect of a DMSO control (0.1%),
conjugate 17 (5 mm), and 18 (5 mm) on the alkaline phosphatase
activity of iPS colonies and partially differentiated iPSCs. The cells were
treated with the conjugates for 72 h, followed by colorimetric alkaline
phosphatase staining (blue). C–E) qPCR analysis of nanog (C), oct3/4
(D), and sox2 (E) in partially differentiated iPSC cells treated with
DMSO, conjugate 17, or 18. The partially differentiated hiPSCs were
prepared by incubation for 72 h with all-trans retinoic acid (0.5 mm)
followed by treatment with the conjugate (5 mm) for 72 h. The gene
expression levels in hiPSCs, human astrocytes, and hepatocytes are
shown as controls. The expression levels of nanog, sox2, and oct3/4
were normalized to gapdh. Data are shown as meanÆSD. **p<0.01,
n=3. F) Viabilities of hiPSCs (201B7 and 253G1), human primary
cells, and hiPSC-derived cardiomyocytes after 72 h treatment with
conjugate 17. Data are shown as meanÆSD. *p<0.05, **p<0.01,
&
&
*
n=3. iPS201B7 ( ), iPS253G1 ( ), adrenal microvascular cells ( ),
*
^
astrocytes ( ), brain microvascular cells ( ), prostate epithelial cells
!
~
*
(
), hepatocytes ( ), and hiPSC-derived cardiomyocytes ( ).
G) Colony formation assay of partially differentiated hiPSCs with 72 h
treatment of DMSO (0.1%) or conjugate 17 (1 mm). Undifferentiated
cells were immunostained with an anti-Oct4 antibody, and visualized
by a horseradish peroxidase conjugated immunoglobulin polyclonal
antibody and 3,3’-diaminobenzidine (DAB). The whole cell populations
were visualized by crystal violet.
The intrinsic selectivity of SN38 might originate from its
mode of action.[18] SN38 exerts anti-proliferation activity by
inhibiting topoisomerase I (Topo I).[19] High proliferation
rates are important for the pluripotency and self-renewal of
PSCs.[20] In contrast, differentiated cells usually exhibit low
proliferation rates, possibly providing some level of resistance
to Topo I inhibitors. To examine whether conjugate 17 inhibits
Topo I, we carried out in vitro biochemical assays with
recombinant human topoisomerase I (Figure S9). Conjugate
17 inhibited Topo I even more strongly than SN38 did. This
seems somewhat surprising because conjugate 17 is less
cytotoxic than SN38. However, the in vitro potency of Topo I
inhibition does not completely correlate with the cytotoxicity,
which is influenced by many other factors, including mem-
brane permeability.[21]
The high selectivity of conjugate 17 appears to stem from
a synergy between its ABCB1/G2-mediated efflux and Topo I
inhibition. In fact, the synergy makes conjugate 17 more
selective than the okadaic acid derivative that we previously
reported as a cytotoxic substrate for both ABCB1 and
ABCG2 (Figure S10).[22] Clinical application of this marine
natural product derivative was hampered by its moderate
selectivity, limited supply, and low chemical tractability.[23]
The high selectivity and chemical tractability of conjugate 17
makes it more attractive for further development than the
okadaic acid derivative.
conjugate 17 to the resistant somatic cells (Figure S7),
suggesting involvement of ABCB1 and ABCG2 in the
resistance.
It was surprising that astrocytes, which lack in ABCB1 and
ABCG2 expression, showed resistance to conjugate 17
although the resistance was weaker than for other somatic
cells. Moreover, addition of transporter inhibitors to the
medium restored the cytotoxicity of conjugate 17 for human
primary somatic cells, but not completely. Hepatocytes,
prostate epithelial cells, and brain microvascular cells, which
are known to express high levels of ABC transporters,[8]
retained resistance, even in the presence of CsA or Ko143
(Figure S7). These two observations suggest that the conju-
gate has a selective cytotoxicity mechanism other than ABC-
transporter-mediated efflux. To examine whether SN38 itself
is selective for hiPSCs, we compared the effect of SN38 alone
with the effect of conjugate 17 on various human primary cells
and hiPSCs. Interestingly, SN38 exhibited moderate selectiv-
ity for hiPSCs; however, it was not as selective as conjugate 17
(Figure S8). These results suggest that although SN38 is
intrinsically somewhat selective for hiPSCs, conjugation with
KP-1 enhances that selectivity.
To further validate the utility of conjugate 17, we carried
out colony formation assays. Partially differentiated hiPSCs
were treated with conjugate 17 (1 mm) or DMSO (0.1%) for
72 h. The residual cells were then washed and cultured in
a growth medium for another seven days, followed by fixation
and immunostaining with an anti-Oct4 antibody. Whereas
4
www.angewandte.org
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 7
These are not the final page numbers!

Angewandte
Communications
Chemie
DMSO-treated cells formed Oct4-positive cell colonies, cells
tractable reagents for the preparation of non-tumorigenic
treated with conjugate 17 were Oct4-negative and failed to
form colonies (Figure 3G), indicating that 72 h treatment
with conjugate 17 (1 mm) eliminated colony-forming undiffer-
entiated cells. This complete removal was somewhat surpris-
ing, considering that the FACS analysis in Figure S6 showed
a leftover (5.7%) of hiPSCs after three-day treatment with
conjugate 17 of the partially differentiated hiPSC population.
To investigate the proliferation rate of the leftover iPSCs, we
followed their viability over seven days after removal of
conjugate 17 (Figure S11). The result indicated that the
residual iPSCs failed to proliferate, displaying delayed cell
death.
We next examined whether the cells treated with con-
jugate 17 were not only alive but also healthy. As mitochon-
dria play an important role in the function of cardiomyocytes
that require high levels of energy production,[24] we evaluated
the effect of conjugate 17 on the mitochondria membrane
potentials of hiPSC-derived cardiomyocytes. Three-day treat-
ment with 5 mm of conjugate 17 had no apparent effect on the
mitochondria membrane potentials of the hiPSC-derived
cardiomyocytes (Figure S12). The effect of conjugate 17 on
hepatocytes was also investigated in terms of their prolifer-
ation and transportation of indocyanine green (ICG), a cya-
nine dye used for measuring hepatic function. Hepatocytes
treated with conjugate 17 (5 mm for 3 days) maintained
a proliferative rate comparable to that of non-treated cells
(Figure S13A), and their capability of cellular uptake and
excretion of ICG was unaffected (Figure S13B). These results
suggest that the three-day treatment with 5 mm of conjugate
17 is within an acceptable range for maintaining the functions
of differentiated cells. However, further studies are needed to
ensure the safety of conjugate 17.
clinical cell samples.
Acknowledgements
We thank Dr. Masato Nakagawa (Center for iPS Cell
Research and Application, Kyoto University) for providing
SNL feeder cells. This research is supported by the project
“Development of Cell Production and Processing Systems for
Commercialization of Regenerative Medicine” from the
Japan Agency for Medical Research and Development
(AMED). This work is also supported in part by the JSPS
(26220206 to M.U. and 25221203 to K.U.). iCeMS is supported
by the World Premier International Research Center Initia-
tive (WPI), MEXT, Japan. This work was inspired by the
international and interdisciplinary environment 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.
Conflict of interest
The authors declare no conflict of interest.
Keywords: ABC transporters · cytotoxicity · drug conjugates ·
fluorescent probes · stem cell therapy
Once tumor-prone undifferentiated cells have been elim-
inated by conjugate 17, the conjugate needs to be cleared
from cell samples before transplantation. To monitor residual
amounts of conjugate 17, we took advantage of its fluores-
cence (Figure S14). A culture medium containing 5 mm of
conjugate 17 showed strong fluorescence emission at 548 nm.
After simply washing with a fresh medium, the fluorescence
intensity decreased significantly to approximately 2%.
Another brief wash further reduced the fluorescence to the
background level (medium alone). The fluorescence of
conjugate 17 or other KP-1–drug conjugates might provide
an advantage in future applications by ensuring the clearance
of toxic reagents in resulting cell samples.
[1] a) K. Miura, Y. Okada, T. Aoi, A. Okada, K. Takahashi, K.
Okita, M. Nakagawa, M. Koyanagi, K. Tanabe, M. Ohnuki, D.
Ogawa, E. Ikeda, H. Okano, S. Yamanaka, Nat. Biotechnol.
2009, 27, 743 – 745; b) U. Ben-David, N. Benvenisty, Nat. Rev.
Cancer 2011, 11, 268 – 277.
[2] a) C. Tang, A. S. Lee, J. P. Volkmer, D. Sahoo, D. Nag, A. R.
Mosley, M. A. Inlay, R. Ardehali, S. L. Chavez, R. R. Pera, B.
Behr, J. C. Wu, I. L. Weissman, M. Drukker, Nat. Biotechnol.
2011, 29, 829 – 834; b) K. Schriebl, G. Satianegara, A. Hwang,
H. L. Tan, W. J. Fong, H. H. Yang, A. Jungbauer, A. Choo, Tissue
Eng. Part A 2012, 18, 899 – 909.
[3] A. B. Choo, H. L. Tan, S. N. Ang, W. J. Fong, A. Chin, J. Lo, L.
Zheng, H. Hentze, R. J. Philp, S. K. Oh, M. Yap, Stem Cells 2008,
26, 1454 – 1463.
In conclusion, we have described fluorescent KP-1–drug
conjugates that efficiently reduce the number of hiPSCs in
cell mixtures. Although it remains unknown why ABCB1 and
ABCG2 share few substrates, the present study offers an
approach to designing cytotoxic substrates for both ABCB1
and ABCG2. The KP-1 conjugate with SN38 (conjugate 17)
exhibited exceptionally high selectivity for hiPSCs. The high
selectivity appears to stem from a synergy between its
ABCB1/G2-mediated efflux and Topo I inhibition. The high
selectivity and chemical tractability of conjugate 17 make it an
attractive molecule for further development. This work
indicates that conjugation of KP-1 with an anticancer drug
is a new approach for developing highly selective, chemically
[4] a) S. Tohyama, F. Hattori, M. Sano, T. Hishiki, Y. Nagahata, T.
Matsuura, H. Hashimoto, T. Suzuki, H. Yamashita, Y. Satoh, T.
Egashira, T. Seki, N. Muraoka, H. Yamakawa, Y. Ohgino, T.
Tanaka, M. Yoichi, S. Yuasa, M. Murata, M. Suematsu, K.
Fukuda, Cell Stem Cell 2013, 12, 127 – 137; b) K. Matsuura, F.
Kodama, K. Sugiyama, T. Shimizu, N. Hagiwara, T. Okano,
Tissue Eng. Part C 2015, 21, 330 – 338.
[5] H. Tateno, Y. Onuma, Y. Ito, F. Minoshima, S. Saito, M. Shimizu,
Y. Aiki, M. Asashima, J. Hirabayashi, Stem Cell Rep. 2015, 4,
811 – 820.
[6] U. Ben-David, Q. F. Gan, T. Golan-Lev, P. Arora, O. Yanuka,
Y. S. Oren, A. Leikin-Frenkel, M. Graf, R. Garippa, M.
Boehringer, G. Gromo, N. Benvenisty, Cell Stem Cell 2013, 12,
167 – 179.
Angew. Chem. Int. Ed. 2017, 56, 1 – 7
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

Angewandte
Communications
Chemie
[7] M. O. Lee, S. H. Moon, H. C. Jeong, J. Y. Yi, T. H. Lee, S. H.
Shim, Y. H. Rhee, S. H. Lee, S. J. Oh, M. Y. Lee, M. J. Han, Y. S.
Cho, H. M. Chung, K. S. Kim, H. J. Cha, Proc. Natl. Acad. Sci.
USA 2013, 110, E3281 – E3290.
[8] N. Hirata, M. Nakagawa, Y. Fujibayashi, K. Yamauchi, A.
Murata, I. Minami, M. Tomioka, T. Kondo, T. F. Kuo, H. Endo,
H. Inoue, S. Sato, S. Ando, Y. Kawazoe, K. Aiba, K. Nagata, E.
Kawase, Y. T. Chang, H. Suemori, K. Eto, H. Nakauchi, S.
Yamanaka, N. Nakatsuji, K. Ueda, M. Uesugi, Cell Rep. 2014, 6,
1165 – 1174.
[17] a) J. Cai, J. Chen, Y. Liu, T. Miura, Y. Luo, J. F. Loring, W. J.
Freed, M. S. Rao, X. Zeng, Stem Cells 2006, 24, 516 – 530; b) M.
Stadtfeld, K. Hochedlinger, Genes Dev. 2010, 24, 2239 – 2263.
[18] a) C. P. Garcꢁa, G. A. Videla Richardson, L. Romorini, S. G.
Miriuka, G. E. Sevlever, M. E. Scassa, Stem Cell Res. 2014, 12,
400 – 414; b) O. Momcilovic, L. Knobloch, J. Fornsaglio, S.
Varum, C. Easley, G. Schatten, PLoS One 2010, 5, e13410.
[19] B. L. Staker, K. Hjerrild, M. D. Feese, C. A. Behnke, A. B.
Burgin, Jr., L. Stewart, Proc. Natl. Acad. Sci. USA 2002, 99,
15387 – 15392.
[9] a) K. Ueda, C. Cardarelli, M. M. Gottesman, I. Pastan, Proc.
Natl. Acad. Sci. USA 1987, 84, 3004 – 3008; b) S. P. C. Cole, G.
Bhardwaj, J. H. Gerlach, J. E. Mackie, C. E. Grant, K. C.
Almquist, A. J. Stewart, E. U. Kurz, A. M. V. Duncan, R. G.
Deeley, Science 1992, 258, 1650 – 1654; c) J. I. Fletcher, M.
Haber, M. J. Henderson, M. D. Norris, Nat. Rev. Cancer 2010, 10,
147 – 156.
[20] S. Ruiz, A. D. Panopoulos, A. Herrerias, K. D. Bissig, M. Lutz,
W. T. Berggren, I. M. Verma, J. C. Izpisua Belmonte, Curr. Biol.
2011, 21, 45 – 52.
[21] a) A. Chevalier, M. Dubois, V. Le Joncour, S. Dautrey, C.
Lecointre, A. Romieu, P. Y. Renard, H. Castel, C. Sabot,
Bioconjugate Chem. 2013, 24, 1119 – 1133; b) D. E. Uehling
et al., J. Med. Chem. 1995, 38, 1106 – 1118.
[10] Y. Taguchi, K. Kino, M. Morishima, T. Komano, S. E. Kane, K.
Ueda, Biochemistry 1997, 36, 8883 – 8889.
[11] K. Nagata, M. Nishitani, M. Matsuo, N. Kioka, T. Amachi, K.
Ueda, J. Biol. Chem. 2000, 275, 17626 – 17630.
[12] J. J. Strouse, I. Ivnitski-Steele, A. Waller, S. M. Young, D. Perez,
A. M. Evangelisti, O. Ursu, C. G. Bologa, M. B. Carter, V. M.
Salas, G. Tegos, R. S. Larson, T. I. Oprea, B. S. Edwards, L. A.
Sklar, Anal. Biochem. 2013, 437, 77 – 87.
[13] L. M. Greene, S. M. Nathwani, S. A. Bright, D. Fayne, A. Croke,
M. Gagliardi, A. M. McElligott, L. OꢀConnor, M. Carr, N. O.
Keely, N. M. OꢀBoyle, P. Carroll, B. Sarkadi, E. Conneally, D. G.
Lloyd, M. Lawler, M. J. Meegan, D. M. Zisterer, J. Pharmacol.
Exp. Ther. 2010, 335, 302 – 313.
[22] T. F. Kuo, D. Mao, N. Hirata, B. Khambu, Y. Kimura, E. Kawase,
H. Shimogawa, M. Ojika, N. Nakatsuji, K. Ueda, M. Uesugi, J.
Am. Chem. Soc. 2014, 136, 9798 – 9801.
[23] a) A. B. Dounay, R. A. Urbanek, S. F. Sabes, C. J. Forsyth,
Angew. Chem. Int. Ed. 1999, 38, 2258 – 2262; Angew. Chem.
1999, 111, 2403 – 2406; b) C. J. Forsyth, S. F. Sabes, R. A.
Urbanek, J. Am. Chem. Soc. 1997, 119, 8381 – 8382.
[24] a) G. Santulli, W. Xie, S. R. Reiken, A. R. Marks, Proc. Natl.
Acad. Sci. USA 2015, 112, 11389 – 11394; b) A. Mathur, Y. Hong,
B. K. Kemp, A. A. Barrientos, J. D. Erusalimsky, Cardiovasc.
Res. 2000, 46, 126 – 138; c) V. Lukyanenko, A. Chikando, W. J.
Lederer, Int. J. Biochem. Cell Biol. 2009, 41, 1957 – 1971.
[14] L. Austin Doyle, D. D. Ross, Oncogene 2003, 22, 7340 – 7358.
[15] L. D. Weidner, S. S. Zoghbi, S. Lu, S. Shukla, S. V. Ambudkar,
V. W. Pike, J. Mulder, M. M. Gottesman, R. B. Innis, M. D. Hall,
J. Pharmacol. Exp. Ther. 2015, 354, 384 – 393.
[16] M. D. OꢀConnor, M. D. Kardel, I. Iosfina, D. Youssef, M. Lu,
M. M. Li, S. Vercauteren, A. Nagy, C. J. Eaves, Stem Cells 2008,
26, 1109 – 1116.
Manuscript received: October 20, 2016
Revised: November 29, 2016
Final Article published: && &&, &&&&
6
www.angewandte.org
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 7
These are not the final page numbers!

Angewandte
Communications
Chemie
Communications
Stem Cells
Wiping out stem cells: A major obstacle
in stem cell therapy is the tumorigenic
risk of residual undifferentiated stem
cells. Fluorescent hybrid molecules were
synthesized for the selective removal of
human induced pluripotent stem cells
from cell mixtures. The high selectivity
results from the synergy between the
transporter-mediated efflux and the cyto-
toxicity mode of action.
D. Mao, S. Ando, S. Sato, Y. Qin,
N. Hirata, Y. Katsuda, E. Kawase,
T.-F. Kuo, I. Minami, Y. Shiba, K. Ueda,
N. Nakatsuji, M. Uesugi*
&&& — &&&
A Synthetic Hybrid Molecule for the
Selective Removal of Human Pluripotent
Stem Cells from Cell Mixtures
Angew. Chem. Int. Ed. 2017, 56, 1 – 7
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7
These are not the final page numbers!