Identification of a selective small molecule inhibitor of breast cancer stem cells

Article abstract of DOI:10.1016/j.bmcl.2012.01.035

A high-throughput screen (HTS) with the National Institute of Health-Molecular Libraries Small Molecule Repository (NIH-MLSMR) compound collection identified a class of acyl hydrazones to be selectively lethal to breast cancer stem cell (CSC) enriched pop

Full text of DOI:10.1016/j.bmcl.2012.01.035

Bioorganic & Medicinal Chemistry Letters 22 (2012) 3571–3574
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Identification of a selective small molecule inhibitor of breast cancer stem cells
Andrew R. Germain a, Leigh C. Carmody a, Barbara Morgan a, Cristina Fernandez a, Erin Forbeck a,
Timothy A. Lewis a, Partha P. Nag a, Amal Ting a, Lynn VerPlank a, Yuxiong Feng b, Jose R. Perez a,
Sivaraman Dandapani a, Michelle Palmer a, Eric S. Lander a, Piyush B. Gupta b,c, Stuart L. Schreiber a,d
,
Benito Munoz a,
⇑
a Chemical Biology Platform and Probe Development Center, Broad Institute of MIT and Harvard, 7 Cambridge Center, Cambridge, MA 02142, USA
b Whitehead Institute for Biomedical Research, Cambridge, MA 02142, USA
c Department of Biology, Massachusetts Institute of Technology, Cambridge, MA 02142, USA
d Howard Hughes Medical Institute, Broad Institute of Harvard and MIT, 7 Cambridge Center, Cambridge, MA 02142, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A high-throughput screen (HTS) with the National Institute of Health–Molecular Libraries Small Molecule
Repository (NIH–MLSMR) compound collection identified a class of acyl hydrazones to be selectively
lethal to breast cancer stem cell (CSC) enriched populations. Medicinal chemistry efforts were undertaken
to optimize potency and selectivity of this class of compounds. The optimized compound was declared as
a probe (ML239) with the NIH Molecular Libraries Program and displayed greater than 20-fold selective
inhibition of the breast CSC-like cell line (HMLE_sh_Ecad) over the isogenic control line (HMLE_sh_GFP).
Ó 2012 Elsevier Ltd. All rights reserved.
Received 16 November 2011
Revised 9 January 2012
Accepted 11 January 2012
Available online 25 January 2012
Keywords:
Cancer stem cells
Breast cancer
Molecular libraries probe production center
Network (MLPCN)
Cancer stem cells (CSCs), which drive tumor growth, are known
to be resistant to standard chemotherapy and radiation treat-
ments.1 In addition, CSCs within tumors are responsible for recur-
rence, the primary cause of patient mortality. Hence finding small
molecules that target CSCs can lead to new therapies.
in order to screen large chemical libraries to identify novel poten-
tial therapeutics, it is imperative that new HTS strategies be devel-
oped. Herein, we describe our efforts which utilize a novel recently
reported HTS approach which was validated in a proof of principle
pilot screen of 15,000 compounds.6 Here, we report on the results
of our screening campaign of >300,000 compounds from the NIH–
MLSMR library against human mammary epithelial cells (HMLE)
that have been forced to undergo an epithelial to mesenchymal
transition (EMT) to produce a stable stem-like phenotype. This
screen has led to the identification of a class of compounds that
may selectively kill breast CSCs. The chemical compounds identi-
fied by the experiments in this report may serve as useful probes
to study CSC biology, which are currently lacking.
A HTS of >300,000 compounds from the NIH–MLSMR collection
assessing the viability of the breast CSC-like population
(HMLE_sh_Ecad) was completed.6 HMLE_sh_Ecad cells have
undergone an EMT by introduction of a short hairpin RNA (shRNA)
targeting the CDH1 gene, which encodes E-cadherin into human
mammary epithelial cells (HMLEs). These cells provide a uniform
population of stable breast stem-like cells that can be used to con-
duct a large-scale screen.6 A total of 3190 compounds showed
greater than 75% inhibition of HMLE_sh_Ecad cells. The list of
active hits was narrowed by discarding compounds that hit in
>10% of MLPCN assays (presumed to be promiscuous hitters) or
scored positive in mammalian cell toxicity screens.
In principle, it would be desirable to apply high-throughput
screening (HTS) technologies to facilitate the identification of
chemical compounds that specifically target CSCs. However, CSCs
generally comprise only a small minority within cancer cell popu-
lations and standard high-throughput cell viability assays applied
to bulk populations of cancer cells cannot identify agents with
CSC-specific toxicity.2 Accordingly, screening for agents that pref-
erentially kill CSCs depends, at present, on the ability to obtain
and to stably propagate highly pure CSC populations in vitro.3
However, since CSCs cannot currently be stably propagated in
culture, it has not been possible to leverage HTS to identify CSC-
targeting compounds. Several compounds have been identified,
through moderate throughput screens, that target cancer stem cell
pathways responsible for self-renewal and survival4 including
compounds that inhibit tumorsphere formation in vitro and
tumors in vivo.5 While these approaches have been fruitful in
providing chemical starting points for medicinal chemistry efforts,
⇑
Corresponding author. Tel.: +1 617 714 7725.
E-mail addresses: bmunoz@broadinstitute.org, benmz101@yahoo.com (B. Munoz).
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2012.01.035

3572
A. R. Germain et al. / Bioorg. Med. Chem. Lett. 22 (2012) 3571–3574
Table 2
O
CH3
O
H3C
H3C
SAR of phenyl group for breast CSC inhibition
O
N
O
O
N
N
H
N
H
N
H
N
H
O
2
R
Cl
Cl
O
N
1
2
N
H
N
H
7
4
6
Figure 1. Prioritized scaffold identified from HTS screen.
a
Analog
R
EC50
(lM)
HMLE_shEcad HMLE_Shcontrol Selectivity
O
O
MgSO4
EtOH
60°C
Inactiveb
Inactive
3.33 versus
inactive
33.5 versus
inactive
NA
O
NH2
R2
+
N
R2
1
2-Cl, 4-Cl
2-Cl
3.33
33.5
R1
N
H
R1
N
H
3
4
5
7a
Scheme 1. Synthesis of analogs.
7b
7c
7d
2-CN
2-OMe
2-Me
Inactive
Inactive
40.7
Inactive
Inactive
Inactive
NA
40.7 versus
inactive
22.5 versus
inactive
126 versus
inactive
9.2
Table 1
7e
7f
4-Cl
22.5
126
2.89
Inactive
Inactive
26.7
SAR of the pyrrole for breast CSC inhibition
2-Cl, 6-Cl
O
O
7g
2-Me, 4-
Cl
R
Cl
Cl
6
7h
7i
2-Br, 4-Cl 3.16
20.9
6.61
2-Me, 4-
Br
2,4,6-Cl
2,4,6-Me
3.76
Inactive
3.76 versus
inactive
23.8
9.42 versus
inactive
a
Analog
R
EC50
(
l
M)
7j
7k
1.16
9.42
27.6
Inactive
HMLE_shEcad HMLE_Shcontrol Selectivity
3.33
versus
inactive
N
N
N
N
H
1
3.33
Inactiveb
16.93
N
H
a
Average of at least three runs.
No inhibition at highest dose (20
b
l
M).
Me
N
H
6a
6b
14.35
Inactive
1.18
NA
N
H
Dichlorophenyl hydrazone 1 was found to have an IC50 of
M and was inactive against the control cell line. Dioxane
M)
M against
N
Inactive
N
H
Me
NH
3.3
l
hydrazone 2 was found to have similar potency (IC50 of 2.51
l
N
N
6c
Inactive
Inactive
inactive
Inactive
NA
NA
N
but showed lower selectivity with an IC50 of 14.95
l
N
H
the control cell line. Since both hit compounds contained the same
pyrrole portion on the eastern side of the molecule it appeared that
selectivity could be controlled by varying the western portion of
the molecule.
NH
6d
N
N
H
The acyl hydrazones 3 could be easily accessed in one step
through condensation of commercially available hydrazides 4
and aldehydes 5 as shown in Scheme 1.
Screening results for the series are described in the tables
below.
6e
N
N
N
125
23.5
0.19
N
H
N
H
6f
Inactive
Inactive
inactive
Inactive
NA
NA
O
S
N
H
As can be seen in Table 1, attempts to replace the pyrrole on the
eastern portion of the molecule, was not tolerated. Introduction of
a methyl group onto the pyrrole (6a) results in a significant drop in
potency. N-methylation of the pyrrole (6b) completely eliminated
the activity of the molecule. Replacement of the pyrrole with other
aromatic groups (6c–6h) resulted in at least a greater than 10-fold
decrease in potency. Ketone derived hydrazone (6k) was inactive.
Overall, the SAR for the eastern part of the molecule suggested
the pyrrole was optimal.
Next, we evaluated the western part of the molecule and fo-
cused our efforts on a series of substitutions on the phenyl ring.
The results are summarized in Table 2. Halogen substitution on
the phenyl ring is favored over other substituents and multiple
substitutions improve potency. Substitution at the 4-position
seems to be vital to the activity of the molecule 2,4- and 2,5-substi-
tuted compounds resulted in more potent compounds than other
substitution patterns. Substitutions at the 2-, 4-, and 6-positions
resulted in the most potent compound that retains selectivity.
Table 3 illustrates the SAR of the linker between the hydrazone
and the aromatic group. Conversion of the ether linker to a thioe-
ther (compare 1 and 8a) resulted in a slight increase in potency
but a loss of selectivity. Conversion to a carbon linker (compare
8b and 8c), resulted in a slight decrease in potency, while
6g
N
H
N
N
N
N
6h
6i
Inactive
Inactive
79.55
Inactive
Inactive
56.30
NA
N
H
NA
N
H
N
6j
0.71
NA
N
H
OH
N
H
6k
N
H
Inactive
Inactive
Me
a
Average of at least three runs.
No inhibition at highest dose (20
b
lM).
The remaining 2244 compounds were retested in dose in the
primary assay using the HMLE_sh_Ecad line and in a counterscreen
against the HMLE_sh_eGFP line. The latter cell line has not been in-
duced to go through an EMT and provide a mammary epithelial cell
control. Two acyl hydrazones (1 and 2), were found to be both po-
tent and selective, and were prioritized for further optimization
(Fig. 1).

A. R. Germain et al. / Bioorg. Med. Chem. Lett. 22 (2012) 3571–3574
3573
Table 3
Table 4
SAR of the linker for breast CSC inhibition
SAR of bicyclic compounds for breast CSC inhibition
O
N
O
N
N
H
N
H
R
N
H
R
N
H
8
8
a
a
Analog
R
EC50
(
lM)
Analog
R
EC50
(lM)
HMLE_shEcad HMLE_Shcontrol Selectivity
HMLE_shEcad HMLE_
Selectivity
5.96
shcontrol
14.95
3.33
versus
inactive
O
1
3.33
2.30
36.2
53.4
40.7
Inactiveb
18.0
O
O
Cl
Cl
Cl
S
2
2.51
8a
8b
8c
8d
7.83
0.58
Br
Cl
O
O
8m
8n
8o
8p
1.11
3.03
2.73
4.49
O
O
21.0
O
O
4.21
18.9
53.4
O
O
Inactive
Inactive
versus
inactive
40.7
versus
inactive
O
6.32 versus
inactive
6.32
Inactiveb
Inactive
O
Me
O
Inactive
NA
H
N
8e
Inactive
Inactive
NA
a
Average of at least three runs.
No inhibition at highest dose (20
Me
b
l
M).
3.39
versus
inactive
Cl
Cl
O
8f
8g
8h
8i
3.39
33.4
8.86
35.4
12.9
0.79
59.1
inactive
97.7
Cl
O
2.92
3.79
Me
The solubility of 7j was experimentally determined to be
M in phosphate buffered saline (PBS, pH 7.4, 23 °C) solution.
Cl
6.3
l
O
Plasma protein binding (PPB) studies showed that 7j was 98.8%
bound to human plasma. The probe is modestly stable in human
plasma, with approximately 49% remaining after a 5-hour incuba-
tion period. The probe compound was found to be stable in gluta-
thione (GSH) with 85% remaining after 48 h. The stability of 7j in
PBS (0.1% DMSO) was measured over 48 h. From these results,
the probe seems to be stable in PBS since more than 76% is still
present after 48 h of incubation.
The current work evaluated over 300,000 compounds of the
NIH–MLSMR collection to identify agents capable of selectively
inhibiting breast cancer stem cells. After triage through a series
of counterscreens and an orthogonal assay, hydrazone 1 was se-
lected for further SAR investigation and optimization efforts. As a
result of these synthetic studies, compound 7j emerged as the most
potent and selective inhibitor, able to achieve inhibition CSCs with
33.6
F
F
Cl
O
35.4
Inactive
Inactive
2.14
versus
inactive
12.9
versus
inactive
Me
Cl
8j
8k
8l
2.71
t-Bu
59.1
versus
inactive
Inactive
Cl
Cl
a
Average of at least three runs.
No inhibition at highest dose (20
b
lM).
an IC50 of 1.16
l
M with ꢀ24-fold selectivity against the control cell
line. This compound has been registered with NIH Molecular
Libraries Program (probe ML239) and is available upon request.
Studies are currently underway to determine the ability of this
compound to inhibit other breast cancer stem cell-like cell lines
along with tumor cell lines with naturally high populations of can-
cer stems cells and will be reported in due course.
conversion to an amine linker (compare 8d and 8e) resulted in a
complete loss of activity. Substitution at the -position to the ether
a
(compare 8f and 8g; 8h and 8i) yielded a significant decrease in po-
tency. A cyclopropyl linker (8j) resulted in moderate loss of po-
tency; however, substitution at the 4-position with a t-butyl
group at the 4-position 8k resulted in a potent compound but with
poor selectivity. The 2,4-dichloro phenyl containing cyclopropyl
compound 8l, however was only weakly active possibly due to a
change in conformation from steric hinderance.
Acknowledgments
The authors are grateful to Ms. Patti Aha and Ms. Lori Thomae
for their assistance in manuscript preparation. This work was
funded by the NIH–MLPCN program (1 U54 HG005032-1 awarded
to S.L.S.).
Substitution with bicyclic groups related to 2 as shown in Table
4
produced potent compounds; however these compounds
demonstrated reduced selectivity. Bromonapthalene derived
hydrazone 8m is one of the most potent compounds, but displayed
poor selectivity. Benzofuran derived hydrazone 8o displayed selec-
tivity; however, it was only moderately potent against the
HMLE_sh_Ecad cell line.
Based on the SAR described above, the 2,4,6-trichloro- analog 7j
was designated as a probe with the NIH Molecular Libraries
Program (ML239) due to its potency and excellent selectivity.
Supplementary data
Supplementary data (experimental protocols for cellular assays
and for the preparation of compounds) associated with this article
can be found, in the online version, at doi:10.1016/j.bmcl.2012.
01.035.

3574
A. R. Germain et al. / Bioorg. Med. Chem. Lett. 22 (2012) 3571–3574
Zhang, Y. Breast Cancer Res. Treat. 2009, 115, 269; (c) Kakarala, M.; Brenner, D. E.;
References and notes
Korkaya, H.; Cheng, C.; Tazi, K.; Ginestier, C.; Liu, S.; Dontu, G.; Wicha, M. S.
Cancer Res. Treat. 2010, 122, 777.
1. (a) Bao, S.; Wu, Q.; McLendon, R. E.; Hao, Y.; Shi, Q.; Hjelmeland, A. B.; Dewhirst,
M. W.; Bigner, D. D.; Rich, J. N. Nature 2006, 444, 756; (b) Diehn, M.; Cho, R. W.;
Lobo, N. A.; Kalisky, T.; Dorie, M. J.; Kulp, A. N.; Qian, D.; Lam, J. S.; Ailles, L. E.;
Wong, M. Nature 2009, 458, 780; (c) Diehn, M.; Clarke, M. F. J. Natl. Cancer Inst.
2006, 98, 1755; (d) Eyler, C. E.; Rich, J. N. J. Clin. Oncol. 2008, 17, 2839; (e) Li, X.;
Lewis, M. T.; Huang, J.; Gutierrez, C.; Osborne, C. K.; Wu, M. F.; Hilsenbeck, S. G.;
Pavlick, A.; Zhang, X.; Chamness, G. C. J. Natl. Cancer Inst. 2008, 100, 672.
2. Mani, S. A.; Guo, W.; Liao, M. J.; Eaton, E. N.; Ayyanan, A.; Zhou, A. Y.; Brooks, M.;
Reinhard, F.; Zhang, C. C.; Shipitsin, M. Cell 2008, 133, 704.
5. (a) Li, Y.; Zhang, T.; Korkaya, H.; Liu, S.; Lee, H. F.; Newman, B.; Yu, Y.; Clouthier,
S. G.; Schwartz, S. J.; Wicha, M. S.; Sun, D. Clin. Cancer Res. 2010, 16, 2580; (b)
Hirsch, H. A.; Iliopoulos, D.; Tsichlis, P. N.; Struhl, K. Cancer Res. 2009, 69, 7507;
(c) Jiralerspong, S.; Palla, S. L.; Giordano, S. H.; Meric-Bernstam, F.; Liedtke, C.;
Barnett, C. M.; Hsu, L.; Hung, M. C.; Hortobagyi, G. N.; Gonzalez-Angulo, A. M. J.
Clin. Oncol. 2009, 27, 3297; (d) Ginestier, C.; Liu, S.; Diebel, M. E.; Korkaya, H.;
Luo, M.; Brown, M.; Wicinski, J.; Cabaud, O.; Charafe-Jauffret, E.; Birnbaum, D. J.
Clin. Invest. 2010, 120, 485; (e) Kim, J. B.; Ko, E.; Han, W.; Shin, I.; Park, S. Y.; Noh,
D. Y. Planta Med. 2008, 74, 1693.
3. Dontu, G.; Abdallah, W. M.; Foley, J. M.; Jackson, K. W.; Clarke, M. F.; Kawamura,
M. J.; Wicha, M. S. Genes Dev. 2003, 17, 1253.
6. Gupta, P. B.; Onder, T. T.; Jiang, G.; Tao, K.; Kuperwasser, C.; Weinberg, R. A.;
Lander, E. S. Cell 2009, 138, 645.
4. (a) Zhou, J.; Zhang, H.; Gu, P.; Bai, J.; Margolick, J. B.; Zhang, Y. Breast Cancer Res.
Treat. 2007, 111, 419; (b) Zhou, J.; Zhang, H.; Gu, P.; Margolick, J. B.; Yin, D.;