Hydroxyquinoline-derived compounds and analoguing of selective Mcl-1 inhibitors using a functional biomarker

Article abstract of DOI:10.1016/j.bmc.2013.08.017

Anti-apoptotic Bcl-2 family proteins are important oncology therapeutic targets. To date, BH3 mimetics that abrogate anti-apoptotic activity have largely been directed at Bcl-2 and/or Bcl-xL. One observed mechanism of resistance to these inhibitors is increased Mcl-1 levels in cells exposed to such therapeutics. For this reason, and because Mcl-1 is important in the onset of lymphoid, myeloid, and other cancers, it has become a target of great interest. However, small molecule inhibitors displaying potency and selectivity for Mcl-1 are lacking. Identifying such compounds has been challenging due to difficulties in translating the target selectivity observed at the biochemical level to the cellular level. Herein we report the results of an HTS strategy coupled with directed hit optimization. Compounds identified have selective Mcl-1 inhibitory activity with greater than 100-fold reduced affinity for Bcl-xL. The selectivity of these compounds at the cellular level was validated using BH3 profiling, a novel personalized diagnostic approach. This assay provides an important functional biomarker that allows for the characterization of cells based upon their dependencies on various anti-apoptotic Bcl-2 proteins. We demonstrate that cells dependent on Mcl-1 or Bcl-2/Bcl-xL for survival are commensurately responsive to compounds that genuinely target those proteins. The identification of compound 9 with uniquely validated and selective Mcl-1 inhibitory activity provides a valuable tool to those studying the intrinsic apoptosis pathway and highlights an important approach in the development of a first-in-class cancer therapeutic.

Full text of DOI:10.1016/j.bmc.2013.08.017

Bioorganic & Medicinal Chemistry xxx (2013) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Hydroxyquinoline-derived compounds and analoguing of selective
Mcl-1 inhibitors using a functional biomarker
David J. Richard a, Ryan Lena a, Thomas Bannister b, Noel Blake a, William E. Pierceall a, Nicole E. Carlson a,
Christina Eberhart Keller b, Marcel Koenig b, Yuanjun He b, Dmitriy Minond b, Jitendra Mishra b,
Michael Cameron b, Timothy Spicer b, Peter Hodder b, Michael H. Cardone a,
⇑
a Eutropics Pharmaceuticals, 767C Concord Avenue, Cambridge, MA 02138, United States
b The Scripps Research Institute, Department of Chemistry and Lead Identification Division of the Translational Research Institute, 130 Scripps Way, Jupiter, FL 33458, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Anti-apoptotic Bcl-2 family proteins are important oncology therapeutic targets. To date, BH3 mimetics
that abrogate anti-apoptotic activity have largely been directed at Bcl-2 and/or Bcl-xL. One observed
mechanism of resistance to these inhibitors is increased Mcl-1 levels in cells exposed to such therapeu-
tics. For this reason, and because Mcl-1 is important in the onset of lymphoid, myeloid, and other cancers,
it has become a target of great interest. However, small molecule inhibitors displaying potency and selec-
tivity for Mcl-1 are lacking. Identifying such compounds has been challenging due to difficulties in trans-
lating the target selectivity observed at the biochemical level to the cellular level. Herein we report the
results of an HTS strategy coupled with directed hit optimization. Compounds identified have selective
Mcl-1 inhibitory activity with greater than 100-fold reduced affinity for Bcl-xL. The selectivity of these
compounds at the cellular level was validated using BH3 profiling, a novel personalized diagnostic
approach. This assay provides an important functional biomarker that allows for the characterization
of cells based upon their dependencies on various anti-apoptotic Bcl-2 proteins. We demonstrate that
cells dependent on Mcl-1 or Bcl-2/Bcl-xL for survival are commensurately responsive to compounds that
genuinely target those proteins. The identification of compound 9 with uniquely validated and selective
Mcl-1 inhibitory activity provides a valuable tool to those studying the intrinsic apoptosis pathway and
highlights an important approach in the development of a first-in-class cancer therapeutic.
Ó 2013 Elsevier Ltd. All rights reserved.
Received 30 May 2013
Revised 30 July 2013
Accepted 6 August 2013
Available online xxxx
Keywords:
Apoptosis
Mcl-1
Biomarkers
Oncology
Cell-based activity
1. Introduction
3131 cancer specimens identified mutations surrounding Mcl-1
as being among the most significant causal factors.18
Modulation of apoptosis has long been of interest to the oncol-
ogy community, as a primary mechanism of cancer cell survival by
evading programmed cell death.1 Bcl-2 family proteins are central
to the regulation of the intrinsic, or mitochondrial, apoptosis path-
way2,3 as family members interactions result in heterodimer for-
mation that modulates the activity of the multidomain pro-
apoptotic proteins Bax and Bak.4 Oligomerization of Bax and Bak
results in mitochondrial outer membrane permeabilization
(MOMP) and release of apoptosis-promoting proteins, including
Smac/DIABLO and cytochrome c, which in turn promote caspase
activation and result in cell death.5 Myeloid cell factor 1 (Mcl-1)
has been identified as an important therapeutic target for the treat-
ment of non-solid tumor6–13 as well as solid tumor malignancies14–
Inhibition of anti-apoptotic Bcl-2 family proteins has been
validated as a therapeutic strategy by the clinical advancement of
the Bcl-2 inhibitors Navitoclax19 and ABT-199.20 These small mole-
cules bind to the hydrophobic groove in Bcl-2 and/or Bcl-xL and mi-
mic the pro-apoptotic ‘‘BH3-only’’ proteins, thereby promoting
activation of Bax and Bak. Cell lines found to be refractory to these
compounds regained sensitivity when Mcl-1 was down-regu-
lated.21,22 These findings strongly support the notion that Mcl-1 is a
key resistance factor to Bcl-2/Bcl-xL targeted therapies and under-
score the importance of developing an Mcl-1 targeted therapy. Two
other purported Mcl-1 inhibitors, obatoclax (GX15-070)23 and gossy-
pol (AT-101)24, have each displayed significant off-target activities
suggesting that their efficacy is largely not derived from Mcl-1 inhibi-
tion but rather from cytotoxicity in a Bax-Bak independent fashion
and induced caspase-9 independent cell death.25,26 Further, inhibi-
tion of certain Bcl-2 family proteins can display adverse clinical con-
sequences. For instance, thrombocytopenia has been observed
following treatment with the Bcl-2/Bcl-xL inhibitor Navitoclax, halt-
16
largely owing to its role as a critical node in intrinsic apoptotic
susceptibility.17 Recently, a study of mutation analyses from
⇑
Corresponding author. Tel.: +1 617 714 4405x112.
E-mail address: mcardone@eutropics.com (M.H. Cardone).
0968-0896/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmc.2013.08.017
Please cite this article in press as: Richard, D. J.; et al. Bioorg. Med. Chem. (2013), http://dx.doi.org/10.1016/j.bmc.2013.08.017

2
D. J. Richard et al. / Bioorg. Med. Chem. xxx (2013) xxx–xxx
ing its clinical development.27 In that case, the activity against Bcl-xL
impacted platelet survival.28 Recent efforts have focused on develop-
ment of selective compounds with diminished Bcl-xL activity, such as
ABT-199, with limited platelet toxicity. Avoiding inhibition of other
anti-apoptotic proteins may be preferred in some cases for patients
comprising a specific malignant disease. A small molecule inhibitor
thatisselective for Mcl-1 wouldprovide animportant chemical probe
todefine thetherapeutic potential of Mcl-1 inhibition, elucidatingthe
significance of Mcl-1 in cancer and determining if tumor cells charac-
terized by elevated Mcl-1 activity can be selectively targeted.
NMR data for novel compound 9 and an analysis of its properties,
and general SAR trends in the series are also shown in Supplemen-
tary data. Vendor information for previously undescribed but com-
mercially available compounds 3–7 are provided as well.
Compound 8 was internally synthesized by the methods used for
compound 9 and its confirmatory LCMS data is shown in Supple-
mentary data. Functional assays of compounds 1–9 were per-
formed using racemic mixtures, therefore stereochemistry has
not been indicated at chiral centers.
Efforts to develop effective Mcl-1 inhibitors have been slowed by
frequent coincident and pronounced off-target activity. Our strategy
of BH3 profiling addresses selectivity by providing a functional bio-
marker, allowing for identification of the mechanism of action of
BH3 mimetics within a cellular context. This approach quantifies
mitochondrial response to any one or any class of BH3 peptides
and indicates a particular dependence upon an anti-apoptotic Bcl-2
family protein. For example, Noxa binds with high affinity only to
Mcl-1, Bad binds to Bcl-xL and Bcl-2 but only weakly to Mcl-1, and
Puma binds strongly to all three targets.29 Each cell line may there-
fore be characterized by its extent of ‘priming’ with respect to a par-
ticular Bcl-2 family member, such as Mcl-1, Bcl-2, or Bcl-xL.29
2.3. Computational chemistry
Docking was performed using FlexX in LeadIT version 2.1.2
(BiosolveIT, St. Augustin, Germany). Compound structure was
drawn in MDL ISIS Draw and minimized coordinates were gener-
ated in PipelinePilot. The binding site was prepared using the coor-
dinates of Mcl-1 (PDB 2NLA) and defined as the amino acids within
a 20 Å radius of Arg263. An essential interaction was defined such
that the side-chain amine of Asn260 was required to donate a
hydrogen bond to all docked poses. The Hybrid Approach was used
for base placement in the docking runs, which generated 200 solu-
tions per iteration and fragmentation. Docked poses were exported
in mol2 format for visualization in PyMOL.
2. Experimental
2.4. Cell lines and growth inhibition assays
2.1. High throughput screening
The mouse leukemia derived cell lines Mcl1-1780 and Bcl2-
186331 have been licensed from the Dana Farber Cancer Institute.
The human lymphoid SUDHL-10, SUDHL-6, myeloid NCI-H929,
acute promyelocytic leukemia HL-60, erythroleukemia TF-1, malig-
nant melanoma A-375, carcinoma A-549 and NCI-H187, and panc-
reactic carcinoma PANC-1 cell lines were obtained from American
Type Culture Collection (ATCC), Manassas, VA. Suspension lines
SUDHL-6, SUDHL-10, NCI-H929, MOLM13, ML2, HL-60, TF-1, and
NCI-H187 were grown in RPMI-1640 with 10% fetal bovine serum
(Life Technologies). Suspension lines Mcl1-1780 and Bcl2-1863
A high throughput screen (HTS), fluorescence polarization (FP)
assay,30 was performed at Scripps Research Institute Molecular
Screening Center (SRIMSC) by peptide binding quantitation. The
NIH screening library (315,100 compounds) was provided by the
National Institutes of Health. FITC-Bim BH3-only peptide (FITC-
AHA-MRPEIWIAQELRRIGDEFNA-[NH2]) was synthesized at the
Tufts University Core Facility. Human Mcl-1, and Bcl-xL-GST (Glu-
tathione-S-Transferase) fusion proteins, with deleted transmem-
brane regions, were cloned into pGEX 4T-1. Proteins were
expressed in BL21 strain and purified using Amersham Hitrap Glu-
tathione column on an ACTA-FPLC.
FP was performed in assay buffer (Dulbecco’s PBS buffer, pH 7.2,
0.001% v/v Brij 35) containing either GST-Mcl-1 or GST-Bcl-xL dis-
pensed into 1536-well microtiter plates (Corning). Test compound,
unlabeled Bim control peptide, or DMSO was added to the appro-
priate test or control wells. Bim-FITC in assay buffer was dispensed
into wells, and plates were centrifuged and incubated at 25 °C for
20 min. Fluorescence polarization was read by Viewlux (PerkinEl-
mer). Percent inhibition for each compound was calculated using
a ratio of the test compound mP to median value of positive control
mP, correcting each with the median value of the negative control
mP. Following primary screening, potential active compounds
were grown in the same media with 3.6 lL/L b-mercaptoethanol
(BME). Adherent cell lines A-375, A-549 and PANC-1 were grown
in DMEM with 10% fetal bovine serum (Life Technologies).
Cells were resuspended at 2.5 Â 105 cells/mL and plated on
384-well plates with seven twofold dilutions of either DMSO (high
concentration: 0.2%) or test compound (high concentration: 5 or
20 lM). Control wells of cells and media only were also included.
Plates were incubated at 37 °C for 48 h and analyzed using Presto
Blue (Invitrogen) on a Tecan Infinite F200 Pro Plate Reader. Back-
ground (media alone) was subtracted from all data. Percent viabil-
ity was normalized to the reading obtained for wells containing
only cells (considered 100% viability).
2.5. Analysis of cell lines by BH3 profiling
were confirmed by repeat of the assay in triplicate at 10
same assays were then performed using a ten-point, three-fold
titration with compound concentrations from 100 M to 0.5 nM.
The resulting response curves were used to calculate IC50 values
for both Mcl-1 and Bcl-xL. Detailed protocols for these assays are
found at the NIH’s MLPCN PubChem website (http://pub-
chem.ncbi.nlm.nih.gov/assay/assay.cgi?aid=2090).
lM. The
Peptides (Bim, Puma, Noxa, Bad) were from New England Pep-
tide. Cells were seeded into a 384 well plate in duplicate with
l
100
l
M peptide, 2.5 Â 104 cells, 1
lM JC-1, and 0.005% digitonin,.
Fluorescence was by continuous read every 5 min for 180 min on
a Tecan Infinite F200 Pro plate reader and data was analyzed using
GraphPad Prism (La Jolla, CA) to determine area under the curve
(AUC). Percent priming was calculated using the AUC measure-
ments as compared to the positive (carbonyl cyanide m-chloro-
phenyl hydrazone, CCCP) and negative (DMSO) controls.
2.2. Analog preparation
Compounds were prepared at the Scripps Research Institute,
Jupiter, FL or where purchased from commercial vendors. Methods
utilized for synthesis are described in the Supplementary data and
described in detail for compound 9, a simple one-step procedure.
Identity and purity of all new compounds were determined by
NMR (Bruker Instruments) and LCMS (Agilent); purity was con-
firmed to be greater than 95% in each case. LCMS, HPLC, and
2.6. Cytochrome c release determination
Cells were collected, washed in PBS, re-suspended at a concentra-
tion of 2 Â 106 cells/ml, and incubated with test and control com-
pounds (0.2% DMSO; 20 lM EU-5346) or peptides (100 lM Bim;
Please cite this article in press as: Richard, D. J.; et al. Bioorg. Med. Chem. (2013), http://dx.doi.org/10.1016/j.bmc.2013.08.017

D. J. Richard et al. / Bioorg. Med. Chem. xxx (2013) xxx–xxx
3
100
l
M Puma) for 2 h at room temperature. Samples were fixed with
assays were used to screen the NIH Molecular Libraries Small Mol-
ecule Repository (MLSMR); 315,000 compounds; 2141 compounds
showing greater than 40% inhibition of Mcl-1 were triaged as pos-
sible hits (0.68% rate). Elimination of compounds with considerable
Bcl-xL binding reduced the hit set to 1720 compounds (0.54% rate).
Confirmation of both Mcl-1 activity and lack of Bcl-xL activity, re-
sulted in identification of 179 validated hits which were advanced
4% Formaldehyde in PBS for 20 min, washed once in PBS, blocked for
15 min with 2% FBS 0.5% TritonX-100 in PBS, then re-suspended in
blocking buffer with 1:250 anti-cytochrome c-Alexa488 (Beckton–
Dickinson) and incubated for 2 h at 4 °C. Samples were washed once
with blocking buffer and mounted on slides using Vectashield
Mounting Media with DAPI (Vector laboratories). A minimum of
100 DAPI positive cells per treatment were visualized by fluores-
cence microscopy and scored as positive or negative for cytochrome
c loss. DMSO was calculated as 0% cytochrome c loss; Bim response
for SUDHL-6 was used to determine 100% cytochrome c loss.
to dose response studies. 52 compounds demonstrated IC50
-
6 10
l
M against Mcl-1; 24 of these small molecules met selectivity
criteria of reduced Bcl-xL inhibitory activity (IC50 P 10 lM against
Bcl-xL).
An assessment of chemical tractability, as well as the potential
for target specificity as indicated by their activity in other HTS
campaigns, resulted in the selection of the substituted 7-hydroxy-
quinoline 1 (Fig. 1a) for additional study. The favorable physico-
2.7. Nuclear apoptosis determination
Cell lines were seeded at 2 Â 105 cells/ml in culture media and
incubated with DMSO (0.1%), Navitoclax (5, 2.5, 1.25, and
chemical properties expected for compound 1 (solubility,
0.625
l
M) or compound 9 (1000, 500, 250, and 125 nM) for 24 h
lipophilicity), its reasonable molecular weight (436), lack of any
apparent toxicophores, and low hit rate in other HTS assays made
this compound a promising candidate. Compound 1 demonstrated
at 37 °C. Following treatment, samples were fixed in 100% metha-
nol on ice for 10 min, washed once with PBS, and mounted on
slides using Vectashield/DAPI. A minimum of 100 cells per treat-
ment were visualized by fluorescence microscopy. Nuclei were
scored as normal (round, normal sized, intact nuclei) or apoptotic
(pyknotic, fragmented). DMSO treatment provided the negative
good Mcl-1 inhibition (IC50 = 2.4
tion of Bcl-xL at 100 M.
lM) with no appreciable inhibi-
l
Upon structural analysis of compound 1, two functional groups
were identified as desirable for elimination or modification; the
carboxylic acid moiety, which could contribute to poor cellular
permeability, and the 4-chloro group, which typically raises cLogP
and reduces aqueous solubility. Examination of a small set of ana-
logs lacking these groups demonstrated that Mcl-1 binding affinity
was maintained upon their removal. Therefore, subsequent efforts
focused upon the generation of derivatives lacking these moieties.
A review of the literature indicated that hydroxyquinoline 2 had
previously been disclosed as an Mcl-1 inhibitor, though without a
reported IC50 value.32 We explored group and substituent effects in
the phenyl region of hit 1 by preparing approximately 100 related
analogs. This work resulted in the conclusion that both the quino-
line nitrogen atom and the 8-hydroxyl group are essential. Repre-
sentative compounds that demonstrate key SAR trends are shown
in Figure 1a and Table 1. Ortho- and meta-Substituted compounds
were much weaker Mcl-1 inhibitors, whereas a bis-substituted
analog also showed reduced potency (3–5, Table 1). Para-Substitu-
tion (methoxy derivative 6) and five-membered heterocycles (fur-
an 7) enhanced potency. An electron-withdrawing F group at the
para position (analog 8), which might be expected to enhance met-
abolic stability, was fivefold more potent than the initial hit.
The next stage of analog efforts focused upon modification of
the aminopyridine ring. Conversion of the aminopyridine to a sec-
ondary amine such as a piperidine, morpholine, or piperazine, re-
sulted in compounds which displayed excellent potency against
Mcl-1 and maintained or enhanced selectivity over Bcl-xL. This
change from an aromatic to a non-aromatic, basic amine also
control. Navitoclax at 5
vided the positive control.
lM treatment on cell line Bcl2-1863 pro-
2.8. Annexin V apoptosis assay
Apoptosis was detected by Annexin V and propidium iodide
staining. Mcl1-1780 and Bcl2-1863 were seeded at 1.0 Â 106 cells/
well in a 24 well plate in culture media containing a range of con-
centrations of compound 9 from 0.15 to 40 lM and incubated for
16 h. Control wells containing DMSO were also included. Cells were
stained with BD Biosciences (San Jose, CA) FITC Annexin V Apoptosis
Detection Kit II according to the manufacturer’s protocol. 1.0 Â 105 -
cells were resuspended in 100
l of Annexin V FITC and 2
mixed and incubated for 15 min at room temperature in the dark.
400 l of 1X binding buffer was added to each tube and analyzed
ll of 1X binding buffer along with
5
l
ll of propidium iodide. Cells were
l
on a FACSCanto II (BD Biosciences). A minimum of 20,000 cells per
sample were analyzed using FACSDiva software (BD Biosciences).
3. Results
3.1. High-throughput screening identified novel, selective Mcl-1
inhibitors
An FP assay provided the primary Mcl-1 biochemical screen. A
second FP assay utilizing Bcl-xL provided the counter screen. Both
O
Cl
OH
CF3
O
S
N
N
N
OH
N
N
OH HN
OH HN
IC50, Mcl-1: 0.31 µM
IC50, Bcl-xL: >40 µM
N
N
1
2
9
(a)
(b)
Figure 1. (a) Structure of compound 1, a 7-hydroxyquinoline identified by high-throughput screening as possessing Mcl-1-selective inhibitory activity relative to the related
Bcl-2 family protein Bcl-xL; (b) Structure of the piperazine-substituted hydroxyquinoline 9, which displays selective inhibition of Mcl-1 relative to Bcl-xL as indicated by
fluorescence polarization assay.
Please cite this article in press as: Richard, D. J.; et al. Bioorg. Med. Chem. (2013), http://dx.doi.org/10.1016/j.bmc.2013.08.017

4
D. J. Richard et al. / Bioorg. Med. Chem. xxx (2013) xxx–xxx
Table 1
Structures and Mcl-1 inhibitory properties of a series of aminopyridine-substituted hydroxyquinoline derivatives
R1
N
OH HN
R2
N
3
4
5
6
7
8
F
O
O
O
H
F
R1
CF3
O
H
H
H
7.5
H
CH3
0.5
R2
IC50, Mcl-1
>25
>25
1.2
0.4
served to reduce LogD and enhance predicted water solubility, as
desired. The combination of this finding with the benefit to Mcl-
1 binding conferred by a para substituent on the phenyl group
(compound 8) led to the identification of lead compound 9. This
position the N-ethylpiperazine moiety to occupy one of the four
primary lipophilic pockets typically filled by side chains of BH3-
only peptides.
analog demonstrated IC50s of 0.31 and 40
l
M for Mcl-1 and Bcl-
3.2. Potent activity in cell-based assays and selectivity
correlation with BH3 profiling
xL, respectively (Fig. 1b). Moreover, compound 9 possesses moder-
ate molecular weight (415), hydrophobicity (cLogD7.4 = 3.7) and
lacks inherent toxicophores. The para-CF3 group and piperazine
ring contribute to a moderately high liver microsome stability for
compound 9 (half-life 55 minutes in human microsomes, 40 min,
in mouse microsomes) and IC50’s for inhibition of 4 isoforms of
The ability of hydroxyquinoline 9 to induce cell death was
assessed in cellular proliferation assays. We focused on cell lines
derived from hematologic malignancies where Mcl-1 has been
shown to play a significant role. Initial screening demonstrated
that compound 9 displayed low micromolar efficacy in the mouse
lymphoma cell lines Mcl1-1780 and Bcl2-1863, the human lym-
phoid cell line SUDHL-6, and the human multiple myeloma line
NCIH929 (Fig. 3a).
While these results provided evidence that compounds were
cytotoxic, we sought to validate the precise mechanism of action
using BH3 profiling. This assay provides a means of measuring
the extent to which a cell relies on the function of any or all of
the anti-apoptotic Bcl-2 family proteins for survival. BH3 profiling
utilizes a panel of BH3-domain peptides (derived from the BH3-
only proteins).34,35 The extent of MOMP observed following
CYP450 tested (1A2, 2D6, 3A4, 2C9) each greater than 10 lM.
Computational chemistry studies were performed to obtain
greater understanding of the nature of the interaction between
hydroxyquinoline 9 and Mcl-1. X-ray crystal structures indicate a
binding site for Mcl-1 and the BH3-only peptides that consist of
a long, hydrophobic region which accommodates four lipophilic
side chains of the BH3-only peptide.33 Our studies indicate that
compound 9 fits compactly within this known Mcl-1 binding
groove (Fig. 2). Modeling of the R-configuration of compound 9
suggests that a hydrogen bond formed between the Mcl-1 residue
Asn260 and the hydroxyquinoline moiety of 9 may serve to
Figure 2. Calculated binding mode of compound 9 in Mcl-1. Modeling suggests a hydrogen bond between the phenol of the hydroxyquinoline and the carbonyl of Asn260
may orient the ethylpiperazine group into one of the four major lipophilic pockets found in the Mcl-1/BH3-only peptide binding groove.
Please cite this article in press as: Richard, D. J.; et al. Bioorg. Med. Chem. (2013), http://dx.doi.org/10.1016/j.bmc.2013.08.017

D. J. Richard et al. / Bioorg. Med. Chem. xxx (2013) xxx–xxx
5
Figure 3. (a) EC50 values from cell proliferation assay experiments (viability determined by PrestoBlue staining) for the Mcl-1 inhibitor 9; Results indicated are the mean of
two independent cell growth inhibition experiments conducted in triplicate (n = 6); (b) results of BH3 profiling assay on the lymphoid cell lines SU-DHL6 and SU-DHL10,
myeloma cell line NCI-H929, and mouse leukemia cell lines Bcl2-1863 and Mcl1-1780. These cell lines display a range of mitochondrial depolarization upon exposure to
various BH3-only peptides (Bim, Puma, Noxa, Bad). For example, SU-DHL6 is highly ‘primed’ for all BH3 peptides tested but SU-DHL10 has low levels of mitochondrial
priming. Priming values are derived from two independent assays conducted in triplicate (n = 6); (c) correlation between antiproliferative activity of 9 (EC50 represented in
l
M) and BH3 profiling response (% mitochondrial depolarization, Bim peptide assayed at 0.3 lM). An excellent correlation is observed between priming state and
antiproliferative response observed in these compound 9—treated cell lines (R2 = 0.81).
treatment of semi-permeablized cells with a specific BH3-only pro-
tein indicates that the cell is near the apoptotic threshold with re-
spect to that binding partner and is mitochondrially ‘primed’.34,35
As determined by BH3 profiling, the lymphoma-derived cell line
SUDHL-6 displays significant priming for a variety of BH3-only
peptides (Fig. 3b).28 In contrast, the related SUDHL-10 line demon-
strates low levels of priming for all Bcl-2 members. The readout of
the BH3 profiling assay in cell lines provides a valuable functional
biomarker for compounds that mimic the BH3 peptides used in the
assay. Cell lines that demonstrate high levels of priming would be
expected to display the greatest response to BH3 mimetics. We
determined response to compound 9 in 11 human cell lines of
varying origin, and with varying dependencies on the different
Bcl-2 family proteins as determined by the BH3 profiling assay (Ta-
ble 2). As shown in Figure 3c, a strong correlation was observed be-
tween the degree of mitochondrial priming with respect to the
BH3-only peptide Bim.
Table 2
EC50 values determined in cell proliferation assays for Mcl-1 inhibitor 9 in a diverse
panel of human-derived cancer cell lines
Cell nine
EC50 (lM)
Mcl1-1780
Bcl2-1863
SUDHL-6
NCI-H929
SUDHL-10
HL60
ML2
MOLM13
TF1
A375
A549
H187
PANC1
0.3
1.1
3.3
1.6
25
6.3
5.2
7.0
4.8
7.1
14.1
9.5
14.5
Data represent the average of three replicate experiments.
peptides, in semi-permeablized cells (Fig. 4a). Treatment of a series
of lymphoma and leukemia cell lines with Mcl-1 inhibitor 9 re-
sulted in significant cytochrome c release at 2 h. The greatest ex-
3.3. Cytochrome c release is correlative to mitochondrial
priming state
tent of cytochrome
c release was observed in the highly
A hallmark of on-target BH3 mimetics is their ability to promote
timely release of cytochrome c compared to competing BH3 pep-
tides. We analyzed the ability of Mcl-1 inhibitor 9 to promote re-
lease of cytochrome c, as compared to the Bim and PUMA BH3
mitochondrially primed SUDHL-6 cell line. The release of cyto-
chrome c from mitochondria in the poorly primed SUDHL-8 cell
line and the Bax/Bak deficient SUDHL-10 cell line was significantly
Please cite this article in press as: Richard, D. J.; et al. Bioorg. Med. Chem. (2013), http://dx.doi.org/10.1016/j.bmc.2013.08.017

6
D. J. Richard et al. / Bioorg. Med. Chem. xxx (2013) xxx–xxx
Figure 4. (a) Compound 9 promotes cytochrome c release in a fashion that may be correlated with the extent of mitochondrial priming in a series of human-derived
lymphoid cell lines (SU-DHL6, SU-DHL8, SU-DHL10) and two mouse leukemia cell lines (Bcl2-1863 and Mcl1-1780); (b) cell viability of mouse leukemia cell lines Bcl2-1863
and Mcl1-1780 following treatment with ABT-263 or compound 9 as measured by DAPI assay; (c) percentage of cells that demonstrate apoptosis, as detected by Annexin V—
positive staining, following treatment with compound 9 at the indicated concentrations. Results depicted are representative of two independent experiments and are
consistent for dose-response trends.
reduced. In the mouse-derived lymphoma cell lines Mcl1-1780 and
Bcl2-1863, cytochrome c release was observed to be significantly
greater in the former cell line (Fig. 4a), which are selectively
dependent on either Bcl-2 or Mcl-1 for survival.31 These two cell
lines are derived from the same mouse lineage and are selectively
dependent on either Mcl-1 or Bcl-2/Bcl-xL for survival, as deter-
mined by the Noxa (Mcl-1 restricted) or Bad (Bcl-2 restricted) sig-
nal by BH3 profiling (Fig. 3b).
Mcl-1 through a high throughput screening effort and validated
these compounds utilizing BH3 profiling. This platform technology
provides a functional biomarker which was utilized to assess the
on-target mechanism of these inhibitors in a cell-based context.
The primary screen was designed and implemented with the
aim of identifying Mcl-1 inhibitors which displayed selectivity over
Bcl-xL. The initial hit (1, Fig. 1) demonstrated affinity for Mcl-1
(2.4 lM) with greater than 40-fold selectivity over Bcl-xL. How-
ever, this analog displayed only modest antiproliferative activity
and therefore required structural modification. An analog effort al-
lowed for the identification of a pharmacophore for Mcl-1 inhibi-
tion. Incorporation of electron-withdrawing groups on the aryl
ring, and a piperazine replacement for the aminopyridine group,
enhanced Mcl-1 inhibition and selectivity over Bcl-xL and resulted
in Mcl-1 inhibitor 9. A proposed binding mode was identified using
computational chemistry methods. The interactions between Mcl-
1 and its various BH3-only peptide binding partners (such as Bim,
Puma, NoxaA) are stabilized through interactions between four
hydrophobic side chains of the BH3-only binding partner and cor-
responding hydrophobic pockets in the binding groove of Mcl-1.
Docking and SAR studies suggest that hydrogen bonding between
3.4. Compound 9 promotes apoptosis as determined by DAPI
and Annexin V staining
Using standard methods we examined Mcl-1 inhibitor 9 for
ability to induce apoptosis; this study was performed in the
Mcl1-1780 and Bcl2-1863 cell lines (see Fig. 4b). The Bcl-2/Bcl-xL
inhibitor Navitoclax (ABT-263) was also examined in this study.
This compound shows a distinct preference in inducing apoptosis
in the highly Bcl-2/Bcl-xL primed cell line Bcl2-1863, and displays
only minimal effectiveness in the highly Mcl1 primed cell line. In
contrast, Mcl-1 inhibitor 9 displays an inverse pattern of apoptotic
induction ability.
The extent of apoptosis conferred by analog 9 was determined
using Annexin V staining. The percent of Annexin V positive cells
displayed dose-dependency and was greater in Mcl1-1780 than
in Bcl2-1863, providing further evidence of Mcl-1 inhibition driven
apoptosis that correlates with the extent of Mcl-1 mediated mito-
chondrial priming (Fig. 4c).
the quinoline C7-hydroxyl group of Mcl-1 inhibitor
9 with
Asn260 of Mcl-1 places the lipophilic alkyl moiety of the pipera-
zine group within a lipophilic pocket of the protein (Fig. 2).
The antiproliferative activity of Mcl-1 inhibitor 9 was studied in
several cell lines derived from various human malignancies. This
lead compound displayed significant activity in a number of cell
lines, with EC50 values in the range of 0.3–15 lM. We then sought
4. Discussion
to further define the mechanism of action of this agent, and did so
by application of BH3 profiling. Using this assay we found a strong
correlation between the degree of priming, with respect to Bim
BH3 peptide, and responsiveness of these cell lines to treatment
with compound 9 (Fig. 3c). These results demonstrate the utility
In this work we utilized a novel approach to select compounds
that have activity against specific members of the Bcl-2 family pro-
teins. We have identified a series of small molecule inhibitors of
Please cite this article in press as: Richard, D. J.; et al. Bioorg. Med. Chem. (2013), http://dx.doi.org/10.1016/j.bmc.2013.08.017

D. J. Richard et al. / Bioorg. Med. Chem. xxx (2013) xxx–xxx
7
of BH3 profiling as a drug discovery tool to validate the mechanism
Conflict-of-interest disclosure
of action of BH3 mimetics.
We then examined the ability of our lead compound to induce
cytochrome c release in cells that were characterized by the BH3
profiling assay. In three leukemia-derived cell lines possessing var-
ious priming states (SUDHL-10, -8, and -6; Fig. 4a), we observed re-
sponse nearly identical to that seen upon treatment with the
peptides comprising the Puma BH3 domain. Cytochrome c release
was observed upon treatment of the highly primed SUDHL-6 cell
line with both compound 9 and Puma, whereas the poorly primed
SUDHL-8 and SUDHL-10 cell lines did not produce such a response
upon exposure to these agents. This indicated that priming was re-
quired for cell response. We also examined the activity of Mcl-1
inhibitor 9 in the mouse leukemia cell lines BCL2-1863 and
Mcl1-1780. These two cell lines have been characterized as mito-
chondrial-primed with respect to Bcl-2 or Mcl-1 (by determination
of the BH3 profiling readout in response to treatment with the Bad
or Noxa BH3-only proteins, respectively, Fig. 3b). In contrast, com-
pound 9 displays selective activation of cytochrome c in Mcl1-
1780. This result provides strong functional support for compound
selectivity in a cellular context.
To further validate the on-target mechanism and selectivity of
our lead analog, we determined the extent to which Mcl-1 inhibi-
tor 9, as well as the Bcl-2/Bcl-xL inhibitor Navitoclax, were able to
promote nuclear apoptosis. Navitoclax demonstrates the ability to
selectively promote cell death in the Bcl2-1863 cell line, which is
characterized as Bcl-2 primed (Fig. 4b). This finding clearly demon-
strates the utility of BH3 profiling as correlative with selectivity in
a cellular context, given the well-validated Bcl-2/Bcl-xL selectivity
of this agent. The response observed for Mcl-1 inhibitor 9 is consis-
tent with its Mcl-1 selectivity; the compound reduces viability
with a marked preference for the Mcl1-1780 cell line. We utilized
Annexin V staining to provide additional confirmation of the ability
of this compound to produce apoptosis. In this study (Fig. 4c), we
observed that the extent of apoptotic induction was greater in
Mcl1-1780 relative to Bcl2-1863, a trend which mirrored the
EC50 values determined for these cell lines in an antiproliferative
David Richard, William Pierceall, Nicole Carlson, Ryan Lena,
Noel Blake, and Michael Cardone are employees of Eutropics, Inc.
The remaining authors declare no conflicts-of-interest.
Acknowledgements
This work was supported by NCI-SBIR grant #2R44CA135915-
02A1. T.B., C.E.K.,M.K., Y.H., D.M., J.M. and P.H. were supported by
NIH MLPCN grant #MH084512.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmc.2013.08.017.
References and notes
1. Hanahan, D.; Weinberg, R. A. Cell 2000, 100, 57.
2. Cory, S.; Adams, J. M. Nat. Rev. Cancer 2002, 2, 647.
3. Gross, A.; McDonnell, J. M.; Korsmeyer, S. J. Genes Dev. 1999, 13, 1899.
4. Deng, J.; Carlson, N.; Takeyama, K.; Dal Cin, P.; Shipp, M.; Letai, A. Cancer Cell
2007, 12, 171.
5. Green, D. R.; Reed, J. C. Science 1998, 281, 1309.
6. Placzek, W. J.; Wei, J.; Kitada, S.; Zhai, D.; Reed, J. C.; Pellecchia, M. Cell Death
Dis. 2010, 1, e40.
7. Zhang, B.; Gojo, I.; Fenton, R. G. Blood 2002, 99, 1885.
8. Cho-Vega, J. H.; Rassidakis, G. Z.; Admirand, J. H.; Oyarzo, M.; Ramalingam, P.;
Paraquya, A., et al Hum. Pathol. 2004, 35, 1095.
9. Zhang, W.; Konopleva, M.; Ruvolo, V. R.; McQueen, T.; Evans, R. L.; Bornmann,
W. G., et al Leukemia 2008, 22, 808.
10. Hussain, S. A.; Cheney, C. M.; Johnson, A. J.; Lin, T. S.; Grever, M. R.; Caligiuri, M.
A., et al Clin. Cancer Res. 2007, 13, 2144.
11. Glaser, S. P.; Lee, E. F.; Trounson, E.; Bouillet, P.; Wei, A.; Fairlie, W. D., et al
Genes Dev. 2012, 26, 120.
12. Inoue, S.; Walewska, R.; Dyer, M. J. S.; Cohen, G. M., et al Leukemia 2008, 22,
819.
13. Pepper, C.; Lin, T. T.; Pratt, G.; Hewamana, S.; Brennan, P.; Hiller, L., et al Blood
2008, 112, 3807.
14. Cavarretta, I. T.; Neuwirt, H.; Zaki, M. H.; Steiner, H.; Hobisch, A.; Nemeth, J. A.,
et al Adv. Exp. Med. Biol. 2008, 617, 547.
15. Zhang, H.; Guttikonda, S.; Roberts, L.; Uziel, T.; Semizarov, D.; Elmore, S. W.,
et al Oncogene 2011, 30, 1963.
assay (0.3 lM and 1.1 lM, respectively, Table 2).
16. Boisvert-Adamo, K.; Longmate, W.; Abel, E. V.; Aplin, A. E. Mol. Cancer Res. 2009,
7, 549.
17. Zhou, P.; Qian, L.; Kozopas, K. M.; Craig, R. W. Blood 1997, 89, 630.
18. Beroukhim, R.; Mermel, C. H.; Porter, D.; Wei, G.; Raychaudhuri, S.; Donovan, J.,
et al Nature 2010, 463, 899.
19. Rudin, C. M.; Hann, C. L.; Garon, E. B.; Ribeiro de Oliveira, M.; Bonomi, P. D.;
Camidge, D. R., et al Clin. Cancer Res. 2012, 18, 3163.
20. Souers, A. J.; Leverson, J. D.; Boghaert, E. R.; Ackler, S. L.; Catron, N. D.; Chen, J.,
et al Nat. Med. 2013, 19, 202.
21. van Delft, M. F.; Wei, A. H.; Mason, K. D.; Vandenberg, C. J.; Chen, L.; Czabotar,
P. E., et al Cancer Cell 2006, 10, 389.
22. Yecies, D.; Carlson, N. E.; Deng, J.; Letai, A. Blood 2010, 114, 3304.
23. O’Brien, S. M.; Claxton, D. F.; Crump, M.; Faderl, S.; Kipps, T.; Keating, M. J., et al
Blood 2009, 113, 299.
24. Oliver, C. L.; Miranda, M. B.; Shangary, S.; Land, S.; Wang, S.; Johnson, D. E., et al
Mol. Cancer Ther. 2005, 4, 23.
25. McCoy, F.; Hurwitz, J.; McTavish, N.; Paul, I.; Barnes, C.; O’Hagan, B., et al Cell
Death Dis. 2010, 1, e108.
26. Vogler, M.; Weber, K.; Dinsdale, D.; Schmitz, I.; Schulze-Osthoff, K.; Dyer, M. J.
S., et al Cell Death Differ. 2009, 16, 1030.
27. Wilson, W. H.; O’Connor, O. A.; Czuczman, M. S.; LaCasce, A. S.; Gerecitano, J. F.;
Leonard, J. P.; Tulpule, A.; Dunleavy, K.; Xiong, H.; Chiu, Y. L.; Cui, Y.; Busman,
T.; Elmore, S. W.; Rosenberg, S. H.; Krivoshik, A. P.; Enschede, S. H.;
Humerickhouse, R. A. Lancet Oncol. 2010, 11, 1149.
28. Zhang, H.; Nimmer, P. M.; Tahir, S. K.; Chen, J.; Fryer, R. M.; Hahn, K. R.; Iciek, L.
A.; Morgan, S. J.; Nasarre, M. C.; Nelson, R.; Preusser, L. C.; Reinhart, G. A.;
Smith, M. L.; Rosenberg, S. H.; Elmore, S. W.; Tse, C. Cell Death Differ. 2007,
2007, 943.
29. Certo, M.; Del Gaizo Moore, V.; Nishino, M.; Wei, G.; Korsmeyer, S.; Armstrong,
S. A., et al Cancer Cell 2006, 9, 351.
30. Degterev, A.; Lugovskoy, A.; Cardone, M.; Mulley, B.; Wagner, G.; Mitchison, T.,
et al Nat. Cell Biol. 2001, 3, 173.
31. Ryan, J. A.; Burnelle, J. K.; Letai, A. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 12895.
32. Nikolovska-Coleska, Z.; Qiu, S.; Long, J.; Yang, C-Y.; Larsen, M.; Tomita, Y. et al.
Abstract 4148, Discovery and characterization of new small-molecule inhibitors
5. Conclusion
Here we have described the discovery and characterization of a
selective inhibitor of the Mcl-1 protein, a target central to intrinsic
apoptosis that has been implicated in a cancer. We demonstrated
that Mcl-1 inhibitor 9 displays efficacy in promoting death in a pa-
nel of cell lines derived from a variety of malignancies. Impor-
tantly, we highlighted the utility of the BH3 profiling assay in
confirming the mechanism of action of agents that modify the
Bcl-2 pathway. We showed that Mcl-1 inhibitor 9 functions in
the cellular context as a BH3 mimetic and that antiproliferative re-
sponse in a number of cell lines may be correlated with the extent
of mitochondrial priming. We also demonstrated that the Mcl-1
selectivity observed in biochemical assays may be translated to
cell-based assays by correlating Mcl-1 and Bcl-2 priming with
the degree of cytochrome c release and apoptosis induction. Com-
pounds of this class represent promising chemical probes for the
interrogation of Mcl-1 biology and we anticipate such tools will
be of great utility to the chemical biology community.
Identification of relevant companion diagnostics at the preclin-
ical research stage has been recognized as a critical component of
facilitating the development of personalized cancer therapeutics.
In addition to providing utility as a discovery tool by guiding
SAR, BH3 profiling may be utilized as a companion diagnostic test
for Mcl-1 inhibitors and other compounds which modulate the
intrinsic apoptosis pathway.
Please cite this article in press as: Richard, D. J.; et al. Bioorg. Med. Chem. (2013), http://dx.doi.org/10.1016/j.bmc.2013.08.017

8
D. J. Richard et al. / Bioorg. Med. Chem. xxx (2013) xxx–xxx
of Mcl-1 by high-throughput screening. AACR Annual Meeting, San Diego, CA,
April, 2008.
34. Chonghaile, T.; Sarosiek, K. A.; Vo, T.; Ryan, J. A.; Tammareddi, A.; Moore, V. D.
G., et al Science 2011, 334, 1129.
33. Acoca, S.; Cui, Q.; Shore, G. C.; Purisima, E. O. Struct. Funct. Bioinform. 2011, 79,
35. Vo, T.; Ryan, J.; Carrasco, R.; Neuberg, D.; Rossi, D. J.; Stone, R. M., et al Cell 2012,
151, 344.
2624.
Please cite this article in press as: Richard, D. J.; et al. Bioorg. Med. Chem. (2013), http://dx.doi.org/10.1016/j.bmc.2013.08.017