Combined Scaffold Evaluation and Systems-Level Transcriptome-Based Analysis for Accelerated Lead Optimization Reveals Ribosomal Targeting Spirooxindole Cyclopropanes

Article abstract of DOI:10.1002/cmdc.201900266

With evolutionary drug resistance impacting efforts to treat disease, the need for small molecules that exhibit novel molecular mechanisms of action is paramount. In this study, we combined scaffold-directed synthesis with a hybrid experimental and transc

Full text of DOI:10.1002/cmdc.201900266

Accepted Article
Title: Combined Scaffold Evaluation and Systems-level Transcriptome-
Based Analysis for Accelerated Lead Optimization Reveals
Ribosomal Targeting Spirooxindole Cyclopropanes
Authors: Kevin Rodriguez, Erin Howe, Emily Bacher, Miranda Burnette,
Jennifer Meloche, Jayda Meisel, Patricia Schnepp, Xuejuan
Tan, Mayland Chang, Jeremiah Zartman, Siyuan Zhang, and
Brandon Lee Ashfeld
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemMedChem 10.1002/cmdc.201900266
Link to VoR: http://dx.doi.org/10.1002/cmdc.201900266
A Journal of
www.chemmedchem.org

10.1002/cmdc.201900266
ChemMedChem
FULL PAPER
Combined Scaffold Evaluation and Systems-level Transcriptome-
Based Analysis for Accelerated Lead Optimization Reveals
Ribosomal Targeting Spirooxindole Cyclopropanes
Kevin X. Rodriguez,[a]† Erin N. Howe,[b]† Emily P. Bacher,[a] Miranda Burnette,[c] Jennifer L. Meloche,[a] Jayda
Meisel,[a] Patricia Schnepp,[b] Xuejuan Tan,[b] Mayland Chang,[a] Jeremiah Zartman,[c]* Siyuan Zhang[b]* and
Brandon L. Ashfeld[a]*
[a]
[b]
[c]
K. X. Rodriguez, E. P. Bacher, J. L. Meloche, J. Meisel, Prof. Dr. M. Chang, Prof. Dr. B. L. Ashfeld
Department of Chemistry and Biochemistry
University of Notre Dame
305 McCourtney Hall, Notre Dame, IN 46556 (USA)
E-mail: bashfeld@nd.edu
E. N. Howe, P. Schnepp, X. Tan, Prof. Dr. S. Zhang
Department of Biological Sciences
University of Notre Dame
100 Galvin Life Science Centere, Notre Dame, IN 46556 (USA)
Email: szhang8@nd.edu
M. Burnette, Prof. Dr. J. Zartman
Department of Chemical and Biological Engineering
University of Notre Dame
182 Fitzpatrick Hall, Notre Dame, IN 46556 (USA)
Email: jzartman@nd.edu
†These authors contributed equally to this work.
Supporting information for this article is given via a link at the end of the document.
Abstract: With evolutionary drug resistance impacting efforts to treat
disease, the need for small molecules that exhibit novel molecular
mechanisms of action is paramount. In this study, we have combined
identification of bis-spirooxindole cyclopropanes, inspired by 2-
oxindole alkaloid natural products.
Oxindole alkaloids are
a ubiquitous class of plant
scaffold-directed synthesis with
a
hybrid experimental and
metabolites that exhibit a diverse array of potential therapeutic
applications, including anti-mitotic activity in a number of cancer
models.7 Of these, the C3-spirooxindole has emerged as a
promising scaffold for chemotherapeutic design.8 Whether
naturally occurring or designed, carbo- and heterocyclic
spirooxindoles exhibit anti-viral, anti-cancer, and anti-
inflammatory properties (Figure 1). Owing to the functionally
diverse periphery and stereochemical density, many groups have
sought to establish efficient synthetic strategies for the rapid
construction of structurally diverse spirooxindole compound
collections.9
transcriptome analysis to identify bis-spirooxindole cyclopropanes
that inhibit cancer cell proliferation through disruption of ribosomal
function. These findings demonstrate the value of an integrated,
biologically-inspired synthesis and assay strategy for the accelerated
identification of first-in-class cancer therapeutic candidates.
Introduction
Although target-based drug discovery is the most widely
employed pharmacological strategy today,1 recent retrospective
analyses have revealed that more first-in-class small molecule
leads are discovered through phenotypic screening.1b,2 However,
a major challenge with this strategy is the lack of novel chemical
space, which ultimately is of greater importance than the size of
the library.3 Compared to in silico drug design, privileged core
scaffolds of evolutionarily conserved active natural products4 can
serve as a guide in diverting total synthesis to create structurally
related small molecule libraries for functionally targeted
phenotypical screens.4a,5 Therefore, phenotypic drug discovery
based on pharmacologically active natural products remains a
powerful strategy for rapidly evaluating the relevant chemical
landscape in search of novel therapeutic mechanisms of action.6
However, lead identification and guided chemical optimization
based on a defined molecular mechanism of action remains a
time-consuming process. Herein, we present an integrative
approach toward the synthesis and mechanism of action
Figure 1. Biologically active spirooxindole small molecular targets.
1
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900266
ChemMedChem
FULL PAPER
Results and Discussion
Scaffold Validation
To evaluate the chemical space within 2-oxindoles, we
synthesized an initial library of selected spirooxindoles bearing
dihydrobenzofuranyl 1, benzopyranyl 2, and cyclopropyl rings 3
and 4 at the C3-spirocenter (Figure 2). Oxindoles 1-4 were
screened for their anticancer activity using two isogenic human
triple negative breast cancer cell lines, MDA-MB-231, and the
brain metastatic derivative, MDA-MB-231-Br. Triple negative
breast cancers are so named because they fail to express
estrogen and progesterone receptors, and do not over-express
Her2. Thus, these cancer cell lines were chosen as they often do
not respond to hormone or anti-Her2 therapies, leaving
chemotherapy as the only treatment option for patients suffering
from this breast cancer subtype.
Representative dihydrobenzofuranyl and benzopyranyl
spirooxindoles 1a (R1 = R3= Me, R2 = H) and 2a (R1 = R3= Me, R2
= H) showed low activity (63-87 µM) at inducing MDA-MB-231- Br
cell death (Table 1). However, bisoxindole 2a proved more
effective against the less robust MDA-MB-231 cell line.
Comparatively, cyclopropyl oxindoles 3a (R1 = Ts, R2 = H, R3= o
BrC6H4, R4= Ph) and 4a (R1 = Ts, R2 = H, R3= o-BrC6H4, R4= Ts,
R5 = H) were more active against MDA-MB-231-Br cells than
either 1a or 2a, but showed poor activity against MDA-MB-231
Table 1. Representative lead spirooxindole activity.
Compound
R1 R2 R3
R4 R5 EC50 (µM)
MDA-MB-
231-Br
MDA-MB-
231
1a
2a
3a
4a
Me
Me
Ts
H
H
H
H
Me
–
–
–
–
–
48.8
54.0
30.2
30.1
45.9
8.2
Me
–
o-BrC6H4
o-BrC6H4
Ph
Ts
52.8
66.8
Ts
comparable to 1a. Given the more aggressive, metastatic nature
of MDA-MB-231-Br cells, subsequent efforts focused on the
spirocyclopropyl oxindole cores of 3a and 4a. Our second iteration
(Figure 2a, below dashed line) revealed that compounds bearing
the bis-spirooxindole cyclopropane
4 motif proved most
efficacious against both the MDA-MB-231-Br and MDA-MB-231
cell lines. While less cytotoxic than most current anticancer
therapeutics, this initial survey enabled lead identification of the
bisspirooxindole cyclopropane architectural core, and prompted
further investigation of compounds derived from this subset.
Chemistry and Pharmacology
CO2Me
R4
While synthetic efforts have focused on constructing C3-
spirofused 5- and 6-membered rings,10 not until recently has the
spirooxindole cyclopropane emerged as a scaffold for guided
drug design.11 In contrast to the rich history of 2-oxindoles, reports
of spirooxindole cyclopropanes as architectural templates for
therapeutic exploitation are limited.12 Based on parallels drawn to
known spirooxindole leads, as well as our initial compound
screening, we targeted a series of derivatives around a central
cyclopropane with (i) the 2-oxindole motif, (ii) an aromatic
substituent, and (iii) geminal donor/acceptor functionality as three
main points of variability (Figure 3).
R3
O
R2
R2
R3
R5
O
O
R4N
O
N
R1
N
R1
R3
1
3
R3
O
R2
R2
NR1
O
O
O
N
R1
N
R1
2
4
To access the structural analogs of cyclopropyl oxindoles 4,
we sought a convergent fragment coupling synthetic strategy that
enabled direct access to the core ring system while maintaining
the flexibility required for architectural diversification. To
accomplish this goal, we exploited a modification of the Kukhtin-
Ramirez condensation reaction utilizing a phosphine-mediated
cyclopropanation of β-aryl substituted alkylidene oxindoles with
Figure 2. Oxindole framework with various spirofused rings being investigated.
Identification of lead bis-spirooxindole framework 4 and log(EC50) values
observed when screened against MDA-MB-231-Br human breast cancer cells.
Figure 3. General molecular framework of the 3-spirocyclopropyl-2-oxindole
core architecture with highlighted points of structural diversity. Primary lead
compounds and EC50 values observed when screened for cytotoxicity against
the human breast cancer (brain-seeking) MDA-MB-231-Br cell line.
2
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900266
ChemMedChem
FULL PAPER
Additionally, 4b-e are stable in mouse plasma and rat liver S9
fractions, indicating that the compounds are metabolically stable.
RNA Sequencing and Molecular Mechanism of Action
Prediction
The characterization of the molecular mechanism of action for
novel compounds is a critical step in the drug discovery process.
The development of high-throughput genomic analyses,
combined with recently developed bioinformatic methods, can be
harnessed to yield insight into the putative mechanism of action
of a compound,1a as well as a global assessment of off-target
effects.2a,3b To establish an mechanism of action-guided
framework for chemical optimization, we selected oxindole 4b for
mechanism of action analysis (Figure 4). To identify time-
dependent transcriptome shifting induced by 4b, we treated MDA-
MB-231-Br cancer cells with vehicle control (DMSO) or 4b and
performed RNA-sequencing. Given that prolonged treatment with
4b induces apoptosis, we performed sequencing at 6, 12, and 24
hrs of treatment with 4b (Figure 4A). These early timepoints
provide insight into the transcriptome changes that occur in cells
prior to activation of apoptotic pathways. Early transcriptome
changes will be driven by the specific mechanism of action of the
compound, allowing bioinformatic analysis to provide insight into
the mechanism of action of 4b.
The identification of robust transcriptome level changes in
the majority of cells at early stages of treatment was ensured by
conducting compound exposure at five times the EC50 value
reported at 5 days post-treatment (Table 2; 62 µM). After 24 h
post-exposure, we identified 802 genes that were significantly
down-regulated, while 928 genes were significantly up-regulated
(p < 0.05). Expression of the top 400 differentially regulated genes
shows a time-dependent progressive change from vehicle to 24 h
treatment with 4b, with some gene clusters being progressively
down-regulated (Figure 4B, green) and others progressively up-
regulated (Figure 4B, pink). This suggests that treatment with 4b
induces a time-dependent transcriptome shift over the course of
24 hours.
Scheme 1. Syntheses of bisspirooxindole cyclopropanes.
isatin derivatives.13 Employing this strategy, we synthesized
bisoxindoles 4b and 4c in four steps beginning with the
condensation of 2-oxindole (5a) and aryl aldehydes 6a (X = F)
and 6b (X = Cl) to yield a mixture of alkylidenes 7a (E/Z = 1.5:1)
and 7b (E/Z = 1.8:1) respectively (Scheme 1). The resulting E/Z
mixtures were N-alkylated, and the alkylidene isomers separated
to provide the desired E-alkylidene oxindoles 8a (X = F) and 8b
(X = Cl). Separate treatment with N-Boc isatin (9) and P(NMe2)3
afforded bis-spirooxindole cyclopropanes 10a and 10b in 62%
and 90% yield respectively as separable mixtures of only two
diastereomers. Removal of the Boc group proceeded in
quantitative yield to provide racemic 4b and 4c. The bis-N-methyl
spiroxindole cyclopropanes 4b and 4e were synthesized following
a similar three step sequence starting from either 2-oxindole (5a)
or 5-methoxy-2-oxindole (5b) and employing N-methyl isatin in
the P(NMe2)3-mediated cyclopropanation. This design strategy
enabled us to assemble a collection of ~80 2-oxindole derivatives
with site specific point substitutions on the benzenoid ring of the
oxindole, donor/acceptor functional group variability, and
substitution around the remaining aryl ring. Of these derivatives,
compounds 4b-e were identified as optimized hits in subsequent
biological screenings.
Bioinformatic analysis of the transcriptome provides an
unbiased view of how cells respond to drug treatment, which
potentially infers the mechanism of action of
a
drug
candidate.4b,5f,6 Most small molecule therapeutics target proteins
directly, as opposed to moderating protein activity at the
transcriptional, or mRNA level. However, recent advances in gene
network analysis have demonstrated the predictive power of
transcriptome readouts for identifying the mechanism of action of
a compound. One such approach – detecting mechanism of
action by network dysregulation (DeMAND) analysis - predicts
Table 2. Representative lead spirooxindole activity
4b
4c
4d
4e
Compounds 4b-e have drug-like properties, as assessed by
Lipinsky’s rule of five (molecular weight <500, ClogP <5,
hydrogen-bond donors <5, hydrogen-bond acceptors <10).14
Substitution of fluorine (4b) for chlorine (4c) lowered ClogP and
increased the water solubility almost two-fold, and improved the
potency against metastatic breast cancer cells (MDA-MB-231-Br)
(Table 2).15 To evaluate blood-brain barrier permeability of
oxindoles analogs, we calculated the topical polar surface area
(tPSA). A tPSA range between 40-49 Å2 indicates that the
compounds may penetrate the blood-brain barrier.16 The average
tPSA for the top 25 central nervous system drugs is 47 Å.2,17
clogP
2.91
49.4
384
>2
3.48
49.4
401
>2
3.86
40.6
415
>2
3.78
49.8
445
>2
tPSA
MW (g/mol)
Plasma stability t1/2 (h)
Rat S9 t1/2 (h)
EC50 MDA-231-Br (µM)
EC50 MDA-231 (µM)
>1
>1
>1
>1
12
17
19
15
22
3
2
18
3
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900266
ChemMedChem
FULL PAPER
of 4b predicted by DeMAND (Figure 4D) consistently identified
the same set of effector genes across all time points. The top 25
ranked effectors within the dysregulated regulons were evident as
early as 6 h and persisted through 24 h post treatment (Figure
4B), suggesting a 4b-specific mechanism of action.
DeMAND identified 9831 genes as possible mechanism of
action mediators with a false discovery rate (FDR) of < 0.05, and
1732 genes with FDR q-value < 1E-5 (Table S7). To explore
common functionality of mechanism of action mediators, we
performed Gene Set Enrichment Analysis (GSEA).20 GSEA
utilizes lists of genes (gene sets), which are grouped by biological
activity, including common function, sequence homology, or
chromosomal location. Each gene set is compared to a ranked
input file, to determine the rank of each gene in the gene set. If
the input genes fall near the top of the list, the gene set is
positively enriched, if they fall near the bottom of the list, the gene
set is negatively enriched. For our data, genes were ranked by
their DeMAND predicted significance (inverse p-value) and
assigned directionality by their differential expression. GSEA
analysis identified 125 positively enriched gene sets (induced by
4b treatment, Table S8) and 231 negatively enriched gene sets
(inhibited by 4b treatment, Table S9) with FDR q-value < 1E-3.
Interestingly, the KEGG Ribosome pathway had the highest
negative enrichment score, -9.45 (Figure 5A), suggesting that this
gene set is significantly negatively regulated by 4b treatment.
Several of the top 15 negatively enriched gene sets represent
either the structural components of the ribosome (Figure 5A, dark
green bars), the function of the ribosome (Figure 5A, medium
green bars), or activities that are related to or dependent on
ribosomal function (Figure 5A, light green bars). Indeed, the top
25 DeMAND predicted genes exclusively encode ribosomal
proteins (Figure 4D), and 75% of the top 100 ranked genes
encode ribosomal proteins.21
To further explore the specificity of the predicted
mechanism of action of 4b, we curated custom gene sets that
represent distinct mechanism of action classes of various, well-
characterized anti-cancer therapeutics, namely 1) apoptosis, 2)
cell cycle regulation, 3) regulation of the ribosome, 4) disruption
of microtubule function, 5) DNA damage, 6) kinase activation, 7)
G-protein coupled receptor (GPCR) signaling, and 8) ion channel
signaling.22 Given that the majority of cytotoxic agents exhibit a
mechanism of action covered by one of these broad categories,
we determined the enrichment of each of these gene sets in the
DeMAND results. Analysis revealed that the top 100 predicted 4b
mechanism of action effector genes cluster strongly in the
ribosome regulation class, without overlapping with other
common drug mechanism of action gene sets (Figure 5B). This
strongly suggests that 4b cytotoxicity is due to inhibition of
ribosomal function.
Figure 4. RNA-sequencing and mechanism of action Prediction. A. Schematic of
experimental set-up. B. Heatmap depicting top 400 differentially up- (red) or down-
(blue) regulated genes between DMSO and 4b treatment conditions in MDA-MB-231-
Br cells. Gene clusters are annotated as being progressively up-regulated (green), or
progressively down-regulated (pink). C. Schematic of bioinformatic analysis method –
detecting mechanism of action by network dysregulation (DeMAND). D. Top 25 ranked
genes in the DeMAND output for each paired comparison. All: DMSO vs all 4b
treatment groups averaged; 6 h, 12 h, 24 h: paired comparison between each
individual time point and DMSO.
the mechanism of action by examining differential expression of
gene networks upon drug treatment, with the underlying
assumption that although a drug’s target protein may not be
changed at the transcriptional level, the network of interacting
proteins will be sufficiently dysregulated as to allow
identification.18 First, DeMAND defines the regulon of each gene
as a network of established interactions based on prior biological
knowledge, such as the STRING network (Figure 4C, each brown
square represent a single gene-level regulon).19 Next, DeMAND
maps the observed transcriptome changes from RNA-sequencing
analysis to the regulon of each gene (Figure 4C, Gn expression
represented by red/blue colorbar). DeMAND then uses the
dysregulation of regulons, rather than individual genes, to identify
putative cellular targets (effectors). In the example in Figure 4C,
although G10 is up-regulated with treatment, the entire regulon is
relatively unaffected, while G0 has a strongly dysregulated
regulon. Here, G0 would be ranked more highly as being involved
in the mechanism of action due to the dysregulation of effectors.
In an effort to predict the mechanism of action of the
spirocyclopropyl oxindoles, we input the RNA-sequencing data
obtained for 4b into the DeMAND algorithm. The DeMAND
analysis returns a ranked list of genes that corresponds to the
probability that a particular gene is an effector, based on the
dysregulation of that gene’s network. As 4b is a novel compound,
we had no knowledge a priori regarding the speed with which 4b
would induce transcriptional changes. Thus, analysis was
conducted in two ways: 1) each individual 4b treatment time point
(e.g. 6 h) was compared to the DMSO control; 2) the average
expression across all 4b treatment timepoints was compared to
the DMSO (labeled All in Figure 4D). Depicted based on rank, with
rank 1 being most likely to be an effector, the top 25 effector genes
Next, we sought to validate the effect of 4b on ribosomal
function by evaluating its impact on protein translation. The breast
cancer cell line MDA-MB-231-Br was treated separately with
DMSO (vehicle), doxorubicin (DNA intercalating agent), paclitaxel
(microtubule stabilizer), cyclohexamide (protein synthesis
inhibitor), and compounds 4b and 4c for 4 h, then subjected to a
fluorescence-based methionine incorporation assay to evaluate
the impact of each on protein synthesis (Figure 5C). Those cells
treated with 4b exhibited a decrease in protein synthesis
comparable to cells treated with the positive control
cycloheximide, and a substantial decrease in comparison to the
negative controls doxorubicin and paclitaxel (Figure 5C and 5D).
4
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900266
ChemMedChem
FULL PAPER
approach toward drug design, we identified
a series of
bisspriooxindole cyclopropanes that exhibit a ribosomal-targeting
mechanism of action toward limiting cancer cell proliferation.
These results suggest that our lead compounds exhibit an
intriguing mechanism of action of targeting ribosomal processes.
Since translation of mRNA into protein is an incredibly energy-
consumptive process, inhibition of translation is under
investigation as a promising cancer therapeutic strategy.23
Historically, defining the structure and function relationship of
pharmacologically active natural products has been a largely
empirical process that lacks molecular level insight. Biologically
inspired total synthesis, coupled with global bioinformatics
analysis of transcriptome networks, constitutes an alternative
strategy for the rapid identification of potent bioactive lead
candidates with clear molecular level details of mechanism of
action. Our study illustrates how an accelerated phenotypic drug
discovery process incorporating elements of target-directed
scaffold assembly and transcriptome bioinformatics analysis can
lead to the identification and biomolecular characterization of
therapeutic lead compounds.
Experimental Section
General. Solvents and reagents were ACS reagent grade and used
without further purification unless noted below. Dimethylformamide (DMF),
tetrahydrofuran (THF), dichloromethane (CH2Cl2) and diethyl ether (Et2O)
were passed through a column of molecular sieves and stored under argon.
All reactions were carried out in flame-dried glassware under an argon
atmosphere unless otherwise specified. Spirocyclopropyl oxindoles 3a, 4a,
alkylidene oxindoles 813a,24 and isatin derivatives 924b,25 were prepared
according to literature procedures and spectral data (1H NMR and 13C
NMR) were consistent with reported data. The relative stereochemistry of
4d was unambiguously assigned using X-ray crystallography, and the
relative stereochemistry for 10a, 10b, 4b-4e were assigned by comparison
of 1H NMR and 13C NMR data.
Figure 5. Inhibition of ribosomal function and translation by 4b. A. The rank of
DeMAND predicted effectors of 4b treatment in MDA-231-Br cells was used as input
for gene set enrichment analysis (GSEA). Top fifteen gene sets predicted by GSEA to be
negatively enriched following 4b treatment. Dark green bars, gene sets representing
components of the ribosome. Medium green bars, gene sets representing the function
of the ribosome. Light green bars, gene sets representing cellular processes related to
translation. White bars, gene sets not related to ribosomal structure or function. B.
Gene sets were compiled to represent the common drug MoA of current cancer
treatment options, and each MoA is assigned a color and symbol as shown top. For all
genes represented by these 8 common MoA classes that were in the top 10,000
DeMAND predicted effectors of 4b MoA, the DeMAND rank is shown. C. MDA-MB-231-
Br cells were treated for four hours and subjected to a fluorescent methionine
incorporation assay. Representative images are shown. Red: methionine incorporation
for protein synthesis. Blue: DAPI staining for nuclei. D. Quantitation of data shown in
C, methionine signal is normalized to number of cells per field as indicated by DAPI
staining. Boxes, first to third quartile range of 10 images per condition, line, mean,
whiskers, full range of data points. One-way ANOVA with Dunnett’s multiple
comparison test.
1H Nuclear magnetic resonance (NMR) spectra were obtained at 600 MHz,
and 13C NMR spectra at 100, 125 or 150 MHz. Chemical shifts are reported
in parts per million (ppm, δ), and referenced to residual solvent or
tetramethylsilane (TMS). Coupling constants are reported in Hertz (Hz).
Spectral splitting patterns are designated as s, singlet; d, doublet; t, triplet;
q, quartet; p, pentet; m, multiplet; comp, complex; app, apparent; hom,
higher order multiplet; and br, broad. Infrared (IR) spectra were obtained
using a Thermo Electron Nicolet 380 FT-IR using a silicon (Si) crystal in
an attenuated total reflectance (ATR) tower and reported as wavenumbers
(cm-1). High and Low resolution electrospray ionization (ESI)
measurements were made with a Bruker MicroTOF II mass spectrometer.
Analytical thin layer chromatography (TLC) was performed using EMD 250
micron 60 F254 silica gel plates, visualized with UV light and stained with
a p-anisaldehyde solution. Flash column chromatography was performed
according to Still’s procedure (Still, W. C.; Kahn, M.; Mitra, A. J. Org. Chem.
1978, 43, 2923) using EMD 40-63 μm 60Å silica gel.
The combination of advanced DeMAND transcriptome analysis
and data obtained from the protein synthesis assay strongly
suggests that spirocyclopropyl oxindole 4b rapidly targets
ribosome function, resulting in an mechanism of action that
includes inhibition of mRNA protein translation, which negatively
impacts the proliferation and subsequent survival of rapidly
dividing cancer cells.
General procedure for the synthesis of benzofuranylspirooxindoles.
Methyl lithium (0.45 mmol, 0.28 mL, 1.6 M) was dropwise added to a
solution of Boc-salicylaldehyde (66.7 mg, 0.30 mmol) in THF (3 mL) at -
78º C. After 5 minutes, hexamethylphosphoroustriamide (53.8 mg, 0.30
mmol) was slowly added after which a solution of N-isatin derivative (48.3
mg, 0.30 mmol) in 1 mL of THF was dropwise added to the reaction mixture.
The mixture was allowed to warm slowly to rt over a period of 2-2.5 h by
removal of the dry ice/acetone bath. The reaction was concentrated under
reduced pressure and the crude material was run under a small plug of
silica gel eluting with ethyl acetate and concentrated. The residue was
Conclusions
In summary, by employing an integrated spirooxindole scaffold
synthesis, phenotypic evaluation, and transcriptome analysis
5
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900266
ChemMedChem
FULL PAPER
purified by flash chromatography eluting with hexanes/ethyl acetate (3:1)
to provide 43.7 mg (55%) of 1a in a 2:1 diastereometric ratio as a light red
solid and 30.8 mg (25%) of 2a as a red solid in a 4:1 diastereometric ratio.
Diastereometric ratios were determined by 1H NMR or HPLC analysis.
(d, J = 7.8 Hz, 1 H), 7.77 (d, J = 7.8 Hz, 1 H), 7.33 (td, J = 7.2, 1.2 Hz, 1
H), 7.26 (comp, 1 H), 7.12 (td, J = 7.2, 1.2 Hz, 1 H), 7.09 (dd, J = 7.8, 1.2
Hz, 1 H), 7.05-7.03 (m, 2 H), 7.00-6.95 (m, 3 H), 6.85 (d, J = 7.8 Hz, 1 H),
4.40 (s, 1 H), 3.14 (s, 3 H), 1.60 (s, 9 H); 13C NMR (150 MHz, CDCl3) δ
169.8, 168.7, 163.1, 161.4, 148.7, 143.8, 140.1, 132.1, 132.1, 128.9, 128.4,
128.3, 125.7, 124.7, 124.7, 123.8, 122.8, 122.0, 120.2, 115.1, 115.0, 113.7,
107.6, 84.6, 48.3, 46.7, 39.5, 28.1, 26.5; IR (neat) 2989, 2933, 1763, 1726,
1712, 1609, 1513, 1466, 1353, 1294, 1251, 1148, 1089, 748 cm-1; HRMS
(ESI) m/z 507.1703 [C29H25FN2O4 (M+Na) requires 507.1691]; m.p. >
200˚C.
1´,3-Dimethyl-3H-spiro[benzofuran-2,3´-indolin]-2´-one
(1a).
Diastereoselectivity was determined by analysis of the 500 MHz 1H NMR
spectra (A: 4.01 (q, 1 H); B: 3.92 (q, 1 H)). 1H NMR (500 MHz, CDCl3)
Diastereomer A: δ 7.35 (t, J = 8 Hz, 1 H), 7.20-7.16 (comp, 2 H), 7.09 (d,
J = INSERT Hz, 1 H), 7.00-6.96 (m, 2 H), 6.88-6.86 (comp, 2 H), 4.01 (q,
J = 7.2 Hz, 1 H), 3.24 (s, 3 H), 1.13 (d, J = 7.2 Hz, 3 H), Diastereomer B:
δ 7.35 (t, J = 8 Hz, 1 H), 7.20-7.16 (comp, 2 H), 7.09 (d, J = 8 Hz, 1H),
7.00-6.96 (m, 2 H), 6.88-6.86 (comp, 2 H), 3.92 (q, J = 7.1 Hz, 1 H), 3.18
(s, 3 H), 1.34 (d, J = 7.1 Hz, 3 H); 13C NMR (125 MHz, CDCl3)
Diastereomer A: δ 175.75, 158.86, 143.89, 131.27, 130.65, 128.66,
125.39, 123.92, 122.97, 121.64, 110.13, 108.90, 90.15, 43.82, 26.74,
15.21, Diastereomer B: δ 175.75, 158.86, 143.89, 130.77, 128.76, 126.06,
124.34, 123.60, 123.53, 121.59, 110.13, 108.90, 90.15, 46.37, 27.90,
13.78; IR (neat) 2961, 2927, 2872, 2853, 1725, 1612, 1595, 1494, 1468,
1373, 1224, 1120, 970, 862 cm-1; HRMS (ESI) m/z 266.1173 [C17H15NO2
(M+H) requires 266.1176]; m.p. = 100-105 ˚C.
tert-Butyl-3´-(4-chlorophenyl)-1´´-methyl-2,2´´-dioxodispiro[indoline-
3,1´-cyclopropane-2´,3´´-indoline]-1-carboxylate
(10b).
Cyclopropanation of (E)-8b24b (53.9 mg, 0.20 mmol) with 9a25b (48.3 mg,
0.30 mmol), purified by flash chromatography eluting with hexanes/EtOAc
(4:1), provided 51 mg (51%) in a 2.3:1 diastereometric ratio of 10b as a
light yellow solid. Diastereoselectivity was determined by 1H NMR (500
MHz) analysis (A: 4.34 (s, 1 H); B: 4.36 (s, 1 H)). Major diastereomer: 1H
NMR (500 MHz, CDCl3) δ 8.03 (d, J = 7.8 Hz, 1 H), 7.67 (d, J = 8.2 Hz, 1
H), 7.35 (t, J = 7.8 Hz, 1 H), 7.29-7.22 (m, 4 H), 7.05 (d, J = 6.0 Hz, 1 H),
7.01 (d, J = 8.2 Hz, 2 H), 6.90 (t, J = 7.8 Hz, 1 H), 6.85 (d, J = 7.8 Hz, 1 H),
4.37 (s, 1 H), 3.22 (s, 3 H), 1.54 (s, 9 H); 13C NMR (125 MHz, CDCl3) δ
171.1, 167.3, 148.9, 144.6, 139.9, 133.8, 131.9, 129.0, 128.8, 128.6, 128.5,
127.9, 125.8, 123.6, 123.3, 121.9, 120.9, 114.2, 107.8, 84.6, 49.1, 46.4,
39.5, 28.3, 27.0; IR (neat) 3018, 2980, 1787, 1761, 1716, 1611, 1493,
1466, 1350, 1151 cm-1; HRMS (ESI) m/z 523.1384 [C29H25ClN2O4 (M+Na)
requires 523.1395]; m.p.= 199˚C.
1,1´´,4´-Trimethyldispiro[indoline-3,2´-chromane-3´,3´´-indoline]-
2,2´´-dione (2a). Diastereoselectivity was determined by analysis of the
500 MHz 1H NMR spectra (A: 4.49 (q, 1 H); B: 4.00 (q, 1 H)). 1H NMR (500
MHz, CDCl3) Diastereomer A: δ 7.37-7.31 (m, 2 H), 7.24-7.18 (m, 2 H),
7.06-7.00 (m, 2 H), 6.94 (t, J = 8.1 Hz, 2 H), 6.74 (t, J = 8.1 Hz, 2 H), 6.57
(t, J = 7.6 Hz, 1 H), 5.78 (d, J = 7.6 Hz, 1 H), 4.49 (q, J = 6.9 Hz, 1 H), 3.20
(s, 3 H), 2.88 (s, 3 H), 0.92 (d, J = 6.9 Hz, 3 H), Diastereomer B: δ 7.39-
7.28 (m, 4 H), 7.13 (t, J = 7.6 Hz, 1 H), 7.08-6.97 (m, 5 H), 6.87 (d, J = 7.8
Hz, 1 H), 6.83 (d, J = 7.8 Hz, 1 H), 4.00 (q, J = 6.5 Hz, 1 H), 3.14 (s, 3 H),
3.07 (s, 3 H), 0.92 (d, J = 6.5 Hz, 3 H), 13C NMR (125 MHz, CDCl3)
Diastereomer A: δ 174.85, 172.99, 152.56, 144.96, 144.57, 130.89,
129.03, 128.15, 127.28, 126.23, 126.06, 125.87, 124.90, 124.60, 123.07,
122.07, 122.02, 116.65, 108.33, 108.18, 78.94, 55.55, 30.68, 26.39, 26.30,
13.47, Diastereomer B: δ 173.82, 170.97, 152.77, 143.81, 143.64,
130.42, 128.74, 128.66, 127.75, 127.23, 126.32, 125.72, 125.53, 125.30,
123.12, 123.09, 122.65, 117.97, 108.31, 107.94, 80.75, 54.65, 33.21,
26.56, 26.54, 13.90;IR (neat) 2965, 1724, 1709, 1610, 1469 cm-1; HRMS
(ESI) m/z 411.1738 [C26H23N2O3 (M+H) requires 411.1703]; m.p. = >260
˚C.
3´-(4-Chlorophenyl)-1,1´´-dimethyldispiro[indoline-3,1´-
cyclopropane-2´,3´´-indoline]-2,2´´-dione (4d). Cyclopropanation of (E)-
8b24b (37.6 mg, 0.14 mmol) with N-methylisatin25a 9b (32.2 mg, 0.20 mmol),
purified by flash chromatography eluting with hexanes/EtOAc (3:1),
provided 52 mg (90%) in a 6:1 diastereometric ratio of 4d as a light yellow
solid. Diastereoselectivity was determined by 1H NMR (500 MHz) analysis
(A: 4.36 (s, 1 H); B: 4.73 (s, 1 H)). Major diastereomer: 1H NMR (500 MHz,
CDCl3) δ 8.06 (d, J = 10 Hz, 1 H), 7.34 (td, J = 1.0, 7.5 Hz, 1 H), 7.29-25
(m, 3 H), 7.18-7.13 (comp, 2 H), 7.02 (d, J = 10 Hz, 2 H), 6.91 (td, J = 1.0,
7.5 Hz, 1 H, 6.85 (dd, J = 7.5, 1.5 Hz, 2 H), 4.36 (s, 1 H), 3.23 (s, 3 H),
3.15 (s, 3 H); 13C NMR (125 MHz, CDCl3) 172.0, 169.6, 144.5, 144.0, 133.7,
132.0, 129.3, 128.6, 128.5, 128.4, 128.4, 126.0, 124.7, 121.9, 121.5, 121.5,
107.7, 47.8, 45.7, 39.8, 27.0, 26.7; IR (neat) 2981, 2923 2855, 1764, 1710,
1708, 1609, 1467, 1343, 1088, 748 cm-1; HRMS (ESI) m/z 437.1027
[C25H19ClN2O2 (M+Na) requires 437.1027]; m.p. = 100-105 ˚C.
General procedure for the synthesis of bisspirooxindole
cyclopropanes. A two-dram scintillation vial, equipped with a magnetic
stir bar, was charged with alkylidene oxindole (E)-8 (1 equiv) and isatin 9
(1.5 equiv). Dry, degassed CH2Cl2 (0.25 M) was added, the resulting
solution cooled to –78 ˚C and stirred for 5 min. Then P(NMe2)3 (1.5 equiv)
was added dropwise, and the mixture allowed to warm slowly to rt over a
period of 2-2.5 h by removal of the dry ice/acetone bath. The solution was
immediately concentrated under reduced pressure, and the resulting crude
mixture was purified by flash chromatography eluting with hexanes/EtOAc
at the indicated ratio to provide the title bis-spirooxindole cyclopropane 10.
3´-(4-Chlorophenyl)-5-methoxy-1,1´´-dimethyldispiro[indoline-3,1´-
cyclopropane-2´,3´´-indoline]-2,2´´-dione (4e). Cyclopropanation of (E)-
8b24b (37.6 mg, 0.14 mmol) with 5-methoxy-1-methylindoline-2,3-dione26
9c (32.2 mg, 0.20 mmol), purified by flash chromatography eluting with
hexanes/EtOAc (3:1), provided 30 mg (46%) in a 10:1 diastereometric ratio
of 4e as a green-gray solid. Diastereoselectivity was determined by 1H
NMR (400 MHz) analysis (A: 4.32 (s, 1H); B: 4.37 (s, 1H)). Major
diastereomer: 1H NMR (400 MHz, CDCl3) δ 7.81 (d, J = 4.0 Hz, 1H), 7.31-
7.27 (comp, 2H), 7.23 (dd, J = 0.8, 6.8 Hz, 1H), 7.06-7.04 (comp, 2H), 6.94
(td, J = 1.2, 8.0 Hz, 1H), 6.90 (d, J = 2.8 Hz, 1H), 6.88 (d, J = 2.4 Hz, 1H),
6.86 (d, J = 7.6 Hz, 1H), 6.76 (d, J = 8.4 Hz, 1H), 4.32 (s, 1H), 3.88 (s, 3H),
3.24 (s, 3H), 3.12 (s, 3H); 13C NMR (100 MHz, CDCl3): δ 171.96, 169.17,
155.42, 144.53, 137.59, 133.73, 132.01, 132.01, 129.39, 128.67, 128.41,
128.41, 128.37, 125.98, 121.63, 121.48, 113.48, 133.14, 107.84, 107.69,
56.21, 47.82, 45.69, 40.07, 27.05, 26.78; IR (neat) 2925, 2854, 1708, 1610,
1466 cm-1; HRMS (ESI) m/z 445.1328 [C26H21ClN2O3 (M+H) requires
445.1313]; m.p. >220˚C.
tert-Butyl-3´-(4-fluorophenyl)-1´´-methyl-2,2´´-dioxodispiro[indoline-
3,1´-cyclopropane-2´,3´´-indoline]-1-carboxylate
(10a).
Cyclopropanation of (E)-8a13a (152 mg, 0.60 mmol) with 9a25b (178 mg,
0.72 mmol), purified by flash chromatography eluting with hexanes/EtOAc
(4:1), provided 145 mg (50 %) in a 1:1 diastereometric ratio of 10a as a
light yellow solid. Diastereoselectivity was determined by 1H NMR (500
MHz) analysis (A: 4.38 (s, 1 H); B: 4.40 (s, 1 H)). Diastereomer A: 1H
NMR (600 MHz, CDCl3) δ 8.03 (d, J = 7.8, 1,2 Hz, 1 H), 7.76 (d, J = 8.4
Hz, 1 H), 7.33 (td, J = 7.8, 1.2 Hz, 1 H), 7.26 (td, J = 7.8, 1.2 Hz, 1 H), 7.23
(td, J = 7.8, 1.2 Hz, 1 H), 7.05-7.02 (comp, 2 H), 6.98-6.96 (comp, 2 H),
6.89 (td, J = 8.4, 1.2, Hz, 1 H), 6.84 (d, 7.2 Hz, 1 H), 4.38 (s, 1 H), 3.21 (s,
3 H), 1.54 (s, 9 H); 13C NMR (150 MHz, CDCl3) δ 171.0, 167.1, 163.0,
161.4, 148.7, 144.3, 139.6, 132.0, 131.9, 128.8, 128.5, 128.3, 125.6, 124.8,
124.8, 123.4, 123.1, 121.6, 120.7, 115.1, 115.0, 114.0, 107.6, 84.4, 49.0,
46.2, 39.3, 28.1, 26.8; Diastereomer B: 1H NMR (600 MHz, CDCl3) δ 7.93
3´-(4-Fluorophenyl)-1-methyldispiro[indoline-3,1´-cyclopropane-
2´,3´´-indoline]-2,2´´-dione (4b). A solution of trifluoroacetic acid (57 mg,
0.5 mmol, 38 μL) in CH2Cl2 (0.2 mL) was added dropwise to a stirred
solution of 10a (18.6 mg, 0.04 mmol) in CH2Cl2 (0.10 mL) at rt and stirred
for 3 h. The resulting solution was then diluted with CH2Cl2 (1 mL) and
neutralized with saturated aqueous NaHCO3 (2 mL), transferred to a
6
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900266
ChemMedChem
FULL PAPER
separatory funnel, and the layers separated. The aqueous phase was
extracted with CH2Cl2 (3 x 3 mL) and the combined organic extracts were
washed with saturated aqueous NaCl (4 mL) then concentrated under
reduced pressure to provide 15 mg (quant.) of 4b as a light yellow solid.
1H NMR (500 MHz, CDCl3) δ 8.24 (s, 1 H), 8.01 (d, J = 7.8 Hz, 1 H), 7.26-
7.23 (m, 2 H), 7.11-7.7.08 (m, 2 H), 7.07-7.05 (m, 2 H), 6.97 (m, 2 H), 6.88
(td, J = 7.8, 1.2 Hz, 1 H), 6.83 (d, J = 7.8 Hz, 1 H), 6.80 (d, J = 7.8 Hz, 1
H), 4.36 (s, 1 H), 3.22 (s, 3 H); 13C NMR (150 MHz, CDCl3) δ 171.6, 171.5,
163.0, 161.4, 144.3, 140.7, 132.1, 132.1, 128.9, 128.2, 128.2, 126.1, 125.3,
125.3, 125.1, 121.6, 121.2, 121.1, 115.0, 114.9, 109.1, 107.6, 48.0, 45.7,
39.6, 26.8; IR (neat) 3143, 3079, 3002, 1706, 1608, 1512, 1469, 1342,
1085, 833, 745 cm-1; HRMS (ESI) m/z 385.1369 [C24H18FN2O2 (M+1)
requires 385.1347]; m.p. > 220˚C.
RNA-sequencing and molecular mechanism of action prediction.
MDA-231-Br cells were plated in 6 well plates such that they would be 70%
confluent and allowed to adhere overnight. Cells were treated with 4b at
the five times the EC50 value, 62.5 µM. Cells were collected in triplicate at
6, 12 and 24 hours post treatment, and RNA extracted with the PureLink
kit (Invitrogen). Libraries for RNA sequencing were prepared according to
the SmartSeq2 protocol.28 Briefly, mRNA was specifically amplified using
a primer complimentary to the polyA tail, and a template-switching
oligonucleotide. This primer combination makes it possible to perform
subsequent polymerase chain reaction (PCR) amplifications with a single
primer set. Following first round amplification of whole mRNA sequences,
a
transposase-based system (Illumina Nextera XT DNA sample
preparation kit) is used for tagmentation, where each library is broken into
300-800 bp fragments, which are tagged on both ends. This product is
then PCR amplified using a bar-coded primer, so that multiple libraries can
be pooled for a single sequencing run. All 12 sequencing libraries were run
together in a single run on the Illumina MiSeq machine, using the V3 150
cycle kit, programmed for dual-end 75 bp reads. Following RNA
sequencing, transcripts were aligned to the genome, and read counts were
normalized using the TopHat and Cufflinks suite of tools.29 DeMAND
analysis was performed in R using the BioConductor software provided
with the original manuscript.18 One modification was made to the RNA-
sequencing data to facilitate the use of DeMAND. The DeMAND algorithm
was developed for use on microarray data, which is significantly noisier
than RNA-sequencing data due to the nature of the technology – reading
fluorescence versus counting transcripts. Microarray data will never yield
three identical values across biological replicates, while RNA-sequencing
can yield identical replicates, particularly for undetectable transcripts,
which will have zero counts across all replicates. To estimate the
probability that the change in gene expression between vehicle and
treatment is greater than would be expected randomly, DeMAND
compares probability density functions. The probability density functions
are generated by performed kernel density smoothing on the experimental
replicates for each gene. Kernel density smoothing cannot be applied to
three identical values; therefore, to facilitate the use of DeMAND to
analyze our RNA-sequencing data we added noise to our data in the form
of random values drawn from a normal distribution with mean = 0.1% of
the maximum read count, and standard deviation = 0.01% of the maximum
read count. We examined the robustness of the DeMAND algorithm to
0.01%, 0.1%, 1% and 10% noise, and verified that the addition of noise
did not alter the results. For analysis of broad drug mechanism of action
classes, multiple gene annotation sets were combined. Apoptosis: GO
biological process – Regulation of apoptosis, GO biological process –
Apoptotic program. Cell cycle: GO biological process – Regulation of the
cell cycle, Biocarta – Cell cycle pathway. Ribosome: KEGG – Ribosome,
GO biological process – Ribosome biogenesis and assembly. Microtubule
disruption: GO biological processes – Microtubule polymerization or
depolymerization, Reactome – Formation of tubulin folding intermediates,
GO biological processes – Microtubule cytoskeleton organization and
biogenesis. DNA damage: KEGG mismatch repair, KEGG non-
homologous end joining, KEGG nucleotide excision repair. Kinase
activation: GO biological processes – Activation of protein kinase activity,
GO molecular function – Kinase regulator activity. GPCR activation:
Biocarta – GPCR pathway, Reactome – Signaling by GPCR. Ion channel
signaling: GO molecular function – Extracellular ligand gated ion channel
activity, GO molecular function – Ion channel activity, GO molecular
function – Ligand gated channel activity. For each gene in the combined
gene set, the rank in the DeMAND. The genes predicted by DeMAND to
be affected by 4b treatment were ranked, with 1 being the gene whose
regulon was most significantly altered by treatment. The rank of each gene
in the combined gene sets was plotted.
3´-(4-Chlorophenyl)-1-methyldispiro[indoline-3,1´-cyclopropane-
2´,3´´-indoline]-2,2''-dione (4c). A solution of trifluoroacetic acid (35.1 mg,
0.31 mmol, 24 μL) in CH2Cl2 (0.12 mL) was added dropwise to a stirred
solution of 10b (11.2 mg, 0.02 mmol) in CH2Cl2 (0.10 mL) at rt and stirred
for 3 h. The resulting solution was then diluted with CH2Cl2 (1 mL) and
neutralized with saturated aqueous NaHCO3 (2 mL), transferred to a
separatory funnel, and the layers separated. The aqueous phase was
extracted with CH2Cl2 (3 x 3 mL) and the combined organic extracts were
washed with saturated aqueous NaCl (4 mL) then concentrated under
reduced pressure to provide 9 mg (quant.) of 4c as a white solid. 1H NMR
(500 MHz, CDCl3) δ 8.17 (s, 1 H), 8.05 (d, J = 7.8 Hz, 1 H), 7.32-7.28 (m,
4 H), 7.15 (d, J = 7.5 Hz, 2 H), 7.07 (d, J = 7.5 Hz, 2 H), 6.94 (td, J = 7.65,
1.0 Hz, 1 H), 6.89-6.85 (comp, 2 H), 4.39 (s, 1 H), 3.26 (s, 3 H); 13C NMR
(150 MHz, CDCl3) δ 171.7, 171.6, 144.5, 140.9, 133.7, 132.0, 129.0, 128.4,
128.4, 128.4, 128.3, 126.2, 125.9, 125.2, 124.1, 121.8, 121.3, 121.3, 112.4,
109.3, 107.8, 48.0, 45.9, 39.7, 29.8, 27.0 ;IR (neat) 3245(b), 3084, 2924,
1707, 1612, 1492, 1469, 1339, 1087; HRMS (ESI) m/z 423.0886
[C24H17ClN2O2 (M+Na) requires 423.0871]; m.p. > 220˚C.
Cell culture. Human breast cancer cell lines MDA-MB-231 and MDA-MB-
231-Br were cultured in DME/F12 with 10% fetal bovine serum (FBS), 2.5
mM L-Glutamine, 15 mM HEPES and penicillin (100 unit/mL)/streptomycin
(100 µg/mL). All cell lines were maintained and grown in a 37°C incubator
with 5% CO2. Parental MDA-MB-231 cells were purchased from ATCC.
The MDA-MB-231-Br cell line was developed by in vivo selection of brain
trophic sub-lines by a minimum of three rounds of intracardiac injection.
EC50 determination. To generate a randomized plate for compound
screening, all compounds were first solubilized in DMSO at 10 mM. 10 mM
compounds were serially 1:10 diluted into DMSO to span a concentration
range from 10 mM to 1 nM. An Eppendorf EpMotion 5075 robotic pipettor
followed a randomization program to dilute compounds 1:10 into DME/F12
medium without FBS. The randomized plate contained compounds in 10%
DMSO spanning a concentration range from 1 mM to 100 pM. Cells were
plated in 96 well tissue culture plates at a concentration of 1,000 cells/well
in 90 µL of medium, and given overnight to adhere. 10 µL of compound
was pipetted from the randomized plates to the cell plates, yielding a
working solution of 1% DMSO, with compound concentrations ranging
from 100 µM to 10 pM. Plates were incubated for four days, then fixed and
stained using a sulforhodamine B assay.27 Briefly, medium was removed
and cells were fixed in 10% tricholoracetic acid (Sigma-Aldrich) for one
hour. Plates were rinsed repeatedly in deionized water and dried, prior to
staining with a solution of 0.4% sulforhodamine B (Sigma-Aldrich) in 1%
acetic acid (Sigma-Aldrich). Following one hour of staining, plates were
rinsed 5 times in 1% acetic acid, and dried. 100 µL of 10 mM tris base was
used to solubilize each well. Plates were incubated on an orbital shaker
for 10-20 minutes to ensure full solubilization, and absorption at 554 nm
was read an a BioTek Synergy H1 plate reader. Deconvolution of the
randomization was performed in MATLAB (MathWorks), and EC50 values
were calculated by performing a non-linear four parameter curve fit in
Prism 6 (GraphPad Software). Three technical replicates were performed
for each experiment, and all experiments were repeated three times.
Protein assay. MDA-231-Br cells were plated in 96 well tissue culture
plates at a concentration of 10,000 cells/well in 90 µL of medium, and given
overnight to adhere. Treatments were prepared in medium at 10x EC25
concentrations and 10 µL was pipetted to the cell plates, yielding a working
solution of 1% DMSO, with EC25 concentrations – doxorubicin (1.14 µM),
paclitaxel (0.05 µM), 4b (62.3 µM). Cells were lysed in RIPA lysis buffer
(ThermoFisher, 89900), with the addition of 2% SDS, and protease and
7
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900266
ChemMedChem
FULL PAPER
phosphatase inhibitor (Life Technologies, 88668) according to the
manufacturer’s instructions. Lysate was assayed for protein concentration
using the Pierce BCA protein assay kit (ThermoFisher, 23225) according
to the manufacturer’s instructions.
[1]
(a) Heath, J. R.; Ribas, A.; Mischel, P. S. Nat. Rev. Drug Discovery 2016,
15, 204-216. (b) Swinney, D. C.; Anthony, J. Nat. Rev. Drug Discovery
2011, 10, 507-519.
[2]
[3]
(a) Scannell, J. W.; Blanckley, A.; Boldon, H.; Warrington, B. Nat. Rev.
Drug Discovery 2012, 11, 191-200. (b) Kotz, J. SciBX 2012, 5, 1-3.
(a) Harvey, A. L.; Edrada-Ebel, R.; Quinn, R. J. Nat. Rev. Drug Discovery
2015, 14, 111-129. (b) Wehling, M. Nat. Rev. Drug Discovery 2009, 8,
541-546. (c) Sukuru, S. C. K.; Jenkins, J. L.; Beckwith, R. E. J.; Scheiber,
J.; Bender, A.; Mikhailov, D.; Davies, J. W.; Glick, M. J. Biomol. Screen.
2009, 14, 690-699.
Phalloidin staining. At the protein assay endpoint, cells were stained
with Alexa Fluor 594 Phalloidin (ThermoFisher, A12381) according to the
manufacturer’s instructions. Briefly, cells were washed twice with PBS,
fixed in 4% PFA for ten minutes, washed twice with PBS, and
permeablized with 0.1% Triton-X for 5 minutes. Cells were washed twice
with PBS, blocked with 1% BSA in TBST for 5 minutes, and stained with
phalloidin in 1% BSA in TBST for 20 minutes. Cells were counterstained
with DAPI to show nuclei, washed twice with PBS and imaged in PBS.
Imaging was performed on an EVOS FL system (Life Technologies) with
a 20x objective.
[4]
[5]
(a) Wetzel, S.; Bon, R. S.; Kumar, K.; Waldmann, H. Angew. Chem., Int.
Ed. 2011, 50, 10800-10826. (b) Yildirim, M. A.; Goh, K.-I.; Cusick, M. E.;
Barabási, A.-L.; Vidal, M. Nat. Biotechnol. 2007, 25, 1119-1126.
(a) Kellenberger, E.; Hofmann, A.; Quinn, R. J. Nat. Prod. Rep. 2011, 28,
1483-1492. (b) Hert, J.; Irwin, J. J.; Laggner, C.; Keiser, M. J.; Shoichet,
B. K. Nat. Chem. Biol. 2009, 5, 479-483. (c) Quinn, R. J.; Carroll, A. R.;
Pham, N. B.; Baron, P.; Palframan, M. E.; Suraweera, L.; Pierens, G. K.;
Muresan, S. J. Nat. Prod. 2008, 71, 464-468. (d) McArdle, B. M.;
Campitelli, M. R.; Quinn, R. J. J. Nat. Prod. 2006, 69, 14-17. (e) Wess,
G.; Urmann, M.; Sickenberger, B. Angew. Chem., Int. Ed. 2001, 40,
3341-3350. (f) Debouck, C.; Metcalf, B. Annu. Rev. Pharmacool. Toxicol.
2000, 40, 193-207.
Protein synthesis assay. MDA-231 cells were plated on cover slips at
a concentration of 5,000 cells/slip in 90 µL of medium, and given overnight
to adhere. Treatments were prepared in medium at 5x EC50 concentrations
and 10 µL was pipetted to the cell plates, yielding a working solution of 1%
DMSO, with EC50 concentrations – doxorubicin (2 µM), paclitaxel (0.05
µM), 4b (100 µM). After treatment, the drug-containing medium was
removed and replaced with a solution of methionine-free media and Click-
iT® HPG (an alkyne-containing methionine derivative). The cells were
incubated for 30 minutes. Then, the Click-iT® HPG solution was removed,
and the cells were washed once with PBS. Then the cells were fixed with
3.7% formaldehyde in PBS, incubated for 20 minutes, washed twice with
3% BSA in PBS, permeabilized with 0.5% Triton in PBS, and incubated for
20 minutes. Newly synthesized protein was detected using the Click-iT™
HPG Alexa Fluor™ 488 Protein Synthesis Assay Kit (ThermoFisher
C10428) according to the manufacturer’s instructions.
[6]
[7]
Dai, Y.-F.; Zhao, X.-M. Biomed Res Int 2015, 3, 1-9.
(a) Imamura, S. Cell Mol. Neurobiol. 2013, 33, 129-135. (b) Yu, B.; Yu,
D.-Q.; Liu, H.-M. Eur. J. Med. Chem. 2015, 97, 673-698.
[8]
(a) Antonchick, A. P.; Gerding-Reimers, C.; Catarinella, M.; Schürmann,
M.; Preut, H.; Ziegler, S.; Rauh, D.; Waldmann, H. Nat. Chem. 2010, 2,
735-740. (b) Galliford, C. V.; Scheidt, K. A. Angew. Chem., Int. Ed. 2007,
46, 8748-8758. (c) Marti, C.; Carreira, Erick M. Eur. J. Org. Chem. 2003,
2003, 2209-2219. (d) Monteiro, Â.; Gonçalves, L. M.; Santos, M. M. M.
Eur. J. Med. Chem. 2014, 79, 266-272. (e) Rajesh, S. M.; Perumal, S.;
Menéndez, J. C.; Yogeeswari, P.; Sriram, D. MedChemComm 2011, 2,
626-630. (f) Schönhaber, J.; Müller, T. J. J. Org. Biomol. Chem. 2011, 9,
6196-6199. (g) Yeung, B. K. S.; Zou, B.; Rottmann, M.; Lakshminarayana,
S. B.; Ang, S. H.; Leong, S. Y.; Tan, J.; Wong, J.; Keller-Maerki, S.;
Fischli, C.; Goh, A.; Schmitt, E. K.; Krastel, P.; Francotte, E.; Kuhen, K.;
Plouffe, D.; Henson, K.; Wagner, T.; Winzeler, E. A.; Petersen, F.; Brun,
R.; Dartois, V.; Diagana, T. T.; Keller, T. H. J. Med. Chem. 2010, 53,
5155-5164.
Protein detection. At the protein assay endpoint, the permeabilization
solution was removed, and cells were washed twice with 3% BSA in PBS.
Click-iT® reaction cocktail (prepared according to the manufacturer’s
instructions) was added to each coverslip, and the cells were allowed to
incubate for 30 minutes, protected from light. The reaction cocktail was
removed, and the cells were washed once with Click-iT® reaction rinse
buffer. Cell DNA was stained with HCS NuclearMask™ Blue Stain, and the
cells were allowed to incubate for 30 minutes, protected from light. The
staining solution was removed, and the cells were washed twice with PBS.
Imaging with performed on an EVOS FL system (Life Technologies) with
a 20x objective.
[9]
(a) Ball-Jones, N. R.; Badillo, J. J.; Franz, A. K. Org. Biomol. Chem. 2012,
10, 5165-5181. (b) Cheng, D.; Ishihara, Y.; Tan, B.; Barbas, C. F. ACS
Catalysis 2014, 4, 743-762. (c) Xia, M.; Ma, R.-Z. J. Heterocycl. Chem.
2014, 51, 539-554.
[10] Pavlovska, T. L.; Redkin, R. G.; Lipson, V. V.; Atamanuk, D. V. Mol.
Divers. 2016, 20, 299-344.
[11] (a) Jiang, T.; Kuhen, K. L.; Wolff, K.; Yin, H.; Bieza, K.; Caldwell, J.;
Bursulaya, B.; Tuntland, T.; Zhang, K.; Karanewsky, D.; He, Y. Bioorg.
Med. Chem. Lett. 2006, 16, 2109-2112. (b) Jiang, T.; Kuhen, K. L.; Wolff,
K.; Yin, H.; Bieza, K.; Caldwell, J.; Bursulaya, B.; Wu, T. Y.-H.; He, Y.
Biorg. Med. Chem. Lett. 2006, 16, 2105-2108. (c) Sampson, P. B.; Liu,
Y.; Patel, N. K.; Feher, M.; Forrest, B.; Li, S.-W.; Edwards, L.; Laufer, R.;
Lang, Y.; Ban, F.; Awrey, D. E.; Mao, G.; Plotnikova, O.; Leung, G.;
Hodgson, R.; Mason, J.; Wei, X.; Kiarash, R.; Green, E.; Qiu, W.;
Chirgadze, N. Y.; Mak, T. W.; Pan, G.; Pauls, H. W. J. Med. Chem. 2015,
58, 130-146. (d) Stevens, F. C.; Bloomquist, W. E.; Borel, A. G.; Cohen,
M. L.; Droste, C. A.; Heiman, M. L.; Kriauciunas, A.; Sall, D. J.; Tinsley,
F. C.; Jesudason, C. D. Biorg. Med. Chem. Lett. 2007, 17, 6270-6273.
(e) Ye, N.; Chen, H.; Wold, E. A.; Shi, P.-Y.; Zhou, J. ACS Infect. Dis.
2016, 2, 382-392.
Acknowledgements
This work was supported by a Walther Cancer Foundation
Advancing Basic Cancer Research Program grant and the
National Science Foundation (CAREER CHE-1056242 and CHE-
1665440). E.P.B was supported by a K.X.R. was supported by a
Walther Cancer Foundation ENSCCII Training Grant. E.N.H was
supported by Indiana Clinical and Translational Sciences Institute
Training Grant TL1R001107, and 1F32CA210583. M.B. was
supported an American Cancer Society Institutional Research
Grant IRG-14-195-01. We thank the Mike and Josie Harper
Cancer Research Institute for their support.
[12] Reddy, C. N.; Nayak, V. L.; Mani, G. S.; Kapure, J. S.; Adiyala, P. R.;
Maurya, R. A.; Kamal, A. Biorg. Med. Chem. Lett. 2015, 25, 4580-4586.
[13] (a) Wilson, E. E.; Rodriguez, K. X.; Ashfeld, B. L. Tetrahedron 2015, 71,
5765-5775. (b) Zhou, R.; Yang, C.; Liu, Y.; Li, R.; He, Z. The Journal of
Organic Chemistry 2014, 79, 10709-10715.
Keywords: spirooxindoles • transcriptome network • drug
discovery • natural products • mechanism of action
[14] Lipinski, C. A. Drug Discov. Today Technol. 2004, 1, 337-341.
[15] Hewage, H. S.; Anslyn, E. V. J. Am. Chem. Soc. 2009, 131, 13099-13106.
[16] Pajouhesh, H.; Lenz, G. R. NeuroRx 2005, 2, 541-553.
[17] Hitchcock, S. A. Curr. Opin. Chem. Biol. 2008, 12, 318-323.
References:
8
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900266
ChemMedChem
FULL PAPER
[18] Woo, J. H.; Shimoni, Y.; Yang, W. S.; Subramaniam, P.; Iyer, A.; Nicoletti,
P.; Rodríguez Martínez, M.; López, G.; Mattioli, M.; Realubit, R.; Karan,
C.; Stockwell, B. R.; Bansal, M.; Califano, A. Cell 2015, 162, 441-451.
[19] Szklarczyk, D.; Franceschini, A.; Wyder, S.; Forslund, K.; Heller, D.;
Huerta-Cepas, J.; Simonovic, M.; Roth, A.; Santos, A.; Tsafou, K. P.;
Kuhn, M.; Bork, P.; Jensen, L. J.; von Mering, C. Nucleic Acids Res. 2015,
43, D447-D452.
[20] Subramanian, A.; Tamayo, P.; Mootha, V. K.; Mukherjee, S.; Ebert, B.
L.; Gillette, M. A.; Paulovich, A.; Pomeroy, S. L.; Golub, T. R.; Lander, E.
S.; Mesirov, J. P. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 15545-15550.
[21] Hess, S. T.; Girirajan, T. P. K.; Mason, M. D. Biophys. J. 2006, 91, 4258-
4272.
[22] (a) Hoelder, S.; Clarke, P. A.; Workman, P. Mol. Oncol. 2012, 6, 155-176.
(b) Moffat, J. G.; Rudolph, J.; Bailey, D. Nat. Rev. Drug Discovery 2014,
13, 588-602.
[23] Bhat, M.; Robichaud, N.; Hulea, L.; Sonenberg, N.; Pelletier, J.;
Topisirovic, I. Nat. Rev. Drug Discovery 2015, 14, 261-278.
[24] (a) Ivashkin, P.; Couve-Bonnaire, S.; Jubault, P.; Pannecoucke, X. Org.
Lett. 2012, 14, 2270-2273. (b) Li, C.; Guo, F.; Xu, K.; Zhang, S.; Hu, Y.;
Zha, Z.; Wang, Z. Org. Lett. 2014, 16, 3192-3195.
[25] (a) Li, J.; Wang, N.; Li, C.; Jia, X. Chem. Eur. J. 2012, 18, 9645-9650. (b)
Wille, G.; Steglich, W. Synthesis-Stuttgart 2001, 759-762. (c) Jie, H. H.;
Li, J.; Li, C. J.; Jia, X. S. Synlett 2012, 23, 2274-2278.
[26] Beauchard, A.; Ferandin, Y.; Frère, S.; Lozach, O.; Blairvacq, M.; Meijer,
L.; Thiéry, V.; Besson, T. Biorg. Med. Chem. 2006, 14, 6434-6443.
[27] (a) Voigt, W. Sulforhodamine B assay and chemosensitivity, 2005; Vol.
110. (b) Vichai, V.; Kirtikara, K. Nat Protoc 2006, 1, 1112-1116.
[28] Picelli, S.; Faridani, O. R.; Bjorklund, A. K.; Winberg, G.; Sagasser, S.;
Sandberg, R. Nat Protoc 2014, 9, 171-181.
[29] (a) Trapnell, C.; Roberts, A.; Goff, L.; Pertea, G.; Kim, D.; Kelley, D. R.;
Pimentel, H.; Salzberg, S. L.; Rinn, J. L.; Pachter, L. Nat Protoc 2012, 7,
562-578. (b) Trapnell, C.; Hendrickson, D. G.; Sauvageau, M.; Goff, L.;
Rinn, J. L.; Pachter, L. Nat. Biotechnol. 2013, 31, 46-53.
9
This article is protected by copyright. All rights reserved.

10.1002/cmdc.201900266
ChemMedChem
FULL PAPER
Entry for the Table of Contents
This study demonstrates the value of an integrated, biologically-inspired synthesis and assay strategy for the identification of potential
small molecule drug candidates. The combination of scaffold-directed synthesis in conjunction with a hybrid
experimental/transcriptome analysis successfully identified a subset of bis-spirooxindole cyclopropanes that were shown to inhibit
cancer cell proliferation through the disruption of ribosomal function.
10
This article is protected by copyright. All rights reserved.