Novel vitexin-inspired scaffold against leukemia

Article abstract of DOI:10.1016/j.ejmech.2018.01.004

Acute lymphoblastic leukemia (ALL) is the most common type of leukemia in children. Up to a quarter of ALL patients relapse and face poor prognosis. To identify new compound leads, we conducted a phenotypic screen using terrestrial natural product (NP) fr

Full text of DOI:10.1016/j.ejmech.2018.01.004

European Journal of Medicinal Chemistry 146 (2018) 501e510
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Research paper
Novel vitexin-inspired scaffold against leukemia
Taotao Ling, Walter Lang, Xiang Feng, Sourav Das, Julie Maier, Cynthia Jeffries,
Anang Shelat, Fatima Rivas*
Department of Chemical Biology and Therapeutics, St. Jude Children's Research Hospital, Memphis, TN 38105-3678, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Acute lymphoblastic leukemia (ALL) is the most common type of leukemia in children. Up to a quarter of
ALL patients relapse and face poor prognosis. To identify new compound leads, we conducted a
phenotypic screen using terrestrial natural product (NP) fractions against immortalized ALL cellular
models.
Received 23 June 2017
Received in revised form
4 December 2017
Accepted 1 January 2018
We identified vitexin, a flavonoid, as a promising hit with biological activity (EC50 ¼ 30
mM) in pre-B
cell ALL models with no toxicity against normal human tissue (BJ cells) at the tested concentrations.
To develop more potent compounds against ALL and elucidate its potential mode of action, a vitexin-
inspired compound library was synthesized. Thus, we developed an improved and scalable protocol
for the direct synthesis of 4-quinolone core heterocycles containing an N-sulfonamide using a one-pot
condensation reaction protocol. The newly generated compounds represent a novel molecular scaffold
against ALL as exemplified by compounds 13 and 15, which demonstrated EC50 values in the low
Keywords:
Vitexin
Quinolin-4(1H)-one
Cdk inhibitor
Reactive oxygen species
micromolar range (0.3e10 mM) with little to no toxicity in normal cellular models. Computational studies
support the hypothesis that these compounds are potential CDK inhibitors. The compounds induced
apoptosis, caused cell arrest at G0/G1 and G2/M, and induced ROS in cancer cells.
© 2018 Elsevier Masson SAS. All rights reserved.
1. Introduction
with anti-inflammatory, anti-histamic, anti-cancer and anti-
bradykinin among a broad range of other biological activities
Acute lymphoblastic leukemia (ALL) is the most common type of
leukemia in children [1]. Up to a quarter of ALL patients relapse and
face poor prognosis [2]. Relapsed ALL cases are associated with
initial poor response and chemotherapy resistance [1e4]. New
potent compounds are required to effectively treat relapsed ALL.
Our drug discovery program objective is to identify new chemical
scaffolds to advance the bench-to-bedside therapeutic agent
development process in high-risk and drug-resistant leukemia.
Based on a phenotypic screen using stable B cellular ALL models
(representatives of this high-risk patient cohort), we expected to
identify specific molecular scaffolds to expand our knowledge on
current treatment modalities against ALL. We identified the natural
product vitexin (1, Fig. 1) as a hit compound. Vitexin has been
isolated from several sources including Acer palmatum [5]. Vitexin
was reported as a potent hypotensive inhibitor of ganglion activity,
including antioxidant properties [5e7]. Some glycosylated flavo-
noids have a direct bond between the sugar and the anomeric
carbon (O-C bond), while others such as vitexin feature the sugar
bond at C6 or C8 (C-C bond). Studies of natural products have led to
the development of clinical candidates such as flavopiridol (Alvo-
cidib, 4, with FDA orphan drug designation to treat AML, Fig. 1),
flavonoid-like and FDA approved ATP-mimetic compound 5 (Fig. 1)
[8]. Mechanistic studies of flavonoid compounds suggest a two-
prone mode of action, CDK inhibition (particularly CDK9) and
various cellular effects from metabolic changes to antioxidant ac-
tivity [8]. Recently, the treatment of chronic lymphocytic leukemia
by flavopiridol in clinical trials was successfully reported [9].
Compounds 4 and 5 were used as control compounds in this study.
The antioxidant activity of flavonoids depends on the molecular
structure, the degree of hydroxylation/glycosylation, and the po-
sitions of hydroxyl groups as they provide resonance effects on the
aromatic ring for radical atom engagement. Some of the most
widely studied isoflavones are genistein (2, Fig. 1) and baicalein (3,
Fig. 1), which reacts with topoisomerase II via radical mechanism,
disrupting the maintenance of DNA stability, and wogonin induces
autophagy [10]. However, their biological effects are observed in
* Corresponding author. Department of Chemical Biology and Therapeutics, St.
Jude Children's Research Hospital, 262 Danny Thomas Place, Memphis, TN 38105-
3678, USA.
E-mail address: Fatima.rivas@stjude.org (F. Rivas).
https://doi.org/10.1016/j.ejmech.2018.01.004
0223-5234/© 2018 Elsevier Masson SAS. All rights reserved.

502
T. Ling et al. / European Journal of Medicinal Chemistry 146 (2018) 501e510
by acid or base (i.e. Conrad-Limpach or Gould-Jacobs reactions).
Once we identified 3, 5-dimethoxy aniline and 4-nitroaniline as the
synthetic building blocks for our 4-quinolone libraries, reaction
conditions were evaluated to access the desired core compounds
[11a]. Nucleophilic addition of 3, 5-dimethoxyaniline with aryl acid
chlorides in the presence of triethyl amine provided intermediate
A, which was treated with acyl chloride in the presence of tin
chloride, followed by condensation reaction mediated by potas-
sium t-butoxide under refluxing conditions (Scheme 1). The reac-
tion sequence involves the amide bond formation via Schotten-
Baumann reaction or aza-Michael addition followed by acylation/
condensation reaction to provide the functionalized cores for series
I-III in good yield. The intramolecular cyclization/condensation
reaction of the aniline with an electrophile (intermediate B, Scheme
1), was mediated by modified thermal cyclization reaction condi-
tions (diphenyl ether at 250 ꢀC, in one-pot reaction sequence).
Although, polyphosphoric acid can mediate such cyclization re-
actions, its corrosive nature and its high viscosity render work-up
and the purification process challenging [11].
Fig. 1. Vitexin and related molecular scaffolds.
For the synthesis of the core quinolones for series II and III,
substrate 3,5-dimethoxyaniline and 4-nitroaniline respectively
were treated with dimethyl but-2-ynedioate in the presence of
diphenyl ether to afford intermediate B, which was heated to form
the desired product via intramolecular cyclization reaction in one-
pot reaction sequence [11]. Consequently, the challenges posed by
the purification of the products were circumvented by cooling the
reaction vessel to RT, and rapidly adding hexane. The net effect of
this binary solvent mixture enabled the precipitation of the desired
product as a dark red powder, which was filtered and dried. Thus,
our one-pot protocol afforded the quinolone product in nearly
quantitative yield and avoided purification steps. To complete the
synthesis of series III core, the corresponding nitro compound was
reduced with Pd/C under a hydrogen atmosphere to yield the ex-
pected aniline product in excellent yield.
A preliminary survey of SAR was designed to assess potential
liabilities related to the following concerns a) similar quinolone
compounds display abroad range of biological properties, namely
antimicrobial, antiviral, antimalarial activity, and mediated cyto-
and geno-toxicity. Thus, our compounds were required to be
evaluated in normal tissue (BJ cellular model) to avoid globally
cytotoxicity, and b) comprehensive information regarding steric
and electronic factors at the C2-C3 centers to provide guidance in
generating a more appropriate quinolone system (Fig. 2). Com-
pounds 7e16 were synthesized to evaluate the role of an aromatic
versus the methyl ester at C-2, and the relevance of the introduc-
tion of the sulfonamide at N-1, followed by the evaluation of
halogen groups at C3 center. Compounds 16e21 were synthesized
to evaluate whether protecting groups influence biological activity.
To expand our evaluation of the molecular diversity that could
be introduced at N-1, we proceeded to focus on sulfonamide syn-
thesis (Fig. 3). Although numerous efforts have been made to
modify quinolones [11], there are currently few reports concerning
electrophilic substitutions at C-3 and direct N-modifications, which
are governed by the C2-C3 substituents of the quinolone. Direct N-
sulfonylation upon 2-aryl quinolone compound 7 was challenging
using conventional conditions such as sulfonyl chloride in the
presence of triethyl amine, as it provided O-sulfonylated compound
7b as the sole product. After extensive experimentation, we were
pleased to identify that copper mediated N-sulfonylation was
highly efficient for this quinolone substrate 7 with great functional
group tolerance, enabling the synthesis of quinolones with electron
withdrawing or donating groups on the aryl sulfonamide moiety.
Our findings extend the scope of utility for N-sulfonylation of
quinolones without C-3 electron withdrawing directing groups,
and also enable the direct N-sulfonylation on substituted anilines in
the high micromolar range as solubility/cellular permeability are
some of the obstacles for the chromen core compounds [6]. While
natural products (NPs) show potent bioactivities, their modest
bioavailability properties can be affected by the presence and
number of hydrophilic moieties (i.e. sugars, which can modulate
water solubility and cell permeability). Thus, the introduction of
heteroatoms and polar functional groups could lead to modified
NPs with improved bioavailability properties.
The lack of scaffold diversity among standard care agents
against leukemia and relapsed patients strongly supports the
investigation of new chemical matter. We developed a vitexin-
inspired library based on the hypothesis that a quinolone core (6,
Fig. 1) instead of the chromen core (2, Fig. 1) would display
improved biological activity against ALL since such compounds
would a) decrease overall electronegativity due to the nitrogen, b)
enable hydrogen bonding via donor rather than acceptor, and c)
provide a handle to introduce functional groups to survey the
chemical space. Thus, the resulting compounds would be more
potent than vitexin, potentially enabling the development of a new
class of chemical agents against high-risk and relapsed ALL.
Our studies describe an efficient synthesis, structure-activity
relationship (SAR) studies of quinolin-4(1H)-one derivatives
inspired by vitexin against ALL cellular models and the potential
biological target(s).
2. Results and discussion
The study began by developing suitable synthetic conditions to
generate the core synthons required for biological evaluation. On
the basis of their chemical structures, the quinolone derivatives
were divided into 3 group series (Scheme 1). Series I consisted of
derivatives with 2-(4-methoxyphenyl) quinolone and 6, 8-
dimethoxy on the A ring to explore the biological effect of the N-
1 substituents and potentially improve biological potency. Series II
derivatives focused on the role of C2 substituents via amide for-
mation and evaluation of the electronic effects of these sub-
stituents. Series III included the evaluation of C6 to explore
alternative functional groups to the methoxy groups and reduce
potential metabolic liabilities, while maintaining biological activity.
Thus, the combined group series would enable the generation of
the corresponding N-sulfonamides, 2-amino-benzenethiazoles
connected at C2 and urea moiety at C6 for a thorough biological
evaluation.
Synthetic methodologies to access 4-quinolone derivatives are
frequently based on intramolecular cyclization reactions mediated

T. Ling et al. / European Journal of Medicinal Chemistry 146 (2018) 501e510
503
Scheme 1. General synthesis to 4-quinolone core series I-III.
conditions provided poor yields so the acid chloride of compound
16 was generated following the classic Schotten-Baumann reaction
conditions [12]. Compound 16 was hydrolyzed and then treated
with thionyl chloride to generate the corresponding acyl chloride,
which was treated with the corresponding secondary amine sub-
strate to provide the desired compound 54 in good yield.
Next, an exploration of solubility-enhancing substituents (such
as the urea functionality) [13] at the C6 position of the aromatic ring
was an appropriate step to enable cellular permeability, and eval-
uate the influence of the substituents of this scaffold. In addition,
the role of the methoxy groups could be compared with other
heteroatoms. Compound 21 was treated with the corresponding
electrophiles in DMF at RT for 12 h, affording the desired quinolone-
urea compound in good yield (Fig. 5).
Fig. 2. Initial SAR evaluation: compounds 7e21.
2.1. Inhibition of proliferation by new compounds
a quinolone motif. The optimal experimental conditions for the
solvolysis of copper oxide required acetonitrile under refluxing
conditions for 16 h. Thus, compound 7 was treated with the cor-
responding aromatic sulfonyl chloride reagents to provide 7a as a
single compound; 7b was not detected by TLC or NMR analysis. The
reaction proceeded in good yields for compounds 22e38, which
were purified as single products.
A heat map of the anti-proliferative activity in ALL models
(KOPN8, SEM, SUP-B15 and UoCB-1) in comparison to normal tissue
models (BJ, THLE-2) using established CelTiter-Glo® cell viability
protocol [13] is shown in Table 1. While vitexin has been disclosed
to have activity against U937 lymphoma cellular model [7], we did
not observe cytotoxicity against Raji (Burkitt lymphoma cellular
Amides are one of the most ubiquitous and important functional
groups in natural and synthetic organic compounds so direct
amidation of the ester 16 was a synthetic efficient option. Several
conditions for the direct amidation were evaluated and found that
the sodium bis(trimethylsilyl) amide (NaHMDS) amidation condi-
tions were mild with high substrate generality for aromatic amines
[12]. Thus, we employed this nucleophilic amidation upon com-
pound 16 with a series of commercially available anilines. The
substrates were treated with sodium bis(trimethylsilyl) amide
(NaHMDS) in THF at RT. The base catalyzed amidation favored
electronically rich anilines, with modest yields were obtained for
electron deficient anilines. A plausible reaction mechanism is
shown in Fig. 4. The reaction was sensitive to sterically demanding
substrates or non-aromatic substrates, presumably not only due to
the electronics, but also by the encumbered approach of the
branched nucleophile. The reaction was allowed to stir for 24 h
under an inert atmosphere to provide the desired products in good
yields (39e53, Fig. 4). However, for compound 54, this amidation
model at the tested concentrations up to 100 mM using CelTiter-
Glo®). Several of these analogues displayed promising biological
activity, while compounds 19e25, 30e42, 46e53, 64e66, and 73
showed no cytotoxicity at the tested concentrations (see SI).
While several leukemia cellular models were utilized for our
viability assays, we focused our attention on compounds with
cytotoxicity against a particular subset of acute lymphoblastic
leukemia: KOPN-8, SEM, and SUP-B15 models, all of which carry
specific genomic lesions. KOPN-8 carries the MLL-ENL fusion gene,
SEM is a carrier of t(4, 11) fusion gene (MLL-AF4), and SUP-B15 was
established from a pediatric ALL relapsed patient (with the m-BCR
ALL variant of the BCR-ABL1 fusion gene) [14]. The gene for the
histone methyltransferase MLL participates in chromosomal
translocations that eventually create MLL-fusion proteins associ-
ated with very aggressive forms of childhood acute leukemia. The
presence of some MLL rearrangements is an independent dismal
prognostic factor and patients are usually treated according to
high-risk protocols [3,4]. Therefore, the identification of

504
T. Ling et al. / European Journal of Medicinal Chemistry 146 (2018) 501e510
Fig. 3. Synthesis of N-1 substituted compounds 22e38.
Fig. 4. Synthesis of C2 substituted quinolone compounds 39e54.
compounds against models with MLL gene fusions is necessary for
the discovery of new therapies.
While quinolones have been reported to exert various biological
activities, but mainly antibiotic properties [15], we questioned the
possibility that they were inhibiting CDK9. The selectivity and po-
tency of flavopiridol, 4 towards CDK9 has been attributed to the 4-
keto, 6-phenyl ether interacting at the CDK9 hinge as a hydrogen
donor and acceptor between amino acid Asp104 and Cys106. The

T. Ling et al. / European Journal of Medicinal Chemistry 146 (2018) 501e510
505
Fig. 5. Synthesis of C6 substituted urea compounds 55e73.
Table 1
Representative heat map of selected compounds.
remaining of the molecule can engage in
p
-stacking from the B ring
moderately small groups, or bulky groups (9-methyl-9H-fluorene)
to the urea, such as alkyl phenyl, heterocycles (furan, indole),
introduction of ethers (trimethoxy, dioxane) did not substantially
and hydrophobic interactions [13]. With a dihedral angle of 122ꢀ,
the chromen C ring is slightly different from our quinolone core
with an angle of 118ꢀ, which is forced to maintain the pyramidal
improve their biological profile (EC50 > 15 mM) against these
structure of the amine (Fig. 10A). However, the
a
-halogens influ-
cellular models.
ence the binding affinity of the enone to accept protons. In fact, the
dehalogenated compounds had weaker activity. However, the sul-
fonamides such as compound 18, are active in the low micromolar
The results indicate the methoxy groups at the A ring of the
molecule might not play a significant role in the biological activity
observed. It is also possible that our compounds (series I-III) might
not act on the same target as the SAR among them is not clearly
defined. Gratifyingly, our compound series provided compounds
(13, 15, 18 from series I, 54 from series II and 56e58 from series III)
with improved biological activity than vitexin. Further mechanistic
studies and fine tuning to improve solubility on these compounds
will be conducted.
range (5e6
mM) against KOPN-8 and SEM, and it showed less ac-
tivity (10 M) against SUP-B15. The urea compounds 54, 56e58 also
m
showed promising activity against ALL cellular models and further
mechanistic studies will be conducted to evaluate which amino
acids are interacting with the carbamide group. Current SAR among
these compounds shows an activity trend based on the placement
and nature of the substituents of the aromatic ring at the NH at C6
against KOPN-8 and SEM. A few of these compounds showed
selectivity for SUP-B15, such as compounds 29, 60, and 72. Com-
2.2. Compound 15 induces G0/G1 cell cycle arrest
pound 60 showed more promising activity (EC50 ¼ 3.2
mM) than
M) against SUP-B15. An interesting
Fluorochrome-labeled Annexin V and DNA content [16] with PI
were studied in treatment with compounds 15, 54, and 56 for a 24 h
period in KOPN-8 and SUP-B15 (see SI) to confirm apoptosis.
Representative images for negative control (DMSO), positive con-
compound 72 (EC50 ¼ 9.1
m
observation for these latter compounds is their lack of SAR. Thus,
further biochemical assays combined with structural biological
studies will be pursued to answer this question. For instance,
compound 55 having meta-fluoride substituents showed no ac-
trol (2.0 mM staurosporine) and compound 4 (5.0 mM) for 24 h
treatment are shown. Data presented on compounds 15, 54, and 56
(D-F, Fig. 6) clearly captured the increase of late apoptotic state (Q2)
upon compound treatment with a modest amount observed for
compound 54.
tivity against SUP-B15 at the tested concentrations (ꢁ30
mM) yet
the para-fluoride substituent 60, showed potency, selectivity to-
wards this cell line and therapeutic window in BJ cells. Addition of

506
T. Ling et al. / European Journal of Medicinal Chemistry 146 (2018) 501e510
Fig. 6. Representative of FITC Annexin V apoptosis, and Propidium Iodide (PI) staining after 24 h in KOPN-8 cellular model. A. vehicle control (DMSO). B. positive control (2
staurosporine) C. Compound 4 (5 M). D. Compound 15 (10 M). E. Compound 54 (10 M). F. Compound 56 (10 M).
mM
m
m
m
m
Next, DNA content of KOPN-8 cells was measured to interrogate
cell cycle effects compared with flavopiridol, 4. Proliferating cells
progress through various phases of the cell cycle (G0, G1, S, G2, and
M phase) and CDKs are important regulators of cell cycle [17]. No
serum starvation or chemical-induced synchronization methods
were used for ALL cellular models to avoid potential secondary
effects due to such manipulation [18], however, synchronization
was performed in a validated cellular model (MDA-MB231 cell
lines, see SI for data) to capture any differences. Graphical quanti-
tation of the respective cell cycle phases for compound 15, 54, and
cytotoxic staurosporine, but showed a slightly increase at the G2/M
phase, a phenomenon also observed for compound 5. Data are
shown from at least three independent experiments. Statistical
analysis of data was performed using Graph Pad Prism 5 and
Microsoft Excel software. The differences between the groups were
analyzed by t-test. Standard error bars represented the standard
deviation of the mean ( SD).
2.3. Cell death validation and potential biological targets
56 at 10e20
osporine and compound 4 as positive controls) for 24 h is shown in
Fig. 7. Cell cycle arrest occurred in the G0/G1 phase at 10 M con-
centration for compounds 15 and 56 as compared to controls. At
this concentration, ~15e20% of cells accumulated at G0/G1 phase,
indicating inhibition of the pre-replicative phase, and initiation of
apoptosis was also observed by Annexin V. The DMSO control in-
dicates that the cells were primarily at the S phase at the time of the
experiment and only a small number (10%) were entering the G2/
M. Compound 15 distributed the cells in a similar manner to the
mM versus controls (DMSO negative control, staur-
To further examine the effects of compound 15 in comparison to
clinical compounds 4, and 5, and to gain understanding on how
compound 15 causes cell death, additional cellular experiments
were conducted (Figs. 8 and 9). Because quinolone compounds
have the potential to react with glutathione, the precursor N-acetyl
cysteine (NAcCys) was also added to a set of cells to investigate
whether 1, 4-Michael addition reactions would prevent cell death.
m
As shown in Fig. 8, compound 15 (10e100 mM final concentra-
tion) caused cell death in a dose dependent manner as shown by
PARP-cleavage after 3 and 24 h. In the NAcCys treated cells, little to
no PARP cleavage was observed at 3 h for the higher concentrations,
presumably due to its anti-oxidant effect. However, it was detected
at the higher dose at 24 h.
Similarly compound 5 induced PARP cleavage after 3 h, but not
at 24 h. The addition of NAcCys in combination to compound 5 had
similar effects as compound 15. Compound 4 treated cells showed
consistent levels of PARP cleavage over time with or without anti-
oxidant. Then, the expression levels of the short-lived anti-
apoptotic factor Mcl-1 were investigated as blocking CDK9 removes
its ability to inhibit apoptosis [19]. Both compounds 4 and 5 sup-
pressed expression of Mcl-1, and compound 15 showed a similar
pattern in a dose dependent manner.
To interrogate the signaling of cyclin dependent kinases, we
turned our attention to the positive transcription elongation factor
b (P-TEFb, which consists of CDK9 and cyclin T1 or T2) complex. P-
TEFb is recruited to RNA polymerase II transcription initiation
complexes, where P-TEFb phosphorylates the transcription elon-
gation factors DRB sensitivity inducing factor (DSIF), the negative
elongation factor (NELF) and the C-terminal domain (CTD) of the
Fig. 7. KOPN-8 cell cycle distribution upon compound treatment. Compound 15 effects
on cell cycle were similar to compound 4, while compounds 54 and 56 had less effects
on the cycle compared to controls.

T. Ling et al. / European Journal of Medicinal Chemistry 146 (2018) 501e510
507
Fig. 8. Western analysis of KOPN-8 cellular line treated with compound 15 and control compound 4e5 for 3, and 24 h respectively.
Fig. 9. Reactive oxidative species (ROS) studies of compound 15. A. MCF7-Ruby red labeled (Cox8A-mitochondria) cells treated with 50 mM 15 alone. B. 50 mM 15 along with 50 mM
NAcCys for 1 h, followed by staining with CellROX® green reagent and nuclear stain Hoechst 33342 (blue). C. ROS relative quantification. (For interpretation of the references to
colour in this figure legend, the reader is referred to the Web version of this article.)
Fig. 10. Predicted binding mode of compound 15 to CDK9 using MOE program. A. Protein surface and compound 15. B. Top close view of interactions between compound 15
(magenta) and the hinge region of the ATP binding site, (ATP gray) and interacting amino acids (cyan). C. Surface representation of ATP pocket of CDK9, with compound 15 bound
(magenta). Red arrows indicate regions for chemical space exploration. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version
of this article.)
largest subunit of RNA polymerase II, thereby enabling transcrip-
tion complexes to transition into the productive elongation phase
[20e22]. Work by Chao has demonstrated inhibition of RNAPII
elongation by compound 4, resulting in abortive transcription of
most protein coding genes [22]. Both compounds 4 and 5 caused a
decrease in the levels of phosphorylated Ser-2 of the CTD after 3 h
with recovery recorded for compound 5, but not for compound 4
after 24 h. Compound 15 showed a decrease in Ser-2 phosphory-
lation CTD levels in a dose dependent manner at 24 h. The cMyc
expression decreased in dose dependent manner for compound 15,
with no substantial change for compound 4.
Finally, we evaluated the phosphorylation levels of the Retino-
blastoma tumor suppressor protein (Rb) critical to multiple cellular
activities related to cell cycle. Rb is rarely mutated in human can-
cers, instead regulatory pathways that alter Rb activity by phos-
phorylation at CDK sites are altered [22]. Recent studies indicate
that CDK4/6 activates Rb for binding cellular targets during early G1
phase, which results in Rb mono-phosphorylation to drive

508
T. Ling et al. / European Journal of Medicinal Chemistry 146 (2018) 501e510
quiescent G0 cells into a more metabolically active, but early G1
phase arrested phenotype [16]. The constitutively mono-
phosphorylated Rb neoplastic cell avoids cell cycle exit and differ-
entiation, so we evaluated changes in Rb phosphorylation in our
compound treated KOPN-8 cells. Little to no change was recorded
after 3 h, while less expression of total Rb and pRb was observed in
a dose dependent manner for compound 15 after 24 h treatment.
After 72 h, extensive cell death was recorded, limiting the detection
of these proteins (data not shown). At the 24 h time point, com-
pound 15 showed reduction in pRb in a dose dependent manner
similarly to the observed effects of compound 4. The combined
findings suggest that CDK9 is a potential target (s) of the described
compounds. However, further mechanistic studies are warranted to
confirm this speculated biological target.
Eleven Nineteen Leukemia (ENL) and AF4 proteins, common
associating partners of MLL in childhood acute leukemia are known
to bind and utilize P-TEFb for their transformation properties [21],
thus highlighting why compound such as 15 have better potency in
subtypes of ALL cellular models. This suggests that therapies tar-
geting P-TEFb activity in leukemia might be a direction to pursue
for refining precision therapeutics.
location as observed in the 1qmz model. A local minimization of
nearby residues was performed to facilitate the octahedral coor-
dination geometry between protein residues and magnesium ion.
The CDK2 with a derivative of flavopiridol (L868276) complex
shows that the inhibitor binds at the ATP binding site. The protein is
folded into the typical bilobal structure with the N-terminal
domain consisting predominantly of b-sheet structures, and the C-
terminal domain consisting mainly of alpha helices [26].
We predicted the interactions of compound 15 with the CDKs
could be driven by hydrophobic and van der Waals interactions
with the amino residues that form the pocket for the adenine base
in the ATP-CDK2 protein complex. To test such hypothesis, com-
pound 15 and compound 13 were docked into the cleft of CDK4 and
CDK6 using the Glide docking tool in Maestro [27]. However, both
CDK4 and CDK9 provided low docking scores indicating additional
effects from solvent or protein rearrangement might be important
contributors to the binding modes, which were not captured by our
solvent free flexible ligand docking studies.
Compounds 13 and 15 showed similar binding mode for CDK9,
which was lacking in the CDK4 studies, supporting our experi-
mental data. The docked compound 15 (magenta) in the binding
cleft of CDK9 overlaid on the ATP from our working template 1qmz
[26], is depicted in Fig. 10A. A close-up image (Fig. 10B) showcases
the docked ligand interactions with the hydrophobic residues L156,
V79, A166, A153 and the aliphatic chain of LYS48 in the CDK9 ho-
2.4. Compound 15 induces oxidative stress in live cellular models
A high level of oxidative phosphorylation represents a liability
to tumor cells due to the role of mitochondria in apoptosis, and
generation of ROS is carefully regulated [23]. Live cell CellROX®
experiments were conducted to evaluate if our lead compounds
trigger intrinsic apoptotic pathways by inducing reactive oxygen
species (ROS) since reduced Mcl-1 expression and CDK inhibition
had been recorded. CellROX® green reagent is a cell-permeable dye
with weak fluorescence while in a reduced state, and exhibits
bright green photostable fluorescence upon oxidation by ROS (with
absorption/emission maxima of 485/520 nm) [24]. A representative
experiment is shown in Fig. 9. The mRuby-Cox8A [25] labeled MCF-
7 cellular model was treated with compound 15 (Fig. 9A) or treated
with 15 along with ROS quencher, NAcCys for 1 h, followed by
CellROX® green reagent addition, and relative quantification
(Fig. 9B and C). Compound 15 clearly increased total ROS, particu-
larly mitochondria-derived ROS (orange, Fig. 9A). The cells exposed
to NAcCys were distinctly protected against ROS, as it presumably
increases glutathione levels, which bind to either the toxic break-
down products generated from 15, or directly to 15. To further
assess the latter possibility, LC-MS analysis of incubation of com-
pound 15 in DMSO with NAcCys or with glutathione for 1 h at 37 ꢀC
did not detect any potential 1,2 or 1,4 Michael addition products.
mology model. A hydrogen-p interaction between Ala153 and the
quinolone motif was observed while no significant overlap be-
tween the ATP adenosine ring and compound 15 were observed.
Furthermore, the sulfonyl group of this compound interacts
favorably with the magnesium ion. Additional molecular in-
teractions could be exploited at the indicated sites (red arrows,
Fig. 10C) to increase binding affinity. Our combined molecular
studies highlight the restricted ATP pocket as a potential region to
explore in order to improve selectivity and potency for our lead
compounds (13/15) as they bind differently to CDK9 from CDK4 in
the docked models. The differences in pose and docking scores
indicate the divergence in how these compounds interact with the
CDK9 cleft region, emphasizing opportunities for selectively tar-
geting these isoforms. Our studies also suggest the 2-methoxy
phenyl ring could grow into the pocket made by Thr62, Phe30
and Thr29 to improve its binding affinity (Fig. 10C). Furthermore,
the study shows that coordinating groups on the sulfonamide could
improve potency through chelation with the magnesium ion. The
methoxy group of the aromatic sulfonamide fits in the hydrophobic
pocket lined by F103 and V79, which can accommodate larger alkyl/
Ar groups to increase favorable interactions in this region.
Hydrogen bond acceptors or donors may also provide interactions
with the backbone of F105 and D104. Molecular dynamic studies
with explicit solvent are warranted to better understand their
binding mode and develop more potent compounds.
2.5. Molecular docking studies
Protein kinase structural studies have shown that the adenine
moiety of adenosine triphosphate (ATP) docks into a hydrophobic
cleft between the two lobes of the kinase domain of CDK4/6 and 9,
interacting with the kinase hinge region through hydrogen
bonding. Several inhibitors of these CDKs are known to bind at this
conserved nucleotide binding site. For instance, compound 4 es-
tablishes ATP-like hydrogen bond interactions with CDK2, CDK4/6,
and CDK9 in complex with cyclin T domain binding cleft by
computational docking studies coupled with co-crystallographic
data [26].
To elucidate whether our compounds were interacting with the
CDKs in a similar binding mode to compound 4 and 5, molecular
docking studies of these compounds with CDK4/6 and CDK9 were
performed. Homology models of CDK4 and CDK9 were built with
MOE from the ATP bound conformation of CDK2 (PDB code: 1qmz)
[26]. In addition, a magnesium ion was included in the same
2.6. Conclusion
In summary, a new series of 4-quinolone compounds inspired
by the natural product vitexin was developed using a simplified and
scalable synthetic protocol using readily available reagents. Com-
pounds 12e15 represent a novel anticancer sulfonamide scaffold
class with similar mechanism of action to CDK inhibitors and share
similar chemical properties to flavopiridol. As depicted in Table 1,
the halogenated 4-quinolon-N-sulphonamides displayed prom-
ising biological activity and selectivity towards specific ALL cellular
models. Our cellular evaluation indicate halogenated sulfonamides
(12e15) show anti-proliferative effects preferentially against MLL-
ALL cellular models, while the urea derivatives (55e73) displayed
anti-proliferative effects across most ALL cellular models tested

T. Ling et al. / European Journal of Medicinal Chemistry 146 (2018) 501e510
509
alleviating heat stress-induced cytotoxicity: regulatory effect of Hsp90 on ER
stress-mediated autophagy, Life Sci. 142 (2015) 36e48.
with therapeutic window in non-tumor cellular mode (BJ) of at
least 10 fold.
[6] a) H. Takahashi, M.C. Chen, H. Pham, E. Angst, J.C. King, J. Park, E.Y. Brovman,
H. Ishiguro, D.M. Harris, H.A. Reber, O.J. Hines, A.S. Gukovskaya, V.L. Go, G. Eibl,
Baicalein, a component of Scutellaria baicalensis, induces apoptosis by Mcl-1
down-regulation in human pancreatic cancer cells, Biochim. Biophys. Acta 13
(2011) 1465e1474;
This study is significant as: (a) we have developed a protocol to
access 4-quinolon-core scaffolds in one-pot reaction, followed by a
direct N-sulfonylation reaction of these quinolones which enables
the synthesis of a diverse group of compounds (b) while chroman-
flavones are well documented from a biological perspective, our 4-
quinolon-core offers a new scaffold that will serve as a chemical
probes to better understand high risk ALL cellular subtypes and
better understand the biological mechanisms that drive them, and
(c) our study illustrates the importance of natural products as
valuable tools for drug discovery as they have provide the foun-
dation for a targeted-driven medicinal chemistry program against
high risk ALL subtypes. The combined results warrant future
studies to elucidate the exact mechanism of action of these novel 4-
quinolones as selective agents against high risk ALL.
b) G. Wei, B. Yu, Isoflavone glycosides: synthesis and evaluation as
sidase inhibitors, Eur. J. Org. Chem. (2008) 3156e3163;
a-gluco-
c) G. Lewin, A. Maciuk, A. Moncomble, J.-P. Cornard, Enhancement of the
water solubility of flavone glycosides by disruption of molecular planarity of
the aglycone moiety, J. Nat. Prod. 76 (2013) 8e12.
[7] a) J. Zhou, H. Hu, J. Long, F. Wan, L. Li, S. Zhang, Y.E. Shi, Y. Chen, Vitexin 6, a
novel lignan, induces autophagy and apoptosis by activating the Jun N-ter-
minal kinase pathway, Anti Canc. Drugs 24 (2013) 928e936;
b) C.Y. Lee, Y.S. Chien, T.H. Chiu, W.W. Huang, C.C. Lu, J.H. Chiang, J.S. Yang,
Apoptosis triggered by vitexin in U937 human leukemia cells via a mito-
chondrial signaling pathway, Oncol. Rep. 28 (2012) 1883e1888;
c) Y. Zhou, Y.E. Liu, J. Cao, G. Zeng, C. Shen, Y. Li, M. Zhou, Y. Chen, W. Pu,
L. Potters, Y.E. Shi, Vitexins, nature-derived lignan compounds, induce
apoptosis and suppress tumor growth, Clin. Canc. Res. 15 (2009) 5161e5169;
d) J. Lee, T. Park, S. Jeong, K.-H. Kim, C. Hong, Hydroxychromones as cyclin-
dependent kinase inhibitors: synthesis and biological evaluation, Bioorg.
Med. Chem. Lett 17 (2007) 1284e1287.
Conflicts of interest
[8] J.K. Brunelle, et al., Loss of Mcl-1 protein and inhibition of electron transport
chain together induce anoxic cell death, Mol. Cell. Biol. 27 (2007) 1222e1235.
[9] C. LaCerte, V. Ivaturi, J. Gobburu, J.M. Greer, L.A. Doyle, J.J. Wright, J.E. Karp,
M.A. Rudek, Exposure-response analysis of alvocidib (Flavopiridol) treatment
by bolus or hybrid administration in newly diagnosed or relapsed/refractory
acute leukemia patients, Clin. Canc. Res. (2017). CCR-16e2629.
The authors declare that they have no competing interests.
Acknowledgments
[10] a) M. Russo, G.L. Russo, M. Daglia, P.D. Kasi, S. Ravi, S.F. Nabavi, S.M. Nabavi,
Understanding genistein in cancer: the “good” and the “bad” effects: a review,
Food Chem. 196 (2016) 589e600;
This study was supported by the American Lebanese Syrian
Associated Charities (ALSAC). The authors thank the Analytical
Technologies Center Core Facility for analytical support, the Flow
and Cell Cycle Facility and the Cell and Tissue Imaging Center Mi-
croscopy for technical assistance at St. Jude Children's Research
Hospital.
b) S.-J. Li, S.-L. Sun, J. Gao, F.-B. Sun, Wogonin induces Beclin-1/PI3K and
reactive oxygen species-mediated autophagy in human pancreatic cancer
cells, Oncol. Lett. 12 (2016) 5059e5067;
ꢀ
c) A. Uifalean, S. Schneider, C. Ionescu, M. Lalk, C.A. Iuga, Soy isoflavones and
breast cancer cell lines: molecular mechanisms and future perspectives,
Molecules 21 (2015), E13;
€
d) G. Polier, J. Ding, B.V. Konkimalla, D. Eick, N. Ribeiro, R. Kohler, M. Giaisi,
T. Efferth, L. Desaubry, P.H. Krammer, M. Li-Weber, Wogonin and related
natural flavones are inhibitors of CDK9 that induce apoptosis in cancer cells by
transcriptional suppression of Mcl-1, Cell Death Dis. 21 (2011), 2:e182.
[11] a) D.H.A. Rocha, D.C.G.A. Pinto, A.M.S. Silva, Synthesis and cyclisation studies
of (E)-2-aryl-1-methyl-3-styrylquinolin-4(1H)-ones, Tetrahedron 71 (2015)
7717e7721;
Appendix A. Supplementary data
Supplementary data related to this article can be found at
https://doi.org/10.1016/j.ejmech.2018.01.004.
b) X. Gu, G.I. Georg, Regioselective C5-alkylation and C5-methylcarbamate
formation of 2,3-dihydro-4-pyridones and C3- alkylation and C3-
methylcarbamate formation of 4-(pyrrolidin-1-yl)furan-2(5H)-one, Tetrahe-
dron Lett. 56 (2015) 5874e5877;
c) R. Cincinelli, L. Musso, G. Beretta, S. Dallavalle, 4-Quinolone fused hetero-
cyclic ring systems byintramolecular reactions of 4-quinolone-2-
carboxamides, Tetrahedron 70 (2014) 9797e9804;
d) J.-G. Kim, D.O. Jang, Mild and efficient indium metal catalyzed synthesis of
sulfonamides and sulfonic esters, Synlett 16 (2007) 2501e2504;
e) M.E. Rodriguez, D.A. Mareska, J.J. Hans, D.M. Harvey, R.D. Groneberg, M.
O'sullivan (2010). Piperidine-containing compounds and use thereof, Google
Patents WO 2010080864 A1. f) G.A. Meshram, D. Patil, Vishvanath, A simple
and efficient method for sulfonylation of amines, alcohols and phenols with
cupric oxide under mild conditions, Tetrahedron Lett. 50 (2009) 1117e1121;
g) D. Zewge, C.-Y. Chen, C. Deer, P.G. Dormer, D.L. Hughes, A mild and efficient
synthesis of 4-quinolones and quinolone heterocycles, J. Org. Chem. 72 (2007)
4276e4279;
References
[1] a) G. Tallen, R. Ratei, G. Mann, et al., Long-term outcome in children with
relapsed acute lymphoblastic leukemia after time-point and site-of-relapse
stratification and intensified short-course multidrug chemotherapy: results
of trial ALL-REZ BFM 90, J. Clin. Oncol. 28 (2010) 2339e2347;
b) Cancer Trends Progress Reporte 2009/2010 Update, National Cancer
Institute, NIH, DHHS, Bethesda, MD, April 2013. http://progressreport.cancer.
gov;
c) The Leukemia & Lymphoma Society, (Accessed 10 February 2017). http://
www.lls.org/.
[2] a) C.H. Pui, W.E. Evans, Treatment of acute lymphoblastic leukemia, N. Engl. J.
Med. 354 (2006) 166e178;
b) C.G. Mullighan, et al., Genomic analysis of the clonal origins of relapsed
acute lymphoblastic leukemia, Science 322 (2008) 1377e1380.
[3] a) C. Meyer, J. Hofmann, T. Burmeister, et al., The MLL recombinome of acute
leukemias in 2013, Leukemia 27 (2013) 2165e2176;
h) C.P. Jones, K.W. Anderson, S.L. Buchwald, Sequential Cu-catalyzed amida-
tion-base-mediated camps cyclization:
a two-step synthesis of 2-Aryl-4-
b) A.K. Anderson, J. Ma, J. Wang, et al., The landscape of somatic mutations in
infant MLL-rearranged acute lymphoblastic leukemias, Nat. Genet. 47 (2015)
330e337;
c) R. Marschalek, Mechanisms of leukemogenesis by MLL fusion proteins, Br. J.
Haematol. 152 (2010) 141e154;
d) A.K. Anderson, J. Ma, J. Wang, et al., The landscape of somatic mutations in
infant MLL-rearranged acute lymphoblastic leukemias, Nat. Genet. 47 (2015)
330e337.
quinolones from o-halophenones, J. Org. Chem. 72 (2007) 7968e7973;
ꢁ
ꢁ
i) I. Borza, S. Kolok, K. Galgoczy, A. Gere, C. Horvath, S. Farkas, I. Greiner,
ꢁ
G. Domany, Kynurenic acid amides as novel NR2B selective NMDA receptor
antagonists, Bioorg. Med. Chem. Lett. 17 (2007) 406e409;
j) D. Ding, X. Li, X. Wang, Y. Du, J. Shen, Microwave-assisted rapid and
straightforward synthesis of 2-aryl-4-quinolones from acylated 20-amino-
acetophenones, Tetrahedron Lett. 47 (2006) 6997e6999.
[12] a) D.A. Nugiel, J.R. Krumrine, D.C. Hill, J.R. Damewood Jr., P.R. Bernstein,
C.D. Sobotka-Briner, J.W. Liu, A. Zacco, M.E. Pierson, De novo design of a
picomolar nonbasic 5-HT1B receptor antagonist, J. Med. Chem. 53 (2010)
1876e1880;
[4] a) E. Etienne De Braekeleer, et al., Identification of MLL partner genes in 27
patients with acute leukemia from a single cytogenetic laboratory, Mol. Oncol.
5 (2011) 555e563;
b) M. Dordelmann, A. Reiter, et al., Prednisone response is the strongest
predictor of treatment outcome in infant acute lymphoblastic leukemia, Blood
94 (1999) 1209e1217.
b) R.S. Varma, K.P. Naicker, Solvent-free synthesis of amides from non-
enolizable esters and amines using microwave irradiation, Tetrahedron Lett.
40 (1999) 6177;
[5] a) M. He, J.W. Min, W.L. Kong, X.H. He, J.X. Li, B.W. Peng, A review on the
pharmacological effects of vitexin and isovitexin, Fitoterapia 115 (2016)
74e85;
ꢁ
c) L. Perreux, A. Andre Loupy, M. Delmotte, Microwave effects in solvent-free
esters aminolysis, Tetrahedron 59 (2003) 2185e2189.
[13] a) I.H. Kim, H.J. Tsai, K. Nishi, T. Kasagami, C. Morisseau, B.D. Hammock, 1,3-
disubstituted ureas functionalized with ether groups are potent inhibitors of
the soluble epoxide hydrolase with improved pharmacokinetic properties,
J. Med. Chem. 50 (2007) 5217e5226;
b) J.H. Kim, B.C. Lee, J.H. Kim, G.S. Sim, D.H. Lee, K.E. Lee, Y.P. Yun, H.B. Pyo, The
isolation and antioxidative effects of vitexin from Acer palmatum, Arch Pharm
Res. 28 (2005) 195e202;
c) M. Bhardwaj, S. Paul, R. Jakhar, S.C. Kang, Potential role of vitexin in

510
T. Ling et al. / European Journal of Medicinal Chemistry 146 (2018) 501e510
b) K. Mitachi, Y.G. Salinas, M. Connelly, N. Jensen, T. Ling, F. Rivas, Synthesis
b) V. Krystof, S. Baumli, R. Fürst, Perspective of cyclin-dependent kinase 9
(CDK9) as a drug target, Curr. Pharmaceut. Des. 18 (2012) 2883e2890.
[20] S.H. Chao, D.H. Price, Flavopiridol inactivates P-TEFb and blocks most RNA
polymerase II transcription in vivo, J. Biol. Chem. 276 (2001) 31793e31799.
[21] a) A. Yokoyama, M. Lin, A. Naresh, I. Kitabayashi, M.L. Cleary, A higher-order
complex containing AF4- and ENL-family proteins with P-TEFb facilitates
oncogenic and physiologic MLL-dependent transcription, Canc. Cell 17 (2010)
198e212;
and structure-activity relationship of disubstituted benzamides as a novel
class of antimalarial agents, Bioorg. Med. Chem. Lett. 22 (2012) 4536e4539;
c) J.H. Zhang, T.D. Chung, K.R. Oldenburg, A simple statistical parameter for
use in evaluation and validation of high throughput screening assays,
J. Biomol. Screen 4 (1999) 67e73.
[14] H.G. Dexler, Guide to Leukemia and Lymphoma Cell Lines, second ed., 2010.
German Collection of microorganisms and cell cultures Braunschweig,
Germany.
[15] a) M.L. Bolognesi, R. Banzi, M. Bartolini, A. Cavalli, A. Tarozzi, V. Andrisano,
A. Minarini, M. Rosini, V. Tumiatti, C. Bergamini, R. Fato, G. Lenaz, P. Hrelia,
A. Cattaneo, M. Recanatini, C. Melchiorre, Novel class of quinone-bearing
polyamines as multi-target-directed ligands to combat Alzheimer's disease,
J. Med. Chem. 50 (2007) 4882;
b) E.A. Bowman, W.G. Kelly, RNA polymerase II transcription elongation and
Pol II CTD Ser2 phosphorylation: a tail of two kinases, Nucleus 5 (2014)
224e236.
[22] A.M. Narasimha, M. Kaulich, G.S. Shapiro, Y.J. Choi, P. Sicinski, S.F. Dowdy,
Cyclin D activates the Rb tumor suppressor by mono-phosphorylation, Elife 3
(2014) 1e21.
b) Y.X. Yu, J.M. Hill, G. Yu, Y. Yang, A.F. Kluge, D. Keith, J. Finn, P. Gallant,
J. Silverman, A. Lim, A series of quinoline analogues as potent inhibitors of C.
albicans prolyl tRNA synthetase, Bioorg. Med. Chem. Lett. 11 (2001) 541.
[16] a) M.B. Kastan, J. Bartek, Cell-cycle checkpoints and cancer, Nature 432 (2004)
316e323;
[23] W.C. Burhans, M. Weinberger, M.A. Marchetti, L. Ramachandran, G. D'Urso,
J.A. Huberman, Apoptosis-like yeast cell death in response to DNA damage
and replication defects, (ROS) Mutat. Res. 532 (2003) 227e243.
[24] C.A. Davison, et al., Antioxidant enzymes mediate survival of breast cancer
cells deprived of extracellular matrix, Canc. Res. 73 (2013) 3704e3715.
[25] MRuby-Mito-7 was obtained from Addgene plasmid # 55874 (a gift from
Michael Davidson).
b) C. Riccardi, I. Nicoletti, Analysis of apoptosis by propidium iodide staining
and flow cytometry, Nat. Protoc. 1 (2006) 1458e1461;
c) L. Galluzzi, et al., Cell Death Differ. 19 (2012) 107e120.
[17] a) C.J. Sherr, F. McCormick, The RB and p53 pathways in cancer, Canc. Cell 2
(2002) 103e112;
b) Y.J. Choi, L. Anders, Signaling through cyclin D-dependent kinases, Onco-
gene 33 (2014) 1890e1903.
[18] U. Kim, C.W. Shu, K.Y. Dane, P.S. Daugherty, J.Y. Wang, H.T. Soh, Selection of
mammalian cells based on their cell-cycle phase using dielectrophoresis, Proc.
Natl. Acad. Sci. U. S. A 104 (2007) 20708e20712.
[26] a) S. Baumli, G. Lolli, E.D. Lowe, S. Troiani, L. Rusconi, A.N. Bullock,
J.E. Debreczeni, S. Knapp, L.N. Johnson, The structure of P-TEFb (CDK9/cyclin
T1), its complex with flavopiridol and regulation by phosphorylation, EMBO J.
27 (2008) 1907e1918;
b) N.R. Brown, M.E. Noble, J.A. Endicott, L.N. Johnson, The structural basis for
specificity of substrate and recruitment peptides for cyclin-dependent ki-
nases, Nat. Cell Biol. 1 (1999) 438.
[27] R.A. Friesner, R.B. Murphy, M.P. Repasky, L.L. Frye, J.R. Greenwood,
T.A. Halgren, P.C. Sanschagrin, D.T. Mainz, Extra Precision Glide: docking and
scoring incorporating a model of hydrophobic enclosure for protein-ligand
complexes, J. Med. Chem. 49 (2006) 6177e6196.
[19] a) K. Wang, P. Hampson, J. Hazeldine, V. Krystof, M. Strnad, et al., Cyclin-
dependent kinase 9 activity regulates neutrophil spontaneous apoptosis, PLoS
One 7 (2012), e30128;