Structure guided design and synthesis of furyl thiazolidinedione derivatives as inhibitors of GLUT 1 and GLUT 4, and evaluation of their anti-leukemic potential

Article abstract of DOI:10.1016/j.ejmech.2020.112603

Cancer cells increase their glucose uptake and glycolytic activity to meet the high energy requirements of proliferation. Glucose transporters (GLUTs), which facilitate the transport of glucose and related hexoses across the cell membrane, play a vital ro

Full text of DOI:10.1016/j.ejmech.2020.112603

Journal Pre-proof
Structure guided design and synthesis of furyl thiazolidinedione derivatives as
inhibitors of GLUT 1 and GLUT 4, and evaluation of their anti-leukemic potential
Kalpana Tilekar, Neha Upadhyay, Jessica D. Hess, Lucasantiago Henze Macias,
Piotr Mrowka, Renato J. Aguilera, Franz-Josef Meyer-Almes, Cristina V. Iancu, Jun-
yong Choe, C.S. Ramaa
PII:
S0223-5234(20)30575-4
DOI:
https://doi.org/10.1016/j.ejmech.2020.112603
EJMECH 112603
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 6 May 2020
Revised Date: 14 June 2020
Accepted Date: 18 June 2020
Please cite this article as: K. Tilekar, N. Upadhyay, J.D. Hess, L.H. Macias, P. Mrowka, R.J. Aguilera,
F.-J. Meyer-Almes, C.V. Iancu, J.-y. Choe, C. Ramaa, Structure guided design and synthesis of furyl
thiazolidinedione derivatives as inhibitors of GLUT 1 and GLUT 4, and evaluation of their anti-leukemic
potential, European Journal of Medicinal Chemistry, https://doi.org/10.1016/j.ejmech.2020.112603.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
© 2020 Elsevier Masson SAS. All rights reserved.

Structure guided design and synthesis of furyl thiazolidinedione derivatives as inhibitors of
GLUT 1 and GLUT 4, and evaluation of their anti-leukemic potential.
Kalpana Tilekar1, Neha Upadhyay1, Jessica D. Hess2, Lucasantiago Henze Macias2, Piotr
Mrowka3, Renato J. Aguilera2, Franz-Josef Meyer-Almes4, Cristina V. Iancu5, Jun-yong
Choe5,6*, C S Ramaa1*,
1Department of Pharmaceutical Chemistry, Bharati Vidyapeeth’s College of Pharmacy, Navi
2
Mumbai, Maharashtra, India; The Cytometry, Screening and Imaging Core Facility & Border
Biomedical Research Center & Department of Biological Sciences, The University of Texas at
3
El Paso, El Paso, Texas, USA; Department of Biophysics and Human Physiology, Medical
University of Warsaw, Chalubinskiego, Warsaw, Poland; 4Department of Chemical Engineering
5
and Biotechnology, University of Applied Sciences, Darmstadt, Germany; East Carolina
Diabetes and Obesity Institute, Department of Chemistry, East Carolina University, Greenville,
6
North Carolina, USA; Department of Biochemistry and Molecular Biology, The Chicago
Medical School, Rosalind Franklin University of Medicine and Science, North Chicago, Illinois,
USA.
Corresponding Authors
1. Dr C S Ramaa
2. Jun-yong Choe
Email: sinharamaa@yahoo.in
Email: choej18@ecu.edu
Abstract
Cancer cells increase their glucose uptake and glycolytic activity to meet the high energy
requirements of proliferation. Glucose transporters (GLUTs), which facilitate the transport of
glucose and related hexoses across the cell membrane, play a vital role in tumor cell survival and
are overexpressed in various cancers. GLUT1, the most overexpressed GLUT in many cancers,
is emerging as a promising anti-cancer target. To develop GLUT1 inhibitors, we rationally
designed, synthesized, structurally characterized, and biologically evaluated in-vitro and in-vivo
a novel series of furyl-2-methylene thiazolidinediones (TZDs). Among 25 TZDs tested, F18 and
F19 inhibited GLUT1 most potently (IC50 11.4 and 14.7 µM, respectively). F18 was equally
selective for GLUT4 (IC50 6.8 µM), while F19 was specific for GLUT1 (IC50 152 µM in
GLUT4). In-silico ligand docking studies showed that F18 interacted with conserved residues in
GLUT1 and GLUT4, while F19 had slightly different interactions with the transporters. In in-
vitro antiproliferative screening of leukemic/lymphoid cells, F18 was most lethal to CEM cells
(CC50 of 1.7 μM). Flow cytometry analysis indicated that F18 arrested cell cycle growth in the
subG0-G1 phase and lead to cell death due to necrosis and apoptosis. Western blot analysis
exhibited alterations in cell signaling proteins, consistent with cell growth arrest and death. In-
vivo xenograft study in a CEM model showed that F18 impaired tumor growth significantly.
1

Keywords Antiproliferative; Thiazolidinedione (TZD); Furan; Leukemia; Cytotoxicity; GLUT1,
GLUT4; GLUT5; Inhibitor specificity
1. Introduction
A major challenge of cancer research is to identify target therapies that kill cancer cells while
sparing normal ones. To keep pace with the rapid rate of proliferation and high energy
requirement, tumor cells significantly increase their glucose dependence and metabolism[1–3].
Interestingly, cancer cells can shift their metabolism from aerobic oxidation to anaerobic
glycolysis, irrespective of the oxygen levels. A direct consequence of this metabolic
reprogramming is lower ATP production, leading to an increase in facilitative glucose
transporters (GLUT, gene family SLC2) on the membrane surface[4]. Since GLUTs are
upregulated in many cancer cell types, including leukemia[5,6], they provide a selective
mechanism to kill tumor cells.
The human GLUT family has 14 members, which vary in substrate affinity and specificity, tissue
localization and expression level[7–9]. For instance, GLUT1, 3, 4, 8, and 12 transport glucose,
GLUT5 only transports fructose, GLUT2, 7, 9, and 11 transport both glucose and fructose,
GLUT13 transports myo-inositol. Literature also reveals a correlation between inhibition of
GLUTs and cancer suppression[10–13]. Given the number of GLUTs and their involvement in
various types of cancers, the process is still very complex and hence the need for GLUT specific
ligands[14].
Leukemia is a cancer of the bone marrow, a dynamic tissue consisting of different types of cells,
interdependent for their proper functioning, growth, and survival. Since leukemic cells are
rapidly growing, they exhibit high glycolytic activity. Several mutational changes, such as those
in IDH1/2, have also been found in leukemia[15]. The high glycolytic activity, along with
mutations, leads to microenvironment modifications that can alter the support capacity of stromal
cells for normal blood-forming cells. The development of therapeutic agents that can reduce the
glycolytic rate directly or indirectly in leukemic cells is expected to both control leukemic cell
proliferation and also reverse changes in other cells within the bone marrow, returning them to
normal hematopoiesis. Several reports indicate overexpression of GLUT1 and GLUT4 in various
types of leukemia[16–18]. Fasentin binds directly to GLUT1 inhibiting glucose uptake of U-937
leukemia cells[19]. Nilotinib and imatinib, inhibitors of a tyrosine kinase, downregulate GLUT4
in chronic myeloid leukemia (CML)[20,21]. IL-3 stimulation is involved in the translocation of
glucose transporter protein (GLUT1) to the cell surface in CML cells[22]. Small molecules have
been found to inhibit GLUTs in acute myeloid leukemia (AML) cells, HL60 cells, and T-Cell
lymphomas[17,23].
2

Several natural and synthetic inhibitors of the GLUT members have been reported[24–28],
including Thiazolidinediones (TZD)[29,30]. TZD scaffold containing antidiabetic agents such as
pioglitazone(PG), ciglitazone(CG) and troglitazone(TG) which are PPAR-γ agonists, exhibits
moderate antiproliferative effects[31,32], but their usage was limited due to mounting evidences
on side effects and toxicity of these drugs, which were attributed to the full agonistic activity at
binding sit of PPARγ[33,34]. Thus, TZDs devoid of PPAR- γ agonistic activity retaining
antiproliferative effects or exhibiting PPAR-γ independent antitumor effects, came in limelight.
Recent studies of PPAR- γ inactive analogs of Ciglitazone (CG), Δ2CG, proved ability to
mimic glucose starvation through the inhibition of glucose uptake[35]. Thus, TZD analogs
possess the capacity to create glucose deficiency state in cancer cells and can act as energy
restriction mimetics to achieve antitumor effects[36,37]. Interestingly, a couple of compounds
possessing the TZD scaffold have been reported as dual PPARγ and GLUT4 activators of their
protein expression (mRNA level), and found to have therapeutic applications as anti-diabetic
agents[38,39]. Close observation of their structures reveals that the TZD ring location may
determine the GLUT activity of the compound: activator of GLUT expression or GLUT
inhibitor. Thus, we hypothesized that when the TZD ring is at one end of the molecule, the
compounds exhibit GLUT upregulation, whereas placing it in the center (Fig. 1., Compound
2)[29] of the molecule leads to GLUT inhibitory potential. For more than a decade, we have been
working on TZD scaffold modifications to achieve partial PPARγ agonistic[40,41], anti-
angiogenic[42], and HDAC inhibitory activity[43,44]. In the current work, we have designed a
novel series of TZD derivatives targeting GLUT1 based on insights obtained from the literature
regarding desirable structural features for GLUT inhibition, and our knowledge of structure-
based drug discovery (SBDD).
2. Results and Discussion
2.1. Designing aspects of Novel series of compounds F1-F25.
Not much literature is available about the fingerprints of structural features needed to develop
inhibitors for GLUTs. Also, only a few crystal structures of human GLUTs liganded with
inhibitors are available[45]. Some TZD-containing molecules are energy restriction mimetic
agents, which selectively inhibit GLUT1[29,30,36]. A US patent by Chen. et al. have described
novel TZD-based GLUT1 inhibitors with an IC50 value of 2 μM[29]. Our work has also been
focused for more than a decade on TZD-based antiproliferative agents in anticancer drug
discovery. We noticed the structural similarity of some of our previously reported TZDs (3)[46]
(Fig. 1) with the commercially available GLUT1 inhibitor, Fasentin (1) [19] (GLUT1 IC50 =10
μM), and with a TZD molecule (2) reported by chen. et al.[29], which induced anticancer effects
by inhibiting GLUT1. Therefore, we decided to explore TZD-based agents for GLUT1
inhibition.
3

Intriguingly, the above two classes of GLUT1 inhibitors and our previous TZD derivatives share
the following pharmacophoric features:
a. Aryl acetyl amido (Ar-NH-CO-CH2) functionality as present in Fasentin (1), and
b. 5-benzylidene-2,4-TZD scaffold similar to the compound 2[29] molecule reported by Chen. et
al
Thus, after carefully analyzing all three structures, we designed a novel series of potential
GLUT1 inhibitors by modifying our previously reported TZDs to meet the structural
requirements for GLUT1 inhibitory activity (Fig. 1). We kept the TZD ring in the middle of the
molecule, as in the structures of GLUT inhibitors reported by Chen. et al.[29], since having it at
the compound terminus may lead to GLUT upregulation[38,39]
Figure 1. Designing a novel series of GLUT1 inhibitors. F1-F25 compounds were designed to
incorporate features from two potent GLUT1 inhibitors, shared with our previously reported
TZD-based antiproliferative series[46]: the aryl acetyl amido moiety (red) and the 5-
benzylidene-2,4-TZD scaffold (blue).
3.2 Chemistry
4

Figure 2 outlines the synthetic pathway for the target compounds F1-F25. Their construction
involved the synthesis of 5-(furan-2-ylmethylene) thiazolidine-2,4-dione, 5, by refluxing
furfuraldehyde, 3, with 2,4-TZD, 4, in the presence of acetic acid and a catalytic amount of
sodium acetate. Compound 5, on further treatment with ethanolic KOH, yielded potassium salt 6,
which was a common intermediate for all the target compounds. Both 5 and 6 were obtained in
good yield. Compounds 2a-2y resulted from N-chloroacetylation of various aromatic and
heteroaromatic amines (1a-1y), using a weak base like K2CO3 in organic solvents such as
dichloromethane (DCM). The final step involved the condensation between 2a-2y and 6 in
acetone to produce the target compounds F1-F25. The reactions were monitored for completion
by thin layer chromatography. The structures of newly synthesized compounds were confirmed
spectrally by 1H NMR,13C NMR, FTIR, and mass spectrometry (Supplementary Data).
FT-IR spectra of synthesized derivatives displayed characteristic absorption in the following
domains: 3340-3166 cm-1 corresponding to NH(s) vibration of the amide group; 1687-1660 cm-1
corresponding to C=O(s) of the amide group, and 1080-1120 and 1200-1280 cm-1 corresponding
to C-O-C(s) ether linkage of the furan ring.
Proton NMR spectra of F1-F25 compounds showed a characteristic singlet of -NH of amide
linkage in the region of δ=10.01-12.87 ppm. A characteristic singlet of furylidene -C=CH- was
observed in the region of δ= 8.00-8.10 ppm for intermediates 5 and 6, and final compounds F1-
F25. The singlet of the two protons from methylene-CH2 was observed at δ= 4.40-4.80 ppm,
confirming the formation of the target compounds by condensation of 2a-2y and 6. The spectra
also showed various peaks in the aromatic and aliphatic regions corresponding to the protons of
variously substituted aromatic and heteroaromatic groups attached to the N atom of the amide
linkage. Peaks of the furan ring appeared in the aromatic region at δ= 6.50-8.00 ppm.
13
Furthermore, C NMR spectra of F1-F25 compounds elicited characteristic peaks of carbonyl
carbon, C=O, at δ=162-169 ppm, and of linker -CH2 at δ=43-45 ppm. Furylidene carbon-C=CH-
was found at δ= 147-148 ppm. They also had three peaks at δ=113-114, 115-118, and 149-150
ppm corresponding to the four carbon atoms of the furan ring. Various other peaks
corresponding to carbon atoms of aromatic or heteroaromatic substitution at the N atom of the
amide linkage were also observed.
The mass spectrum was recorded for both positive and negative ionization. There was extensive
ionization in positive scanning, but in negative scanning, the characteristic [M-H+] peak was
observed at 40-100% intensity for all the compounds.
Details of the synthesis procedure and additional spectral observations are presented in Materials
and Methods and Supplementary Data.
5

Figure 2. Synthetic route for the synthesis of intermediates and F1-F25. a) K2CO3, DCM, 24
hrs; b) AcOH, Sodium acetate, reflux 5 hrs; c) EtOH, KOH, reflux 3 hrs; d) Acetone, reflux 6
hrs. ‘Het’ and ‘Ar’ are abbreviations for heteroaromatic and aromatic, respectively. The specific
aromatic moiety (Ar) in the final product is defined for each compound.
2.3. Screening for transport activity inhibition of GLUT1, GLUT4, and GLUT5.
To assay the transport activity and determine the effect of designed compounds on a singular
GLUT, we used GLUT-specific assay systems: hexose transporter deficient yeast cells (hxt0)
expressing a particular human GLUT[47]. In hxt0 strains, yeast cell growth in hexose-based
media relies solely on the transport activity of the expressing human GLUT for bringing the
6

carbon source into cells. Thus, GLUT transport activity is determined as the C14-hexose uptake
into whole cells[47]. The primary inhibition screening was carried out at 100 μM compound
concentration on three glucose transporters: GLUT1, GLUT4, and GLUT5. For compounds that
decreased the relative transport activity by 50% or more, inhibitor concentration was varied in
the assay from 0.1 to 100 µM to determine inhibitor IC50. In the primary screen, none of the
tested compounds inhibited GLUT5 significantly, but the compounds F8, F18, and F19
decreased the relative activity of GLUT1 and GLUT4 by at least 50% (Fig. 3a-c). F8 was a weak
inhibitor (Fig. 3d, g; IC50 72 ± 3 µM and 123.3 ± 1.5 µM for GLUT1 and GLUT4, respectively).
F18 inhibited both GLUT1 and GLUT4 similarly, with IC50 values 11.4 ± 1.2 µM and 6.8 ± 1.1
µM, respectively (Fig. 3e, h). F19 was more specific for GLUT1; its IC50 for GLUT1 inhibition
(IC50 14.7 ± 1.2 µM) was approximately ten times lower than that for GLUT4 (IC50=152.3 ± 1.5
µM) (Fig. 3f, i). Nevertheless, F19 was less efficacious than F18 in inhibiting GLUT1; at 100
µM F19 concentration GLUT1 retained ~ 30% of its relative activity, while the same
concentration of F18 inhibited GLUT1 almost completely (Fig. 3e, f). Thus, with experimental
proof of GLUT1 and GLUT4 inhibition, our observation of differentiating TZD-containing
analogs as activators of GLUT protein level or GLUT transport activity inhibitors, based on the
position of the TZD ring terminally or centrally, may serve as a critical milestone in the
discovery of TZD analogs as GLUT modulating agents and their application as antidiabetic and
anticancer agents.
We had designed the molecules to target GLUT1 but found that F18 and F19 inhibited the
activity of both GLUT1 and GLUT4, though with different affinities and specificities.
Nevertheless, neither F18 nor F19 affected the transport activity of GLUT5. Based on sequence
homology, GLUTs are divided into three classes; GLUT1 and GLUT4 belong to class I, GLUT5
to class II [48]. While both F18 and F19 seem to distinguish between class I and class II GLUTs,
F19 also discriminates between GLUT1 and GLUT4, making F19 a promising starting point for
the design of future GLUT1-specific probes.
7

Figure 3. Screening of F1-F23 for inhibition of the transport activity by GLUT1, GLUT4
and GLUT5. Relative transport activity of GLUT1 (a), GLUT4 (b) and GLUT5 (c) in the
presence of 100 μM concentration of F1-F23. Dose response curve of GLUT1 for F8 (d), F18
(e), and F19 (f). Dose response curve of GLUT4 for F8 (g), F18 (h), and F19 (i). Error bars
represent standard deviation from three independent measurements. Transport activity was
determined in hxt0 yeast cells expressing a single GLUT. Transport assay was initiated by the
addition of 5 mM C14- glucose for GLUT1 or GLUT4 or 10 mM C14-fructose for GLUT5 and
stopped after 10 minutes (see Materials and Methods for details).
2.4. Ligand docking studies.
Consistent with the alternating transport mechanism of the major facilitator superfamily (MFS)
proteins, GLUTs have two major states for their substrate cavity: inward-facing (open towards
cytoplasm) and outward-facing (open towards the extracellular space); these states have been
captured in the crystal structures of GLUTs and their homologs[49]. Mutagenesis, functional,
and structural data have established the location of the substrate binding site and the residues
essential for substrate binding. Fig. 4 shows the ligand docking of F18 and F19 to the inward-
facing and outward-facing conformations of GLUT1. F18 docks to the substrate binding site
formed by residues Q282, Q283, W388, W412, and others, in both the inward- and outward-
facing conformations of GLUT1. However, F19 docks to the substrate binding site only in the
8

inward-facing conformation. In GLUT1 and GLUT4, the residues forming the substrate binding
site in the outward-facing conformation are conserved but those forming the inward-facing
conformation have two differences: T137GLUT1 vs. S153GLUT4 and H160GLUT1 vs. N176GLUT4
(Supplementary Data, Table S3). In the inward-facing conformation, F19 docks close to H160
(4 Å between Cl of F19 and the imidazole of H160), which is N176 in GLUT4. Interestingly,
H160 is adjacent to the highly conserved Q161[50] and varies in all other GLUTs, except for
GLUT2 (Supplementary Data, Table S3). On the other hand, F18 does not interact with either
T137 or H160. The validation of modeled structures was done by docking n-nonyl-β-D-
glucopyranoside to GLUT1 crystal structure (PDB ID 4PYP) (Supplementary Data, Fig. S1).
Thus, F18 can bind to both the inward- and outward-facing conformations of GLUT1 and
GLUT4, interacting with conserved residues in these transporters, whereas F19 binds only to the
inward-facing conformation of the transporters, and may interact differently with GLUT1 and
GLUT4. The increased stabilization of the chlorine charge by the pyrrole nitrogen of H160
compared to the carboxamide of N176 may account for the 10-fold difference in the IC50 of F19
for GLUT1 and GLUT4, and suggests a way to increase ligand specificity in GLUT1: optimize
ligand interactions with H160 versus its substitution in GLUT4 or other GLUTs, especially given
the low conservation of this residue in other GLUTs. Supplementary Data, Fig. S2 shows the
two-dimensional representation of GLUT1 interactions with F18 or F19 generated with MOE
(https://www.chemcomp.com) ligand interaction function.
9

Figure 4. Docking of F18 and F19 to GLUT1. Inward- (a) and outward-facing (b)
conformations of GLUT1. Red circles show the substrate binding site. F18 and F19 are shown as
ball and stick model. Close-up view of F18 docking to the inward- (c) and outward-facing (d)
conformations of GLUT1. Close-up view of F19 docking to the inward- (e) and outward-facing
(f) conformations of GLUT1. F19 docks away from the substrate binding site in the outward-
facing conformation and does not have significant interactions with protein residues. Labeled
GLUT1 residues Q161, Q282, Q283, W388, N411, and W412 are conserved in GLUT4;
however, T137 and H160 are S153 and N176 in GLUT4, respectively. H1, H2, H4, H5, H7, H10
and H11 indicate the transmembrane helices surrounding the substrate cavity. The figure was
drawn using UCSF Chimera version 1.12 (http://rbvi.ucsf.edu/chimera/)
10
.

2.5. Prediction of physicochemical, pharmacokinetic, and ADME properties with
SwissADME.
To predict the physicochemical and pharmacokinetic properties of the compounds, in silico
calculations were performed using the SwissADME server[51,52] (Supplementary data, Table
S1). Results showed that the molecular weights of F1-F25 ranged between 328.34 and 486.13
g/mol, octanol–water partition coefficient ilogP[51] varied from 1.97 to 3.20, topological polar
surface area (TPSA) was between 104.92 and 150.74 Å2. The number of hydrogen bond
acceptors was <10, and of hydrogen bond donors was <5. Sixty percent of the compounds were
predicted to be soluble, and the rest moderately soluble. The solubility of the latter could be
mitigated using solubility enhancement techniques such as salt formation. Eighty percent of the
compounds exhibited high gastrointestinal tract (GIT) absorption. All the compounds were
predicted not to penetrate the blood–brain barrier (BBA), indicating no CNS adverse effect.
Also, compounds were predicted not to be substrates for P glycoprotein (P gp). P-gp, present
in different organs, acts as an efflux transporter and pumps xenobiotics out of the cells leading to
increased clearance of drugs[53]. All 25 compounds had a bioavailability score of 0.55 and
fulfilled the drug-likeness as indicted by Lipinski’s rule of five[54], Veber’s rule[55], Ghose
filter[56], Egan rule[57] and Muegge’s filter[58], except for molecular weight in some cases.
The bioavailability radar plot depicts the predicted oral bioavailability, and its pink zone shows
the ideal zone for lead-like molecules. This optimum radar zone appraises six physicochemical
properties: size, polarity, lipophilicity, solubility, saturation, and flexibility. Supplementary
Data, Fig. S3 shows the bioavailability radar plot of compounds F18 and F19. All compounds
F1-F25 were in the pink zone, with the degree of saturation slightly deviating out of the radar as
the fraction of sp3 hybridized carbons was less than 0.25. For F8, F14, F16, F18, and F25,
polarity slightly deviated out of the pink zone as TPSA was >140 Å2. Thus, the results of the in
silico ADME prediction analysis suggest that this series of compounds have favorable
physicochemical and pharmacokinetic properties indicative of drug likeness.
2.6. Cell viability assay of TZD derivatives.
To evaluate the cytotoxic potential of this compound series, we performed the Differential
Nuclear Staining (DNS) assays after exposing human hematologic cancer cell lines to each
compound, individually, for 48 hours. The DNS assay uses two fluorescent nuclear stains,
Hoechst 33342 and Propidium iodide (PI), to selectively label living and dead cells[59]. In the
initial high-throughput screening of all compounds, K-562, KCL-22, and CEM leukemia cells,
and triple-negative breast cancer MDA-MB-231 cells were treated with 50 μM compound
concentration. Each compound solubilized in DMSO at the desired concentration was directly
added to cell-containing wells at a final concentration of 1% (v/v) DMSO. Out of 25 compounds,
11

seven induced considerable cell death (>50%) in one or more cancer cell lines (Supplementary
Data, Fig. S4). These seven compounds were further assessed in a concentration range of 0.5-
100 μM via a secondary DNS assay on all three leukemic cell lines to estimate their cytotoxic
concentration 50% (CC50) values (Table 1); CC50 indicates the concentration of a given
compound needed to kill 50% of the cell population. CC50 values were estimated by linear
interpolation of the two concentrations nearest 50% cell death. Except for F18 and F19, all other
compounds had moderate antiproliferative effects in all three leukemic cell lines tested with
CC50 values in the mid-micromolar range in one or more leukemia cell lines. F19 CC50 values
were in the mid-micromolar range in the K-562 cell line. Interestingly, F18 and F19
demonstrated potent cytotoxicity at low-micromolar concentration toward CEM acute
lymphoblastic leukemia cells (CC50 = 1.7 μM and 7.8 μM, respectively), but were less effective
against K-562 and KCL-22 chronic myelogenous leukemia cell lines. CC50 value of F18 in
MDA-MB-231 was in mid-micromolar range (Table 1). Notably, except for F18, substantial
dose-dependent crystallization was observed at concentrations over 50 μM. These crystals were
roughly cell-sized but were easily excluded when quantifying cell death via IN Cell Analyzer
software.
The cytotoxic effect of the two most active compounds, F18 and F19, against non-cancerous
human HS-27 fibroblasts was assessed after 48 hours of treatment with concentrations ranging
from 1 to 100 µM. The CC50 value for F18 was 44.0 ± 2.75 µM, while F19 elicited an average of
just 27.32% cell death at 100 µM. When compared to the CC50 values of F18 and F19 against
CEM leukemia cells, 1.7 µM and 7.8 µM respectively, these derivatives demonstrated significant
cancer-selectivity. Selective cytotoxic index (SCI) values of F18 and F19 were calculated to
indicate the capacity of these compounds to preferentially kill cancer cells in vitro. F18 was
highly selective toward CEM (CC50 1.7 µM) with an SCI value of 25.8. F19 was also highly
selective, with an SCI greater than 12.8. These findings suggest that F18 and F19 could be safe
for normal cells at therapeutic dosage.
Table 1 Cytotoxic concentration 50% (CC50) values of active F-series compounds in human
cancerous and non-tumorigenic cell lines. CEM, K-562, and KCL-22 are leukemia cell lines,
MB-231 is an abbreviation for MDA-MB-231, a triple-negative breast cancer cell line. HS-27 is
a normal foreskin fibroblast cell line.
CC50 (µM)#
Cell
Line
F7
F8
F15
F18
F19
F21
F23
102.1
81 ± 1.12 ±0.16
79.9 ±
3.74
96.8 ±
0.61
CEM
>100
1.7 ± 0.03 7.8 ± 1.6
98.6 ±
1.04
91.9 ±
2.61
43.3 ±
3.27
K-562
>100
97.1 ± 0.5
94 ± 1.28 >100
KCL-22 >100
97 ± 1.94 >100
15.3 ±
72.2 ±
98.2
>100
12

2.53
1.63
±1.79
45.1 ±
0.24
44.0 ±
2.75
MB-231 nd
HS-27 nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
>100
#Values are reported as the mean ± the SD of 4 replicates per treatment. *nd - not determined.
2.7. HDAC enzyme activity assay.
We have reported an N-substituted TZD analog as an HDAC inhibitor in our earlier work[44],
and current F1-F25 series compounds are also N-substituted TZD derivatives. Even though the
designing of the molecules is entirely different, to ascertain the selectivity of these compounds
for GLUTs over HDACs, we undertook HDAC screening of F1-F25 on two subtypes, HDAC4
and HDAC8. We selected these two isoforms as they possess high deacetylation activity and
represent two different classes of HDAC, class1 (HDAC 8) and class II (HDAC4). The primary
screening was performed with 50 µM of F1-F25 (Fig. 5, Supplementary Data-Table S4 and
Table S5). Only one compound, F14, produced a significant change in HDAC8 activity (residual
enzyme activity: (17±3) % ). The IC50 of F14 was subsequently be determined to be 13 µM.
However, F14 did not inhibit GLUT1, 4, or 5, and F25 was not evaluated for GLUT inhibition.
All remaining compounds retained HDAC activity greater than 50% for HDAC8 and HDAC4,
respectively, indicating that these derivatives may not interact with HDACs to such an extent to
elicit therapeutic effect.
Figure 5. Histone Deacetylase (HDAC) activity assay. Percent relative activity of HDAC4 and
HDAC8 after treatment with F1-F25 at 50 μM concentration. Values were determined in
triplicates and presented as Mean±SD.
13

2.8. Apoptosis studies
Apoptotic cell death of cells treated with GLUT inhibitors has been observed in various
cancerous cells; it has been attributed to activation of caspase 8 and 3[60], sensitizing cells to
FAS, a death receptor belonging to the TNF family[24]. GLUT1 inhibitors can decrease the
levels of GLUT1, intracellular ATP, and glycolytic enzymes, leading to cell-cycle arrest,
senescence, and necrosis through an increase in expression of AMPK, an ATP sensing enzyme,
and decrease in cyclin E2[61]. OSU-CG5, a TZD derivative, has been shown to induce apoptosis
through induction of ER stress[37]. Thus, the consequences of GLUT inhibition, which lead to
anticancer effects, are the induction of apoptosis, necrosis, and cell cycle arrest. In agreement
with these observations, our GLUT1/GLUT4 inhibitor, F18, exhibited apoptotic and necrotic cell
death in lymphoid leukemic CEM cells (Supplementary Data, Fig. S5).
To examine whether the cytotoxicity induced by F18 is linked to apoptosis, we used CEM cells
treated with annexin V-FITC/PI double staining and determined the rate of apoptosis by flow
cytometry. Treatment of F18 in CEM cells at its CC50 concentration caused an increase of cell
population in the apoptotic region as compared to control (Fig. 6b). For F18, the percentage of
apoptotic cells in the late and early apoptosis region was 55.87% and 1.38%, respectively. The
percentage of live cells was only 26.14% compared with 92.15% of control (refer to
Supplementary Data for cytograms, Fig. S6). A considerable number of cells (16.60%) were
observed in the necrotic region, in agreement with observations by Liu et al., who have reported
that WZB117, a GLUT1 inhibitor, induces cell-cycle arrest and necrosis as a consequence of
GLUT1 inhibition[61]. Thus, our results suggest that compound F18 can cause apoptotic death
in the CEM cell line by induction of apoptosis and necrosis. This drug-induced cell death is, at
least in part, attributable to apoptosis, as demonstrated by an increase in caspase 3 cleavage in
response to compound F18 (see below in western blotting section).
14

Figure 6. In vitro cytotoxicity screening and in vivo impact of F18 on a CEM xenograft
tumor model. (a) Flow cytometry analysis of CEM cells treated with F18 (n=3, P = ≤ 0.001).
Controls were: untreated cells, cells treated with 1% DMSO, or 1 mM hydrogen peroxide
(H2O2). Concentrations for F18 treatment were 10, 25 and 50 µM. (b) Graphical representation
of apoptotic events of untreated and F18-treated CEM cells. L-Live cells, EA- Early apoptotic,
LA-Late apoptotic. (n=3, P =≤ 0.001). (c) Dose response curves of F18 in CEM cells
(continuous line) and WBCs (dotted line); the mean cell viability is plotted versus increasing
concentrations of F18. Human leukemic cell line CEM and normal WBC cells were treated with
various F18 concentrations for 48 hours. (d) Western blot analysis of markers associated with:
apoptosis (Caspase 3), cell cycle regulation (p-mTOR and CDK2), and ER stress (GRP78) in
F18-treated cells (F18) and controls (UT – untreated cells, Stauro - staurosporin positive
control).
2.9. Cell cycle analysis
Cell cycle progression was assessed via a Beckman Coulter Gallios flow cytometer to investigate
if compound F18 perturbs this process. CEM cells were exposed to 50 μM, 25 μM, or 10 μM of
F18 for 72 hours, then permeabilized and stained in a single step with a NIM-DAPI reagent
15

(Beckman Coulter). The fluorescent signal of DAPI was then analyzed to quantify DNA content
in each sample[62]. A dose-dependent increase in DNA fragmentation, as indicated by the sub
G0/G1 population, was observed in cells treated with F18 (Supplementary Data, Fig. S5). The
G0/G1, S, and G2-M phases were also significantly disrupted in a dose-dependent manner (Fig.
6a). Treatment with F18 caused cell cycle disruption, closely mirroring that of the hydrogen
peroxide positive control and displaying arrest of the cell cycle in the sub G0/G1phase.
2.10. Western blotting
After identifying changes related to glucose transport activity, apoptosis, and changes in cell
cycle regulation, Western blotting was performed to identify the consequences of the treatment
with F18 on cell signaling proteins such as p-mTOR, CDK2, GRP78 and Caspase 3. GLUT1
inhibitors can downregulate the levels of mTOR and Akt, phosphorylated enzymes involved in
cell growth signaling pathway, and energy homeostasis[63–65]. Endoplasmic reticulum (ER)
stress studies of GLUT inhibitors found increased levels of ER stress markers such as
GRP78/BiP and PKR-like endoplasmic reticulum kinase (PERK)[66], indicating induction of ER
stress leading to apoptosis. Phosphorylated retinoblastoma (pRb) regulates G1 arrest, and
phosphorylation of pRb is controlled by the cyclin-dependent kinase (CDK)2/cyclin E2 complex.
Thus, downregulation of CDK2 or cyclin E2 leads to decreased level of phosphorylated pRb and
arrest of cells in early G1 phase[61,67–69]. The relationship between cell cycle progression and
caspase 3 activation in leukemic cells has been demonstrated and found to be associated with G1
phase arrest. Activation of caspase 3 cleaves the substrate DNA triggering fragmentation leading
to apoptosis[70].
In agreement with these observations, Western blot analysis of lysates from cells treated with 10
μM F18 for 6 hours showed a decrease in levels of phosphorylated mTOR and CDK2 proteins,
an increase in endoplasmic stress marker, GRP78, and activation of caspase 3 (Fig. 6d).
Together, data revealed that after exposure to F18, cancer cells experienced a drop in glucose
levels due to GLUT inhibition and maybe a concomitant decrease in some key glycolytic
enzymes and metabolites like ATP and lactate[61]. This further leads to changes in key enzymes,
particularly mTOR and CDK2. All these changes cause cell cycle arrest, accompanied by
induction of ER stresses and caspase 3 activation, further inhibiting cancer cell growth. Thus,
early G1 arrest, mediated by downregulation of CDK2 and phosphorylation of mTOR, induction
of ER stress, DNA fragmentation by activation of caspase 3, and subsequent necrosis/apoptosis
was the major mechanism underlying the inhibitory action of F18 on cancer cell growth.
2.11. F18 effect on the viability of non-transformed WBCs.
We evaluated the cytotoxic effects of F18 on normal WBCs by MTT assay after treatment of
WBCs with F18, at concentrations ranging from 2.5 to 100 µM, for 48 h (Fig. 6c). The IC50
value (187.2 μM) was 107 times greater than that for the CEM cell line (1.7 μM). More than 62%
16

of the WBCs were viable even after treatment with the highest concentration of F18
Supplementary Table S2). Hence, these findings suggest that F18 could not be interfering much
(
with the metabolic activities of normal blood cells and could be safer to them.
2.12. In-vivo effect of F18 on the CEM xenograft tumor.
To examine if F18 has in-vivo anticancer activity in CEM cells, we established an in-vivo
xenograft model with CEM cells. A total of 18 SCID mice were used, and tumors were generated
using CEM cell lines. When the tumor reached palpable size, mice were randomly divided into
three groups: the untreated group received the saline vehicle, the standard drug treated group
received 20 mg/kg of doxorubicin, and the test drug group received 50 mg/kg of F18 for 18 days
(details in Supplementary Information). Tumor volume was measured every two days. Treatment
with F18 impaired the growth of tumors, as evidenced by the tumor growth curve (Fig. 7a) and
tumor size (Fig. 7b). In the untreated group, all xenografts continued growing, and the volumes
of the xenografts were approximately 1.5 and 2 times higher compared with those of xenografts
from the doxorubicin and F18 treated groups, respectively. Thus, F18 exhibited potent in-vivo
antitumor activity, which was comparable to doxorubicin but at a higher dose and far better than
the untreated control. Moreover, throughout the experiment, no considerable weight loss or any
other noticeable signs of toxicity were observed in any of the mice treated with F18.
Figure 7. Tumor regression by F18 in CEM xenograft model. (a) Graph of tumor volume vs.
time in CEM xenografts treated with doxorubicin 20 mg/kg i.p. (red), F18 50 mg/kg i.p. (green),
and saline i.p. (blue). Error bars represent SEM n=6. p<0.001. (b) Excised tumor image of the
animals treated with doxorubicin 20mg/kg i.p. , F18 50 mg/kg i.p. ,untreated saline i.p.
3. Conclusion
This series of furyl TZDs was designed for GLUT1 inhibition, wherein we found that two
compounds, F18 and F19, inhibited both GLUT1 and GLUT4. F19 had higher selectivity for
17

GLUT1, and ligand docking analysis showed that F19 could only bind to the inward-facing
conformation of the transporters, with slightly different interactions in GLUT1 and GLUT4. F18
inhibited potently and comparably both GLUT1 and GLUT4, as indicated by the IC50 values,
consistent with the ligand docking analysis showing that F18 had similar interactions with both
transporters, irrespective of the transporter state (i.e., substrate cavity opened to either side of the
membrane). The compound F18 was taken further for in-vitro and in-vivo antiproliferative
evaluation. It exhibited promising cytotoxic effects in the CEM cell line, in sub-micromolar
range, and flow cytometric analysis indicated that F18 arrested cell cycle growth in the sub G0-
G1 phase and lead to cell death due to necrosis and apoptosis. On the other hand, normal cells
withstood 100-fold higher F18 concentrations than CEM cells, indicating that F18, besides its
drug-likeness shown in the ADME analysis, may also be safe on non-cancerous cells. Western
blot analysis data supported apoptosis and cell growth arrest in F18-treated cells, as evidenced
by changes in the levels of proteins involved in cell growth signaling: downregulation of p-
mTOR and CDK3, activation of caspase 3, and induction of ER stresses by increased levels of
ER stress marker, GRP78. These changes suggest that GLUT inhibition could lead to cell growth
inhibition and apoptosis (Fig. 8). F18 demonstrated its antiproliferative effects in the in-vivo
xenograft study too, wherein it impaired the growth of tumors much more effectively compared
to untreated tumors. Thus, based on our observations, it appears that the beneficial anticancer
effects of F18 are due, at least in part, to inhibition of GLUT1 and GLUT4. F19 also looks
promising as it inhibits the growth of CEM cells at the sub-micromolar range and may exert
beneficial effects if evaluated further. Furthermore, given the increased selectivity of F19 for
GLUT1, the interactions of F19 with GLUT1, particularly involving residue H160, can be
exploited in designing even more potent and specific GLUT1 inhibitors.
Figure 8. Schematic depiction of the proposed mode of action for compound F18 in CEM
cells.
4. Materials and methods
18

4.1. Chemistry
4.1.1. General
Commercial reagents were from S D Fine Chem (Chennai, India), Research Lab (Mumbai,
India), or Sigma Aldrich (St. Louis, USA), and were procured from suppliers in India. Thin layer
chromatography was performed on Merck pre-coated Silica Gel 60 F254. Melting points (M.P.)
were determined by the open capillary method on a VEEGO melting point apparatus and are
uncorrected. Infrared spectra were recorded on Schimadzu FT/IR-8400S by direct sampling
1
13
technique. H and C NMR spectra were recorded at 400 MHz on a Bruker instrument using
tetramethylsilane (TMS) as internal standard, and chemical shifts (δ) are reported in ppm. J
values for coupling constants are in hertz (Hz). Abbreviations used in NMR interpretation are: bs
- broad singlet, s - singlet, d - doublet, t - triplet and m - multiplet. Mass spectra were recorded
on a LC-MS Agilent Technologies 1260 Infinity instrument. The purity of final derivatives
(≥95%) was confirmed by HPLC (high-performance liquid chromatography) on an Agilent 1100
system. The conditions for chromatography were: column - Hemochrome C18, 15 cm; detection
wavelength - 300 nm; detector - UV visible detector; flow rate - 1 mL/min; oven temperature 30
°C; gradient elution with a run time of 15 min on mobile phase - Methanol: (0.1%) formic acid
(FA) in a ratio 70:30. The chemical synthesis scheme is in Fig. 2.
4.1.2. Synthesis of intermediate, [5-(Furan-2-ylmethylene)thiazolidine-2,4-dione] (5).
Furfuraldehyde 3 (5 ml, 0.07 mol), thiazolidine-2,4-dione 4 (5 g, 0.04 mol) and sodium acetate
(3 g, 0.03 mol) were added to 5 ml of acetic acid and refluxed for 5 h. The reaction mixture was
then cooled to room temperature (RT), and precipitated solid was collected by filtration under
vacuum, then washed several times with water and dried at RT. Crude solid was purified by
recrystallization from appropriate solvent to obtain brown shiny crystals of 5.
Yield 7.7 g (68%); M.P. 238.4 °C; Brown shiny crystals; IR (cm-1) 3234.73, 3140.22, 3010.98,
1674.27, 1604.83, 1543.10, 1512.24, 1263.42, 1147.68; 1H NMR (400 MHz, DMSO-d6, δ ppm)
6.72-6.73 (m, 1H), 7.06 (d, J = 3.6 Hz, 1H), 7.5 (s, 1H), 12.43 (s, 1H).
4.1.3. Synthesis of intermediate, Potassium-5-(furan-2-ylmethylene)-2,4-dioxothiazolidin-3-
ide (6).
To the solution of potassium hydroxide (2.5 g, 0.044 mol) in 25 ml of ethanol, 4 (5 g, 0.025 mol)
was added with stirring, and the reaction mixture was refluxed for 3 h. A fine light brown solid
was obtained after cooling the reaction mixture, which was collected by filtering and washed
with cold ethanol to obtain potassium salt 6.
Yield 4.9 g (94 %); M.P. 269.9 °C (charred); Light brown solid; IR (cm-1) 3037.99, 1662.69,
1626.05, 1585.54, 1531.53, 1139.97, 1226.77; 1H NMR (400 MHz, DMSO-d6, δ ppm) 6.56-6.59
(m, 2H), 7.06 (s, 1H), 7.75-7.76 (m, 1H).
19

4.1.4. Condensation of 6 with 2a-2y to produce F1-F25 (Figure 2).
Differently substituted chloroacetylated amides (2a-2y) were prepared according to our previous
work[40,46,71]. Acetone (10 ml) was transferred in RBF, into which variously substituted
chloroacetylated amides (2a-2y) and 6 (in equimolar ratio) were added and refluxed for 3-6 hrs
and monitored by TLC for completion. After completion, the reaction was stopped and cooled.
Acetone layer was evaporated to obtain a crude product, which was purified by either
recrystallization from an appropriate solvent or by column chromatography (Ethyl acetate:
Hexane, 1:4) to produce F1-F25.
2-(5-(Furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl)-N-phenylacetamide (F1)
Yield 42.5%; M.P. 234.7 °C (Charred); Buff white color solid; IR (cm-1) 3269, 1741, 1689,
1614, 1599, 1111, 1255; 1H NMR (400 MHz, DMSO-d6, δ ppm) 4.41 (s, 2H), 6.78-6.79 (m, 1H),
7.06-7.10 (t, J = 7.2 Hz, 1H), 7.18-7.19 (d, J = 3.6 Hz, 1H), 7.30-7.34 (t, J = 7.8 Hz, 2H), 7.54-
7.56 (d, J = 8 Hz, 2H), 7.82 (s, 1H), 8.10 (s, 1H), 10.39 (s, 1H); 13CNMR (400 MHz, DMSO-d6)
43.84, 113.68, 117.50, 149.113, 119.60, 119.96, 123.70, 128.85, 138.32, 147.97, 163.81, 165.07,
167.85; UV spectrum (10 ppm) λmax 349 nm; MS (m/z) 327.0, 328.0 [M-H] +; HPLC Purity %
Area 99.61, retention time (RT) 4.58 mins.
N-(2-fluorophenyl)-2-(5-(furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl)acetamide (F2)
Yield 44.5%; M.P. 248.4 °C (Charred); Buff white color solid; IR (cm-1) 3269, 1741, 1668,
1
1616, 1602, 1112, 1263, 1383; H NMR (400 MHz, DMSO-d6, δ ppm) 4.46 (s, 2H), 6.77-6.79
(m, 1H), 7.14-7.18 (m, 3H), 7.25-7.30 (m, 1H), 7.81-7.86 (m, 2H), 8.09 (s, 1H), 10.26 (s, 1H);
13CNMR (400 MHz, DMSO-d6) 43.68, 113.680, 115.471, 149.107, 115.661, 117.510, 119.590,
119.967, 123.679, 123.800, 124.474, 125.332, 125.628, 147.959, 164.443, 165.041, 167.832; UV
spectrum (10 ppm) λmax 350.3 nm;. MS (m/z) 345.0, 346.2 [M-H] +; HPLC Purity % Area 99.71,
RT 4.52 mins.
N-(3-fluorophenyl)-2-(5-(furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl) acetamide (F3)
Yield 50.2%; M.P. 250.2 °C (Charred); Brown color solid; IR (cm-1) 3269, 1737, 1681, 1612,
1599, 1111, 1263, 1371; 1H NMR (400 MHz, DMSO-d6, δ ppm) 4.50 (s, 2H), 6.74-6.79 (m, 1H),
6.89-6.94 (m, 1H), 7.18-7.19 (d, 1H), 7.27-7.29 (m, 1H), 7.34-7.40 (m, 1H), 7.50-7.53 (d, 1H),
7.82 (s, 1H), 8.10 (s, 1H), 10.62 (s, 1H); 13CNMR (400 MHz, DMSO-d6) 43.86, 113.693,
114.910 149.100, 117.428, 119.652, 120.036, 130.523, 130.617, 139.933, 140.043, 147.995,
160.897, 163.298, 164.273, 165.031, 167.826; UV spectrum (10 ppm) λmax 349.8nm; MS (m/z)
345.0, 346.0 [M-H] +; HPLC Purity % Area 98.85, RT 5.61 mins.
N-(4-bromo-2,6-difluorophenyl)-2-(5-(furan-2-ylmethylene)-2,4-dioxothiazolidin-3-
yl)acetamide (F4)
20

Yield 48.7%; M.P. 279.5 °C (Charred); Brownish solid; IR (cm-1) 3269, 1745, 1680, 1695,
1618, 1545, 1114, 1379, 665; 1H NMR (400 MHz, DMSO-d6, δ ppm) 4.51 (s, 2H), 6.77 (s, 1H),
7.17 (d, J = 3.2 Hz, 1H), 7.43-7.48 (m, 1H), 7.58 (d, J = 8 Hz, 1H), 7.81 (s, 1H), 8.08 (s, 1H),
10.19 (s, 1H); 13CNMR (400 MHz, DMSO-d6) 43.172, 104.256, 104.515, 104.775, 113.671,
115.725, 116.011, 117.557, 119.554, 119.933, 123.217, 123.370, 147.948, 149.095, 164.579,
164.899, 167.689; UV spectrum (10 ppm) λmax 349.5 nm; MS (m/z) 440.9, 442.9 [M-H] +; HPLC
Purity % Area 97.97, RT 4.61 mins.
N-(2,4-difluorophenyl)-2-(5-(furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl)acetamide (F5)
Yield 44.4%; M.P. 227.8 °C (Charred); Buff white color solid; IR (cm-1) 3267, 1737, 1664,
1
1612, 1537, 1099, 1259, 1141; H NMR (400 MHz, DMSO-d6, δ ppm) 4.55 (s, 2H), 6.77-6.78
(m, 1H), 7.07 (t, J = 8.2 Hz, 1H), 7.17 (d, J = 3.6 Hz, 1H), 7.31-7.36 (m, 1H), 7.77-7.85 (m, 2H),
8.09 (s, 1H), 10.26 (s, 1H); 13C NMR (400 MHz, DMSO-d6) 43.578, 103.977, 104.216, 104.483,
111.098, 111.316, 113.681, 117.496, 119.599, 119.971, 121.844, 125.276, 147.968, 149.100,
164.509, 165.019, 167.816; UV spectrum (10 ppm) λmax 351.3 nm; MS (m/z) 363.0, 364.0 [M-H]
+; HPLC Purity % Area 99.60, RT 4.97 mins.
2-(5-(Furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl)-N-(3-(trifluoromethyl)phenyl)
acetamide (F6)
Yield 46.3%; M.P. 216.7 °C (Charred); Yellowish brown solid; IR (cm-1) 3271, 1691, 1658,
1618, 1554, 1116, 1155; 1H NMR (400 MHz, DMSO-d6, δ ppm) 4.53 (s, 2H), 6.78 (t, J = 1.6 Hz,
1H), 7.18 (d, J = 3.2 Hz, 1H), 7.44 (d, J = 7.6 Hz, 1H), 7.58 (t, J = 8 Hz, 1H), 7.73 (d, J = 8 Hz,
1H), 7.82 (s, 1H), 8.04 (s, 1H), 8.10 (s, 1H), 10.26 (s, 1H); 13CNMR (400 MHz, DMSO-d6)
43.875, 113.692, 115.209, 117.413, 119.665, 120.052, 122.745, 125.316, 129.399, 129.713,
130.177, 139.040, 148.000, 149.094, 164.567, 165.031, 167.839; UV spectrum (10 ppm) λmax
349.9 nm; MS (m/z) 395.0, 396.0 [M-H] +; HPLC Purity % Area 97.79, RT 8.95 mins.
N-(4-chloro-2-(trifluoromethyl)phenyl)-2-(5-(furan-2-ylmethylene)-2,4-dioxothiazolidin-3-
yl)acetamide (F7)
Yield 41.5%; M.P. 274.4 °C (Charred); Buff white color solid; IR (cm-1) 3275, 1739, 1680,
1618, 1545, 1124, 1273, 1159, 746; 1H NMR (400 MHz, DMSO-d6, δ ppm) 4.51 (s, 2H), 6.77 (d,
J = 1.6 Hz, 1H), 7.17 (d, J = 3.6 Hz, 1H), 7.51 (d, J = 8.4 Hz, 1H), 7.76-7.83 (m, 3H), 8.08 (s,
1H), 10.17 (s, 1H); 13CNMR (400 MHz, DMSO-d6) 43.416, 113.673, 117.503, 119.575, 119.948,
121.114, 123.837, 126.396, 131.352, 131.797, 133.114, 133.448, 147.955, 149.093, 164.934,
165.442, 167.703; UV spectrum (10 ppm) λmax 350.9 nm; MS (m/z) 429.0, 430.9, 429.9 [M-H] +;
HPLC Purity % Area 98.72, RT 7.93 mins.
2-(5-(Furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl)-N-(4-methylbenzo[d]thiazol-2-
yl)acetamide (F8)
21

Yield 47.5%; M.P. 220.2 °C (Charred); Pale yellow solid; IR (cm-1) 3302, 2941, 1712, 1666,
1
1610, 1548, 1147, 1284; H NMR (400 MHz, DMSO-d6, δ ppm) 2.57 (s, 3H), 4.66 (s, 2H), 6.78
(d, J = 1.2 Hz, 1H), 7.18 (d, J = 3.6 Hz, 1H), 7.22 (d, J = 7.6 Hz, 1H), 7.26 (d, J = 6.8 Hz, 1H),
13
7.78 (d, J = 7.6 Hz, 1H), 7.82 (s, 1H), 8.09 (s, 1H), 12.92 (s, 1H); CNMR (400 MHz, DMSO-
d6) 17.843, 43.345, 113.698, 117.402, 119.125, 119.707, 120.141, 123.785, 126.715, 130.065,
131.103, 147.503, 148.026, 149.089, 164.924, 167.823; UV spectrum (10 ppm) λmax 350.2 nm;
MS (m/z) 398.0, 399.0 [M-H] +; HPLC Purity % Area 999.4, RT 14.09 mins.
2-(5-(Furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl)-N-(4-phenoxyphenyl)acetamide (F9)
Yield 45.6%; M.P. 163.5 °C (Charred); Brown color solid; IR (cm-1) 3269, 1739, 1678, 1618,
1529, 1109, 1153, 1230, 1267; 1H NMR (400 MHz, DMSO-d6, δ ppm) 4.53 (s, 2H), 6.78 (s, 1H),
6.87 (d, J = 3.6Hz, 1H), 7.04 (d, J = 8 Hz, 2H), 7.11 (s, 2H), 7.17 (d, J = 4 Hz, 2H), 7.41(t, J =
7.4 Hz, 2H), 7.80 (s, 1H), 7.98 (s, 1H), 8.09 (s, 1H), 10.05 (s, 1H); 13CNMR (400 MHz, DMSO-
d6) 43.772, 113.677, 117.516, 118.431, 118.713, 119.561, 119.927, 123.150, 123.535, 123.632,
125.207, 129.012, 129.947, 147.340, 147.946, 149.114, 156.449, 164.354, 165.048, 167.804; UV
spectrum (10 ppm) λmax 349.7 nm; MS (m/z) 418.8, 420.1 [M-H] +; HPLC Purity: % Area 97.67,
RT 10.62 mins.
N-(2-chloro-5-(trifluoromethyl)phenyl)-2-(5-(furan-2-ylmethylene)-2,4-dioxothiazolidin-3-
yl)acetamide (F10)
Yield 44.5%; M.P. 242.6 °C (Charred); Buff white color solid; IR (cm-1) 3261, 1739, 1678,
1614, 1587, 1126, 1265, 1265, 819; 1H NMR (400 MHz, DMSO-d6, δ ppm) 4.65 (s, 2H), 6.77 (s,
1H), 7.18 (d, J = 2.8 Hz, 1H), 7.56 (d, J = 8 Hz, 1H), 7.77 (d, J = 8.4 Hz, 1H), 7.82 (s, 1H), 8.09
13
(s, 1H), 8.16 (s, 1H), 10.33 (s, 1H); CNMR (400 MHz, DMSO-d6) 43.727, 113.676, 117.414,
119.640, 120.041, 121.373, 122.726, 124.841, 127.918, 128.240, 130.885, 135.135, 147.985,
149.089, 164.997, 165.250, 167.815; UV spectrum (10 ppm) λmax 350.4 nm; MS (m/z) 431.0,
429.9 [M-H]+; HPLC Purity % Area 96.76, RT 9.95 mins.
N-(4-bromo-2-fluorophenyl)-2-(5-(furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl)acetamide
(F11)
Yield 48.2%; M.P. 257.6 °C (Charred); Buff white color solid; IR (cm-1) 3261, 1739, 1668,
1614, 1593, 1112, 1273, 1381, 761; 1H NMR (400 MHz, DMSO-d6, δ ppm) 4.57(s, 2H), 6.77 (s,
1H), 7.17 (d, J = 3.2 Hz, 1H), 7.38 (d, J = 8.8 Hz, 1H), 7.61 (d, J = 10.4Hz, 1), 7.81-7.89 (m,
2H), 8.08 (s, 1H), 10.37 (s, 1H); 13CNMR (400 MHz, DMSO-d6) 43.701, 113.673, 115.975,
116.059, 117.448, 118.796, 119.022, 119.611, 120.005, 124.876, 125.056, 125.170, 127.534,
127.568, 147.962, 149.092, 164.601, 165.003, 167.811; UV spectrum (10 ppm) λmax 349.3nm;
MS (m/z) 424.9, 423.0 [M-H] +; HPLC Purity % Area 98.50, RT 8.79 mins.
2-(5-(Furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl)-N-(4-methoxyphenyl) acetamide (F12)
22

Yield 52.5%; M.P. 246 °C (Charred); Pale yellow solid; IR (cm-1) 3286, 1737, 1667, 1640, 1546,
1
1111, 1284, 1149; H NMR (400 MHz, DMSO-d6, δ ppm) 3.72 (s,3H), 4.46 (s, 2H), 3.78-3.79
(m, 1H), 6.90 (d, J = 9, 2H), 7.19 (d, J = 3.48, 1H), 7.46(d, J = 9.04, 2H), 7.82 (s, 1H), 8.10 (d, J
= 1.56 Hz, 1H), 10.22 (s, 1H); 13C NMR (400 MHz, DMSO-d6) 43.331, 54.953, 111.396,
124.271, 125.202, 127.980, 129.492, 129.711, 137.617, 139.449, 149.322, 150.965, 159.528,
163.947, 165.532, 171.024; UV spectrum (10 ppm) λmax 350.1 nm; MS (m/z) 357.0, 357.8 [M-H]
+; HPLC Purity % Area 99.53, RT 4.33 mins.
N-(4-fluorophenyl)-2-(5-(furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl) acetamide (F13)
Yield 42.5%; M.P. 240.8 °C (Charred); Brownish solid; IR (cm-1) 3286, 1743, 1676, 1612, 1545,
1105, 1217, 1149; 1H NMR (400 MHz, DMSO-d6, δ ppm) 4.48 (s, 2H), 6.77-6.78 (m, 1H), 7.14-
13
7.18 (m, 3H), 7.54-7.58 (m, 2H),7.80 (s, 1H), 8.08 (s, 1H), 10.46 (s, 1H); CNMR (400 MHz,
DMSO-d6) 43.737, 113.695, 115.333, 115.554, 117.478, 119.612, 119.984, 120.976, 121.055,
134.647, 134.670, 147.946, 149.079, 157.007, 159.396, 163.803, 165.067, 167.881; UV
spectrum (10 ppm) λmax 345.2 nm; MS (m/z) 345.0, 345.9 [M-H] +; HPLC Purity % Area 99.5,
RT 5.09 mins.
2-(5-(Furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl)-N-(4-nitrophenyl) acetamide (F14)
Yield 47.5%; M.P. 277.5 °C (Charred); Pale yellow solid; IR (cm-1) 3267, 1735, 1670, 1614,
1
1595, 1103, 1217, 1496; H NMR (400 MHz, DMSO-d6, δ ppm) 4.57 (s, 2H), 6.78 (d, J = 1.2
Hz, 1H), 7.18 (d, J = 3.2 Hz, 1H), 7.79-7.82 (m, 3H), 8.09 (s, 1H), 8.23 (d, J = 9.2 Hz, 2H),
11.02 (s, 1H); 13CNMR (400 MHz, DMSO-d6) 44.017, 113.710, 117.336, 119.004, 119.725,
120.129, 125.046, 142.587, 144.355, 148.026, 149.078, 164.968, 164.990, 167.818; UV
spectrum (10 ppm) λmax 348.8 nm; MS (m/z) 371.9, 373.0 [M-H] +; HPLC Purity % Area 96.95,
RT 6.94 mins.
2-(5-(Furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl)-N-(5-methylisoxazol-3-yl)acetamide
(F15)
Yield 41%; M.P. 253.6 °C (Charred); Buff white color solid; IR (cm-1) 3288, 2985, 1732, 1689,
1610, 1564, 1105, 1267, 1377; 1H NMR (400 MHz, DMSO-d6, δ ppm) 2.36 (s, 3H), 4.50 (s, 2H),
6.56 (s, 1H), 6.77 (s, 1H), 7.16 (d, J = 3.2 Hz, 1H), 7.8 (s, 1H), 8.07 (s, 1H), 11.02 (s, 1H);
13CNMR (400 MHz, DMSO-d6) 12.038, 43.531, 96.114, 113.688, 117.426, 119.642, 120.034,
147.958, 149.060, 157.534, 164.961, 167.828, 169.956; UV spectrum (10 ppm) λmax 351.1 nm;
MS (m/z) 332.0, 333.0 [M-H] +; HPLC Purity % Area 99.48, RT 3.66 mins.
2-(5-(Furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl)-N-(thiazol-2-yl) acetamide (F16)
Yield 41.5%; M.P. 223.4 °C (Charred); Buff white color solid; IR (cm-1) 3290, 1726, 1680,
1608, 1564, 1109, 1280; 1H NMR (400 MHz, DMSO-d6, δ ppm) 4.60 (s, 2H), 6.78-6.79 (m, 1H),
23

7.18 (d, J = 3.2 Hz, 1H), 7.25 (d, J = 3.6 Hz, 1H), 7.49 (d, J = 3.6 Hz, 1H), 7.82 (s, 1H), 8.10 (s,
1H), 12.58 (s, 1H); 13CNMR (400 MHz, DMSO-d6) 43.182, 113.710, 114.027, 117.413, 119.709,
120.099, 137.759, 148.028, 149.085, 164.956, 167.825; UV spectrum (10 ppm) λmax 351.8 nm;
MS (m/z) 333.9, 335.0 [M-H] +; HPLC Purity % Area 96.04, RT 3.88 mins.
N-(3-chloro-4-methylphenyl)-2-(5-(furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl)acetamide
(F17)
Yield 40.3%; M.P. 228.1 °C (Charred); Light brown solid; IR (cm-1) 3263, 3041, 1743, 1687,
1618, 1585, 1112, 1274, 815; 1H NMR (400 MHz, DMSO-d6, δ ppm) 2.26 (s, 3H), 4.48 (s, 2H),
6.77 (t, J = 1.6 Hz, 1H), 7.17 (d, J = 3.2 Hz, 1H), 7.27-7.34 (m, 2H), 7.72 (d, J = 0.8 Hz, 1H),
7.81 (s, 1H), 8.08 (s, 1H), 10.50 (s, 1H); 13CNMR (400 MHz, DMSO-d6) 18.870, 43.812,
113.677, 117.455, 117.795, 119.166, 119.612, 119.995, 130.442, 131.296, 133.066, 137.377,
147.956, 149.093, 164.034, 165.039, 167.846; UV spectrum (10 ppm) λmax 350.2 nm; MS (m/z)
375.0, 375.8 [M-H] +; HPLC Purity % Area 96.53, RT 10.01 mins.
N-(benzo[d]thiazol-2-yl)-2-(5-(furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl)acetamide (F18)
Yield 38.5%; M.P. 220.2 °C (Charred); Buff white color solid; IR (cm-1) 3223, 1735, 1662,
1602, 1545, 1112, 1257; 1H NMR (400 MHz, DMSO-d6, δ ppm) 4.67 (s, 2H), 6.79 (t, J = 1.6 Hz,
1H), 7.19 (d, J = 3.2 Hz, 1H), 7.33 (t, J = 7.4 Hz, 1H), 7.46 (t, J = 7.8 Hz, 1H), 7.77 (d, J = 8 Hz,
1H), 7.83 (s, 1H), 7.98 (d, J = 8 Hz, 1H), 8.10 (s,1H), 12.87 (s, 1H); 13CNMR (400 MHz,
DMSO-d6) 43.423, 113.723, 117.381, 119.754, 120.180, 121.787, 123.853, 126.275, 148.050,
149.083, 164.945, 167.840; UV spectrum (10 ppm) λmax 349.4 nm; MS (m/z) 384.0, 384.8 [M-H]
+; HPLC Purity % Area 97.36, RT 8.82 mins.
N-(3,4-dichlorophenyl)-2-(5-(furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl) acetamide (F19)
Yield 46%; M.P. 246.2 °C (Charred); Buff white color solid; IR (cm-1) 3302, 1735, 1668, 1608,
1
1587, 1288, 817; H NMR (400 MHz, DMSO-d6, δ ppm) 4.51 (s, 2H), 6.77-6.78 (m, 1H), 7.17
(d, J = 3.2 Hz,1H), 7.45 (dd, J = 8.8 Hz, 1H), 7.57 (d, J = 8 Hz, 1H), 7.81 (s, 1H), 7.91 (d, J = 2.4
Hz, 1H), 8.08 (s, 1H), 10.71 (s, 1H); 13CNMR (400 MHz, DMSO-d6): 43.863, 113.684, 117.391,
119.252, 119.655, 120.061, 120.450, 125.272, 130.800, 131.118, 138.324, 147.980, 149.083,
164.471, 164.998, 167.818; UV spectrum (10 ppm) λmax 350.1 nm; MS (m/z) 395.0, 396.9 [M-H]
+; HPLC Purity % Area 99.13, RT 13.18 mins.
N-(1,5-dimethyl-3-oxo-2-phenyl-2,3-dihydro-1H-pyrazol-4-yl)-2-(5-(furan-2-ylmethylene)-2,4-
dioxothiazo lidin-3-yl)acetamide (F20)
Yield 42.5%; M.P. 254.7 °C (Charred); Brownish solid; IR (cm-1) 3201, 1735, 1708, 1680,
1
1616, 1593, 1109, 1288; H NMR (400 MHz, DMSO-d6, δ ppm) 2.10 (s, 3H), 3.04 (s, 3H), 4.45
(s, 2H), 6.77 (s, 1H), 7.16 (d, J = 3.2 Hz, 1H), 7.34 (d, J = 8.4 Hz, 3H)7.50 (t, J = 7.6 Hz, 2H),
24

13
7.81 (s, 1H), 8.08 (s, 1H) 10.71(s, 1H); CNMR (400 MHz, DMSO-d6) 11.117, 35.825, 43.347,
106.474, 113.658, 117.653, 119.492, 119.834, 123.682, 126.399, 129.086, 134.839, 147.910,
149.122, 152.062, 161.423, 164.732, 165.046, 167.819; UV spectrum (10 ppm) λmax 349.3 nm;
MS (m/z) 439.1 [M-H] +; HPLC Purity % Area 98.25, RT 3.41 mins.
N-(4-bromophenyl)-2-(5-(furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl)acetamide (F21)
Yield 45.5%; M.P. 242.2 °C (Charred); Buff white color solid; IR (cm-1) 3254, 1743, 1687,
1614, 1591, 1111, 1274, 815; 1H NMR (400 MHz, DMSO-d6, δ ppm) 4.49 (s, 2H), 6.78 (s, 1H),
7.18 (s, 1H), 7.51 (s, 4H), 7.81 (s, 1H), 8.09 (s, 1H), 10.53 (s, 1H); 13CNMR (400 MHz, DMSO-
d6) 43.868, 113.681, 115.338, 117.463, 119.620, 120.007, 121.131, 131.676, 137.672, 147.974,
149.102, 164.056, 165.035, 167.824; UV spectrum (10 ppm) λmax 350.3 nm; MS (m/z) 406.9,
405.1 [M-H] +; HPLC Purity % Area 96.4, RT 8.53 mins.
2-(5-(Furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl)-N-(pyridin-2-yl) acetamide (F22)
Yield 35.5%; M.P. 230.1 °C (Charred); Pale yellow solid; IR (cm-1) 3117, 1730, 1674, 1614,
1583, 1105, 1205; 1H NMR (400 MHz, DMSO-d6, δ ppm) 4.56 (s, 2H), 6.77-6.79 (m, 1H), 7.12
(dd, J = 6.8 Hz, 1H), 7.18 (d, J = 3.6 Hz, 1H), 7.77 (d, J = 1.2 Hz, 1H), 7.81 (s, 1H), 7.96 (m,
1H), 8.09 (d, J = 0.8 Hz, 1H), 8.34 (d, J = 4 Hz, 1H), 10.53 (s, 1H); 13CNMR (400 MHz, DMSO-
d6) 43.827, 113.489, 113.688, 117.494, 119.613, 119.846, 119.986, 138.397, 147.978, 148.085,
149.107, 151.309, 165.047, 167.848; UV spectrum (10 ppm) λmax 350.3 nm; MS (m/z) 330.0,
331.0 [M-H] +; HPLC Purity % Area 97.43, RT 3.79 mins.
2-(5-(Furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl)-N-(p-tolyl)acetamide (F23)
Yield 49.5%; M.P. 268.2 °C (Charred); Buff white color solid; IR (cm-1) 3267, 3041, 1687,
1658, 1618, 1597, 1111, 1257; 1H NMR (400 MHz, DMSO-d6, δ ppm) 2.25 (s, 3H), 4.47 (s, 2H),
6.78 (t, J = 1.6 Hz, 1H), 7.12 (d, J = 8.4 Hz, 2H), 7.18 (d, J = 3.2 Hz, 1H), 7.43 (d, J = 8.4 Hz,
13
2H), 7.81 (s, 1H), 8.09 (s, 1H), 10.29 (s, 1H); CNMR (400 MHz, DMSO-d6) 43.803, 113.675,
117.543, 119.165, 119.565, 119.927, 129.214, 132.680, 135.822, 147.948, 149.116, 163.547,
165.083, 167.854; UV spectrum (10 ppm) λmax 350.5 nm; MS (m/z) 340.9, 341.2, 342.0 [M-H] +;
HPLC Purity % Area 96.44, RT 5.75 mins.
N-(3,4-dibromophenyl)-2-(5-(furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl)acetamide (F24)
Yield 41.2%; M.P. 247.2 °C (Charred); Buff white color solid; IR (cm-1) 3255, 1735, 1672,
1610, 1585, 1111, 1288, 536; 1H NMR (400 MHz, DMSO-d6, δ ppm) 4.50 (s, 2H), 6.78 (s, 1H),
7.18 (d, J = 2.8 Hz, 1H), 7.46 (d, J = 8.8 Hz, 1H), 7.58 (d, J = 8.8 Hz, 1H), 7.81 (s, 1H), 7.91 (s,
1H), 8.09 (s, 1H), 10.71 (s, 1H); 13CNMR (400 MHz, DMSO-d6) 43.862, 113.699, 117.389,
119.269, 119.673, 120.068, 120.457, 125.273, 130.817, 131.114, 138.321, 147.992, 149.078,
25

164.481, 165.003, 167.824; UV spectrum (10ppm) λmax 350nm; MS (m/z) 394.9, 396.9 [M-H] +;
HPLC Purity % Area 99.27, RT 13.08 mins.
2-(5-(Furan-2-ylmethylene)-2,4-dioxothiazolidin-3-yl)-N-(4-methylthiazol-2-yl)acetamide
(F25)
Yield 36%; M.P. 268.7 °C (Charred); Buff white color solid; IR (cm-1) 3117, 2972, 1734, 1672,
1612, 1579, 1109, 1286, 1369; 1H NMR (400 MHz, DMSO-d6, δ ppm) 2.33 (s, 3H), 4.57 (s, 2H),
6.77-6.78 (m, 1H), 7.15 (d, 1H), 7.18 (d, 1H), 7.81 (s, 1H), 8.09 (s, 1H), 12.37 (s, 1H); 13CNMR
(400 MHz, DMSO-d6) 11.045, 43.147, 113.708, 117.435, 119.681, 120.078, 126.854, 148.005,
149.079, 164.956, 167.824; UV spectrum (10 ppm) λmax 350.8 nm; MS (m/z) 348.1, 349.0 [M-H]
+; HPLC Purity % Area 95.74, RT 13.1 mins.
4.2. GLUT1, GLUT4, and GLUT5 transport assay.
Synthesized compounds were examined for their effect on transport activity of human
GLUT1[47], GLUT4[47], and GLUT5[72] expressed in hexose transporter null yeast cells (hxt0).
Yeast cell culturing was done at 30 ºC with shaking (180 rpm). VW4000fgy1 yeast cells
expressing GLUT1[47] were cultured for 2-3 days in the Synthetic Complete media without
uracil (SC-uracil) with 2% (w/v) maltose. Cells were washed once in SC-uracil with 2% (w/v)
glucose media, transferred in the same media at OD600nm~0.5, and grown further for 1-2 days.
VW4000fgy1erg4 yeast cells expressing GLUT4[47] were cultured like GLUT1, but in media
with lower concentrations of maltose and glucose: SC-uracil with 1% (w/v) maltose media for
the initial cell culture, and SC-uracil with 0.2% (w/v) glucose media for the final cell culture.
VW4000 yeast cells expressing GLUT5[72] were cultured for one day in YEP [1% (w/v) yeast
and 2% (w/v) peptone] media with 2% (w/v) maltose and 100 µg/ml geneticin G418. Cells were
washed once in YEP media with 2% (w/v) fructose and 100 µg/ml geneticin G418, transferred in
the same media at OD600nm~0.5, and grown further for 1-2 days. For transport assay, cells in the
hexose media were centrifuged (1000xg, 5 minutes), washed once in PBS buffer (10 mM
Na2HPO4, 1.8 mM KH2PO4, 2.7 mM KCl, 137 mM NaCl, pH 7.4), then resuspended in PBS
buffer at an OD600nm ~ 10; each assay contained 100 µl of this cell solution. Transport assay was
started by the addition of C14-hexose to a final concentration of 5 mM glucose for GLUT1 or
GLUT4 and 10 mM fructose for GLUT5. Transport assay was stopped after 10 minutes by
adding 3-ml ice-chilled Quench buffer (0.1 M KPi, 0.1 M LiCl, pH 5.5), followed by filtration on
a glass fiber filter (GC50; Advantec, Tokyo, Japan) under vacuum, and another wash with 3-ml
Quench buffer and filtration. The filtration membranes were transferred to scintillation vials, and
10 ml of Scintillation buffer (BioSafeII; Research Products International, Mount Prospect, IL,
USA) was added in each vial. After brief vortexing, radioactivity was measured with a
scintillation counter (Tri-carb 2900TR, Perkin Elmer, USA). As all synthesized compounds were
solubilized in dimethyl sulfoxide (DMSO), controls for the determination of the relative activity
included 1% (v/v) DMSO - to account for DMSO concentration in the transport assay due to
26

inhibitor addition - as well as saturating concentrations of known inhibitors for GLUTs [200 µM
phloretin for GLUT1 and GLUT4[73], and 100 µM N-[4-(methylsulfonyl)-2-nitrophenyl]-1,3-
benzodioxol-5-amine (MSNBA) for GLUT5[74]]. All inhibitors were added to the assay from
stock concentrations so that the final concentration in the assay was 1% (v/v). Initial inhibition
screening for TZD compounds was done at 100 µM concentration; for the compounds that
decreased the relative transport activity by 50% or more, inhibitor concentration was varied in
the transport assay from 0.1 to 100 µM to determine inhibitor IC50. Data were analyzed with
GraphPad Prism (San Diego, CA, USA).
4.3. Docking studies.
The SiteFinder function of Molecular Operation Environment (MOE, Chemical Computing
Group, Montreal, Canada) was used to model the binding sites of compounds F18 and F19 in the
3D crystal structure of GLUT1 inward-facing conformation (PDB ID: 4PYP, www.rcsb.org),
and a GLUT1 outward-facing homology model based on the GLUT3 crystal structure (PDB ID:
4ZWC, www.rcsb.org). The latter model was generated in MOE with the Homology Model
function. Possible conformations of the compounds were generated with MOE Conformational
Generation function; F18 had 29 conformations, F19 had 104 conformations. Virtual docking
was carried out with the MOE Dock function with the Triangle Matcher placement scoring
London dG and Rigid Receptor refinement scoring GBVI/WSA dG against the dummy atoms
around the substrate binding site. Potential docking positions were selected with a selection
criterion of the lowest energy scoring algorithm and verifying that the docking pocket has
sufficient space for the compound binding and reasonable interactions with the protein residues.
The homology models and docking were validated with Ramachandran plots and docking of the
octyl glucoside, the original ligand in the crystal structure, to the GLUT1 crystal structure
(Supplementary Data, Figure S1).
4.4. Cell cultures.
Three human leukemia/lymphoma cancer cell lines (KCl-22, K562, and CCRF-CEM), one breast
cancer cell line (MDA-MB-231), and one normal, non-cancerous, fibroblast cell line (HS-27)
were used to test the cytotoxic potential of the synthetic derivatives. MDA-MB-231 and HS-27
cells were grown in DMEM medium (HyClone, Logan UT, USA) supplemented with 10% (v/v)
FBS, 100 U/ml penicillin and 100 μg/ml streptomycin. The culture medium for the
leukemic/lymphoid cancer cells was RPMI, supplemented in the same manner. Cell cultures
were maintained at 37 °C in a humidified 5% CO2 atmosphere. Cells were counted and seeded
for experimental use upon reaching ≥90% viability and 65–75% confluence.
4.5. Cell viability assay by differential nuclear staining assay.
27

The Differential Nuclear Staining (DNS) assay is a live cell imaging-based assay, which utilizes
two nuclear dyes to easily label living and dead cells[59]. An initial assessment of cytotoxicity
on this series of 25 compounds was conducted via the DNS assay upon exposure of CCRF-CEM,
K-562, KCL-22, and MDA-MB-231 human cancer cell lines to 50 μM of a compound for 48
hours. Compounds were dissolved and diluted in dimethyl sulfoxide (DMSO) to reach the
desired concentration. DMSO, hydrogen peroxide, and untreated cells were the vehicle, positive
and negative controls, respectively, in all experiments. Cells were seeded in 96-well plates at a
density of 10,000 cells/well in 100 μL of complete culture media, with the exception of HS-27,
for which only 3,000 cells/well were used to compensate for their large size. Cells were seeded
in 96-well plates at a density of 10,000 cells/well in 100 μL of complete culture media. Images
of stained cells were taken with the GE Healthcare Life Sciences IN Cell Analyzer 2200. Two
hours before imaging, Hoechst, and propidium iodide (PI) stains were added to each well to
distinguish living and dead/dying cells. Hoechst is a dye capable of permeating the membranes
of all cells within a sample, whereas PI will only penetrate cells with compromised membranes.
Colocalization of Hoechst (blue) and PI (red) signals indicate the dead cell population [59]. In
this initial screening, experimental samples were assessed singularly, and control samples in
quadruplicate. Subsequent DNS assays, using CCRF-CEM, K-562, KCl-22, and MDA-MB-231
cell lines, were then performed to determine the 50% cytotoxic concentration (CC50) values of
the seven compounds with the highest cytotoxicity in the initial screening. Similarly, the
selective toxicity of the two most active compounds, F18 and F19, was assessed using the non-
cancerous cell line HS-27. Cells were seeded identically but were treated with test compounds at
concentrations ranging from 0.5 to 100 μM. Experimental samples and controls were evaluated
for cytotoxicity in quadruplicate after 48 hours of incubation. Selectivity Index (SCI) values
were calculated by dividing the CC50 for non-cancerous cell line HS-27 by the CC50 of each
cancer cell line after 48 hours of incubation (SCI = Normal cell CC50 / Cancer cell CC50). High
SCI scores are suggestive of selective toxicity toward cancer cells, an effect which may or may
not be conferred in vivo.
4.6. Apoptosis studies by flow cytometry.
The induction of apoptosis by compound F18 was studied by flow cytometry, as described
earlier[75]. Cells were seeded in a 24-well flat-bottom microplate and incubated overnight at 37
0C in a CO2 incubator for 24 hrs. The cell media was replaced with fresh media, and then the
cells were treated with IC50 concentration of F18 for 24 hrs. Untreated cells were used as
negative control. Post incubation, cells were washed with PBS, then centrifuged for 5 minutes at
500 g at 4 °C, and the supernatant was discarded. Cell pellets were resuspended in ice-cold 1X
Binding Buffer and then mixed with 1 µL of annexin V-FITC solution and 5 µL PI (propidium
iodide). Tubes were kept on ice and incubated for 15 minutes in the dark, then 400 µL of ice-
cold 1X Binding Buffer was added, and cell preparations were analyzed by flow cytometry (BD
Accuri C5 flow cytometer, BD Biosciences, CA, USA). Cytometry data were analyzed with
28