Design, synthesis characterization and biological evaluation of novel multi-isoform ALDH inhibitors as potential anticancer agents

Article abstract of DOI:10.1016/j.ejmech.2019.111962

The aldehyde dehydrogenases (ALDHs) are a family of detoxifying enzymes that are overexpressed in various cancers. Increased expression of ALDH is associated with poor prognosis, stemness, and drug resistance. Because of the critical role of ALDH in cancer stem cells, several ALDH inhibitors have been developed. Nonetheless, all these inhibitors either lack efficacy or are too toxic or have not been tested extensively. Thus, the continued development of ALDH inhibitors is warranted. In this study, we designed and synthesized potent multi-ALDH isoform inhibitors based on the isatin backbone. The early molecular docking studies and enzymatic tests revealed that 3(a–l) and 4(a–l) are the potent ALDH1A1, ALDHA2, and ALDH3A1 inhibitors. ALDH inhibitory IC₅₀s of 3(a–l) and 4(a–l) were 230 nM to >10,000 nM for ALDH1A1, 939 nM to >10,000 nM for ALDH2 and 193 nM to >10,000 nM for ALDH3A1. The most potent compounds 3(h–l) had IC₅₀s for killing melanoma cells ranged from 2.1 to 5.7 μM, while for colon cancer cells, it ranged from 2.5 to 5.8 μM and for multiple myeloma cells ranging from 0.3 to 4.7 μM. Toxicity studies of 3(h–l) revealed that 3h to be the least toxic multi-ALDH isoform inhibitor. Mechanistically, 3(h–l) caused increased ROS activity, lipid peroxidation, and toxic aldehyde accumulation, secondary to potent multi-ALDH isoform inhibition leading to increased apoptosis and G2/M cell cycle arrest. Together, the study details the design, synthesis, and evaluation of potent, multi-isoform ALDH inhibitors to treat cancers.

Full text of DOI:10.1016/j.ejmech.2019.111962

Journal Pre-proof
Design, synthesis characterization and biological evaluation of novel multi-isoform
ALDH inhibitors as potential anticancer agents
Saketh S. Dinavahi, Raghavendra Gowda, Christopher G. Bazewicz, Madhu Babu
Battu, Jyh Ming Lin, Robert J. Christen, Manoj K. Pandey, Shantu Amin, Gavin P.
Robertson, Krishne Gowda
PII:
S0223-5234(19)31114-6
DOI:
https://doi.org/10.1016/j.ejmech.2019.111962
EJMECH 111962
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 11 September 2019
Revised Date: 3 December 2019
Accepted Date: 9 December 2019
Please cite this article as: S.S. Dinavahi, R. Gowda, C.G. Bazewicz, M.B. Battu, J.M. Lin, R.J. Christen,
M.K. Pandey, S. Amin, G.P. Robertson, K. Gowda, Design, synthesis characterization and biological
evaluation of novel multi-isoform ALDH inhibitors as potential anticancer agents, European Journal of
Medicinal Chemistry (2020), doi: https://doi.org/10.1016/j.ejmech.2019.111962.
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.
© 2019 Published by Elsevier Masson SAS.

Design, synthesis characterization and biological evaluation of novel multi- isoform ALDH
inhibitors as potential anticancer agents
Saketh S. Dinavahia,f, Raghavendra Gowdaa,f, Christopher G. Bazewiczf,g, Madhu Babu Battuj, Jyh
Ming Line, Robert J. Christeni, Manoj K. Pandeyi, Shantu Amina,h, Gavin P. Robertsona,b,c,d,f
Krishne Gowdaa,h*
,
aDepartment of Pharmacology, bPathology, cDermatology, dSurgery, eBiochemistry and Molecular
Biology, fMelanoma Skin Cancer Center, gCollege of Medicine, hPenn State Hershey Cancer
Institute, The Pennsylvania State University College of Medicine, Hershey, PA 17033, United
States; iDepartment of Biomedical Sciences, Cooper Medical School of Rowan University,
Camden, NJ 08103, United States; jLaboratory of Molecular Cell Biology, Centre for DNA
Fingerprinting and Diagnostics, Uppal, Hyderabad, 500039, India.
*Corresponding author: Krishne Gowda, Department of Pharmacology, The Pennsylvania State
University College of Medicine, 500 University Drive, Hershey, PA 17033, United States. Phone:
(717) 531-0003 Ext. 285014; Fax: (717) 531-0244; E-mail: kgowda@pennstatehealth.psu.edu
Funding sources: G.P.R. acknowledges the Penn State Chocolate Tour Cancer Research fund and
Melanoma Research Alliance fund.
Short title: Novel multi-ALDH isoform inhibitors.
Word count: 6513
Key words: Aldehyde dehydrogenases, Cancer therapeutics, Colon cancer, Multiple Myeloma,
Drug design, Drug development, Molecular docking.
Conflicts of interest: The authors declare no conflicts of interest.

Abstract
The aldehyde dehydrogenases (ALDHs) are a family of detoxifying enzymes that are overexpressed
in various cancers. Increased expression of ALDH is associated with poor prognosis, stemness, and
drug resistance. Because of the critical role of ALDH in cancer stem cells, several ALDH inhibitors
have been developed. Nonetheless, all these inhibitors either lack efficacy or too toxic or not been
tested extensively. Thus, the continued development of ALDH inhibitors is warranted. In this study,
we designed and synthesized potent multi-ALDH isoform inhibitors based on the isatin backbone.
The early molecular docking studies and enzymatic tests revealed that 3(a-l) and 4(a-l) are the potent
ALDH1A1, ALDHA2, and ALDH3A1 inhibitors. ALDH inhibitory IC50s of 3(a-l) and 4(a-l) were
230 nM to >10,000 nM for ALDH1A1, 939 nM to >10,000 nM for ALDH2 and 193 nM to >10,000
nM for ALDH3A1. The most potent compounds 3(h-l) had IC50s for killing melanoma cells ranged
from 2.1 to 5.7 µM, while for colon cancer cells ranged from 2.5 to 5.8 µM and for multiple
myeloma cells ranged from 0.3 to 4.7 µM. Toxicity studies of 3(h-l) revealed that 3h to be the least
toxic multi-ALDH isoform inhibitor. Mechanistically, 3(h-l) caused increased ROS activity, lipid
peroxidation, and toxic aldehyde accumulation, secondary to potent multi-ALDH isoform inhibition
leading to increased apoptosis and G2/M cell cycle arrest. Together, the study details the design,
synthesis, and evaluation of potent, multi-isoform ALDH inhibitors to treat cancers.

1. Introduction
The human aldehyde dehydrogenases (ALDHs) are a family of 19 enzymes that oxidize endogenous
and exogenous aldehydes to less reactive, more soluble carboxylic acids in a NAD(P)+-dependent
reaction [1, 2]. The ALDHs protect cells against oxidative stress and reactive oxygen species (ROS),
which trigger lipid peroxidation and the accumulation of toxic aldehydes, such as acrolein,
malondialdehyde (MDA), 4-hydroxy-2-nonenal (4-HNE) and 4-hydroxy-2-hexanal (4-HHE) [3].
Collection of such aldehydes damages cells by forming protein adducts through non-enzymatic,
covalent bonds with lysine, cysteine, and histidine residues, as well as through increased ROS
generation and lipid peroxidation [3-5]. Consequently, cancer cells often overexpress ALDH
isoforms to combat oxidative stress secondary to high metabolic demands, radiation, and ROS-
generating chemotherapeutics, including paclitaxel, doxorubicin, staurosporine, and sorafenib [6].
Overexpression of ALDH isoforms is associated with cancer progression, resistance and poor
prognosis in numerous cancers including breast, lung, colorectal, esophageal, Ewing’s sarcoma, head
and neck squamous cell carcinoma (HNSCC), melanoma, mesothelioma, multiple myeloma,
neuroblastoma, prostate, and pancreatic cancer [7-18]. ALDH1A1 and ALDH3A1 directly
metabolize the active metabolites of cyclophosphamide, contributing to resistance against this
chemotherapeutic drug [19]. Cancer cells with a stem-cell-like phenotype, which are a small subset
of multipotent cells within tumors responsible for tumor initiation, progression, resistance, and
metastasis, often express high levels of ALDH [20, 21]. Further, retinoic acid generated by
ALDH1A1, ALDH1A2, and ALDH1A3 can lead to the induction and function of tumor-infiltrating
regulatory T cells, which create an immunosuppressive tumor microenvironment and impair tumor
immunity [22].
Due to the role of ALDH overexpression in cancer progression and therapy resistance, many ALDH
inhibitors have been developed [23]. Current ALDH inhibitors can be classified into multi-ALDH
isoform inhibitors and isoform-specific inhibitors, which primarily inhibit one isoform [23]. Multi-
ALDH isoform inhibitors include N,N-diethylaminobenzaldehyde (DEAB), 4-dimethylamino-4-
methyl–pent-2-ynthioic acid-S-methylester (DIMATE), citral, ALDH inhibitors (aldis)-1, -2, -3, -4
and -6 and dyclonine [23-33]. Each inhibitor exhibits antiproliferative effects on cultured cancer
cells, particularly as combinatorial therapy, but only DEAB, DIMATE, citral, aldi-6, and dyclonine
have been tested in in vivo cancer models [23-33]. DEAB, DIMATE, citral, and aldi-6 show

antitumor efficacy as monotherapy; however, variations in ALDH expression hinder the efficacy of
DEAB, the oral bioavailability of DIMATE has not been reported, and citral has a range of off-target
effects leading to toxicity [24, 26-30, 32]. Aldi-6 at 24 mg/kg/day inhibits HNSCC xenograft
development by 60% with no toxicity observed [32]. However, it was administered via an osmotic
mini-pump, and due to its recent development, has not been tested further [32].
Isoform-specific ALDH inhibitors include the ALDH1A1 specific inhibitors Cpd 3, CM037,
CM026, NCT-501, NCT-505, NCT-506, 13g and 13h, the ALDH2 specific inhibitor CVT10216, and
the ALDH3A1 specific inhibitors CB7 and CB29 [23, 34-40]. Minimal in vivo data exist for these
compounds, either due to limited efficacy, poor bioavailability, or recent development [23, 34-40].
NCT-501 has been tested in vivo, and leads to a 78% reduction in HNSCC xenograft growth;
however, it was administered via intratumoral injection [36]. Thus, the development of novel, potent,
non-toxic, bioavailable ALDH inhibitors is still needed [41]. Targeting more than one ALDH
isoform is likely necessary for optimal anticancer therapy, as ALDH activity within cancer cells is
generally a composite of the activity of multiple ALDH isoforms [1, 2, 41].
In the present study, a series of novel, potent, multi-ALDH isoform inhibitors were designed and
synthesized. Ligand-based docking studies were used to identify 3(a-l) and 4(a-l). 3(a-l) and 4(a-l)
indicated strong docking scores for ALDH1A1, ALDH2, and ALDH3A1 and were subsequently
synthesized and ALDH inhibitory activity evaluated. Compounds 3(h-l) had the highest potency
against ALDH1A1, ALDH2, and ALDH3A1 and showed efficacy and selectivity for killing cultured
colon cancer, multiple myeloma, and melanoma cell lines. Toxicity in animals was significant for
3(i-l), whereas 3h was not toxic. Thus, 3h may be useful for the treatment of cancer.
Mechanistically, 3(h-l) led to a significant increase in toxic aldehyde accumulation, ROS activity,
and lipid peroxidation mediated by potent ALDH1A1, ALDH2, and ALDH3A1 inhibition leading to
increased apoptosis and G2/M arrest of the cell cycle. N-acetyl cysteine (NAC), a ROS scavenger
reduced the activity of the active compounds demonstrating the significance of ROS activity in
ALDH pathway.

2. Results and Discussion
2.1. Molecular docking studies
Multi-ALDH isoform inhibition may be more useful for cancer treatment since ALDH activity
within cancer cells is a composite of multiple ALDH isoforms [1, 2]. Cpd 3, an isatin analog, is one
of the most potent ALDH1A1 specific inhibitors (Fig. 1) [34]. We have previously identified KS99
(Fig. 1), a similar isatin analog, to be a potent ALDH inhibitor to reduce cancer growth [42-44]. It
exhibits multi-ALDH isoform inhibition with IC50s of 350 nM, 1.5 µM, and 850 nM for ALDH1A1,
ALDH2, and ALDH3A1, respectively (unpublished data), but is toxic in animals starting at 5
mg/kg/day [44]. Thus, we attempted to optimize the structure of KS99 using molecular docking
studies to develop novel, non-toxic, potent, multi-ALDH isoform inhibitors.
A series of compounds 3(a-l) and 4(a-l) were designed and tested for the ability to bind in the active
site pockets of ALDH1A1, ALDH2, and ALDH3A1 using molecular docking studies. 1,4-
bis(bromomethyl) benzene was selected as a linker to connect the isatin scaffold and isothiourea
moieties. The protein structures of ALDH1A1, ALDH2, and ALDH3A1 co-crystallized with the
corresponding potent, isoform-specific ALDH inhibitors CM037 (ALDH1A1), psoralen (ALDH2)
and CB7 (ALDH3A1) were selected. The designed compounds were first docked into the ligand-
binding pocket of ALDH1A1. Significant interactions identified between the crystal ligand, CM037,
and ALDH1A1 were a π-π interaction with the W178 residue and an H-bond interactions with the
Gly458 and Ser121 residues along with interactions with Cys303. Isatin did not exhibit any of these
interactions with ALDH1A1; however, KS99 had a similar π-π interaction with the W178 residue
and H-bond interactions with the Gly458 and Ser121 residues of ALDH1A1. Cpd 3 had interactions
with W178 and S121. Importantly, compounds 3(a-l) and 4(a-l) shared similar interactions (Fig. 2
and Fig. S1) with residues in the ligand-binding pocket of ALDH1A1 compared to CM037 and
KS99, suggesting they could potentially be inhibitors of ALDH1A1.
Docking studies were similarly conducted with compounds and the ALDH2 and ALDH3A1 protein
structures. π-π interactions with the F459 residue and H-bond interactions with the L269 residues
occurred between ALDH2 and the crystal ligand, psoralen. Similarly, π-π interactions with the T115
residue occurred between ALDH3A1 and the crystal ligand, CB7. Importantly, compounds 3(a-l)
and 4(a-l) shared similar interactions with residues in the ligand-binding pockets of ALDH2 and
ALDH3A1 compared to psoralen and CB7, respectively (Fig. 2 and Fig. S1), suggesting they could

be inhibitors of ALDH2 and ALDH3A1. As Cpd 3 is not an inhibitor of ALDH2 and ALDH3A1, it
did not have these interactions. Docking scores for 3(a-l) and 4(a-l) ranged from -7.495 to -11.938
for ALDH1A1, -6.756 to -11.205 for ALDH2 and -12.119 to -14.564 for ALDH3A1 (Table 1).
Based on these strong docking scores, 3(a-l) and 4(a-l) were synthesized for further analysis of
ALDH enzyme inhibitory activity, anticancer efficacy, and toxicity.
2.2. Chemistry
The synthesis of target compounds, substituted 2-[4-(2,3-dioxo-2,3-dihydroindol-1-
ylmethyl)benzyl]isothiourea hydrobromides 3(a-l) and 2-[4-(2,3-dioxo-2,3-dihydroindol-1-
ylmethyl)benzyl]isoselenourea hydrobromide analogs 4(a-l) are illustrated in Scheme 1. The key
intermediates 2(a-l) were prepared in one step. Initially, unsubstituted, 5 or 7 mono substituted, and
5,7-disubstituted isatins were reacted with 1,4-bis(bromomethyl)benzene in the presence of
potassium carbonate in DMF to the corresponding N-(p-bromomethylbenzyl) isatins 2(a-l) in a good
yield. These intermediates 2(a-l) were then refluxed with thiourea in ethanol to produce the
corresponding 2-[4-(2,3-dioxo-2,3-dihydroindol-1-ylmethyl)benzyl]isothiourea hydrobromides 3(a-
l) and refluxed with selenourea in ethanol to yield 2-[4-(2,3-dioxo-2,3-dihydroindol-1-
ylmethyl)benzyl]isoselenourea hydrobromide analogs 4(a-l) in an excellent yield. The structures of
all isatin derivatives were confirmed by 1H NMR, 13C NMR, and HRMS analysis. The compound
purity (>98%) was confirmed by analytical high-performance liquid chromatography (HPLC) before
proceeding for in vitro biological assays.
2.3. ALDH isoform inhibitory activity
All the synthesized compounds 3(a-l) and 4(a-l) were assessed for the inhibition of ALDH1A1,
ALDH2, and ALDH3A1 enzyme activity at various concentrations, and the results were summarized
in Table 2. The enzymes inhibition were evaluated by measuring the conversion of NAD+ to NADH
following the addition of isoform-specific aldehydes in the presence of 3(a-l) and 4(a-l). ALDH
inhibitory IC50s activity of 3(a-l) and 4(a-l) were 230 nM to >10,000 nM for ALDH1A1, 939 nM to
>10,000 nM for ALDH2 and 193 nM to >10,000 nM for ALDH3A1. (Table 2; dose-response
curves in Fig. S2). 3(h-l), 4b, and 4(j-l) had the most potent inhibition of ALDH1A1, ALDH2, and
ALDH3A1 at the concentrations tested and were considered potent, multi-ALDH isoform inhibitors.
The most potent multi-ALDH isoform inhibitor, on average, was 3j, which had IC50s of 230 nM,
1542 nM, and 193 nM for ALDH1A1, ALDH2, and ALDH3A1 enzyme activity (Table 2).

Several trends in the structure-activity relationship of compounds 3(a-l) and 4(a-l) were noted
(Tables 1 and 2). Compounds (X=S, series 3) with isothiourea moiety generally had greater multi-
ALDH isoform inhibitory activity than that of corresponding isoselenourea compounds (X=Se,
series 4). The lower inhibitory activity of selenium analogs may vary due to the larger size of
selenium than a sulfur atom, which may interfere with the binding in the active-site pocket. For
instance, 3h and 3j were more potent ALDH inhibitors than 4h and 4j, respectively. ALDH
inhibitory activity of 3(a-l) and 4(a-l) depended on the halogen substitution at R1 and/or R2.
Specifically, -dibromo substitutions (3j, 4j) led to the best ALDH inhibition, followed by -dichloro
(3k, 4k), -fluoro, bromo (3l, 4l), trifluoromethyl (3h), -fluoro (3f, 4f) and finally un-substituted (3a,
4a) compounds. Also, 5,7-disubstituted compounds (3j, 3k) were more effective compared to 5-
substituted (3b, 3d) or 7-substituted (3c, 3e) compounds. Further, 7-substituted compounds (3c, 3e)
were more effective than 5-substituted compounds (3b, 3d). Finally, 5,7-dibromo substitutions (3j,
4j) had greater ALDH inhibitory activity compared to 5-fluoro,7-bromo substitutions (3l, 4l).
Among all the compounds, 5,7-dibromo substitutions ultimately had the best ALDH inhibitory
activity, may be due to larger size of bromine compared to other halogens and more hydrophobic
nature of bromine, which facilitated the interaction in the hydrophobic binding pocket.
2.4. Cellular activity
Since isothiourea compounds (series 3) were in general, more potent ALDH1A1, ALDH2, and
ALDH3A1 inhibitors compared to the corresponding isoseleno analogs (series 4), only 3(h-l), the
most potent inhibitors in series 3, were tested for antiproliferative effects on cultured cancer cells.
Specifically, 3(h-l) were evaluated for inhibition of proliferation of cultured melanoma cells (UACC
903 and 1205 Lu) as ALDH overexpression is important in melanoma progression [45, 46]. The
range of IC50s against UACC 903 cells was 3 to 5.7 µM, and for 1205 Lu cells was 2.1 to 5.7 µM
(Table 3; dose-response curves in Fig. S3), with 3j and 3k being the most effective across both cell
lines. Cell killing by 3(h-l) was also specific for melanoma cells compared to normal human
fibroblasts (FF2441). Specifically, 3(h-l) were 2- to 3.8-fold more selective for killing melanoma
cells (Table 3). Importantly, Cpd 3 and the inactive compound 3a had IC50s greater than 100 µM in
all the cell lines evaluated, demonstrating the importance of substitutions on the isatin ring of the
compounds synthesized.

Subsequently, 3(h-l) were evaluated for antiproliferative effects in several cancer types. Colon
cancer cells (HCT116 and HT29) were studied as ALDH overexpression is also crucial in colon
cancer progression [11, 47, 48]. Average IC50s for each compound across both cell lines were 5.3
µM for 3h, 5.15 µM for 3i, 2.7 µM for 3j, 2.9 µM for 3k and 5.1 µM for 3l (Table 3; dose-response
curves in Fig. S3). Compounds 3j and 3k were most effective in inhibiting the colon cancer cell
survival likely due to their strong inhibition of ALDH1A1 and ALDH3A1. Multiple myeloma cells
were also examined, as ALDH1A1 overexpression has been associated with stemness, therapy
resistance, and poor outcomes in this cancer type [10, 49-52]. Average IC50s for 3(h-l) across all
multiple myeloma cell lines tested (NCIH929, U266, RPMI8226, MM.1R, MM.1S) were 1.9 µM for
3h, 3.8 µM for 3i, 1 µM for 3j, 1.6 µM for 3k and 2.4 µM for 3l (Table 3; dose-response curves in
Fig. S4). Compounds 3h, 3j, and 3k showed more potent growth inhibition of multiple myeloma
cells when compared to melanoma and colon cancer cells, demonstrating the more effectiveness of
these compounds even in hematological malignancies. Additionally, these compounds displayed
better IC50s at killing melanoma cell with time, and IC50s of 3j were 7.2 µM at 24 hours compared to
4.1 µM at 48 hours and 3.7 µM at 72 hours (Fig. S5). Thus, 3(h-l) were specific to cancer cells and
displayed antiproliferative activity against cultured melanoma, colon cancer, and multiple myeloma
cells, indicating the potential for these compounds to be translated into the clinic. Moreover, 3(h-l)
displayed chemical properties predictive of good solubility, absorption, metabolism, and excretion
(Table 4), indicating the drug-like properties of these compounds. All these compounds adhered to
Lipinski’s rule of five for drug-like compounds.
2.5. Toxicity studies
Since compounds 3(h-l) were identified to be potent, multi-ALDH isoform inhibitors with
antiproliferative activity in multiple cancer types, the toxicity of these compounds were evaluated for
translatability to the clinic. Previously, we identified that KS99 was toxic in animals at 5 mg/kg/day
[44]. Based on this data, we selected 5 mg/kg/day dose for the toxicity study with 3(h-l) compounds.
Specially, Swiss-Webster mice were treated with 5 mg/kg/day of 3(h-l) i.p. for 14 days, and animal
weight was compared to vehicle controls (Table 5). Compounds 3(i-l) led to significant weight loss
(10-15% body weight) after 14 days of treatment, while 3h led to no significant weight loss (toxicity
timeline in Fig. S6). Thus, 3h was identified to be the least toxic agent, which may be due to lesser

ALDH inhibitory activity when compared to compounds 3(i-l). Toxicity of 3(i-l) could be mitigated
using controlled release formulations such as nanoliposomes [53].
2.6. ROS and lipid peroxidation activity and toxic aldehyde accumulation
Accumulation of toxic aldehydes (e.g., 4-HNE, 4-HHE, MDA, acrolein) secondary to ALDH
inhibition can lead to protein adduct formation, increased ROS levels and lipid peroxidation,
ultimately causing cell damage and apoptosis [3-5]. Thus, to evaluate the mechanism by which the
ALDH inhibitors killed the cultured cancer cells, a ROS assay was performed using DCFDA dye
[54]. Specifically, colon cancer cells, HCT116 and SW480 were treated with 5 µM of the most
potent ALDH inhibitor identified, compounds 3j and the least toxic 3h for 24 hours and ROS
activity in treated cells was compared to DMSO. An inactive molecule in the series, 3a was also
included as a negative control, and H2O2 was used as a positive control. As shown in Figure 3A and
3B, compounds 3h and 3j significantly increased ROS levels in both colon cell lines, indicating
elevated ROS levels likely contribute to their anti-proliferative effects. Additionally, compound 3a,
an inactive derivative, did not significantly increase the ROS activity in any of the cell lines
evaluated. Importantly, the ROS-inducing activity of compounds 3h and 3j was abrogated by the
addition of N-Acetyl Cysteine (NAC), a scavenger of ROS activity in cells, indicating that the
compounds affect the ROS-pathway.
To evaluate if 3h and 3j led to increased lipid peroxidation and toxic aldehyde accumulation, a lipid
peroxidation assay was performed using a TBARS assay kit [55]. Specifically, HCT116 cells were
treated with 5 µM of 3h and 3j for 24 hours, and lipid peroxidation activity and toxic aldehyde
accumulation in treated cells were compared to DMSO. As shown in Figure 3C, compounds 3h and
3j significantly increased lipid peroxidation and toxic aldehyde accumulation in HCT116 colon
cancer cell line, likely contributing to their anti-proliferative effects. Additionally, 3a was
ineffective in increasing the lipid peroxidation; while the addition of NAC abrogated the effects of
3h and 3j, indicating the importance of ROS pathway in the accumulation of toxic aldehydes by
these ALDH inhibitors.
Further, the effect of the addition of NAC on the anti-proliferative and apoptotic activity of ALDH
inhibitors was also evaluated. Addition of NAC increased the cell killing IC50s of 3h and 3j in colon
cancer cell line, HCT116 by 6-fold, and 8-fold, respectively (Fig. 3D). Similar activity was
observed in apoptosis and cell cycle assays where NAC abrogated the activity of the potent ALDH

inhibitors. When HCT116 colon cancer cells were treated with 3h and 3j at 5 µM for 24 hours, the
percentage of cells which are both Annexin-V and 7-AAD positive were significantly higher than the
DMSO control-treated cells or inactive compound, 3a treated cells (Fig. 3E; representative dot plots
in Fig. S7). When NAC was added to these compounds, there was no increase in apoptotic cells
compared to DMSO treated cells. Similarly, compounds 3j and 3h caused a G2/M arrest in the cell
cycle activity of the colon cancer cell line HCT116, which was truncated by addition of NAC (Fig.
3F; representative histograms in Fig. S8). These data suggest that the potent ALDH inhibitors
induce ROS activity, lipid peroxidation and accumulation of toxic aldehydes to inhibit cell survival
and induce apoptosis through the modulation of ROS pathway.
3. Conclusion
In summary, this report details the design and synthesis of a series of novel, potent, multi-isoform
ALDH inhibitors. Compounds with the best multi-ALDH isoform inhibition 3(h-l) were evaluated
for antiproliferative effects in multiple cancer types and displayed efficacy and selectivity for
cultured cancer cells. Compound 3h was not toxic in animals and thus may be the most promising
anticancer drug in this study. Compounds 3(i-l) were toxic, however, nanoliposomal formulations of
3(i-l) could be developed to improve toxicity profiles as well as bioavailability. Mechanistically,
potent multi-isoform ALDH inhibition led to significantly increased ROS activity, lipid
peroxidation, and toxic aldehyde accumulation.
4. Experimental
4.1. Chemistry
4.1.1. General methods and materials
Isatin was purchased from Sigma-Aldrich (Sigma-Aldrich, St. Louis, MO, USA), 5,7-dibromo isatin
was synthesized using previously reported methods [42]. Other chemicals were purchased from
commercial vendors. Anhydrous solvents were used for all experiments. Reactions were carried out
using dried glassware and under an atmosphere of nitrogen. Reaction progress was monitored with
analytical thin-layer chromatography (TLC) on aluminum-backed precoated silica gel 60 F254 plates
(E. Merck). The N-benzylisatins were highly colored and would usually be seen on a TLC plate;
colorless compounds were detected using UV light and/or iodine vapor. Column chromatography
was carried out using silica gel 60 (230–400 mesh, E. Merck) with the solvent system indicated in

the individual procedures. NMR spectra were recorded using a Bruker Avance II 500 or 600 MHz
spectrometers. Chemical shifts (δ) were reported in parts per million downfield from the internal
standard. The signals are quoted as s (singlet), d (doublet), t (triplet), m (multiplet), dd (doublet of
doublet), ddd (doublet of doublets of doublets), dt (doublet of triplets). Spectra are referenced to the
residual solvent peak of the solvent stated in the individual procedure. High-resolution mass spectra
(HRMS) were determined in 5600 (QTOF) TripleTOF using a Duospray™ ion source (Sciex,
Framingham, MA). The capillary voltage was set at 5.5 kV in positive ion mode with a declustering
potential of 80V. The mass spectrometer was scanned from 50 to 1000 m/z in operating mode with a
250 ms scan from 50 to 1000 m/z. Melting points were determined on a Fischer-Johns melting point
apparatus and are uncorrected. The purity of the compound was established by HPLC using an HP-
Agilent 1200 HPLC system on a C18 column, and all the compounds had a purity of >98% unless
mentioned.
4.1.2. General procedure for the synthesis of compounds 2(a-l)
Initially, mono (5 or 7) or di-substituted (5,7) or unsubstituted isatins 1(a-l) (10 mmol) were
dissolved in anhydrous DMF (30 mL) and cooled on ice with stirring. Solid dry K2CO3 (11 mmol)
was added in one portion, and the dark-colored suspension was brought to room temperature and
stirred for a further 1 h. 1,4-bis(bromomethyl)benzene (40 mmol) was added slowly with constant
stirring until the mono or di-substituted isatin starting material had been consumed (TLC). The
reaction mixture was poured into cold water and extracted with ethyl acetate. The ethyl acetate layer
was washed with water, brine, and dried over MgSO4. The solvent was removed, and the crude
product was purified by silica gel column chromatography using (hexanes/EtOAc, 80:20) as eluent
to yield the key intermediates (-N-(p-bromomethyl benzyl)isatins 2(a-l) (yield 75-80%) as orange-
red crystals.
4.1.3. General procedure for the synthesis of compounds 3(a-l)
To each unsubstituted, mono, and di-substituted (-N-(p-bromomethyl benzyl)isatins 2(a-l)
(1.02mmol), thiourea (1.02 mmol) and ethanol (25 ml) was added and heated to reflux until the
starting material had been disappeared (TLC). The solvent was removed under vacuum. The crude
product was washed with ethyl acetate to remove unreacted (-N-(p-bromomethyl benzyl)isatins. The
products 3(a-l) were recrystallized by ethanol-ethyl acetate with good yields.

4.1.4. General procedure for the synthesis of compounds 4(a-l)
To each unsubstituted, mono, and di-substituted (-N-(p-bromomethyl benzyl)isatins 2(a-l)
(1.02mmol), selenourea (1.02 mmol) and ethanol (25 ml) was added and heated to reflux until the
starting material had been disappeared (TLC). The solvent was removed under vacuum. The crude
product was washed with ethyl acetate to remove unreacted (-N-(p-bromomethyl benzyl)isatins. The
products 4(a-l) were recrystallized by ethanol-ethyl acetate with good yields.
4.1.5. 2-[4-(2,3-Dioxo-2,3-dihydroindol-1-ylmethyl)benzyl]isothiourea hydrobromide (3a) (KS104).
Yellow solid, Yield: 83%; mp: 208-210 °C; 1H NMR (500 MHz, DMSO-d6): δ 9.18 (s, 2H), 8.98 (s,
2H), 7.61 - 7.57 (m, 2H), 7.46 - 7.38 (m, 4H), 7.13 (dt, J=0.6, 7.5 Hz, 1H), 6.98 (dd, J=0.5, 8.2 Hz,
1H), 4.92 (s, 2H), 4.47 (s, 2H). 13C NMR (150 MHz, DMSO-d6): δ 183.5, 169.5, 158.8, 150.8, 138.5,
135.9, 134.8, 129.8, 128.3, 125.0, 123.9, 118.3, 111.5, 43.1, 34.4. MS (ESI) m/z 326 [M+H]; HRMS
(ESI) m/z for C17H15N3O2S calculated 326.0885, found m/z: 326.0963.
4.1.6. 2-[4-(5-Bromo-2,3-dioxo-2,3-dihydroindol-1-ylmethyl)benzyl]isothiourea hydrobromide (3b)
(KS108).
Orange solid, Yield: 75%; mp: 217-219 °C; 1H NMR (500 MHz, DMSO-d6): δ 9.15 - 8.96 (m, 4H),
7.78 - 7.75 (m, 2H), 7.46 - 7.37 (m, 4H), 6.93 (dd, J=0.5, 8.2 Hz, 1H), 4.91 (s, 2H), 4.46 (s, 2H). 13C
NMR (150 MHz, DMSO-d6): δ 182.3, 169.5, 158.5, 149.6, 140.1, 135.6, 134.8, 129.7, 128.2, 127.2,
120.1, 115.6, 113.6, 43.1, 34.4. MS (ESI) m/z 404 [M+H]; HR-MS (ESI) m/z for C17H14BrN3O2S
calculated 403.9990, found m/z: 404.0075.
4.1.7. 2-[4-(7-Bromo-2,3-dioxo-2,3-dihydroindol-1-ylmethyl)benzyl]isothiourea hydrobromide (3c)
(KS110).
Yellow solid, Yield: 78%; mp: 206-208 °C; 1H NMR (500 MHz, DMSO-d6): δ 9.18 (s, 2H), 8.99 (s,
2H), 7.76 (dd, J=1.2, 8.1 Hz, 1H), 7.65 (dd, J=1.2, 7.3 Hz, 1H), 7.39 (m, 4H), 7.10 (dd, J=7.4, 8.1
Hz, 1H), 5.26 (s, 2H), 4.48 (s, 2H). 13C NMR (150 MHz, DMSO-d6): δ 182.3, 169.5, 159.9, 147.5,
143.3, 137.5, 134.1, 129.6, 127.2, 125.6, 124.5, 122.1, 103.6, 44.4, 34.5. MS (ESI) m/z 404 [M+H];
HRMS (ESI) m/z for C17H14BrN3O2S calculated 403.9990, found m/z: 404.0091.
4.1.8. 2-[4-(5-Chloro-2,3-dioxo-2,3-dihydroindol-1-ylmethyl)benzyl]isothiourea hydrobromide (3d)
(KS112).

Orange solid, Yield: 77%; mp: 226-228 °C; 1H NMR (500 MHz, DMSO-d6): δ 9.17 (s, 2H), 8.97 (s,
2H), 7.65 - 7.63 (m, 2H), 7.42 (dd, J=8.2, 26.3 Hz, 4H), 6.99 (dd, J=1.6, 7.4 Hz, 1H), 4.92 (s, 2H),
4.48 (s, 2H). 13C NMR (150 MHz, DMSO-d6): δ 182.4, 169.5, 158.6, 149.2, 137.3, 135.6, 134.8,
129.7, 128.3, 128.1, 124.5, 119.7, 113.2, 43.2, 34.4. MS (ESI) m/z 360 [M+H]; HRMS (ESI) m/z for
C17H14ClN3O2S calculated 360.0495, found m/z: 360.0570.
4.1.9. 2-[4-(7-Chloro-2,3-dioxo-2,3-dihydroindol-1-ylmethyl)benzyl]isothiourea hydrobromide (3e)
(KS114).
Orange solid, Yield: 73%; mp: 225-227 °C; 1H NMR (600 MHz, DMSO-d6): δ 9.18 (s, 2H), 9.03 (s,
2H), 7.62 (d, J=2.4 Hz, 1H), 7.61 (dd, J=1.1, 4.4 Hz, 1H), 7.40 (m, 4H), 7.17 (t, J=7.7 Hz, 1H), 5.22
(s, 2H), 4.51 (s, 2H). 13C NMR (150 MHz, DMSO-d6): δ 182.3, 169.5, 159.8, 146.0, 139.9, 137.5,
134.2, 129.6, 127.2, 125.3, 124.1, 121.8, 116.2, 44.8, 34.5. . MS (ESI) m/z 360 [M+H]; HRMS (ESI)
m/z for C17H14ClN3O2S calculated 360.0495, found m/z: 360.0584.
4.1.10. 2-[4-(5-Fluoro-2,3-dioxo-2,3-dihydroindol-1-ylmethyl)benzyl]isothiourea hydrobromide (3f)
(KS116).
Orange solid, Yield: 75%; mp: 208-210 °C; 1H NMR (600 MHz, DMSO-d6): δ 9.17 (s, 2H), 9.03 (s,
2H), 7.50 (dd, J=2.4, 7.0 Hz, 1H), 7.47 (d, J=8.3 Hz, 1H), 7.46 (d, J=8.3 Hz, 2H), 7.41 (d, J=8.4 Hz,
2H), 7.00 (dd, J=3.8, 8.6 Hz, 1H), 4.92 (s, 2H), 4.51 (s, 2H). 13C NMR (150 MHz, DMSO-d6): δ
182.9, 169.5, 159.0, 158.9, 147.0, 135.7, 134.8, 129.8, 128.3, 124.3, 119.2, 112.9, 112.0, 43.2, 34.4.
MS (ESI) m/z 344 [M+H]; HRMS (ESI) m/z for C17H14FN3O2S calculated 344.0791, found m/z:
344.0883.
4.1.11. 2-[4-(7- Fluoro -2,3-dioxo-2,3-dihydroindol-1-ylmethyl)benzyl]isothiourea hydrobromide
(3g) (KS118).
Orange solid, Yield: 71%; mp: 210-212 °C; 1H NMR (600 MHz, DMSO-d6): δ 9.20 (s, 2H), 9.06 (s,
2H), 7.52 (dd, J=8.6, 11.6 Hz, 1H), 7.48 (d, J=7.3 Hz, 1H), 7.41 (m, 4H), 7.16 (ddd, J=7.9, 7.9, 3.8
Hz, 1H), 4.97 (s, 2H), 4.51 (s, 2H). 13C NMR (150 MHz, DMSO-d6): δ 182.3, 169.5, 158.9, 147.7,
136.8, 136.5, 134.6, 129.7, 127.7, 126.2, 125.1, 121.5, 121.3, 45.3, 34.4. MS (ESI) m/z 344 [M+H];
HRMS (ESI) m/z for C17H14FN3O2S calculated 344.0791, found m/z: 344.0882.

4.1.12. 2-[4-(2,3-Dioxo-5-trifluoromethyl-2,3-dihydroindol-1-ylmethyl)benzyl]isothiourea
hydrobromide (3h) (KS106).
Yellow solid, Yield: 76%; mp: 225-227 °C; 1H NMR (500 MHz, DMSO-d6): 1H NMR (500 MHz,
DMSO) δ 9.17 (s, 2H), 9.00 (s, 2H), 7.96 (dd, J=1.3, 8.4 Hz, 1H), 7.89 (d, J=1.7 Hz, 1H), 7.49 - 7.39
(m, 4H), 7.15 (d, J=8.4 Hz, 1H), 4.98 (s, 2H), 4.48 (s, 2H). 13C NMR (150 MHz, DMSO-d6): δ
182.0, 169.5, 159.0, 153.4, 135.5, 134.9, 134.9, 129.7, 128.3, 124.4, 124.2, 121.5, 118.9, 112.0, 43.3,
34.4. MS (ESI) m/z 394 [M+H]; HRMS (ESI) m/z for C18H14F3N3O2S calculated 394.0759, found
m/z: 394.0838.
4.1.13. 2-[4-(2,3-Dioxo-7-trifluoromethyl-2,3-dihydroindol-1-ylmethyl)benzyl]isothiourea
hydrobromide (3i) (KS122).
Yellow solid, Yield: 70%; mp: 216-218 °C; 1H NMR (600 MHz, DMSO-d6): 1H NMR (600 MHz,
DMSO) δ 9.23 (s, 2H), 9.05 (s, 2H), 7.95 (d, J=7.6, Hz, 1H), 7.94 (dd, J=1.4, 7.2 Hz, 1H), 7.39 -
7.33 (m, 5H), 5.03 (s, 2H), 4.49 (s, 2H). ¹³C NMR (150 MHz, DMSO-d6): δ 181.4, 169.5, 160.5,
148.4, 136.5, 135.0, 133.8, 129.4, 129.0, 126.5, 123.9, 123.3, 121.7, 112.8, 46.3, 34.4. MS (ESI) m/z
394 [M+H]; HRMS (ESI) m/z for C18H14F3N3O2S calculated 394.0759, found m/z: 394.0852.
4.1.14. 2-[4-(5, 7-Dibromo-2, 3-dioxo-2, 3-dihydroindol-1-ylmethyl)benzyl]isothiourea
hydrobromide (3j) (KS100).
Orange solid, Yield: 84%; mp: 196-198 °C; 1H NMR (500 MHz, DMSO-d6): δ 9.17 (s, 2H), 8.98 (s,
2H), 8.01 (d, J=2.0 Hz, 1H), 7.82 (d, J=2.0 Hz, 1H), 7.42 - 7.36 (m, 4H), 5.25 (s, 2H), 4.47 (s, 2H).
13C NMR (150 MHz, DMSO-d6): δ 181.1, 169.48, 159.6, 146.7, 143.8, 137.4, 134.1, 129.6, 127.2,
126.8, 123.3, 116.2, 104.7, 44.5, 34.5; MS (ESI) m/z 481 [M+H]; HRMS (ESI) m/z for
C17H13Br2N3O2S calculated 481.9173, found m/z: 481.9164.
4.1.15. 2-[4-(5,7-Dichloro-2,3-dioxo-2,3-dihydroindol-1-ylmethyl)benzyl]isothiourea hydrobromide
(3k) (KS102).
Orange solid, Yield: 81%; mp: 203-205 °C; 1H NMR (500 MHz, DMSO-d6): δ 9.22 (s, 2H), 9.04 (s,
2H), 7.78 (d, J=2.1 Hz, 1H), 7.70 (d, J=2.1 Hz, 1H), 7.43 - 7.37 (m, 4H), 5.19 (s, 2H), 4.49 (s, 2H).
13C NMR (125 MHz, DMSO-d6): δ 181.1, 169.5, 159.6, 144.8, 138.0, 137.4, 134.2, 129.6, 128.5,

127.1, 123.7, 122.7, 117.0, 44.8, 34.4. MS (ESI) m/z 394 [M+H]; HRMS (ESI) m/z for
C17H13Cl2N3O2S calculated 394.0106, found m/z: 394.0187.
4.1.16. 2-[4-(7-Bromo-5-fluoro-2,3-dioxo-2,3-dihydroindol-1-ylmethyl)benzyl]isothiourea
hydrobromide (3l) (KS120).
Orange solid, Yield: 67%; mp: 216-218 °C; 1H NMR (500 MHz, DMSO-d6): δ 9.25 (s, 2H), 9.14 (s,
2H), 7.79 (dd, J=2.7, 8.8 Hz, 1H), 7.63 (dd, J=2.7, 6.3 Hz, 1H), 7.37 (m, 4H), 5.23 (s, 2H), 4.50 (s,
2H). 13C NMR (125 MHz, DMSO-d6): δ 181.6, 166.6, 159.3, 158.7, 144.1, 137.0, 136.1, 129.5,
128.8, 127.1, 122.6, 112.0, 103.5, 44.4, 30.5. MS (ESI) m/z 421 [M+H]; HRMS (ESI) m/z for
C17H13BrFN3O2S calculated 421.9895, found m/z: 421.9991.
4.1.17. 2-[4-(2,3-Dioxo-2,3-dihydroindol-1-ylmethyl)benzyl]isoselenourea hydrobromide (4a)
(KS105).
Yellow solid, Yield: 77%; mp: 216-218 °C; 1H NMR (500 MHz, DMSO-d6): δ 9.25 (s, 2H), 9.12 (s,
2H), 7.61 - 7.57 (m, 2H), 7.43 - 7.37 (m, 4H), 7.13 (dt, J=7.5, 7.5, 0.7 Hz, 1H), 6.99 (dd, J=0.6, 8.4
Hz, 1H), 4.90 (s, 2H), 4.50 (s, 2H). 13C NMR (150 MHz, DMSO-d6): δ 183.5, 166.7, 158.8, 150.8,
138.5, 136.8, 135.5, 129.8, 128.2, 125.0, 123.9, 118.2, 111.5, 43.1, 30.4. MS (ESI) m/z 374 [M+H];
HRMS (ESI) m/z for C17H15N3O2Se calculated 374.0329, found m/z: 374.0414.
4.1.18. 2-[4-(5-Bromo-2,3-dioxo-2,3-dihydroindol-1-ylmethyl)benzyl]isoselenourea hydrobromide
(4b) (KS109).
Orange solid, Yield: 70%; mp: 208-210 °C; 1H NMR (500 MHz, DMSO-d6): δ 9.24 (s, 2H), 9.15 (s,
2H), 7.78 - 7.74 (m, 2H), 7.43 - 7.36 (m, 4H), 6.93 (d, J=8.3 Hz, 1H), 4.90 (s, 2H), 4.50 (s, 2H). 13C
NMR (150 MHz, DMSO-d6): δ 182.3, 166.7, 158.5, 149.6, 140.1, 136.8, 135.2, 129.7, 128.2, 127.2,
120.1, 115.6, 113.6, 43.2, 30.4. MS (ESI) m/z 451 [M+H]; HRMS (ESI) m/z for C17H14BrN3O2Se
calculated 451.9435, found m/z: 451.9550.
4.1.19. 2-[4-(7-Bromo-2,3-dioxo-2,3-dihydroindol-1-ylmethyl)benzyl]isoselenourea hydrobromide
(4c) (KS111).
Orange solid, Yield: 72%; mp: 202-204 °C; 1H NMR (500 MHz, DMSO-d6): δ 9.24 (s, 2H), 9.12 (s,
2H), 7.77 (dd, J=1.2, 8.1 Hz, 1H), 7.65 (dd, J=1.2, 7.3 Hz, 1H), 7.36 (m, 4H), 7.10 (dd, J=7.3, 8.1
Hz, 1H), 5.25 (s, 2H), 4.51 (s, 2H). 13C NMR (150 MHz, DMSO): δ 182.3, 166.7, 159.9, 147.5,

143.3, 137.1, 136.1, 129.6, 127.1, 125.6, 124.5, 122.0, 103.6, 44.4, 30.5. . MS (ESI) m/z 451 [M+H];
HRMS (ESI) m/z for C17H14BrN3O2Se calculated 451.9435, found m/z: 451.9543.
4.1.20. 2-[4-(5-Chloro-2,3-dioxo-2,3-dihydroindol-1-ylmethyl)benzyl]isoselenourea hydrobromide
(4d) (KS113).
Orange solid, Yield: 70%; mp: 216-218 °C; 1H NMR (500 MHz, DMSO-d6): δ 9.24 (s, 2H), 9.15 (s,
2H), 7.65 - 7.63 (m, 2H), 7.43 - 7.36 (m, 4H), 6.99 (dd, J=1.8, 7.3 Hz, 1H), 4.91 (s, 2H), 4.50 (s,
2H). 13C NMR (150 MHz, DMSO-d6): δ 182.4, 166.7, 158.6, 149.3, 137.3, 136.9, 135.2, 129.7,
128.2, 128.1, 124.5, 119.7, 113.2, 43.2, 30.4. MS (ESI) m/z 407 [M+H]; HRMS (ESI) m/z for
C17H14ClN3O2Se calculated 407.9940, found m/z: 408.0035.
4.1.21. 2-[4-(7-Chloro-2,3-dioxo-2,3-dihydroindol-1-ylmethyl)benzyl]isoselenourea hydrobromide
(4e) (KS115).
Orange solid, Yield: 70%; mp: 192-194 °C; 1H NMR (500 MHz, DMSO-d6): δ 9.24 (s, 2H), 9.14 (s,
2H), 7.62 (dd, J=1.2, 7.3 Hz, 1H), 7.61 (dd, J=1.2, 8.2 Hz, 1H), 7.37 (m, 4H), 7.16 (dd, J=7.3, 8.1
Hz, 1H), 5.20 (s, 2H), 4.51 (s, 2H). 13C NMR (150 MHz, DMSO-d6): δ 182.3, 166.7, 159.8, 146.0,
140.0, 137.1, 136.2, 129.6, 127.1, 125.3, 124.1, 121.8, 116.2, 44.8, 30.5. MS (ESI) m/z 407 [M+H];
HRMS (ESI) m/z for C17H14ClN3O2Se calculated 407.9940, found m/z: 408.0041.
4.1.22. 2-[4-(5-Fluoro-2,3-dioxo-2,3-dihydroindol-1-ylmethyl)benzyl]isoselenourea hydrobromide
(4f) (KS117).
Orange solid, Yield: 72%; mp: 201-203 °C; 1H NMR (500 MHz, DMSO-d6): δ 9.25 (s, 2H), 9.12 (s,
2H), 7.52 - 7.44 (m, 2H), 7.43 - 7.37 (m, 4H), 6.98 (dd, J=3.8, 8.6 Hz, 1H), 4.90 (s, 2H), 4.50 (s,
2H). 13C NMR (150 MHz, DMSO-d6): δ 182.9, 166.7, 159.0, 158.9, 147.0, 136.9, 135.3, 129.8,
128.2, 124.3, 119.2, 112.9, 112.0, 43.2, 30.4. MS (ESI) m/z 392 [M+H]; HRMS (ESI) m/z for
C17H14FN3O2Se calculated 392.0235, found m/z: 392.0337.
4.1.23. 2-[4-(7- Fluoro -2,3-dioxo-2,3-dihydroindol-1-ylmethyl)benzyl]isoselenourea hydrobromide
(4g) (KS119).
Orange solid, Yield: 69%; mp: 206-208 °C; 1H NMR (500 MHz, DMSO-d6): δ 9.25 (s, 2H), 9.13 (s,
2H), 7.51 (ddd, J=1.0, 8.4, 11.8 Hz, 1H), 7.48 (dd, J=1.0, 7.4 Hz, 1H), 7.38 (m, 4H), 7.15 (ddd,
J=4.0, 7.5, 8.3 Hz, 1H), 4.96 (s, 2H), 4.50 (s, 2H). 13C NMR (150 MHz, DMSO-d6): δ 182.3, 166.7,

158.9, 147.6, 136.6, 136.3, 129.7, 127.6, 126.2, 125.0, 121.4, 121.3, 45.3, 30.5. MS (ESI) m/z 392
[M+H]; HRMS (ESI) m/z for C17H14FN3O2Se calculated 392.0235, found m/z: 392.0334.
4.1.24. 2-[4-(2,3-Dioxo-5-trifluoromethyl-2,3-dihydroindol-1-ylmethyl)benzyl]isoselenourea
hydrobromide (4h) (KS107).
Yellow solid, Yield: 72%; mp: 218-220 °C; 1H NMR (500 MHz, DMSO-d6): δ 9.24 (s, 2H), 9.13 (s,
2H), 7.96 (dd, J=1.2, 8.4 Hz, 1H), 7.89 (d, J=1.2 Hz, 1H), 7.45 - 7.37 (m, 4H), 7.15 (d, J=8.4 Hz,
1H), 4.96 (s, 2H), 4.50 (s, 2H). 13C NMR (150 MHz, DMSO): δ 182.1, 166.7, 159.0, 153.4, 136.9,
135.1, 134.9, 129.7, 128.2, 124.4, 124.2, 121.5, 118.9, 112.0, 43.3, 30.4. MS (ESI) m/z 442 [M+H];
HRMS (ESI) m/z for C18H14F3N3O2Se calculated 442.0203, found m/z: 442.0285.
4.1.25. 2-[4-(2,3-Dioxo-7-trifluoromethyl-2,3-dihydroindol-1-ylmethyl)benzyl]isoselenourea
hydrobromide (4i) (KS123).
Yellow solid, Yield: 62%; mp: 212-214 °C; 1H NMR (600 MHz, DMSO-d6): δ 9.25 (s, 2H), 9.13 (s,
2H), 7.85 (d, J=7.4 Hz, 1H), 7.75 (d, J=7.7 Hz, 1H), 7.36 - 7.31 (m, 5H), 5.05 (s, 2H), 4.50 (s, 2H).
13C NMR (150 MHz, DMSO-d6): δ 181.3, 166.5, 160.4, 158.4, 136.5, 135.0, 133.6, 129.4, 129.1,
126.5, 123.8, 123.3, 121.8, 112.8, 46.2, 30.4. MS (ESI) m/z 442 [M+H]; HRMS (ESI) m/z for
C18H14F3N3O2Se calculated 442.0203, found m/z: 442.0311.
4.1.26. 2-[4-(5, 7-Dibromo-2, 3-dioxo-2, 3-dihydroindol-1-ylmethyl)benzyl]isoselenourea
hydrobromide (4j) (KS101).
Orange solid, Yield: 78%; mp: 193-195°C; 1H NMR (600 MHz, DMSO-d6): δ 9.29 (s, 2H), 9.17 (s,
2H), 8.01 (d, J=2.0 Hz, 1H), 7.81 (d, J=2.0 Hz, 1H), 7.38 (m, 4H), 5.24 (s, 2H), 4.54 (s, 2H). 13C
NMR (150 MHz, DMSO-d6): δ 181.1, 166.7, 159.6, 146.7, 143.8, 136.9, 136.1, 129.6, 127.1, 126.8,
123.2, 116.2, 104.7, 44.5, 30.5. MS (ESI) m/z 529 [M+H]; HRMS (ESI) m/z for C17H13Br2N3O2Se
calculated 529.8617, found m/z: 529.8618.
4.1.27. 2-[4-(5, 7-Dichloro-2, 3-dioxo-2, 3-dihydroindol-1-ylmethyl)benzyl]isoselenourea
hydrobromide (4k) (KS103).
Orange solid, Yield: 76%; mp: 213-215 °C; 1H NMR (500 MHz, DMSO-d6): δ 9.26 (s, 2H), 9.15 (s,
2H), 8.01 (s, 1H), 7.81 (d, J=2.0 Hz, 1H), 7.37 (s, 4H), 5.24 (s, 2H), 4.51 (s, 2H). 13C NMR (150
MHz, DMSO-d6): δ 181.1, 166.7, 159.6, 146.7, 143.8, 136.9, 136.1, 129.5, 127.1, 126.8, 123.2,

116.2, 104.7, 44.5, 30.5. MS (ESI) m/z 441 [M+H]; HR-MS (ESI) m/z for C17H13Cl2N3O2Se
calculated 441.9550, found m/z: 441.9545.
4.1.28. 2-[4-(7-Bromo-5-fluoro-2,3-dioxo-2,3-dihydroindol-1-ylmethyl)benzyl]isoselenourea
hydrobromide (4l) (KS121).
Orange solid, Yield: 65%; mp: 195-197 °C; 1H NMR (600 MHz, DMSO-d6): δ 9.23 (s, 2H), 9.12 (s,
2H), 7.79 (dd, J=2.7, 8.7 Hz, 1H), 7.63 (dd, J=2.6, 6.2 Hz, 1H), 7.37 (m, 4H), 5.24 (s, 2H), 4.51 (s,
2H). 13C NMR (150 MHz, DMSO-d6): δ 181.6, 166.6, 159.2, 158.8, 144.1, 137.0, 136.1, 129.5,
128.9, 127.1, 122.6, 112.0, 103.6, 44.4, 30.5. MS (ESI) m/z 469 [M+H]; HRMS (ESI) m/z for
C17H13BrFN3O2Se calculated 469.9340, found m/z: 469.9437.
4.2. Biology
4.2.1. Cell line and culture conditions
Dr. Craig Myers, Penn State College of Medicine, Hershey, PA, provided normal human fibroblasts
(FF2441). The mutant V600E-BRAF human melanoma cell line 1205 Lu was provided by Dr.
Herlyn; Wistar Institute, Philadelphia, PA, and UACC 903 was provided by Dr. Mark Nelson;
University of Arizona, Tucson, AZ. Colon cancer cells (HCT116, HT29) and multiple myeloma
cells (NCIH929, U266, RPMI8226, MM.1R, MM.1S) were procured from ATCC. Cell lines were
maintained in a 37°C humidified 5% CO2 atmosphere incubator and periodically monitored for
phenotypic, genotypic characteristics, and tumorigenic potential to validate and confirm cell line
identity.
4.2.2. Molecular docking studies
Binding interactions of isatin and isatin derivatives with ALDH1A1 (PDB: 4X4L), ALDH2 (PDB:
5L13), and ALDH3A1 (PDB: 4L2O) proteins were analyzed using the GLIDE (Grid Ligand
Docking with Energetics) docking application in Maestro 10.1 software as described previously [56-
58]. Proteins were prepared using the protein preparation wizard tool (Schrodinger, LLC, 2017) with
default parameters. The proteins were optimized and minimized for spatial conformations. Grids
were generated based on the location of the crystal ligand-binding site (CM037 for ALDH1A1,
psoralen for ALDH2, and CB7 for ALDH3A1), using the GLIDE grid module. Default parameters
were used, and no constraints were included during grid generation. Ligand preparation was then

performed using the ligprep module in Schrodinger as previously described [56-58]. The docking
study was performed using GLIDE 6.6 in Maestro 10.1. The GLIDE algorithm estimates a
systematic search of positions, orientations, and conformations of the ligand in the enzyme-binding
pocket via a series of hierarchical filters. All hits were subjected to the extra precision (XP) mode of
GLIDE. During the docking process, the GLIDE score was used to select the best conformation for
each ligand [56-58].
4.2.3. Predicted drug-like properties of compounds.
Drug-like properties of the designed compounds were predicted by using the Quikprops module of
Maestro 10.1 software [56-58].
4.2.4. ALDH isoform-specific enzyme assays
ALDH1A1, ALDH2, and ALDH3A1 enzyme assays were performed as described by the
manufacturer (R & D systems). Isoform-specific aldehydes were converted to their respective
carboxylic acids along with the conversion of NAD+ to NADH (absorbance at 340 nm).
Specifically, 1 µg/mL of ALDH1A1 was treated with different concentrations of 3(a-l) and 4(a-l) for
15 minutes followed by addition of substrate mixture (10 mM propionaldehyde; 100 mM KCl; 1
mM NAD; 2 mM DTT; 50 mM Tris pH 8.5) and the absorbance of NADH was measured in kinetic
mode for 5 minutes. Similarly, 0.5 µg/mL of ALDH2 was used in the reaction with 2 mM of
acetaldehyde as the substrate, and 0.2 µg/mL of ALDH3A1 was used in the reaction with 1 mM of
4-nitrobenzaldehyde as the substrate.
4.2.5. Cell viability assay
Melanoma, colon cancer, multiple myeloma, and FF2441 cells treated with 3a or 3(h-l) and cell
viability assays were performed as described previously [54, 59, 60]. Briefly, 5,000 cells per well
were plated in a 96-well plate and incubated overnight at 37˚C in a 5% CO2 atmosphere. Cells were
treated with 3a or 3(h-l) at various concentrations and incubated for 72 hours. Additionally, cells
were also treated with 10 mM of NAC. 20 µL of MTS reagent was then added into each well, and
the formation of tetrazolium was measured by absorbance after 1 hour at 492 nm. IC50 values for
each experimental group were measured in 3 independent experiments using GraphPad Prism
version 7.04 (GraphPad Software, La Jolla, CA). Selectivity indices for 3(h-l) were calculated as a
ratio of IC50s in fibroblasts/average of IC50s in melanoma cell lines.

4.2.6. Toxicity studies
To determine the toxicity of 3(h-l), compounds (5 mg/kg/day) were injected i.p. into Swiss-Webster
mice once daily for 14 days [61, 62]. Animals were monitored for changes in body weight,
behavior, and physical distress compared to DMSO control.
4.2.7. ROS assay
ROS activity was measured using DCFDA dye [54]. Briefly, cells were treated with 5 µM of 3a, 3h,
or 3j for 24 hours. Additionally, cells were also treated with 10 mM of NAC. Cells were incubated
with 10 µM of DCFDA for 1 hour, and fluorescence was measured at 485 nm excitation and 510 nm
emission. ROS levels in treated cells were compared to DMSO control.
4.2.8. Lipid peroxidation and toxic aldehyde accumulation
Lipid peroxidation and toxic aldehyde accumulation were measured using the thiobarbituric acid
reactive substances (TBARS) kit according to the manufacturer’s instructions [55]. Briefly, cells
were treated with 5 µM of 3a, 3h, or 3j for 24 hours. Additionally, cells were also treated with 10
mM of NAC. Cell pellets were lysed in PBS by sonication on ice. Lipids in the lysates were
hydrolyzed in the presence of acetic acid and sodium hydroxide. Free MDA released from lipids was
measured by the reaction to TBA colorimetrically at 530 nm. Lipid peroxidation in treated cells was
compared to DMSO control.
4.2.9. Apoptosis assay
Apoptosis assay was measured using Annexin-V-PE/7-AAD kit according to the manufacturer’s
instructions [54]. Briefly, cells were treated with 5 µM of 3a, 3h, or 3j for 24 hours. Additionally,
cells were also treated with 10 mM of NAC. Cells were detached, washed with PBS, and stained
with Annexin-V-PE and 7-AAD. Cells were acquired by BD Fortessa flow cytometer from Penn
State Hershey Flow Cytometry core facility. Cells that are double-positive for both Annexin-V and
7-AAD were considered late-stage apoptotic and are compared to DMSO control.
4.2.10. Cell cycle analysis
Cell cycle analysis was performed using Propidium iodide staining [54]. Briefly, cells were treated
with 5 µM of 3a, 3h, or 3j for 24 hours. Additionally, cells were also treated with 10 mM of NAC.
Cells were detached, washed with PBS, and fixed in 70% ethanol solution. Cells were stained with

Propidium iodide and acquired by BD Fortessa flow cytometer from Penn State Hershey Flow
Cytometry core facility. Cell cycle analysis was performed using FlowJo software.
4.2.11. Statistics
Statistical analysis was undertaken using the one-way/two-way ANOVA GraphPad PRISM Version
7.04 software. Dunnett’s as post hoc analysis was performed when there was a significant difference.
A p-value of <0.05 was considered statistically significant.
Acknowledgements
Authors thank the co-operation of the Penn State Cancer Institute, Organic Synthesis Core, Flow
Cytometry Core, NMR facility in Penn State College of Medicine, Hershey, PA. Authors also thank
the assistance of Mass Spectroscopy Core in Huck Institutes of the Life Sciences, Penn State
University, University Park, PA.

Schemes:
Scheme 1. Synthesis of compounds 3(a-l) and 4(a-l). Reagents and conditions: a) K2CO3, DMF, 1,4-
bis(bromomethyl)benzene; b) Thiourea, EtOH, 90ºC; c) Selenourea, EtOH, 90ºC.
Figure legends:
Figure 1. Isatin based structures ALDH1A1, ALDH2, and ALDH3A1 inhibitors.
Figure 2. Molecular docking studies of compounds in the active site pockets of ALDH1A1,
ALDH2, and ALDH3A1. 3h is shown as a representative compound for 3(a-l) and 4(a-l).
Figure 3. ROS and lipid peroxidation activity and toxic aldehyde accumulation. HCT116 (A) and
SW480 (B) cells were treated with 5 µM of 3a, 3h, or 3j for 24 hours with or without NAC (10
mM). ROS levels were measured using DCFDA dye and compared to DMSO control.
Malondialdehyde (MDA) levels were measured in colon cancer cell line HCT116 using
thiobarbituric acid and compared to DMSO control (C). Cell survival assay was performed by MTS
assay (D), apoptosis by Annexin-V/7-AAD (E) and cell cycle by propidium iodide staining in colon
cancer cell line HCT116 (F).

Table legends:
Table 1. Structures and docking scores of 3(a-l) and 4(a-l). Docking scores were calculated for
compounds against ALDH1A1, ALDH2, and ALDH3A1 using the Glide module of Schrodinger.
Table 2. ALDH enzyme inhibitory IC50s of 3(a-l) and 4(a-l). Compounds 3(a-l) and 4(a-l) were
evaluated for ALDH1A1, ALDH2, and ALDH3A1 inhibitory activity. The % inhibition was
calculated for each compound and compared to DMSO control.
Table 3. in vitro anti-proliferative activity of selected compounds 3(h-l). Compounds 3(h-l) were
evaluated for anti-proliferative effects on melanoma, colon cancer, multiple myeloma, and normal
human fibroblasts (FF2441). Cells were treated with 3(h-l) at various concentrations for 72 hours,
and IC50s were calculated.
a IC50 = Compound concentration required to inhibit cell proliferation by 50%. Data are expressed as the mean
+ SD from the dose-response curve of at least three independent experiments. ND-not determined.
b Melanoma
c Colon cancer
d Multiple myeloma
e Normal human fibroblasts
Table 4. Drug-like properties of 3(h-l). Drug-like properties of 3(h-l) were predicted by
Quikprops module of Schrodinger.
SASA: solvent accessible surface area; donor HB: Estimated number of hydrogen bonds that would be
donated by the solute to water molecules in an aqueous solution; aacptHB: Estimated number of hydrogen
bonds that would be accepted by the solute from water molecules in an aqueous solution; QPlogPo/w:
Predicted octanol/water partition coefficient; QPlogHERG: Predicted IC50 value for blockage of HERG K+
channels; QPPCaco: Predicted apparent Caco-2 cell permeability in nm/sec. Caco2 cells are a model for the
gut-blood barrier; QPlogBB: Predicted brain/blood partition coefficient; QPPMDCK: Predicted apparent
MDCK cell permeability in nm/sec; #metab: Number of likely metabolic reactions; Human Oral Absorption:
Predicted qualitative human oral absorption; Percent Human Oral Absorption: Predicted human oral
absorption on 0 to 100% scale; Rule Of Five: Number of violations of Lipinski’s rule of five.
Table 5. Toxicity of 3(h-l). Compounds 3(h-l) were dosed daily at 5 mg/kg/day via i.p. injection to
Swiss-Webster mice for 14 days. % change in animal weight was compared to DMSO control.

Table 1.
Docking scores
Compound
KS104
KS108
KS110
KS112
KS114
KS116
KS118
KS106
KS122
KS100
KS102
KS120
KS105
KS109
KS111
KS113
KS115
KS117
KS119
KS107
KS123
KS101
KS103
KS121
#
R1
H
Br
H
Cl
H
F
H
CF3
H
Br
Cl
F
H
Br
H
Cl
H
F
H
CF3
H
Br
Cl
F
R2
H
H
Br
H
Cl
H
X
S
S
S
S
S
S
S
S
S
S
S
S
ALDH1A1 ALDH2 ALDH3A1
3a
3b
3c
3d
3e
3f
3g
3h
3i
-10.71
-11.637
-9.727
-8.51
-10.28
-7.308
-13.383
-13.369
-13.358
-11.938
-11.271
-11.056
-11.197
-11.721
-10.836
-10.247
-10.841
-10.432
-11.024 -13.666
-7.061
-11.03
-14.237
-12.157
F
H
-11.205 -14.24
-10.374 -12.229
CF3
Br
Cl
Br
H
H
Br
H
-9.497
-8.716
-8.169
-7.294
-14.564
-13.851
-14.103
-13.039
3j
3k
3l
4a
4b
4c
4d
4e
4f
4g
4h
4i
Se -11.487
Se -10.149
Se -10.144
Se -10.14
Se -10.841
Se -10.144
Se -8.387
Se -9.988
Se -11.741
Se -10.149
Se -10.594
Se -7.495
-10.592 -12.375
-6.82
-14.441
-14.507
-13.268
-13.85
-12.119
-13.943
-6.804
-6.756
-6.978
-7.929
-8.609
Cl
H
F
H
-11.146 -14.229
CF3
Br
Cl
Br
-9.494
-8.975
-13.46
-12.638
4j
4k
4l
-10.841 -13.701
-9.792 -12.567

Table 2.
Enzyme inhibition - IC50 (nM)
Compound ALDH1A1 ALDH2 ALDH3A1
3a
3b
3c
3d
3e
3f
3g
3h
3i
4633
8524
1713
>10000
177
5224
598
334
268
230
279
285
>10000 4205
>10000 4878
>10000 6323
>10000 3067
>10000 3586
>10000 2855
>10000 241
2137
1783
1542
1642
1782
360
246
193
219
219
3j
3k
3l
4a
4b
4c
4d
4e
4f
4g
4h
4i
>10000
333
360
>10000
>10000
5192
1054
6492
657
>10000 4007
939 344
>10000 >10000
>10000 2767
>10000 >10000
>10000 2364
>10000 246
>10000 1520
3491
2012
1809
1917
251
333
327
368
4j
4k
4l
397
384
420

Table 3.
IC50 (µM)a/ 72 h
NCIH929d U266d
Compound UACC 903b 1205 Lub HCT116c
HT29c
RPMI8226d
MM.1Rd
MM.1Sd FF2441e
3h
5.7 + 2.0 5.7 + 1.5 5.7 + 0.1
3.7 + 0.6 3.7 + 0.8 4.5 + 0.2
3.7 + 0.2 2.1 + 0.6 2.9 + 0.1
4.9 + 1.1 1.5 + 0.3
5.8 + 0.1 1.5 + 0.4
2.5 + 0.2 0.3 + 0.1
2.6 + 0.3 0.7 + 0.2
4.4 + 0.7 1.6 + 0.4
2.6 + 0.6 1.6 + 0.4
4.7 + 1.2 4.4 + 1.9
1.0 + 0.3 1.2 + 0.28
2.7 + 0.6 1.3 + 0.17
2.8 + 0.4 1.5 + 0.29
1.7 + 0.4
4.7 + 0.2
1.3 + 0.6
1.6 + 0.3
3.6 + 0.4
ND
2.2 + 0.3 20.7 + 1.3
4.0 + 1.0 13.9 + 0.1
2.1 + 0.6 9.8 + 0.5
1.6 + 0.3 5.4 + 0.2
3.1 + 0.8 10 + 0.9
3i
3j
3k
3 + 0.6
2.4 + 0.3 3.2 + 0.7
3l
3.9 + 1.1 5.6 + 1.5 5.8 + 0.1
3a
>100
>100
>100
>100
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
>100
>100
Cpd 3
ND

Table 4.
Human Oral Percent Human
Absorption Oral Absorption
SASA Donor HB Accpt HB QPlogPo/w QPlogHERG QPPCaco QPlogBB QPPMDCK #metab
Rule of five
Compd
3j
3k
3h
3l
645.543
636.119
638.908
467.728
572.826
3
3
3
0
0
6.5
6.5
6.5
5
2.69
2.545
2.503
1.683
3.86
-5.67
-5.614
-5.499
-5.005
-5.178
88.143
87.592
-1.6
297.149
255.413
187.952
529.159
10000
2
2
2
1
1
3
3
3
3
3
77.508
76.616
74.796
90.976
100
0
0
0
0
0
-1.612
-1.712
-0.432
0.179
71.546
1064.261
1304.932
3i
5