2,4-dihydroxy benzaldehyde derived Schiff bases as small molecule Hsp90 inhibitors: Rational identification of a new anticancer lead

Article abstract of DOI:10.1016/j.bioorg.2015.02.003

Hsp90 is a molecular chaperone that heals diverse array of biomolecules ranging from multiple oncogenic proteins to the ones responsible for development of resistance to chemotherapeutic agents. Moreover they are over-expressed in cancer cells as a complex with co-chaperones and under-expressed in normal cells as a single free entity. Hence inhibitors of Hsp90 will be more effective and selective in destroying cancer cells with minimum chances of acquiring resistance to them. In continuation of our goal to rationally develop effective small molecule azomethines against Hsp90, we designed few more compounds belonging to the class of 2,4-dihydroxy benzaldehyde derived imines (1-13) with our validated docking protocol. The molecules exhibiting good docking score were synthesized and their structures were confirmed by IR, ¹H NMR and mass spectral analysis. Subsequently, they were evaluated for their potential to suppress Hsp90 ATPase activity by Malachite green assay. The antiproliferative effect of the molecules were examined on PC3 prostate cancer cell lines by adopting 3-(4,5-dimethythiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) assay methodology. Finally, schiff base 13 emerged as the lead molecule for future design and development of Hsp90 inhibitors as anticancer agents.

Full text of DOI:10.1016/j.bioorg.2015.02.003

Accepted Manuscript
2,4-dihydroxy benzaldehyde derived Schiff bases as small molecule Hsp90 in-
hibitors: Rational identification of a new anticancer lead
Sayan. Dutta Gupta, B. Revathi, Gisela.I. Mazaira, Mario.D. Galigniana, C.V.S.
Subrahmanyam, N.L. Gowrishankar, N.M. Raghavendra
PII:
S0045-2068(15)00012-7
DOI:
http://dx.doi.org/10.1016/j.bioorg.2015.02.003
YBIOO 1789
Reference:
Bioorganic Chemistry
To appear in:
Received Date:
19 December 2014
Please cite this article as: Sayan. Dutta Gupta, B. Revathi, Gisela.I. Mazaira, Mario.D. Galigniana, C.V.S.
Subrahmanyam, N.L. Gowrishankar, N.M. Raghavendra, 2,4-dihydroxy benzaldehyde derived Schiff bases as small
molecule Hsp90 inhibitors: Rational identification of a new anticancer lead, Bioorganic Chemistry (2015), doi:
http://dx.doi.org/10.1016/j.bioorg.2015.02.003
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. 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.

2,4-dihydroxy benzaldehyde derived Schiff bases as small molecule Hsp90 inhibitors:
Rational identification of a new anticancer lead
Sayan Dutta Guptaa,b B.Revathia, Gisela.I.Mazairad, Mario.D.Galignianad,e, C.V.S.
,
Subrahmanyama , N.L. Gowrishankarc and N.M. Raghavendraa
aDepartment of Pharmaceutical Chemistry, Gokaraju Rangaraju College of Pharmacy, Osmania
University, Hyderabad, India.
bR&D centre, Department of Pharmaceutical Sciences, Jawaharlal Nehru Technological
University, Hyderabad, India.
cSwami Vivekananda Institute of Pharmaceutical Sciences, Nalgonda, Andhrapradesh, India
d Department of Biological Chemistry, Faculty of Natural Sciences, University of Buenos Aires,
Argentina
eInstitute of Experimental Biology and Medicine-CONICET, Argentina.
* Corresponding author. E- mail: sayanduttagupta1@rediffmail.com
Ph: +91-40-3291 2927, Fax: +91-40-2304 0860;
1

1. INTRODUCTION
Heat shock protein (Hsp90) is an ATP dependant molecular chaperone required for the repair
of several signaling proteins which promote cancer cell multiplication and survival [1, 2]. Hence,
the pharmacological inhibition of this chaperone leads to simultaneous blockage of multiple
oncogenic transduction signal pathways, thus making it an attractive anticancer target [3, 4]. The
energy required for its chaperoning function is acquired from the hydrolysis of ATP which binds
to the N-terminal domain of the protein [5, 6]. This ATP binding cleft is structurally different
from most ATP-utilizing vital enzymes of normal cells [7-9]. Moreover, Hsp90 is overexpressed
in cancer cells in association with co-chaperones whereas Hsp90 of normal tissues resides in a
free and uncomplexed state [10, 11]. The above observations were supported by studies
employing various Hsp90 inhibitors wherein it was proved that cancer cells are more sensitive to
Hsp90 inhibition compared to healthy cells [12, 13]. These conclusions further affirm the notion
that suppression of Hsp90 function can lead to selective killing of cancer cells over normal cells.
All the aforementioned research findings gave the impetus to develop novel small molecule
chemotypes against Hsp90 with reduced toxic side effects.
Molecular docking is a fast and efficient computational technique that helps a drug design
process to dramatically widen the chemical space and reduce the number of molecules for
experimental verification [14]. Considering the pivotal role resorcinol moiety plays in Hsp90
suppression [15, 16] and the importance of schiff bases (imines or azomethines) in cancer
chemotherapy [17, 18], we began our Hsp90 research program with structure based drug design
of novel resorcinol containing imines using Malachite green assay which measures the ATPase
activity of human Hsp90 [19]. This effort led to the discovery of a series of novel small molecule
azomethines as Hsp90 inhibitors. Compound A and B with an IC50 value of 3nm and 4nm
respectively were found to be the most potent ones (Fig. 1). Unfortunately, these two compounds
exhibited an IC50 value greater than 15µM in the cell viability assay against PC3 prostate cancer
cell lines performed by adopting 3-(4,5-dimethythiazol- 2-yl)-2,5-diphenyl tetrazolium bromide
(MTT) assay methodology [19]. The large difference between biochemical and cellular potencies
of the compounds may be attributed to the presence of any reactive and/or unstable functional
groups. However this hypothesis can be ascertained by comparing the activity profile of larger
number of resorcinol scaffold bearing imines. Furthermore, the structure of the most promising
imine (Fig. 2) of our study did not offer any scope for further chemical modifications. Hence the
research effort was not able to generate any lead molecule for subsequent optimization.
Moreover not a single compound inhibited 50 percent of cancer cells below 5µM concentration
and thus making them poor candidates for further pre-clinical evaluation in animals. All the
above mentioned limitations of the earlier research work prompted us to rationally design few
more 2,4-dihydroxybenzaldehyde derived Schiff bases (Fig. 3) with our established docking
protocol[20]. The ligands predicted to possess better protein affinity were synthesized and
characterized by IR, NMR and mass spectral analysis. These imines were evaluated for their
Hsp90 inhibition potential by utilizing the colorimetric malachite green assay. The azomethines
were further screened for their antiproliferative effect against PC3 carcinoma cell lines by using
the MTT assay protocol. Therefore the main objective of this study is to rationally identify
Hsp90 selective low molecular weight antineoplastic chemical lead among 2,4-dihydroxy
benzaldehyde derived imines. This will guide medicinal chemist worldwide to discover more
efficacious drug molecules against cancer with fewer toxic side effects. Subsequently the
proposed compounds would add further structural diversity to the field of dihydroxy phenyl
2

imines as Hsp90 inhibitors with an additional insight into the functional groups that can be
accommodated in the Hsp90 N-terminal ATP-binding cleft for this class of compounds.
2. Materials and Methods
2.1. Molecular Modeling methods
Molecular modeling was carried out by docking ligands into the protein active site with
Surflex Geom X tool of Sybyl X-1.2 version software installed on a Windows based Dell
Precision T-1500 workstation. PDB entry 3EKR (a co-crystal structure of 4-{[(2R)-2-(2-
methylphenyl)pyrrolidin-1-yl]carbonyl}benzene- 1,3-diol in complex with human Hsp90 with a
resolution of 2Å)[21] was used for executing the docking studies. Four binding site waters (902,
903, 981 and 1026) were retained in the docking process as they provide key interactions
required for stabilizing the ligand in the active site [20, 22-24]. This was followed by addition of
hydrogens for the receptor and waters. Subsequently, the protein was energy minimized using
AMBER7 FF99 force field and the protomol was generated by maintaining 0.5 Å threshold
factor and 0 Å bloat. [25-27]. The structures of the compounds were built using Chem draw ultra
8.0 and exported as mol files to Schrodinger software (Maestro, 9.1 versions) where they were
converted into SD file format for Sybyl compatibility. Finally a clean, energy minimized, 3D
conformation of the ligands were generated and subsequently docked at the developed protomol
of the prepared protein [28, 29].
2.2. Synthesis of Compounds
The chemical reagents and solvents were commercially available from Hi-media
Laboratories Private Limited and were used without further purification. The termination of the
reaction was confirmed by performing thin layer chromatography on aluminium sheets coated
with silica gel G containing fluorescent indicators, F254. The mobile phase for the development of
the plate was ethyl acetate and petroleum ether (1:1). Melting point values were determined on a
DRK Digital melting point apparatus and are reported uncorrected. IR spectra were recorded
using KBr powder on Shimadzu diffuse reflectance attachment (DRS-800) mounted on a
1
Shimadzu IR-Affinity spectrometer. H NMR spectra were obtained on a Bruker Avance 300
MHz instrument in DMSO-d6 (Tetramethylsilane as the internal standard). Chemical shift values
are reported in parts per million (d, ppm). Mass spectra were run using a atmospheric pressure-
electron spray ionization 6120 Quadrupole LC/MS mass spectrometer (Agilent Technologies,
California, USA).
2.2.1. Synthesis of 2,4-Dihydroxy benzaldehyde [19, 30]
This was synthesized according to a previously described procedure (Fig. 4). Phosphorous
oxychloride(0.026 mol) was added dropwise to a solution of dimethyl formamide (0.030 mol) in
acetonitrile (7ml) at 22°C to 28°C and stirred for 1 h at the above temperature. This was
followed by addition of resorcinol (0.022 mol) in acetonitrile (7ml) at -15 to -17°C. The reaction
mixture was then stirred for 2 h at this temperature and then at 28-32°C for 1h. The mixture was
cooled to 5°C and after stirring for 1h the Vilsmeier formamidinium phosporodichloridate salt
gets precipitated which was filtered and washed with cold acetonitrile. This intermediary salt was
added in two parts to a beaker containing 60 ml of water at 40°C. The solution was then heated
3

to 52°C for 30 min after which it was left to cool at 35°C. Subsequently, 1ml of 0.09M sodium
thiosulfate solution was added until the pink color disappears. The reaction was cooled to 5°C
and stirred for 2 h to precipitate a white solid. The white solid was filtered, washed with cold
water and air dried for 2 h to obtain white crystals of 2,4-dihydroxy benzaldehyde.
2.2.2. General procedure for the synthesis of Schiff base derived from 2,4-Dihydroxy
benzaldehyde [19]
Synthesis of the imines was achieved according to a method outlined in the scheme (Fig. 4).
The appropriate amine (5 mmol) was heated at reflux with 2,4-dihydroxy benzaldehyde (5
mmol) in absolute ethanol for 3 h. The completion of the reaction was confirmed by TLC. It was
followed by addition of water that resulted in the formation of colored precipitate which was
filtered and air dried. The resulting solid was recrystallized from aqueous ethanol to obtain the
pure product.
2.2.3. Synthesis of 2-(4-aminophenyl) benzimidazole (amine used in the preparation of
schiff base 13) [31]
o-phenylenediamine ( 0.00924 moles) and p-amino benzoic acid (0.00924 moles) was
transferred to a round bottom flask containing 6.5 ml of polyphosphoric acid. The reaction
mixture was refluxed at 100°C for 3 h. It was cooled and neutralized with 10% sodium carbonate
solution. The precipitate formed was filtered and washed with cold water. The solid obtained was
dried and recrystallized from methanol/water to obtain the pure product.
The general structure of the ligands along with the numbering system used in this work is
shown in Fig. 3.
2.3 Malachite green assay for Hsp90 ATPase suppression
Histidine-tagged human Hsp90 (hHsp90b) was kindly provided by Dr. Chrisostomos
Prodromou, University of Sussex, United Kingdom. The chaperone was further purified and
expressed according to a previously described procedure[6, 32]. The ATPase assay procedure
was performed by incubating 10µg of pure hHsp90b along with the compounds (diluted in
DMSO) for 10 min at 20oC in a buffer containing 50 mM Hepes at pH 7.5, 6 mM MgCl2, 20 mM
KCl, and 1 mM ATP[32-34]. To stop the incubation, two volumes of malachite green reagent
were added to each well of the plate. The plate was shaken and left to stand at room temperature
for about 25 minutes, and the absorbance at 630 nm was measured[35, 36]. Geldanamycin was
taken as the standard reference compound.
2.4. In vitro cell growth assay
PC3 prostate cancer cells were utilized for the study. Cellular growth in the presence or
absence of experimental agents was determined using the previously reported MTT[3-(4, 5-
dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide] assay[37, 38]. The cell lines was
cultured in RPMI1640 (Invitrogen)/10% fetal bovine serum (Gibco) medium supplemented with
0.007% streptomycin and 0.002% penicillin. The cells were then counted and incubated for 37°C
with 5 % CO2 in a 96 well micro plates. When the cells reach 80% confluence, they were treated
with the test compounds and standard (geldanamycin) prepared in DMSO to furnish the
4

concentration range of study, 0.1µM-5µM. Control wells contained equivalent volume of the
vehicle. This was followed by 48 h incubation after which 5ml of MTT reagent along with 45 ml
of phenol red and FBS free DMEM was added to each well. Thereafter the plates were incubated
at 37 °C with 5 % CO2 for 4 h. Subsequently, 50 ml of solubilization buffer was added to each
well to solubilize the colored formazan crystals produced by the reduction of MTT. After 48 hrs,
the optical density was measured at 550nm using spectrophotometer in a microplate reader (Bio-
Rad, USA)[39].
3. Results and Discussion
3.1. Molecular Design
The docking protocol was validated by comparing the software generated ligand-Hsp90
binding interactions with the crystallographic analysis available from PDB and its corresponding
literature. It was reported that 2'-hydroxyl group of the native ligand interacted with Asp 93 and
water molecule 902 via hydrogen bond. The 4'- hydroxyl group formed hydrogen bond
interaction with amino acid Asn 51 and water molecule 903. The other prominent hydrogen bond
contacts detected was between the carbonyl oxygen and amino acid Thr 184. Further analysis
indicated that the pyrrole moiety occupied a hydrophobic area formed by Met 98, Ile 96, Ala 55
and Lys 58. The other significant hydrophobic affinity was observed between phenyl ring and
amino acid Tyr, Val 136 and Gly 135. The amino acid residue of Leu 107, Val 150 and Gly 135
formed a hydrophopic region where the resorcinol moiety was embedded. These interactions
were found matching with the reported crystallographic data (Fig. 5a and 5b).
The Surflex Geom X docking program predicted thirteen compounds as effective Hsp90
inhibitor. The detailed result of ligand-Hsp90 binding studies is shown in Table 1. These
molecules were found to form hydrogen bond contact with Asp54, Asp93, Asn51, Asn106, Gly
97, Lys58, Thr184 and water molecule 902, 903 981 and 1026. The H- bond contact of 4’-OH
and 2’-OH with Asp93/water 902 and Asn51/water 903 respectively was revealed for majority of
the compounds. The highest scored compound (3) disclosed additional hydrogen bond contact
with amino acid residue Asn 106, Lys 58 via oxygen of para nitro group. A hydrogen bond
between water number 1026 and ortho nitro group was also observed for compound 3 (Fig. 6a).
The hydrophobic interactions of the Schiff bases revealed that the dihydroxy phenyl moiety
along with the imine part is embedded in a hydrophopic pocket of Val 186, Ile 91, Phe 138, Val
150 and Leu 48. The phenyl ring attached to the imine nitrogen was found to occupy a pocket
surrounded by amino acid Ala 55, Met 98 and Leu 107 (Fig. 6b).
3.2. Synthesis
The series of imines derived from 2,4-dihydroxy benzaldehyde were characterized by
1
melting point, Rf values, IR, HNMR and mass spectroscopy. All the relevant data regarding
characterization of compounds is presented in Table 2.
The FT-IR spectra of the synthesized compounds showed absorption bands at 1645-1614 cm-
1 for imines. The broadness of the OH band observed between 3751-3311 cm-1 may be attributed
to the intramolecular hydrogen bond between CH=N (imine nitrogen) and OH (phenolic) group
[40, 41]. The phenolic C-OH stretching vibrations were confirmed by medium intensity bands in
the range of 1284-1049 cm-1. The presence of additional bands between 1598 and 1419 cm-1
5

were characteristic of C=C bands of the aromatic rings. Compound 3 displayed two bands at
1510 and 1130 corresponding to N-O stretching. A stretching vibration band at 1122 cm-1 for
compound 8 was attributed to its C–F bond.
In the 1HNMR spectroscopy, the absence of CHO proton signal at 9.92 and the appearance of
sharp imine singlet at 9.24-8.03 confirmed the formation of schiff bases. The signal
corresponding to the 2’- hydroxyl groups was shifted downfield because of intramolecular
hydrogen bonding between it and the nitrogen of imine (OH----N=C) [42, 43].
The mass spectra of the compounds displayed base peak at M+ and M++1 corresponding to
their respective molecular weight.
3.2.1. Vilsmeier formamidinium phosphoro dichloridate sal [19]
IR (KBr cm-1): 3022-2650(OH, CH, P-OH stretching); 1643(C=N stretching); 1614, 1583,
1517, 1346(P=O stretching, free/bonded and C-O stretching); 1109, 1076 (P-O-C stretching);
555, 528, 493(P-Cl vibration).
3.2.2. 2,4-Dihydroxy benzaldehyde[19]
IR (KBr cm-1): 3103 (OH-stretching); 3037 (C-H aromatic stretching); 2848 (C-H stretching
aldehyde); 1628(C=O stretching, aldehyde); 1496 (C=C aromatic stretching); 1395 (C-H bending
1
aldehyde); 1228 (C–OH stretching). HNMR (DMSO, 400 MHz) δ: 10.84 (s, 1H, OH); 10.83 (s,
1H, OH); 9.92 (s, 1H, O=CH); 7.51 (d, 1H, Ar-H, J = 8.8 Hz); 6.40 (dd, 1H, Ar-H, ABX; JBA =
8.8 Hz, JBX = 2.0 Hz); 6.31(d, 1H, Ar-H, JBX = 2.0 Hz). Mass (m/z): 137 (M-1)- (base peak).
3.3.3. 2-(4-aminophenyl) benzimidazole [31]
IR (KBr cm-1): 3400, 3360 (primary N-H stretching), 3048 (aromatic C-H stretching), 1500,
1446 (C=C aromatic stretching); 1400, 1375 (thiazole skeletal bands).
3.3.4. 4-(hydrazonomethyl)benzene-1,3-diol (1)
IR (KBr cm-1): 3518, 3469 (NH str); 3217(OH-str); 2995 (Ar, C-H str); 1616 (C=N str);
1
1521, 1452 (C=C aromatic stretching) 1255 (C–OH stretching). HNMR (DMSO, 300 MHz) δ:
11.00 (s, 1H, OH); 10.50 (s, 1H, OH); 8.75 (s, 1H, imine C-H); 7.42 (s, 1H, Ar-H, J = 8.6 Hz);
6.40 (dd, 1H, Ar-H, J = 2.4 Hz); 6.32(d, 1H, Ar-H, J = 2.1Hz); 2.4 (s, 2H, amine N-H); Mass
(m/z): 151 (M-1)–.
3.3.5. 4-((2-phenylhydrazono)methyl)benzene-1,3-diol (2)
IR (KBr cm-1): 3504, 3423 (NH str); 3315 (OH-str); 3055, 3034 (Ar, C-H str); 1627 (C=N
1
str); 1508, 1444 (C=C aromatic stretching) 1188, 1187 (C–OH stretching). HNMR (DMSO,
300 MHz): 10.73 (s, 1H, OH); 10.11(s, 1H, OH); 9.66 (s, 1H, amine NH); 8.03 (s, 1H, imine
CH); 7.27 (d, 1H, ArH, J = 8.1 Hz); 7.23 (t, 2H, Ar-H, J = 7.8 Hz); 6.89 (d, 2H, Ar-H, J = 7.8
Hz); 6.74 (t, 1H, Ar-H, J = 7.2 Hz); 6.32 (d, 1H, Ar-H, J = 2.2 Hz); 6.29 (s, 1H, s, Ar-H). Mass
(m/z): 228(M+1)+.
6

3.3.6. 4-((2-(2,4-dinitrophenyl)hydrazono)methyl)benzene-1,3-diol (3)
IR (KBr cm-1): 3743, 3618 (NH str); 3556, 3269(OH str); 3099 (C-H, Ar str); 1614 (C=N
str); 1510, 1330 (N-O str);1419 (C=C aromatic stretching); 1224, 1130 (C–OH stretching); 870
(C-N str of NO2). 1HNMR (DMSO, 300 MHz) δ: 11.40 (s, 1H, OH); 10.05 (s, 1H, OH); 9.66 (s,
1H, NH); 8.94 (s, 1H, imine CH); 8.57 (s, 1H, Ar-H); 8.24 (d, 1H, 1ArH, J = 8.3 Hz); 7.71 (d,
2H, 2Ar-H, J = 6.9 Hz); 6.33 (d, 2H, 2Ar-H, J = 6.1 Hz). Mass (m/z): 319 (M+1)+.
3.3.7. 4-((benzylimino)methyl)benzene-1,3-diol (4)
IR (KBr cm-1): 3741 (OH-str); 3024 (Ar, C-H str); 2873, 2291 (C-H str, methylene); 1643
1
(C=N str); 1535, 1454 (C=C aromatic stretching); 1242, 1203, 1170 (C–OH stretching). HNMR
(DMSO, 300 MHz) δ: 13.78 (s, 1H, OH); 9.96 (s, 1H, OH); 8.49(s, 1H, CH-imine); 7.27-7.38 (m,
5H, Ar-H ); 7.23(d, 1H, Ar-H, J = 8.4 Hz); 6.30 (dd, 1H, Ar-H, J = 6.3 Hz); 6.17(d, 1H, Ar-H, J
= 2.2 Hz); 4.70 (s, 2H, methylene). Mass (m/z): 226 (M-1)–.
3.3.8. 4-((butylimino)methyl)benzene-1,3-diol (5)
IR (KBr cm-1): 3732 (OH str); 3055 (C-H, Ar str); 2964 (C-H str, methyl); 2873 (C-H str,
methylene); 1643 (C=N str); 1539, 1452 (C=C aromatic stretching) 1228, 1174, 1145 (C–OH
1
stretching). HNMR (DMSO, 300 MHz) δ: 13.90 (s, 1H, OH); 10.02 (s, 1H, OH); 8.30 (s, 1H,
CH-imine);7.15 (d, 1H, Ar-H, J = 8.4 Hz); 6.23 (dd, 1H, Ar-H, J = 6.4 Hz); 6.12 (d, 1H, Ar-H,
J = 2.9 Hz); 3.49 (t, 2H, CH2, J = 6.7 Hz); 1.59 (pent, 2H, CH2, J = 6.8 Hz); 1.35(hex, 2H, CH2,
J = 7.4 Hz); 0.91(t, 3H, CH3, J = 7.3 Hz). Mass (m/z): 193 (M+) +.
3.3.9. 4-((phenylimino)methyl)benzene-1,3-diol (6)
IR (KBr cm-1): 3722, 3311 (OH str); 3057 (C-H, Ar str); 1641 (C=N str); 1595, 1453 (C=C
1
aromatic stretching) 1209, 1161, 1126 (C–OH stretching). HNMR (DMSO, 300 MHz) δ: 13.54
(s, 1H, OH); 10.26 (s, 1H, OH); 8.78(s, 1H, CH-imine); 7.39-7.43 (m, 3H, Ar-H); 7.34 (d, 2H,
Ar-H, J = 7.5 Hz); 7.26 (d, 1H, Ar-H, J = 7.2 Hz); 6.41 (dd, 1H, Ar-H, J = 6.3 Hz); 6.28 (d, 1H,
Ar-H, J = 2.3 Hz). Mass (m/z): 213(M+)+.
3.3.10. 4-(((2-hydroxyphenyl)imino)methyl)benzene-1,3-diol (7)
IR (KBr cm-1):3751, 3612, 3475(OH- stretching); 3051(aromatic C-H stretching); 1618
(C=N stretching); 1591, 1489 (C=C aromatic stretching); 1234, 1195, 1161, 1126 (C–OH
1
stretching). HNMR (DMSO, 300 MHz) δ: 14.24 (s, 1H, OH); 10.16 (s, 1H, OH);9.66 (s, 1H,
OH); 8.76(s, 1H, imine C-H); 7.36-7.27(m, 2H, ArH) 7.08–7.02 (m, 1H, Ar-H); 6.92-6.81 (m,
2H, Ar-H); 6.34 (dd, 1H, Ar-H, J = 6.3 Hz); 6.22 (d, 1H, Ar-H, J = 2.1 Hz). Mass (m/z): 229
(M+)+.
3.3.11. 4-(((4-fluorophenyl)imino)methyl)benzene-1,3-diol (8)
IR (KBr cm-1): 3745 (OH- stretching); 3026 (aromatic C-H stretching); 1645 (C=N str);
1
1512, (C=C aromatic stretching); 1240, 1163 (C–OH stretching); 1122(C-F str). HNMR
(DMSO, 300 MHz) δ: 13.34 (s, 1H, OH); 10.25 (s, 1H, OH); 8.76 (s, 1H, imine C-H); 7.53 (d,
7

1H, Ar-H, J = 8.6 Hz); 7.42-7.36 (m, 2H, Ar-H); 7.27–7.21 (m, 1H, Ar-H); 6.41(dd, 1H, Ar-H, J
= 6.3 Hz); 6.22 (d, 1H, Ar-H, J = 2.1 Hz). Mass m/z): 232(M+1)+.
3.3.12. 4-((naphthalen-1-ylimino)methyl)benzene-1,3-diol (9)
IR (KBr cm-1): 3741 (OH stretching); 3051 (aromatic C-H stretching); 1616 (C=N stretching
imine); 1541, 1514, 1481 (C=C aromatic stretching); 1220, 1185, 1157 (C–OH stretching); 839,
796, 769, 623 (C-H oop bending, polynuclear aromatic). 1HNMR (DMSO, 300 MHz) δ: 13.59 (s,
1H, OH); 10.35 (s, 1H, OH); 8.85 (s, 1H, imine C-H); 8.13-7.80 (m, 3H, Ar-H); 7.56–7.37 (m,
5H, Ar-H); 6.45 (d, 2H, Ar-H, J = 6.5 Hz); 6.37 (s, 1H, Ar-H). Mass (m/z): 263(M+)+.
3.3.13 4-((2,4-dihydroxybenzylidene)amino)-3-hydroxynaphthalene-1-sulfonic acid (10)
IR (KBr cm-1): 3745, 3608 (OH stretching); 3093 (aromatic C-H stretching); 1616 (C=N
stretching imine); 1521, 1471, 1411 (C=C aromatic stretching); 1340, 1136 (S=O stretching);
1
1222, 1168, 1107, 1049 (C-OH stretching); 850 (S–O stretch). HNMR (DMSO, 300 MHz) δ:
11.06 (s, 1H, OH); 10.75 (s, 1H, OH); 9.91 (s, 1H, OH); 8.93 (s, 1H, imine C-H); 8.80 (d, 2H,
Ar-H, J= 8.7 Hz); 7.90 (s, 1H, Ar); 7.87-7.76 (m, 2H, Ar-H); 7.59–7.50 (m, 1H, Ar-H); 7.43 (t,
1H, Ar-H, J = 7.6 Hz); 6.56 (d, 1H, Ar-H, J = 1.8 Hz); 3.85 (s, 1H, SO3H). Mass (m/z): 360
(M+1)+.
3.3.14. 4-((thiazol-2-ylimino)methyl)benzene-1,3-diol (11)
IR (KBr cm-1): 3745, 3682 (OH stretching); 3026 (aromatic C-H stretching); 1645 (C=N
stretching imine); 1514 (C=C aromatic stretching); 1336, 1317 (thiazole skeletal bands); 1249,
1
1195, 1151 (C–OH stretching). HNMR (DMSO, 300 MHz) δ: 12.00 (s, 1H, OH); 10.57 (s, 1H,
OH); 9.08 (s, 1H, imine C-H); 7.64 (d, 1H, Ar-H, J = 7.9 Hz); 7.57 (d, 1H, Ar-H, J = 4.3 Hz);
6.91-6.87 (m, 1H, Ar-H); 6.44 (dd, 1H, Ar-H, J = 7.1 Hz); 6.36 (d, 1H, Ar-H, J = 5.7 Hz). Mass
(m/z): 221(M+1)+.
3.3.15. 4-((benzo[d]thiazol-2-ylimino)methyl)benzene-1,3-diol (12)
IR (KBr cm-1): 3739 (OH stretching); 3047 (aromatic C-H stretching); 1629 (C=N stretching
imine); 1598, 1504, 1446 (C=C aromatic stretching); 1344, 1311 (thiazole skeletal bands); 1242,
1
1188, 1116 (C–OH stretching). HNMR (DMSO, 300 MHz) δ:11.95 (s, 1H, OH);10.73 (s, 1H,
OH); 9.24 (s, 1H, CH imine); 8.03 (d, 1H, Ar-H, J= 7.8);7.89(d, 1H, Ar-H, J= 7.8 Hz); 7.75(d,
1H, Ar-H , J = 8.6 Hz); 7.50 (t, 1H, Ar-H, J = 7.2 Hz); 7.40 (t, 1H, Ar-H, J = 7.5 Hz); 6.40 (d,
1H, Ar-H, J = 8.1 Hz); 6.30 (s, 1H, Ar-H). Mass (m/z): 272(M+2)+
3.3.16. 4-(((4-(1H-benzo[d]imidazol-2-yl)phenyl)imino)methyl)benzene-1,3-diol (13)
IR (KBr cm-1): 3545 (N-H str, benzothiazole); 3107(OH stretching); 3064 (aromatic C-H
stretching); 1629 (C=N stretching imine); 1539, 1504, 1448 (C=C aromatic stretching); 1396,
1
1359 (thiazole skeletal bands); 1284, 1255, 1211 (C–OH stretching). HNMR (DMSO, 300
MHz): 13.45 (s, 1H, OH);12.91(s, 1H, OH); 10.35 (s, 1H, amine NH); 8.90 (s, 1H, imine CH);
8.23 (d, 2H, ArH, J = 8.4 Hz); 7.66 (d, 2H, Ar-H, J = 6.6 Hz); 7.54 (d, 2H, Ar-H, J = 8.7 Hz);
7.47 (d, 2H, Ar-H, J = 8.7 Hz); 7.20 (d, 1H, Ar-H, J = 4.8 Hz); 6.44 (dd, 1H, Ar-H, J = 6.3 Hz);
6.31 (d, 1H, Ar-H, J= 2.1 Hz). 328 (M-1)– .
8

3.3. Malachite green assay for Hsp90 ATPase suppression
The malachite green enzyme inhibition assay for validating our docking methodology
revealed that all the azomethines are potent Hsp90 inhibitors with an IC50 value of less than 3.1
µM (Table 3). Compound 9 and 11 showed maximum inhibitory potential with an IC50 value of
0.02 µM. The results summarized in Table 3 further revealed that the software predicted potency
of the compounds (docking score) did not match with their biological results. This mismatch can
be explained from the H-bond interactions of the molecules at the N-terminal ATP binding site
of Hsp90. The schiff base derived from 2,4-dinitro phenyl hydrazine (3) was the top scored
ligand because of more hydrogen bond contact with the protein. However in some cases excess
of hydrogen bond between ligand and protein resulted in reduced activity as few interactions are
necessary for effective inhibition. An insight into the hydrogen bond interaction of the two most
potent compounds (9 and 11) revealed some interesting facts regarding their binding affinity.
Both of them were found to interact with different amino acids and water molecules in the
catalytic site of Hsp90. Ligand 9 formed similar hydrogen bond contact like majority of the
compounds in the series, i.e. 2’-OH with Asn51/water 903 and 4’-OH with Asp93/water 902
(Fig. 7a) . The 2’ and 4’ OH of 11 formed H-bond with Asp 54 whereas its imine nitrogen
demonstrated hydrogen bond with water molecule 1026 (Fig. 7b).
3.4. In vitro cell growth assay
The IC50 values of the MTT assay against PC3 cells are given in Table 3. From the table it is
evident that except 9 all the molecules in the series showed moderate to good cytotoxic effect.
The most effective derivative was found to be 13, which showed high cell kill in PC3 cell at IC50
of 4.85 µM. The schiff base 5 and 6 with an IC50 value of 7.43 µM and 7.15 µM respectively
were the other promising anticancer molecules among the newly synthesized compounds.
However unlike malachite green assay, none of the molecules demonstrated IC50 values in
nanomolar range. The poor cell permeability of compound 3 and 10 (Table 4) can be responsible
for the aforementioned activity difference. For rest of the compounds this discrepancy can be
attributed to rapid efflux of the molecules from the PC3 cells by MRP and GST proteins which
are not clients of Hsp90[44-46]. However the proposed hypothesis needs to be substantiated
practically by various biological experiments.
4. Conclusion
In summary, we have discovered novel 2,4-dihydroxy phenyl imines as potent and selective
inhibitors of molecular chaperone Hsp90 with significant antiproliferative effect against PC3
carcinoma cells. The work was facilitated by a validated molecular docking methodology which
identified additional amino acid and water molecule (Asp 54, water 903, 1026 for H-bond
contact) crucial for effective protein suppression in addition to the reported ones(Asp 93, Asn 51,
water 902 for H-bond contact). Our Hsp90 research effort did not establish any comprehensive
structure-activity relationship for the synthesized imines. However, the rational drug design
approach led to the discovery of high potency compounds with excellent ligand efficiency.
Finally, compound 13 with sub micro-molar target affinity and good cellular potency was
identified as an ideal molecule for pre-clinical evaluation in animals. Moreover this imine offers
diverse synthetic routes for multiple chemical modifications. Furthermore, the simple and facile
three step synthetic procedure for 13 will enable cost-effective scale up and optimization of this
9

molecule. Hence this schiff base can be regarded as a high quality viable lead structure for the
future development of potent Hsp90 antagonists with fewer adverse effects.
Disclosure of interest
The authors declare that they have no conflicts of interest concerning this article.
Acknowledgement
We are thankful to DST (Fast track Scheme: SR/FT/CS- 079/2009) and AICTE (RPS
Scheme: 8023/BOR/RID/RPS- 102/2009-10) for providing drug designing software and other
financial assistance for this project. The authors are also thankful to the President, Gokaraju
Rangaraju Educational Trust for providing high speed workstations along with excellent
infrastructure for carrying out this work. We also thank Indian Institute of chemical Technology
(IICT), Hyderabad, India for carrying out NMR and mass spectral studies. The authors are
grateful to University of Buenos Aires, Argentina and IBYME-CONICET, Buenos Aires,
Argentina for providing infrastructure to carry out the biological activity studies.
References
[1] M.W. Hance, K.D. Nolan, J.S. Isaacs, Cancers. 6 (2014) 1065-1097.
[2] M. Taipale, D.F. Jarosz, S. Lindquist, Nat. Rev. Mol. Cell. Biol. 11 (2010) 515-528.
[3] K. Moulick, J.H. Ahn, H. Zong, A. Rodina, L. Cerchietti, E.M. Gomes DaGama, E. Caldas-
Lopes, K. Beebe, F. Perna, K. Hatzi, L.P. Vu, X. Zhao, D. Zatorska, T. Taldone, P. Smith-Jones,
M. Alpaugh, S.S. Gross, N. Pillarsetty, T. Ku, J.S. Lewis, S.M. Larson, R. Levine, H. Erdjument-
Bromage, M.L. Guzman, S.D. Nimer, A. Melnick, L. Neckers, G. Chiosis, , Nat. Chem. Biol. 7
(2011) 818-826.
[4] P. Workman, Cancer. Lett. 206 (2004) 149-157.
[5] T. Didenko, A.M. Duarte, G.E. Karagoz, S.G. Rudiger, Biochim. Biophys. Acta. 1823 (2012)
636-647.
[6] C.K. Vaughan, P.W. Piper, L.H. Pearl, C. Prodromou, The FEBS J. 276 (2009) 199-209.
[7] C. Ban, W. Yang, Cell. 95 (1998) 541-552.
[8] S.M. Roe, C. Prodromou, R. O'Brien, J.E. Ladbury, P.W. Piper, L.H. Pearl, J. Med. Chem. 42
(1999) 260-266.
[9] A.M. Bilwes, C.M. Quezada, L.R. Croal, B.R. Crane, M.I. Simon, Nat. Struct. Biol. 8 (2001)
353-360.
[10] J.L. Johnson, Biochim. Biophys. Acta. 1823 (2012) 607-613.
10

[11] J.M. Eckl, K. Richter, Int. J. Biochem. Mol. Biol. 4 (2013) 157-165.
[12] S.D. Hartson, R.L. Matts, Biochim. Biophysic. Acta. 1823 (2012) 656-667.
[13] C.L. McDowell, R. Bryan Sutton, W.M. Obermann, Int. J. Biol. Macromol. 45 (2009) 310-
314.
[14] E. Yuriev, P.A. Ramsland, J. Mol. Recogn. 26 (2013) 215-239.
[15] R. Garcia-Carbonero, A. Carnero, L. Paz-Ares, The Lancet. Oncol. 14 (2013) e358-369.
[16] A. Gopalsamy, M. Shi, J. Golas, E. Vogan, J. Jacob, M. Johnson, F. Lee, R. Nilakantan, R.
Petersen, K. Svenson, R. Chopra, M.S. Tam, Y. Wen, J. Ellingboe, K. Arndt, F. Boschelli, J.
Med. Chem. 51 (2008) 373-375.
[17] K. Sztanke, A. Maziarka, A. Osinka, M. Sztanke, Biorg. Med. Chem. 21 (2013) 3648-3666.
[18] A. Ganguly, P. Chakraborty, K. Banerjee, S.K. Choudhuri, Eur. J. Pharm. Sci. 51 (2014) 96-
109.
[19] S. Dutta Gupta, D. Snigdha, G.I. Mazaira, M.D. Galigniana, C.V. Subrahmanyam, N.L.
Gowrishankar, N.M. Raghavendra, Biomed. Pharmacother. 68 (2014) 369-376.
[20] S. Dutta Gupta, C.V.S. Subrahmanyam, N.M. Raghavendra, Biofes-2012, Leonia
International Convention Centre, Hyderabad, India (12-13 December, 2012).
[21] P.P. Kung, L. Funk, J. Meng, M. Collins, J.Z. Zhou, M.C. Johnson, A. Ekker, J. Wang, P.
Mehta, M.J. Yin, C. Rodgers, J.F. Davies, 2nd, E. Bayman, T. Smeal, K.A. Maegley, M.R.
Gehring, Bioorg. Med. Chem. Lett. 18 (2008) 6273-6278.
[22] A.N. Jain, J. Comput. Aided. Mol. Des. 21 (2007) 281-306.
[23] P.P. Kung, P.J. Sinnema, P. Richardson, M.J. Hickey, K.S. Gajiwala, F. Wang, B. Huang,
G. McClellan, J. Wang, K. Maegley, S. Bergqvist, P.P. Mehta, R. Kania, Bioorg. Med. Chem.
Lett. 21 (2011) 3557-3562.
[24] A. Yan, G.H. Grant, W.G. Richards, J. R. Soc. Interface. 6 (2008).
[25] M. Rarey, B. Kramer, T. Lengauer, J. Comput. Aided. Mol. Des. 11 (1997) 369-384.
[26] R. Wang, Y. Lu, S. Wang, J. Med. Chem.46 (2003) 2287-2303.
[27] R. Wang, Y. Lu, X. Fang, S. Wang, J. Chem. Inf. Comp. Sci. 44 (2004) 2114-2125.
[28] T. Cheng, X. Li, Y. Li, Z. Liu, R. Wang, J. Chem. Inf. Model. 49 (2009) 1079-1093.
11

[29] Y. Li, L. Han, Z. Liu, R. Wang, Journal of chemical information and modeling, 54 (2014)
1717-1736.
[30] W.L. Mendelson, S. Hayden, Synth. Commun. 26 (1996) 603-610.
[31] R. Dubey, N.S. Hari Narayana Moorthy, Chem. Pharm. Bull. 55 (2007) 115-117.
[32] T. Scheibel, S. Neuhofen, T. Weikl, C. Mayr, J. Reinstein, P.D. Vogel, J. Buchner, J. Biol.
Chem. 272 (1997) 18608-18613.
[33] M. Rowlands, C. McAndrew, C. Prodromou, L. Pearl, A. Kalusa, K. Jones, P. Workman,
W. Aherne, J. Biomol. Screen. 15 (2010) 279-286.
[34] K.W. Harder, P. Owen, L.K. Wong, R. Aebersold, I. Clark-Lewis, F.R. Jirik, Biochem. J.
298 (1994) 395-401.
[35] C. Avila, B.A. Kornilayev, B.S. Blagg, Bioorg. Med. Chem. 14 (2006) 1134-1142.
[36] M.G. Rowlands, Y.M. Newbatt, C. Prodromou, L.H. Pearl, P. Workman, W. Aherne, Anal.
Biochem. 327 (2004) 176-183.
[37] D.A. Scudiero, R.H. Shoemaker, K.D. Paull, A. Monks, S. Tierney, T.H. Nofziger, M.J.
Currens, D. Seniff, M.R. Boyd, Cancer. Res. 48 (1988) 4827-4833.
[38] P.W. Sylvester, Methods. Mol. Biol. 716 (2011) 157-168.
[39] N.J. Marshall, C.J. Goodwin, S.J. Holt, Growth. Regul. 5 (1995) 69-84.
[40] B.S. Creaven, B. Duff, D.A. Egan, K. Kavanagh, G. Rosair, V.R. Thangella, M. Walsh,
Inorganica Chim. Acta. 363 (2010) 4048-4058.
[41] M.S. Refat, I.M. El-Deen, H.K. Ibrahim, S. El-Ghool, Spectrochim. Acta. Mol. Biomol.
Spectrosc. 65 (2006) 1208-1220.
[42] K.K. Upadhyay, A. Kumar, S. Upadhyay, P.C. Mishra, J. Mol. Struc. 873 (2008) 5-16.
[43] A.J. Abdulghani, A.M. Khaleel, Bioinorg. Chem. App. 2013 (2013) 1-14.
[44] X. Liu, Z. Yan, L. Huang, M. Guo, Z. Zhang, C. Guo, Oncol. Rep. 25 (2011) 1343-1351.
[45] J.P. van Brussel, G.J. van Steenbrugge, J.C. Romijn, F.H. Schroder, G.H. Mickisch, Eur. J.
Cancer. 35 (1999) 664-671.
[46] G.L. David-Beabes, M.J. Overman, J.A. Petrofski, P.A. Campbell, A.M. de Marzo, W.G.
Nelson, Int. J. Oncol. 17 (2000) 1077-1086.
12

Fig. 1. Previously reported potent Hsp90 inhibitors.
Fig. 2. Most promising molecule discovered.
Fig. 3. General structure of synthesized imines along with the numbering system used.
Fig. 4. General scheme of synthesis of Schiff bases derived from 2,4-dihydroxy benzaldehyde.
Fig. 5a. Software generated hydrogen bond contacts (white bold line) of the native ligand with
Hsp90 protein.
Fig. 5b. Hydrophobic interaction (white speheres) between ligand and protein as per the docking
programme.
Fig. 6a. Hydrogen bond contact(white bold line)of 3(ball and stick model) with amino acid
residue (ball and stick model) and water molecule (ball and stick model) of 3EKR protein.
Fig. 6b. Hydrophobic interaction of ligand 3(ball and stick model) and Hsp90 protein (PDB ID:
3EKR). The spheres indicates hydrophobic amino acids.
Fig. 7a. Hydrogen bond contact(white bold line) of 9(ball and stick model) with amino acid
residue (ball and stick model) and water molecule (ball and stick model) of 3EKR protein.
Fig. 7b. Hydrogen bond contact(white bold line) of 11(ball and stick model) with amino acid
residue (ball and stick model) and water molecule (ball and stick model) of 3EKR protein.
13

Fig. 1

Fig.2

Fig.3

Fig.4

Fig.5a

Fig.5b

Fig.6a

Fig.6b

Fig.7a

Fig.7b

Table 1
Structure and molecular modelling results of the synthesized imines.
Compound Structure
Total Crash Polar
score
G
score
PMF D score Chem
score score score
C
1
4.27
-0.69
3.93 -100.40 -10.40 -59.37 -11.06
0
2
3
4
5.67
-0.70
2.78 -153.03 -7.04 -86.80 -15.13
1
2
0
7.65 -0.809 5.91 -128.89 -32.31 -108.92 -16.98
4. 36
-0.61
2.14 -108.24 -3.75
2.15 -149.45 -3.57
-77.06 -15.06
-96.18 -13.66
5
6
6.507 -1.22
2
0
2
0
2
3
0
3
3
5.04
4.38
4.97
5.09
5.91
3.81
4.92
5.18
-0.90
-1.06
-0.86
-0.79
-1.49
-0.90
-0.67
-0.74
2.42 -117.68 -12.32 -88.92 -17.57
0.97 -155.92 -8.09 -97.89 -15.86
7
8
2.38 -118.59 -14.32 -90.51 -17.88
2.45 -128.56 -13.74 -97.39 -20.13
2.68 -163.08 -22.80 -125.43 -14.59
9
10
11
12
13
2.31 -123.91
0.10
-71.04 -12.60
2.37 -143.08 -16.49 -98.40 -1570
3.40 -125.07 -20.26 -100.01 -20.82
Note:
C score: The consensus score: the sum of the number of ‘good’ results for each ligand in each
scoring function. G score: It is based on hydrogen bonding, complex (ligand-protein), and internal
(ligand-ligand). PMF score: It is the free energies of interactions for protein-ligand atom pairs. D
Score: It is based on van der Waals interaction between protein and the ligand. Chem score: It
includes terms for hydrogen bonding, metal-ligand interaction, lipophilic contact and rotational
entropy, along with an intercept term. Total score: Total output of all the scores. Crash: The
ability of the compound to penetrate the active site of the protein. Polar: The polar interaction
between the ligand and the protein

Table 2
Percent yield and physical parameters of the synthesized compoundsa.
Compound
Molecular weight
Melting point
248-250
164-166
202-203
168-170
145-147
168-170
146-147
101-103
156-158
230-233
188-190
204-206
245-248
Rfa
%
69
45
51
80
67
65
79
90
65
78
32
57
45
Color/Physical state
yellow crystals
1
2
152
228
318
227
193
213
229
231
263
359
220
270
329
0.35
0.60
0.75
0.67
0.55
0.70
0.56
0.67
0.80
0.67
0.57
0.64
0.68
Light yellow crystals
Maroon crystals
yellow crystals
3
4
5
Orange crystals
Yellow crystals
Light orange crystals
yellow crystals
6
7
8
9
Orange crystals
Brown crystals
10
11
12
13
Yellow crystals
Orange crystals
Light brown crystals
aDetermined in 50% ethylacetate/petroleum ether solvent system.

Table 3
Binding and growth inhibition assay results for compounds 1-13*.
Compound
Malachite green assay
IC50(µM)
MTT assay
IC50(µM)
1
2
3
4
5
6
7
8
9
10
11
12
13
0.29
0.06
2.50
0.04
0.03
3.09
0.94
0.13
0.02
0.09
0.02
0.08
0.15
0.13
13.37
19.12
22.65
56.16
7.43
7.15
30.03
13.38
>200
11.47
20.17
22.40
4.83
Geldanamycin
2.45
MTT: 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide
*Mean of three independent determinations

Table 4
Predicted cell membrane permeability and partition coefficient of the synthesized molecules*
Compound
MDCK cell
permeability
1608.17
332.99
6.50
Log Po/w
Caco cell permeability
1
2
3
4
5
6
7
8
9
10
11
12
13
1354.92
693.34
18.17
1183.02
1022.79
936.73
343.97
1080.35
1226.60
5.94
0.71
1.81
0.64
2.70
1.99
2.69
1.68
3.01
3.36
1.28
1.36
2.40
3.49
593.26
506.91
460.97
156.09
973.36
616.93
2.53
508.76
391.83
210.66
689.78
510.76
453.91
* Schrodinger software’s QuickProp programme was employed for these calculations
Caco cell: cell line derived from human colorectal carcinoma.
MDCK cell: Madian-Darby canine kideney epithelial cells.
Po/w = partition coefficient calculated for n-octanol/water system.

Graphical abstract
14

Highlights
· Structure based drug design of resorcyladehyde schiff bases as novel Hsp90 inhibitors.
· Potent molecules with IC50 value up to 0.02 µM in the protein inhibition assay.
· Significant antiproliferative effect of the compounds with IC50 value up to 4.83 µM.
· Generation of a lead molecule.
15