Research Paper

Conserved Water Molecule-dependent Docking Strategy

and Atom-Based 3D QSAR Studies to Design Heat Shock
Protein 90 Inhibitors
S. D. GUPTA1,2*, C. V. S. SUBRAHMANYAM1, N. L. GOWRISHANKAR3 AND N. M. RAGHAVENDRA1
1
Department of Pharmaceutical Chemistry, Gokaraju Rangaraju College of Pharmacy, Hyderabad-500 090, 2R&D Centre,
Department of Pharmaceutical Sciences, Jawaharlal Nehru Technological University, Hyderabad-500 085, 3Prime College of
Pharmacy, Erattayal, Palakkad-678 551, India

Gupta et al.: Atom-Based 3D-QSAR Design of Heat Shock Protein 90 Inhibitors

In this study a methodology was described to recognize the conserved water molecules required for
efficient binding of ligands with heat shock protein. Subsequently, a substantiated procedure for the
determination of an effective docking methodology was generated using various programs of Schrodinger
and SYBYL softwares. In order to identify the essential structural features responsible for heat shock
protein inhibitions, an atom based 3D-quantitative structure-activity relationship model with an excellent
surmising ability in both external and internal validation was established. The results of 3D-quantitative
structure-activity relationship and docking studies correlated well with each other. A combined analysis
of docking results and quantitative structure-activity relationship analysis indicates: The key amino acids
and water molecules in the active pocket of heat shock protein. The important structural requirement
in ligands that will lead to enhanced biological activity. The results provide a set of simple and effective
guidelines for rationally designing new heat shock protein inhibitors prior to their synthesis.

Key words: Heat shock protein, cancer, docking, conserved water molecules, atom based 3D-QSAR, Sybyl,
Schrodinger

The use of existing anticancer drugs is hampered molecular chaperone with an average molecular weight
due to development of resistance against them[1]. of 90-kDa[5]. There are two major types of cytosolic
Resistance is attributed to rapid mutation of the isoforms of Hsp90 in eukaryotes, Hsp90α and Hsp90β.
target protein encoding genes which in turn alters the Hsp90α is the inducible form, which is produced
morphology of the active site[1,2]. Additionally, the during stress conditions of the cells, such as elevated
wide-ranging side effects associated with the present temperature, exposure to UV radiation, infection and
cancer chemotherapeutic agents further restricts their inflammation[6]. Hsp90β is the constitutive form,
usage. These adverse effects, such as gastrointestinal which constitutes 1-2 % of total cellular protein under
distress, hair loss and neutropenia were attributed to the non-stress conditions. Moreover Hsp90 α is over
damage of normal cells that divide rapidly similar to the expressed in cancer cells as an activated complex with
cancer cells[3,4]. Therefore, targeted cancer therapy has co-chaperone proteins Hsp 40, Hsp70, Cdc37, p23, Tpr2,
gained momentum in recent years. The most promising Hip and Hop, while the β form of normal cells resides
target for the development of novel, potent and safer in a free state[7]. Therefore, it gave rise to the hypothesis
anticancer agents needs to satisfy the following two that inhibitors against Hsp90 would not affect normal
criteria, multiple oncogenic proteins along with the cells and will be devoid of toxic side effects associated
ones promoting resistance must be dependent upon the
target for maintaining their structure and function. The This is an open access article distributed under the terms of the Creative
target protein should be over-expressed in cancer cells Commons Attribution-NonCommercial-ShareAlike 3.0 License, which
allows others to remix, tweak, and build upon the work non-commercially,
as compared to the healthy ones. as long as the author is credited and the new creations are licensed under
the identical terms
Heat shock protein 90 (Hsp90) is one such exciting
anticancer target, which is an ATP-dependent Accepted 18 January 2020
Revised 19 Decemebr 2019
Received 02 July 2019
*Address for correspondence
E-mail: [email protected] Indian J Pharm Sci 2020;82(2):341-355

March-April 2020 Indian Journal of Pharmaceutical Sciences 341

with cancer chemotherapy. The expression of Hsp90 to this site of the protein. Second functional domain is
is associated with cancers of breast, pancreas, blood, C-terminal domain of ~12 kDa, which contain binding
colon, colorectal, kidney, brain and prostrate[8]. It sites for co-chaperones and the third domain is the
helps in the folding, maturation, stability, trafficking, protein binding domain of ~12 kDa whose contact with
intracellular disposition and proteolytic turnover of the N-terminal domain is necessary for the latter’s proper
proteins responsible for cancer cell development and functioning[13]. Further, small angle X-ray catering
survival. Such proteins are commonly referred to as (SAXS), TROSY NMR and electron microscopy (EM)
the client proteins. These client proteins appear to studies have shown that Hsp90 is highly flexible as
be addicted to Hsp90 for survival. Till today, more a single unit along with the sequences connecting its
than 300 client proteins have been identified[5], some domains. However, individual domains of the protein
of which are dangerous oncogenes like Akt, Raf-1, are less flexible compared to the entire protein. It was
cdk4 and Src. Few are well established cancer targets, experimentally proven that small molecule inhibitors,
Her-2 (trastuzumab), Bcr-Abl (imatinib), estrogen ATP, co-chaperones and client proteins along with
receptor (tamoxifen), androgen receptor (bicalutamide) various post-translational modifications regulate the
and the vascular growth factor endothelial receptor flexibility of the Hsp90 chaperone[14,15].
(sunitinib)[9]. Some mutant proteins like mutant p53 One of the earliest Hsp90 inhibitors discovered were
and imatinib-resistant Bcr-Abl are also dependent on natural products radicicol (1a) and geldanamycin
Hsp90 for preserving their structure and function[10,11]. (1b) (fig. 1)[16]. Radicicol, obtained from the fungus
It was also revealed that many proteins responsible for Monospeorium bonorden was found to be inactive
mutation like PDGFRβ, COT, IGFR1, CRAF, ARAF, in vivo[17]. Geldanamycin is a type of ansamycin isolated
S6, cyclin D1, and AKT are also clients of Hsp90[10,12]. from Streptomyces hygroscopicus. The highly reactive
Hence inhibition of Hsp90 will lead to arrest of multiple 17-methoxyl group of geldanamycin and its high toxic
proteins involved in the development of cancer and profile hindered further development of geldanamycin
resistance. Thus, the two major drawbacks of cancer as such. Therefore, the 17-allylgeldenamycin,
therapy, drug resistance and toxic side effects could be 17-AAG (tanespimycin, 2a (fig. 2) was developed,
overcome by developing agents against Hsp90. which reached phase I clinical trials but eventually
Hsp90 has three functional domains, first of which abandoned due to severe hepatoxicity, poor solubility,
is the N-terminal domain of ~ 25kDa that fulfills all limited bioavailability and extensive metabolism. The
the structural requirement for ATP binding. Structural other analogues of geldanamycin (reduced form of
alterations facilitated by the hydrolysis of ATP to ADP 17-AAG, retaspimycin, 2b and 17-DMAG,
in this domain are thought to play an essential role in the alvespimycin, 2c) shown in fig. 2 developed to
chaperoning activity of the protein. Blocking of ATPase increase water solubility and bioavailability were also
activity hampers the repairing function of Hsp90. rejected after reaching phase-II clinical study due to
Hence, majority of Hsp90 inhibitors competitively bind unsatisfactory safety profile[18,19]. It is believed that the

Fig. 1: First Hsp90 inhibitors discovered

1a. radicicol and 1b. geldanamycin
342 Indian Journal of Pharmaceutical Sciences March-April 2020
 www.ijpsonline.com

Fig. 2: Hsp90 inhibitors that reached various phases of clinical trials

2a. Tanespimycin, 2b. retaspimycin, 2c. alvespimycin and 2d. CNF2024

benzoquinone ring and the chemical complexity of the in more effective binding. The role of conserved water
anamycins contributed to its off target toxicity. The molecules for more effective ligand-protein binding
purine derivative, BllB021 (CNF2024, 2d, fig. 2) was through hydrogen bonding is already established for
one of the synthetic Hsp90 inhibitor to enter clinical Hsp90 as well as for many other proteins[22,23]. Moreover,
trials in 2006. It reached up to clicical phase-II trials the high rate of failure from designing a drug molecule
but did not exhibit any promising result[20]. Till date to its marketing stage warrants the development and
17 candidates are in clinical trials either alone or in validation of a proper docking method by retaining
combination with other anticancer agents, but none of these water molecules.
those has reached approval stage for a market launch[21].
Computational drug design is one of the powerful
Water plays a crucial role in determining the structure techniques used to understand the key features of
and dynamics of biomolecules. In some cases, hydrogen molecules that turn them active against certain targets.
bond forming groups on the protein surface does not Several reports are available in this area, which are
form any H-bond with the ligands due to more distance focused on Hsp90 inhibitor development[24-26]. Although
(> 5Å). This distance can be bridged by some water these studies have disclosed some of the key properties
molecules commonly referred to as the conserved water that turn few compounds into good Hsp90 inhibitors,
molecules. These water molecules form hydrogen bond there is no systematic study that has considered all
with the protein as well as with the ligands. Thus, a the water molecules of the protein. Furthermore, no
protein-water-ligand complex is formed, which results significant attempts have been made to compare various

March-April 2020 Indian Journal of Pharmaceutical Sciences 343

docking methodologies and correlate all of these with Glide Score (G Score). The G Score is calculated
a QSAR model. In view of the above, it appeared as follows, G Score=0.065×vdW+0.130×Coul
of interest to develop a conserved water molecule- +Lipo+Hbond+Metal+BuryP+RotB+Site, where,
dependent validated docking algorithm, which would vdW=van der Waals energy, Coul=coulomb energy,
aid in the discovery of novel Hsp90 inhibitors. The Lipo=lipophilic term derived from hydrophobic
findings of the docking analysis were further correlated grid potential, H bond=hydrogen bonding term,
with a 3D-QSAR model based on the structural features Metal=metal binding term, BuryP=penalty for buried
of known ligands. polar groups, RotB=penalty for freezing rotatable
bonds and Site=polar interactions in the active site.
MATERIALS AND METHODS
SYBYL docking and protein preparation:
Schrodinger docking, protein and grid preparation:
The docking was performed using SYBYL X-1.2
Docking was carried out with Schrodinger docking version software installed on Dell Precision T-1500
software (Maestro, 9.1 versions) installed on Dell workstation (Intel(R) Core(TM) i7 CPU 860 @
Precision T-1500 workstation (Intel(R) Core(TM) 2.80GHz 2.79GHz; 12.0 GB Ram, 1 TB Hard
i7 CPU 860 @ 2.80GHz 2.79GHz; 12.0 GB Ram, disk). The same co-crystal structure of the protein
1 TB Hard disk). The coordinates of the human Hsp90 and ligand as utililized in Schrodinger docking was
in complex with an inhibitor [(2,4-dihydroxyphenyl) obtained from PDB[27]. The protein was processed for
(S)-2-(pyridin-2-yl)pyrrolidin-1-yl)methanone were binding studies by deleting the B chain, removing the
obtained from Protein Data Bank (PDB, PDB ID: undesirable phosphate/water molecules and addition
3EKR, resolution of 2Å)[27,28]. The protein file was of polar hydrogen. The hydrogens of the protein and
prepared for docking by filling missing side chains and ligands were optimized by AMBER7 FF99 force
loops using Prime program of the software, deleting field. The protomol (virtual active site) was developed
the analogous B chain, removing the undesirable from hydrogen-bearing protein mol2 file by keeping a
phosphate and water molecules, followed by addition threshold factor that determined the extent to which the
of polar hydrogens. The protein with ligand having protomol penetrated the protein, of 0.5 and a bloat used
the lowest ionization penalty state was optimized for to dilate the protomol and included the neighbouring
rendering visual inspection of hydrogen bonds. Further cavities of 0 Å[31,32].
the protein–ligand complex was submitted to restrained
molecular mechanics refinement using the optimized Ligand preparation and docking:
potential for liquid simulations (OPLS) 2005 force
The ligands were first drawn in Chem draw, saved
field. The grid (virtual active site) was generated by
as mol files and then converted into SD file format
extracting the native ligand from its binding site. The
using Schrodinger software. The structures were then
GLIDE program’s Grid Generation application was
cleaned by sanitize utility involving filling valencies,
utilized for the above grid generation process[29].
removing duplicates and producing only one molecule
Ligand preparation and docking: per input structure. It was then followed by a general
clean-up, which filtered out compounds containing
The ligands were optimized for docking employing undesirable substructures, metals, or isotopes and
the ligprep program, which creates relevant correctly compounds not obeying Lipinski’s rule of five. A single
protonated 3D forms of the molecule at pH 7±2 by 3D conformation for each input structure was generated
taking into consideration the tautomeric forms and using the Concord program (random perturbation of
chirality of the molecules[30]. The ligand molecules the position and conformation of the supplied ligand
were then docked with standard precision (SP) and followed by bump relaxation)[33]. The SYBYL docking
extra precision (XP) approaches using the GLIDE
software comprises of 4 different methods for the
program of the software. The XP mode of scoring
purpose of docking. The SYBYL Surflex dock method
includes special recognition terms for hydrophobic
docks compounds without preliminary screening
contacts, improved hydrogen bonding, pi-pi/pi-cation
of the ligands. The SYBYL Surflex dock screen
stacking and detection of buried polar groups. These
method takes into account 3 poses per ligand before
extra features are absent in the SP mode of GLIDE[29].
docking. The SYBYL Surflex dock Geom and Geom
The Schrodinger’s Glide program computes the X algorithm computes docking score by considering
binding efficiency of the molecules in terms of 20 poses per ligand. However, the Geom X method is
344 Indian Journal of Pharmaceutical Sciences March-April 2020
 www.ijpsonline.com
more exhaustive because it takes in to consideration biological activities were not taken into consideration
additional 6 conformations whereas the Geom method in the developed QSAR paradigm.
takes only 4. Moreover, the Geom X mode utilizes
a superior alignment method than that of Geom[34]. RESULTS AND DISCUSSION
The binding affinity of the ligands is predicted by the The most comprehensive methods in each software,
software in terms of total score, which is expressed as i.e. Schrodinger’s Glide XP and SYBYL’s Surflex dock
-log Kd, where Kd is binding constant. A high value of Geom X were utilized to identify the water molecules
total score indicates good protein-ligand binding. vital for effective ligand protein binding. The above
Atom-based 3D-QSAR studies, data sets: objective was achieved by including all the water
molecules and then running the docking engine. It was
Phase 3.0 of Schrodinger Maestro 9.1 software version observed that the compounds employed for verifying
was utilized to establish the 3D-QSAR models for docking methodologies formed hydrogen bonds
Hsp90 inhibitors. The experimental information of with water molecule 902, 903, 981 and 1026 as per
33 molecules involved in this study was collected Schrodinger XP method and with water number 902,
from the literature. The inhibitory concentrations of 903, 981 according to SYBYL Geom X method. Hence
the compounds (IC50) were transformed into respective these four water molecules (902, 903, 981 and 1026)
inhibitory activity, pIC50 (−logIC50) values which were were considered important for computing docking
used as dependent variables in the 3D-QSAR study. scores.
The 2D structure was built using Chem draw software
and saved as mol files. The geometry of the molecules
was converted into 3D form for the purpose of QSAR
studies by utilizing the LigPrep module of the software.

Molecular alignment:

In the 3D-QSAR studies, ligand-based molecular

alignment was employed to obtain an authentic and
dependable model. The most potent[33] was chosen
as the reference molecule to alienate with rest of the
training and test set compounds. The above process was
aided by flexible ligand alignment tool of the software.
The common structure of the ligands is shown in fig. 3 Fig. 3: Skeleton structure for analogs utilized in deriving
and the alienated molecules are depicted in fig. 4. the QSAR model

3D-QSAR analysis:

Atom-based 3D-QSAR analysis was executed based on

the molecular alignment as outlined above. The test set
molecules were chosen by considering that they represent
a wide range of ATPase activities (0.02-200 µM) similar
to that of the training set. Additionally, the training set
was utilized to generate the 3D-QSAR paradigm by
partial least square (PLS) method. Subsequently, the
test set was engaged in verifying the quality of the
models using leave-one-out (LOO) method. Sometimes
internal cross-validation is not sufficient to measure the
accuracy of a model in predicting the activities of new
compounds[35]. Hence, external validation with the help
of Phase tool of Schrodinger software was performed
Fig. 4: Alignment of compounds represented as line model
which utilizes a true test set (radicicol, geldanamycin, Alignment of compounds represented as line model
tanespimycin, retaspimycin, alvespimycin and purine with respect to the most potent molecule 33, which is
analogue B11021) whose chemical structures and highlighted in white ball and stick
March-April 2020 Indian Journal of Pharmaceutical Sciences 345
The intensive comparison of the docking scores and Geom X mode of scoring yielded enrichments
calculated by retaining the 4 prominent conserved water superior to other methods. The docking protocol was
molecules (902, 903, 981 and 1026) were carried out further validated by comparing the software generated
among Schrodinger Glide XP and SP methodologies, ligand Hsp90 binding interactions with the previously
SYBYL Surflex dock, Surflex dock Screen, Surflex described crystallographic analysis. In case of SYBYL
dock Geom and Surflex dock Geom X methods. The Geom X method, it was observed that the 2’-hydroxyl
12 most potent compounds (IC 50 values of 0.02) as group of the native ligand, (2,4-dihydroxyphenyl)(R)-
determined by cell based Akt-Lum assay[27] were docked 2-(pyridin-2-yl)pyrrolidin-1-yl)methanone interacted
with all the other less potent ligands using various with Asp 93 and water molecule 902 via hydrogen
docking methods. The accuracy of the methodologies bond. The 4’-hydroxyl group formed hydrogen bond
was validated by evaluating whether the 12 most potent interaction with amino acid Asn 51 and water molecule
compounds are the same as per docking score rankings. 903. The other prominent hydrogen bond contacts
The conclusions of the comparison are summarized detected were between the carbonyl oxygen and
in Tables 1, 2 and fig. 5. From the aforementioned amino acid Thr 184. Further analysis indicated that the
tables and figure it is evident that the SYBYL Geom pyrrole moiety occupied a hydrophobic area formed

TABLE 1: DOCKING RESULTS VS IC50 (HSP90 INHIBITION ASSAY) VALUES OF REPORTED COMPOUNDS
Ligand Code I II III IV V VI IC50* (µM)
1 -8.10 -10.56 6.15 6.15 6.22 6.27 0.40
2 -7.99 -11.38 6.12 6.12 6.12 6.34 1.70
3 -8.33 -12.07 3.67 3.67 5.32 4.28 200
4 -7.01 -11.56 4.01 3.57 6.55 6.49 74.0
5 -8.69 -11.71 7.25 7.25 7.11 7.26 3.20
6 -9.19 -12.21 5.29 5.29 7.38 8.36 1.40
7 -8.65 -11.66 4.62 4.62 5.06 5.02 0.06
8 -8.35 -13.75 8.33 8.33 9.19 9.30 0.50
9 -9.33 -12.77 6.86 6.86 7.12 9.12 0.30
10 -9.16 -12.77 8.95 8.60 8.95 9.00 0.10
11 -9.01 -12.82 7.40 7.40 8.00 8.20 0.08
12 -9.27 -8.96 8.31 8.31 9.12 9.13 0.40
13 -9.27 -12.60 9.04 4.56 7.85 8.14 2.40
14 -9.27 -12.53 8.87 4.56 8.93 9.86 0.06
15 -8.94 -12.60 4.56 8.88 7.85 8.14 5.60
16 -8.51 -12.78 8.00 8.00 9.72 8.95 0.02
17 -7.42 -8.94 7.49 7.49 8.01 7.71 0.30
18 -6.89 -12.44 6.11 6.11 7.72 8.65 1.40
19 -8.57 -12.64 8.34 8.34 9.12 9.58 0.02
20 -6.67 -12.66 9.40 9.40 9.87 10.08 0.02
21 -6.60 -12.34 7.92 7.92 8.41 8.95 0.04
22 -8.99 -12.87 8.88 8.95 9.46 9.61 0.02
23 -7.35 -12.88 8.15 8.15 8.06 8.36 0.20
24 -6.60 -12.11 9.64 9.64 9.97 10.19 0.02
25 -6.50 -12.04 9.84 9.84 9.97 10.37 0.02
26 -6.89 -13.76 8.88 8.88 9.98 10.56 0.02
27 -5.65 -8.94 9.47 9.47 9.97 10.14 0.02
28 -8.72 -8.94 8.88 9.04 9.78 9.76 0.02
29 -8.91 -11.51 9.00 9.00 9.87 9.77 0.02
30 -5.56 -12.04 8.60 9.04 9.38 9.61 1.40
31 -8.71 -8.31 4.56 4.56 7.48 8.14 0.02
32 -8.91 -9.17 3.67 8.88 9.87 8.79 0.30
33 -8.90 -11.43 4.56 4.01 7.85 8.31 0.02
I. Glide SP docking score. II. Glide XP docking score. III. Total score as per SYBYL Dock method. IV. Total score as per SYBYL Dock screen
method. V. Total score as per SYBYL Geom method. VI. Total score as per SYBYL Geom X method. *As per cell-based Akt-lum Hsp90
inhibition assay
346 Indian Journal of Pharmaceutical Sciences March-April 2020
 www.ijpsonline.com
TABLE 2: VALIDATION OF DOCKING SCORES BY DIFFERENT METHODS
Glide SP rank Glide XP rank Surflex dock Surflex dock Surflex dock Surflex dock Geom
Ligand code
(25)* (42)* rank (67)* screen rank (67)* Geom rank (75)* X rank (75)*
16 18 6 17 17 9 16
19 17 10 13 13 14 11
20 27 9 4 4 5 5
22 8 4 8 8 10 9
24 28 17 2 2 3 3
25 30 20 1 1 2 2
26 25 1 9 9 1 1
27 31 31 3 3 4 4
28 13 32 10 5 8 8
29 10 24 6 7 7 7
31 14 33 28 30 25 25
33 12 25 29 31 22 22

(a)

Fig. 5: Percent accuracy of different docking methods

employed

by amino acid Met 98, Ile 96, Ala 55 and Lys 58. The
other significant hydrophobic affinity was observed
between phenyl ring C and amino acid Tyr, Val 136
and Gly 135. The amino acid residue Leu 107, Val 150 (b)

and Gly 135 formed a hydrophopic region where the

resorcinol moiety was embedded (fig. 6a and 6b). These
interactions were found matching with the reported
crystallographic results (fig. 6c)[27]. In other methods
of docking, several important hydrogen bonding and
hydrophobic interactions were not revealed. This was
particularly prominent with SP and XP method of
(c)
Schrodinger docking where the output for hydrophobic
Fig. 6: Ligand interactions with Hsp90 protein (PDB
interaction for all the compounds was calculated as ID:3EKR)
zero. The above discussion indicated that the Geom X (a). Hydrogen bonding of ligand (ball and stick model)
of SYBYL is superior to all the other methods. and Hap90 protein as per SYBYL surflex X Geom
method. The amino acids and the water molecules
In this atom-based 3D-QSAR study, the compounds forming hydrogen bonds are highlighted in green and
were split into a training set of 21 and a test set of violet color respectively. The yellow color lines indicate the
12 molecules. The most potent compound (33) was hydrogen bonds. (b). Hydrophobic interaction of ligand
chosen as a reference to sketch the rest of the molecules. (ball and stick model) and Hsp90 protein as per SYBYL
Atom-based QSAR models were generated by applying surflex X Geom X method. The orange colored spheres
denote hydrophobic amino acids. (c). Hydrogen bonding
PLS regression to a broad set of binary-valued variables
indicated by dotted lines and hydrophobic interactions
that encrypted whether or not atoms of the ligands designated by semicircle arch of native ligand with Hsp90
occupy several cube-shaped part of space. The veracity protein
March-April 2020 Indian Journal of Pharmaceutical Sciences 347
of the models improved with the increasing number to changes in the training set composition was found
of PLS factors until over-fitting starts to occur. In this to be 0.738. The plots for predicted vs. actual pIC50 for
study, the best QSAR model with 5 PLS factors was training and test set are shown in figs 7a and 7b. The
considered and the QSAR visualization was shown by original versus forecasted activities of the training and
cubic 3D grids mapped by atoms of the ligands. test set molecules are listed in Table 3. The estimated
activity of the molecules used for external validation
The statistical specifications of the developed QSAR
(Table 4) was similar to that of their actual values. This
model were as follows: The training group achieved
particular aspect further exemplifies the perfectness of
R2 (coefficient of determination, i.e. percent variance
the derived QSAR model[36-38].
accounted by the model in the observed activity
data) of 0.9869 with an SD (standard deviation of The 3D-QSAR visualization is generated as cubes by
the regression) of 0.1472. The test batch obtained Q2 Phase, in which the green cubes indicate the positive
(value of Q2 for the predicted activities) of 0.7346 structural features and the red ones denote the negative
with an RMSE (ROOT-mean-square error in the test structural elements responsible for the biological
prediction) of 0.43 and Pearson-R (Pearson R value activity spectrum. The 3 parameters that were utilized to
for the correlation between the predicted and observed evaluate the activity profile of the compounds through
activity for the test set) of 0.8697. The constant atom-based 3D-QSAR model are hydrogen bond
F (ratio of the model variance to the observed activity donor, hydrophobic/non-polar and hydrophobicity.
variance) of 286.69 and p value (indication of degree Fig. 8 indicates the areas where hydrogen bond donors
of confidence) of 3.35e-017 further substantiates the in the ligands increase or decrease ATPase activity. The
QSAR protocol. The stability of the model predictions green cubes near the 2-hydroxyl group of the resorcinol
moiety (ring A) for all the molecules suggested that
any H-bond donor substituent in these 2 positions is
favorable for activity. Hence, it is proposed that the
resorcinol group could be retained in the future design
of Hsp90 inhibitors. Additionally, it was revealed that
any H bond group at C-5 of ring A, C-2 of ring B
(pyrrolidine) and C4 of ring C (phenyl) is detrimental
for Hsp90 inhibition.
The favorable and unfavorable cubes for hydrophobic
property of the ligands are presented in fig. 9. The
two red cubes at C2 and C4 of ring A and C4-C6 of
ring C stresses that hydrophobic substitution at these
(a)
positions decreases inhibitor affinity. Further, the model
demonstrated that a large hydrophobic region around
ring B and C is suitable for designing more potent
compounds. Therefore, it is suggested that replacing the
pyrroloidine ring with bulkier hydrophobic rings like
phenyl and naphthalene might be beneficial for activity.
This would particularly aid in the design of potent
small molecular inhibitors of Hsp90. Six hydrophobic
cubes at C5 of ring A implied that a hydrophobic group
at this position is favored. This proposition is verified
by compound 16 with one chlorine at 4 position of
ring A, exhibiting better potency than compound 10.
(b) This region can be explored in future for increased
Fig. 7: Experimental and atom-based 3D-QSAR model activity by substituting a bulkier hydrophobic group
predicted Hsp90 inhibition values
like ethyl or isopropyl. The contribution of electron
Relation between experimental and predicted Hsp90
inhibitory activity values of a) a training set molecules
withdrawing groups in the ligands towards activity
and b) a test set molecules using atom-based 3D-QSAR data is depicted in fig. 10. The 2 green cubes at C2 of
model ring A indicated that the hydroxyl group at this part can
348 Indian Journal of Pharmaceutical Sciences March-April 2020
 www.ijpsonline.com
TABLE 3: PREDICTED VERSES REPORTED PIC50 VALUES OF TRAINING AND TEST SET MOLECULES
Reported activity
Compound Structure QSAR set Phase predicted pIC50*
(pIC50)*

1 Training 6.397 6.222

2 Training 5.769 5.745

3 Training 3.698 3.691

4 Training 4.130 4.136

5 Training 5.494 5.799

6 Test 5.850 6.075

7 Training 7.221 7.291

8 5.608
Test 6.301

Training
9 6.522 6.365

March-April 2020 Indian Journal of Pharmaceutical Sciences 349

 Reported activity
Compound Structure QSAR set Phase predicted pIC50*
(pIC50)*

10 Training 7.000 7.131

11 Training 7.096 6.806

12 Training 6.397 6.471

13 Training 5.619 5.578

14 Test 7.221 7.134

15 Training 5.251 5.290

16 Training 7.698 7.559

Training
17 6.522 6.736

350 Indian Journal of Pharmaceutical Sciences March-April 2020

 www.ijpsonline.com

Reported activity
Compound Structure QSAR set Phase predicted pIC50*
(pIC50)*

18 Test 5.853 6.757

19 Training 7.698 7.562

20 Test 7.698 7.525

21 Training 7.397 7.549

22 Training 7.698 7.740

Training

23 6.698 6.684

March-April 2020 Indian Journal of Pharmaceutical Sciences 351

 Reported activity
Compound Structure QSAR set Phase predicted pIC50*
(pIC50)*

24 Training 7.698 7.758

25 Training 7.698 7.594

26 Test 7.698 7.623

27 Training 7.698 7.738

28 Test 7.698 7.718

29 Training 7.698 7.603

30 Test 5.853 5.471

31 Training 7.698 7.808

32 Training 6.522 6.364

33 Training 7.698 7.725

*pIC50 = -log IC50. IC50 is expressed as µM

be substituted with electron withdrawing hydrophilic with enhanced activity. The study further highlighted
group like COOH and CONH2. Subsequently, it was the importance of alignment of molecules according
disclosed that substitution of C2 portion of ring B, C-4, to their stereochemistry. Compound 14 (the R isomer)
C-5 part of ring C and C2-C5 region of ring C with aligned in a better fashion than the corresponding S
an electron withdrawing group could lead to molecules isomer (15) which resulted in enhanced potency. This
352 Indian Journal of Pharmaceutical Sciences March-April 2020
 www.ijpsonline.com

TABLE 4: PREDICTED AND REPORTED ACTIVITY

(PIC50) VALUES OF COMPOUNDS USED FOR
EXTERNAL VALIDATION
Reported Phase predicted
Compound
activity (pIC50)* pIC50*
Radicicol 7.22 7.19
Geldanamycin 6.55 6.45
Tanespimycin, 5.90 6.05
Retaspimycin, 6.30 6.25
Alvespimycin 6.34 6.45
Purine analogue
5.76 5.68
B11021
*pIC50 = -log IC50. IC50 is expressed as µM

Fig. 10: Atom-based 3D-QSAR model prediction of

electron withdrawing features of molecule 33
Atom-based 3D-QSAR model prediction of electron
withdrawing features for the most active molecule
33. Green and red cubes indicate the favorable and
unfavorable regions

was further explained by the activity of compound 28,

which is more potent than the corresponding S isomer
(30) and equipotent as the R isomer (31).
Fig. 8: Atom-based 3D-QSAR model visualized hydrogen In conclusion, QSAR studies were performed with
bond donor property of molecule 33 significant structural features of various linker regions
Atom-based 3D-QSAR model visualized hydrogen within the molecules. It was observed that hydrophobic,
bond donor properties of the most potent molecule 33. electron withdrawing and H bond donor groups/
The favored regions are depicted in green cubes and
regions between ring A and B plays no significant role
disfavored regions are shown in red cubes
towards enhanced potency of the molecules. Hence,
it was suggested that this amide connector region
can be substituted with easily synthesizable linkers
like imines, enamines, diazines, esters, O and so
on. Consequently, it was revealed that hydrophobic
region is effective as a connector between ring B and C
whereas H-bond donor and electron withdrawing group
in this region has no major role in determining the
Hsp90 inhibition potential. For example, compounds
6 and 8 have linker (-SO2 and -COOCH2-) in this part
which demonstrated lower suppression potential than
compound 19. A definite conclusion was not possible
for the linker between ring C and D as it depends on the
alignment of the molecules. For example, compounds
Fig. 9: Visualization by atom-based 3D-QSAR model of 24-29, 31 and 33 are equipotent. It was observed that
hydrophobicity parameters of molecule 33
for few compounds (24, 26-29 and 31), the linker
Visualization by atom-based 3D-QSAR model of
hydrophobicity parameter with reference to the most region was surrounded by red cubes in terms of electron
active molecule 33. The favorable regions are indicated in withdrawing and H-bond acceptor properties whereas
green cubes and the unfavorable portion is highlighted in in case of 25 and 33 it was shown to be insignificant.
the form of red cubes However, all of them (24-33) demonstrated a large
March-April 2020 Indian Journal of Pharmaceutical Sciences 353
number of green cubes for hydrophobicity in this region. Conflicts of interest
Therefore, it is recommended that a hydrophobic linker
All authors declare no conflict of interest.
group in this region could lead to potent compounds.
The above work not only highlighted the significance REFERENCES
of conserved water molecules in designing docking 1. Gottesman MM, Lavi O, Hall MD, Gillet J-P. Toward a Better
Understanding of the Complexity of Cancer Drug Resistance.
studies but also provided a rapid protocol for
Annu Rev Pharmacol Toxicol 2016;56:85-102.
identifying them. Another significant contribution of 2. Mansoori B, Mohammadi A, Davudian S, Shirjang S,
the study is the development of a robust technique for Baradaran B. The Different Mechanisms of Cancer Drug
comparison and identification of an effective validated Resistance: A Brief Review. Adv Pharm Bull 2017;7:339-48.
docking procedure using 6 different programs. This 3. Ohnishi S, Takeda H. Herbal medicines for the treatment of
cancer chemotherapy-induced side effects. Front Pharmacol
process could be extended to interaction studies of 2015;6:14.
other drug targets with their ligands. The atom-based 4. Tao JJ, Visvanathan K, Wolff AC. Long term side effects of
3D-QSAR model developed by us highlighted the adjuvant chemotherapy in patients with early breast cancer.
important structural features of molecules necessary for Breast 2015;24(Suppl 2):S149-53.
appropriate Hsp90 inhibition. 5. Calderon BP, Beck R, Glorieux C. Targeting hsp90 family
members: a strategy to improve cancer cell death. Biochem
Thus, from the target point of view, the designed Pharmacol 2019;164:177-87.
strategy identified the important amino acids (Asp 93, 6. Chatterjee S, Burns TF. Targeting Heat Shock Proteins in
Cancer: A Promising Therapeutic Approach. Int J Mol Sci
Thr 184, Lys 58, Asn 51 and Gly 97) and conserved 2017;18:E1978.
water molecules (902, 903, 981 and 1026) required 7. Taldone T, Wang T, Rodina A, Pillarsetty NVK, Digwal
for Hsp90 suppression via hydrogen bond interactions CS, Sharma S, et al. A Chemical Biology Approach to the
with the ligands. Further the work also identified Chaperome in Cancer-HSP90 and Beyond. Cold Spring Harb
Perspect Biol 2020;12: a034116.
the hydrophobic amino acids (Phe 138, Val 186, Ile
8. Calderwood SK. Heat shock proteins and cancer: intracellular
91, Met 98, Leu 48, Leu 107, Ala 55, Ile 96 and Val chaperones or extracellular signalling ligands? Philos Trans R
136) essential for effective protein-ligand binding. Soc Lond B Biol Sci 2018;373: 20160524.
From ligand standpoint, the work exemplifies the 9. Taipale M, Krykbaeva I, Koeva M, Kayatekin C, Westover
significant structural requirement of the compounds KD, Karras GI, et al. Quantitative analysis of HSP90-client
interactions reveals principles of substrate recognition. Cell
for suppressing ATPase function of the chaperone. 2012;150:987-1001.
The biologically active ligands were also predicted 10. Zabinsky RA, Mason GA, Queitsch C, Jarosz DF. It’s not
to be potent by the docking and QSAR studies. Thus magic- Hsp90 and its effects on genetic and epigenetic
the receptor dependent studies (docking) correlated variation. Semin Cell Dev Biol 2019;88:21-35.
11. Gorre ME, Ellwood-Yen K, Chiosis G, Rosen N, Sawyers CL.
well with the ligand based drug designing approach
BCR-ABL point mutants isolated from patients with imatinib
(3D QSAR). Our research effort will particularly aid mesylate-resistant chronic myeloid leukemia remain sensitive
in the rational design of small molecules with greater to inhibitors of the BCR-ABL chaperone heat shock protein
specificity and potency against Hsp90. 90. Blood 2002;100:3041-4.
12. Paraiso KHT, Haarberg HE, Wood E, Rebecca VW, Chen YA,
ACKNOWLEDGEMENTS Xiang Y, et al. The HSP90 inhibitor XL888 overcomes BRAF
inhibitor resistance mediated through diverse mechanisms.
Authors thank the DST (Fast track Scheme: SR/FT/ Clin Cancer Res 2012;18:2502-14.
CS- 079/2009) and AICTE (RPS Scheme: 8023/BOR/ 13. Petrikaite V, Matulis D. Binding of natural and synthetic
inhibitors to human heat shock protein 90 and their clinical
RID/RPS-102/2009-10) for providing drug designing application. Medicina (Kaunas) 2011;47:413-20.
software and other financial assistance for this project. 14. Gupta SD. Hsp90 Flexibility and Development of its Inhibitors
The authors also thank the President, Gokaraju for the Treatment of Cancer. Current Chemical Biology.
Rangaraju Educational Trust for providing high speed 2018;12(1):53-64.
15. Schopf FH, Biebl MM, Buchner J. The HSP90 chaperone
workstations along with excellent infrastructure for
machinery. Nature Rev Molecular Cell Biol. 2017;18(6):345-
carrying out this work. The authors are grateful to 60.
the Indian Institute of Chemical Technology (IICT), 16. Thepchatri P, Eliseo T, Cicero DO, Myles D, Snyder JP.
Hyderabad, India for carrying out NMR and mass Relationship among ligand conformations in solution, in the
solid state, and at the Hsp90 binding site: geldanamycin and
spectral studies. The authors are also grateful to the
radicicol. J Am Chem Soc 2007;129:3127-34.
University of Buenos Aires, Argentina and IBYME- 17. Soga S, Neckers LM, Schulte TW, Shiotsu Y, Akasaka K,
CONICET, Buenos Aires, Argentina for providing Narumi H, et al. KF25706, a novel oxime derivative of
infrastructure to carry out the biological activity studies. radicicol, exhibits in vivo antitumor activity via selective

354 Indian Journal of Pharmaceutical Sciences March-April 2020

 www.ijpsonline.com
depletion of Hsp90 binding signaling molecules. Cancer Res benzaldehyde derived Schiff bases as small molecule Hsp90
1999;59:2931-8. inhibitors: rational identification of a new anticancer lead.
18. Kim T, Keum G, Pae AN. Discovery and development of heat Bioorg Chem 2015;59:97-105.
shock protein 90 inhibitors as anticancer agents: a review of 29. Friesner RA, Banks JL, Murphy RB, Halgren TA, Klicic JJ,
patented potent geldanamycin derivatives. Expert Opin Ther Mainz DT, et al. Glide: a new approach for rapid, accurate
Pat 2013;23:919-43. docking and scoring. 1. Method and assessment of docking
19. Taldone T, Gozman A, Maharaj R, Chiosis G. Targeting Hsp90: accuracy. J Med Chem 2004;47:1739-49.
small-molecule inhibitors and their clinical development. Curr 30. Cross JB, Thompson DC, Rai BK, Baber JC, Fan KY, Hu Y, et
Opin Pharmacol 2008;8:370-4. al. Comparison of several molecular docking programs: pose
20. Biamonte MA, Van de Water R, Arndt JW, Scannevin RH, prediction and virtual screening accuracy. J Chem Inf Model
Perret D, Lee W-C. Heat Shock Protein 90: Inhibitors in 2009;49:1455-74.
Clinical Trials. J Med Chem 2010;53:3-17. 31. Jain AN. Surflex-Dock 2.1: robust performance from ligand
21. Yuno A, Lee M-J, Lee S, Tomita Y, Rekhtman D, Moore B, energetic modeling, ring flexibility, and knowledge-based
et al. Clinical Evaluation and Biomarker Profiling of Hsp90 search. J Comput Aided Mol Des 2007;21:281-306.
Inhibitors. Methods Mol Biol 2018;1709:423-41. 32. Spitzer R, Jain AN. Surflex-Dock: Docking benchmarks and
22. Shimomura A, Yamamoto N, Kondo S, Fujiwara Y, Suzuki S, real-world application. J Comput Aided Mol Des 2012;26:687-
Yanagitani N, et al. First-in-Human Phase I Study of an Oral 99.
HSP90 Inhibitor, TAS-116, in Patients with Advanced Solid 33. Guedes IA, de Magalhaes CS, Dardenne LE. Receptor-ligand
Tumors. Mol Cancer Ther 2019;18:531-40. molecular docking. Biophys Rev 2014;6:75-87.
23. Haider K, Huggins DJ. Combining solvent thermodynamic 34. Wang R, Lu Y, Wang S. Comparative evaluation of 11 scoring
profiles with functionality maps of the Hsp90 binding site functions for molecular docking. J Med Chem 2003;46:2287-
to predict the displacement of water molecules. J Chem Inf 303.
Model 2013;53:2571-86. 35. Zhou N, Xu Y, Liu X, Wang Y, Peng J, Luo X, et al.
24. Yan A, Grant GH, Richards WG. Dynamics of conserved Combinatorial Pharmacophore-Based 3D-QSAR Analysis
waters in human Hsp90: implications for drug design. J R Soc and Virtual Screening of FGFR1 Inhibitors. Int J Mol Sci
Interface 2008;5 Suppl 3:S199-205. 2015;16:13407-26.
25. Kung P-P, Sinnema P-J, Richardson P, Hickey MJ, Gajiwala 36. Ge J, Normant E, Porter JR, Ali JA, Dembski MS, Gao Y, et al.
KS, Wang F, et al. Design strategies to target crystallographic Design, synthesis, and biological evaluation of hydroquinone
waters applied to the Hsp90 molecular chaperone. Bioorg Med derivatives of 17-amino-17-demethoxygeldanamycin as
Chem Lett 2011;21:3557-62. potent, water-soluble inhibitors of Hsp90. J Med Chem
26. Verkhivker GM. Computational Modeling of the Hsp90 2006;49:4606-15.
Interactions with Cochaperones and Small-Molecule 37. Howes R, Barril X, Dymock BW, Grant K, Northfield CJ,
Inhibitors. Methods Mol Biol 2018;1709:253-73. Robertson AGS, et al. A fluorescence polarization assay for
27. Kung P-P, Funk L, Meng J, Collins M, Zhou JZ, Johnson MC, inhibitors of Hsp90. Anal Biochem 2006;350:202-13.
et al. Dihydroxylphenyl amides as inhibitors of the Hsp90 38. Lundgren K, Zhang H, Brekken J, Huser N, Powell RE,
molecular chaperone. Bioorg Med Chem Lett 2008;18:6273-8. Timple N, et al. BIIB021, an orally available, fully synthetic
28. Gupta SD, Revathi B, Mazaira GI, Galigniana MD, small-molecule inhibitor of the heat shock protein Hsp90. Mol
Subrahmanyam CVS, Gowrishankar NL, et al. 2,4-dihydroxy Cancer Ther 2009;8:921-9.

March-April 2020 Indian Journal of Pharmaceutical Sciences 355