The Open Nutraceuticals Journal, 2011, 4, 119-124 119
1876-3960/11 2011 Bentham Open
3D QSAR Based Study of Potent Growth Inhibitors of Terpenes as
Neeraja Dwivedia,b,c, Sanjay Mishrab,*, Bhartendu Nath Mishraa, R. B. Singhd and
Vishwa Mohan Katochc
aDepartment of Biotechnology, Institute of Engineering & Technology, Lucknow-226021, India
bDepartment of Biotechnology, College of Engineering & Technology, Moradabad-244001, India
cDepartment of Health Research (Govt. of India) & Indian Council of Medical Research, Ansari Nagar, New
dHalberg Hospital and Research Center, Civil Lines, Moradabad 244 001, UP, India
Abstract: The comparative molecular field analysis (CoMFA) based on three dimensional quantitative structure–activity
relationship (3D-QSAR) studies were carried out employing, natural terpenes as potent antimycobacterial agents. The best
prediction were obtained with a CoMFA standard model (q2 = 0.569, r2 = 0.999) using steric, electrostatic, hydrophobic
and hydrogen bond donor fields. In the current study, a 3D QSAR model of natural product terpenes and their related de-
rivative as antimycobacterial agents was developed. The resulted model exhibits wide-ranging in vitro potency towards
Mycobacterium tuberculosis, with minimum inhibitory concentrations (MIC) from 0.25 ?g/ml saringosterol through 200
?g/ml diaporthein A. In order to establish structure–activity relationships, 3D-QSAR studies were carried out using
CoMFA for natural terpenes (secondary metabolite of plant origin products) as potent antitubercular agents. The in vitro
Minimum Inhibitory Concentration (MIC) data against M. tuberculosis (Mtb) were used. The study was conducted using
twenty four compounds. A QSAR model was developed using a training set of sixteen compounds and the predictive
ability of the QSAR model was assessed employing a test set of eight compounds. The resulting contour maps produced
by the best CoMFA models were used to identify the structural features relevant to the biological activity in this series of
Keywords: CoMFA, Mycobacterium tuberculosis, terpene, MIC, QSAR.
roles in the modern day chemotherapy of tuberculosis. Second-
line natural product or related drugs include capreomycin
and cycloserine is used in tuberculosis chemotherapy. While
rifampicin and streptomycin are part of the front-line treat-
ment regime . Due to a number of factors however, tuber-
culosis still remains a leading cause of death in the world .
The combination of long treatment duration (6–9 months),
increased incidence of (multi or extensive) drug resistance,
co-morbidity with HIV-AIDS and lack of investment in anti-
infectives drug discovery has led to a situation now where
the discovery, development and introduction of new treat-
ments for tuberculosis is critical .
There is currently a re-emerging interest in natural prod-
ucts as being able to provide novel structures for the drug
discovery effort and being particularly effective as antibacte-
rial leads . An excellent minireview by Pauli et al. entitled
“New perspectives on natural products in TB drug research”,
Natural products, or their direct derivatives, play crucial
*Address correspondence to this author at the Department of Biotechnology,
College of Engineering & Technology, Moradabad 244 001, UP, India;
Tel: +91-591-2360818; Fax: +91-591-2360817;
which provides an authoritative account of developments
in both in vitro and in vivo antituberculosis bioassays and
natural product isolation techniques . While the global
economic effects of eradication of tuberculosis have been
reported,  the development of new drugs clearly requires
considerable investment from the public and private sectors.
One of the major worldwide facilitators of research and
development of new antituberculosis drugs and treatment
regimes is the Global Alliance for Tuberculosis Drug Devel-
opment . It is certainly exciting to see natural products
being the focus of such development efforts.
structure and biological activity, quantitative structure–
activity relationships at the three-dimensional level (3D-
QSAR), in general, is considered as a powerful approach
for developing newer drug leads based on small ligand
To gain further insight into the relationship between the
(CoMFA) method, proposed by Cramer et al. in 1988, is
extensively used, in the present practice of drug discovery
. As for as specificity of CoMFA is concerned it enables
to predict biochemical activity of specific molecules by de-
riving a relationship between electrostatic ⁄ steric properties
The 3D-QSAR, comparative molecular field analysis
120 The Open Nutraceuticals Journal, 2011, Volume 4 Dwivedi et al.
and biochemical activities, which can be plotted on contour
maps. Comparative molecular field analysis calculates steric
fields using Lennard–Jones potential and electrostatic fields
using a Coulombic potential. In particular, both of the poten-
tial functions are very steep near the van der Waals surface
of the molecule, causing rapid changes in the surface de-
scriptions and requiring the use of cut-off values so that cal-
culations are not done within the molecular surface. In addi-
tion, a scalar factor is applied to the steric field, so that both
fields can be used in the same partial least-square (PLS)
analysis. Changes in the orientation of the superimposed
molecular set, relative to the calculation grid, can cause sig-
nificant change in the CoMFA results, again probably be-
cause of strict cut-off values. So, the alignment rules are one
of the most sensitive input areas for CoMFA studies. Several
improvements in the alignment methodology like addition of
macroscopic descriptor(s) in the study table and a reverse
method of CoMFA called adaptation of fields for molecular
comparison and topomer CoMFA have been introduced [9-
required for potent inhibitors to inhibit the infection of My-
cobacterium tuberculosis and to obtain predictive 3D-QSAR
models, which may guide rational synthesis of potent novel
compounds. It would have been difficult to perform the
QSAR study at two dimensional levels, as the series under
investigation contain structurally diverse compounds with
variations at different positions. Hence, 3D-QSAR study was
performed to get insight into the structural requirements to
be possessed by molecules possessing Mtb inhibitory activi-
ties. In this study, we report the development of 3D-QSAR
models derived from the most widely used computational
tools, CoMFA, for structurally diverse sets of natural terpe-
nes as inhibitors of Mycobacterium tuberculosis infection
from the literature considering their experimentally reported
in vitro MIC values. The best developed model has been
duly validated. Based on the developed model, some highly
potent compounds could be designed and synthesized.
This study is aimed at elucidating the structural features
penes (Monoterpenes, Diterpenes, Sesquiterpenes, Triterpe-
nes and their related derivatives) as antimycobacterial agents
were developed. The resulted model exhibits wide-ranging in
vitro potency towards M. tuberculosis, with minimum inhibi-
tory concentrations (MIC) from 0.25 ?g/ml saringosterol
through 200 ?g/ml diaporthein A. In order to establish
structure–activity relationships, 3D-QSAR studies were per-
formed using the Comparative molecular field analysis
(CoMFA) of natural terpenes (plant products) as potent anti-
tubercular agents. The in vitro MIC data against Mycobacte-
rium tuberculosis (Mtb) (already available) was used. The
study was performed using Twenty four (24) compounds. A
QSAR model was developed using a training set of sixteen
(16) compounds and the predictive ability of the QSAR
model was assessed using a test set of eight (8) compounds.
In this study, a 3D QSAR model of natural product Ter-
Biological Activity Data
tuberculosis for a group of natural terpenes (mono, di, tri and
sesquiterpenes) containing 24 compounds as antitubercular
agents were used for analysis. The general structure of
The antitubercular activity against Mycobacterium
the compounds, their source and references are indexed
in supplementary table along with their names. The
experimentally calculated MIC value of the compounds
shown in Table 1.
Table 2, the observed and predicted biological activity
in terms of pMIC= -log MIC where, MIC is the minimum
inhibitory concentration, expressed in micro moles per
Dataset for Analysis
from the published work. (References are mentioned in sup-
plementary table). The biological activity used in this study
was expressed as pMIC= -log MIC where, MIC is expressed
in micro moles per milliliter (?M/ml) (Table 2) in in vitro
assay to cause inhibition of Mycobacterium tuberculosis.
Twenty four molecules selected for this study, were taken
Molecular Structures and Optimization
are the natural terpenes taken from an earlier report . The
3d structures of the compounds in the form of sdf files were
taken from pubchem database and their biological activity
data are taken from literature (related references are provided
in Table # 1). The MIC values were converted to the corre-
sponding pMIC (-log MIC) and used as dependent variables
in CoMFA analysis. The MIC values span a range of 0.25
?g/ml to 200 ?g/ml providing a broad spectrum data set for
3D-QSAR study. The 3D QSAR models were generated
using a training set of 16 molecules. Geometrical standardi-
zation followed by specific optimization was performed
using MAXIMIN molecular mechanics and Tripos force
field, Pullman charge supplied with Sybyl7.0 with the
conversion criteria set at 0.05 kcal/(Å mol). Saringosterol is
the most bioactive molecule, was used as template molecule
for alignment. A common backbone structure was created
on the basis of aligned molecules on template molecule.
The structure of the template molecule Saningosterol is
shown in Fig. (1).
For the present study the selected twenty four molecules
Partial Least Squares (PLS) Analysis
(CoMFA interaction energies) and biological activities rela-
tionship. PLS regression method is specifically advantageous
in common cases where the number of descriptors
(independent variables) is comparable to or greater than
the number of compounds (data points) and/or there exist
other factors leading to correlations between variables .
The cross-validation analysis was carried out using Leave-
One- Out (LOO) method where one compound is removed
from the dataset and its activity is predicted using the model
derived from the rest of the dataset. The cross-validated
q2 value and the optimum number of components were
obtained. A minimum column filtering value (?) of
2.00 kcal/ mol was used for the cross-validation. It is
also advantageous to speed up the analysis and reduce noise
. At last non-cross-validated analysis was performed
for calculating non-cross-validated r2 value by using the
optimum number of previously identified components,
employed to analyze the CoMFA result.
PLS algorithm quantifies the structural parameters
QSAR Study of Terpene Natural Inhibitors for Antituberculosis Agent The Open Nutraceuticals Journal, 2011, Volume 4 121
RESULTS AND DISCUSSION
training set of sixteen compounds and a test set of eight
compounds having chemical and biological diversity in both
the training set and the test set molecules. Regardless of the
uncertainty of drug-receptor interactions in general, a statis-
tically significant model was obtained from the CoMFA
analysis. A “cross-validated q2 value” can be defined by
completely analogously to the definition of the conventional
Twenty four molecules were randomly divided into a
Cross-validated q2 = (SD - press)/SD
predictions and SD is the sum of squared deviations of each
biological property value from their mean and press, or pre-
dictive sum of squares, is the sum, over all compounds, of
the squared differences between the actual and “predicted”
biological property values .
Where press is the standard errors of the cross-validated
in the q2 values is observed as the grid spacing is distorted.
The control of the different grid spacing to CoMFA model is
noticeable. The model with the grid spacing of 2.0 Å was
selected as the best model by cross validating value (q2) af-
ter LOO cross-validation.
Often for QSARs model developed by CoMFA a change
piled in Table 3. A cross-validated value (q2) of 0.569 of the
best model was obtained through leave-one-out (LOO) cross
validated PLS analysis, which suggest that the model is
a helpful tool for predicting inhibitory activity of natural
terpenes . The correlation coefficient between the calcu-
lated and experimental activities, non-cross-validated value
(r2) of 0.999 with standard error estimate 0.024. The respec-
tive relative contributions of steric and electrostatic fields
were 0.943 and 0.664, indicating that steric field is more
predominant. Then the condition without electrostatic were
studied and the new q2 and r2 values were found to be 0.739
The statistical parameters of CoMFA analysis is com-
Table 1. Experimentally Calculated MIC Value of the Chemical Sample
S. No. Chemical Sample MIC ?g/ml -log (MIC)
1 CID_10483104_camaric acid 32
2 CID_10007805_rehmanic acid 18
3 CID_10018804_lecheronol A 4
4 CID_10494_oleanolic 28.7
5 CID_14136878_3epiursolic acid 8
6 CID_14161394_saringosterol 0.25
7 CID_3003592_homopseudo 12.5
8 CID_3008606_lecheronol B 128
9 CID_460178_totarol 21.1
10 CID_463811_litosterol 3.13
11 CID_477494_Nephalsterol 12.5
12 CID_485179_lantanolic acid 60
13 CID_485707_3 epioleanolic acid 16
14 CID_6475529_pseudopterazole 12.2
15 CID_6480075_bornyl 25
16 CID_64945_ursolic 41.9
17 CID_64971_betulinic 62.1
18 CID_72293_phorbol ester 25
19 CID_72326_betulin 30
20 CID_72943_heteronemin 6.25
21 CID_9816893_ergog 12.5
22 CID_10248341_diaporthein A 200
23 CID_10473957_diaporthein B 3.1
24 CID_44584761_lantinilic acid 73
122 The Open Nutraceuticals Journal, 2011, Volume 4 Dwivedi et al.
and 0.953 respectively. Practically the electrostatic contribu-
tion was taken to be negligible. The actual and predicted
values of the best model of the training set are given in Table
2. A test set of four compounds (in Table 2) which were not
included in the development of the model, were used for
validation of 3D-QSAR model. This is further validated by
the residual values of the test set (Table 2) on the basis of the
PLS statistics of CoMFA model.
The contour plot representations of the CoMFA results
for tuberculosis inhibitors is presented in Fig. (2) using com-
pound 6 as reference structure. The green-colored regions
indicate areas where steric bulk enhances tuberculosis inhibi-
tory activity, while the yellow contours indicate regions
where steric bulk is detrimental for the biological activity.
Blue-colored regions show areas where electropositive
charged groups enhance inhibitory activity, while red regions
represent where electronegative charged groups improve the
The electrostatic contour map displayed in Fig. (3) shows
a region of red polyhedral space, indicating that the electron-
rich groups are beneficial to the activity. Additionally, a blue
polyhedron in the Fig. (3) indicates that electron-rich sub-
stituent will reduce the biological activity.
Table 2. The Observed and Predicted Biological Activity in Terms of pMIC= -log MIC where, MIC is the Minimum Inhibitory
Concentration, Expressed in Micro Moles Per Milliliter (?M/ml)
S. No. Chemical Sample pMIC (Observed) pMIC (Predicted) Residual
1 *CID_10483104_Camaric acid
0.319 -0.410 0.729
2 CID_10007805_rehmanic acid 0.346 0.342 0.004
3 *CID_10018804_lecheronol A
0.721 0.710 0.011
4 CID_10494_oleanolic 0.298 0.245 0.053
5 *CID_14136878_3epiursolic acid
0.481 0.450 -0.031
6 CID_14161394_saringosterol -0.721 0.695 -0.026
0.396 0.470 0.074
8 *CID_3008606_lecheronol B
0.206 -0.860 1.066
9 CID_460178_totarol 0.328 0.096 0.232
10 CID_463811_litosterol 0.876 0.897 -0.021
11 CID_477494_Nephalsterol 0.396 0.325 0.071
12 CID_485179_lantanolic acid 0.244 0.221 0.023
13 CID_485707_3 epioleanolic acid 0.361 0.730 -0.369
14 CID_6475529_pseudopterazole 0.399 0.403 -0.004
15 CID_6480075_bornyl 0.311 0.317 -0.006
16 CID_64945_ursolic 0.268 0.295 -0.027
17 CID_64971_betulinic 0.242 0.238 0.004
18 CID_72293_phorbol ester 0.311 0.385 -0.074
19 CID_72326_betulin 0.294 0.814 -0.520
0.545 0.435 0.113
0.396 0.237 0.159
22 *CID_10248341_diaporthein A
0.188 -1.232 1.420
23 CID_10473957_diaporthein B 0.884 0.878 0.006
24 CID_44584761_lantinilic acid 0.233 0.229 0.004
Fig. (1). Active structure of Saringosterol.
QSAR Study of Terpene Natural Inhibitors for Antituberculosis Agent The Open Nutraceuticals Journal, 2011, Volume 4 123
Fig. (2). CoMFA contour maps. In the steric contour map to the
left, green contours indicate areas where steric bulk is predicted to
increase antimycobacterial activity, while red contours indicate
regions where steric bulk is predicted to decrease activity. The elec-
trostatic contour map on the right displays yellow polyhedra where
partial negative charge is correlated with antimycobacterial activity;
the blue polyhedra indicating a relationship between partial positive
charge and activity with bioactive compound Saringosterol.
Fig. (3). CoMFA contour map of the steric field and electrostatic
field with alignment of the compounds used in the training set of
green polyhedrons. If a substituent, such as 3-pentyl group,
is attached on molecule 6, it occupies the green contour and
enhances the biological activity. In compounds where the
substituent group R is cyclopentyl and 1-methylcyclohex-1-
yl, can fit into the green contour, and the activity of these
compounds are higher. Additionally, the contour plot shows
a yellow polyhedron in the top left corner of Fig. 2. The yel-
low contour will slow down the biological activity, if a bulky
substituent exist, which represents the disfavored steric re-
gion. Consequently, these compounds almost have no bio-
The steric contour map is displayed in Fig. (3) by big
CoMFA on a series of natural derivatives of terpenoid com-
pounds having inhibitory activity against Mycobacterium
tuberculosis. LOO cross-validation value (q2) and non-cross-
validated value (r2) was obtained 0.569 and 0.999 respec-
tively, shows the acceptable CoMFA model. The developed
CoMFA model also possesses promising foretelling ability
as validated by the testing on the external test set. It could be
significant to elucidate the relationship between compound
structures and biological activities to facilitate designing of
more potent semi-synthetic terpene inhibitors. Our theoreti-
cal prediction may lead to the establishment of developing
new semi-synthetic inhibitors for antimycobacterial activity
using synthetic chemistry approaches.
In present study a 3D QSAR model was developed using
Director) IFTM, Moradabad, U.P., INDIA and Prof. A. Sri-
vastav (Director), College of Engineering and Technology,
Moradabad, U.P., India, for providing research promotional
grant and their generous help and encouragement during the
course of experimental work and manuscript preparation.
The authors are grateful to Prof. R.M. Dubey (Managing
web site along with the published article.
Supplementary material is available on the publishers
Table 3. PLS Statistics of CoMFA 3D-QSAR Model
PLS Statistics CoMFA
q2 (leave-one out cross-validated predicted power of model) 0.569
r2 (correlation coefficient squared of PLS analysis) 0.999
N (optimum number of components obtained from cross-validated PLS analysis and the same used in final non cross-validated analysis) 3
Standard error of estimate (SEE) 0.024
F-test value (F-value) ( n1= 6, n2= 9 ) 1690.169
R2 prediction 0.507
Steric field contribution from CoMFA 0.943
Electrostatic field contribution from CoMFA 0.664
124 The Open Nutraceuticals Journal, 2011, Volume 4 Dwivedi et al.
epidemiology, prevention and treatment. Br Med Bull 2005; 73/74:
The Global Plan to Stop TB, 2006–2015. Document available from
Newman DJ, Cragg GM, Snader KM. The influence of natural
products upon drug discovery. Nat Prod Rep 2000; 17: 215.
Economics of TB drug development. Report available from Global
Alliance for TB Drug Development, www.tballiance.org.
Pauli GF, Case RJ, Inui T, et al. New perspectives on natural prod-
ucts in TB drug research. Life Sci 2005; 78: 485-94.
Economics of TB drug development. Report available from Global
Alliance for TB Drug Development, www.tballiance.org
Web source: www.tballiance.org.
Cramer RD III, Patterson DE, Bunce JD. Comparative molecular
field analysis (CoMFA): Effect of shape on binding steroids to car-
rier proteins. J Am Chem Soc 1988; 110: 5959-67.
Talale TT, Kulkarni SS, Kulkarni VM. Development of pharma-
cophore alignment models as input for comparative molecular field
analysis of a diverse set of azole antifungal agents. J Chem Inf
Comput Sci 1999; 39: 958-66.
Murthy VS, Kulkarni VM. 3D-QSAR CoMFA and CoMSIA on
protein tyrosine phosphatase 1B inhibitors. Bioorg Med Chem
2002; 10: 2267-82
Gohlke H, Klebe G. DrugScore meets CoMFA: adaptation of fields
for molecular comparison (AFMoC) or how to tailor knowledge-
based pair-potentials to a particular protein. J Med Chem 2002; 45:
Cramer RD. Topomer CoMFA: a design methodology for rapid
lead optimization. J Med Chem 2003; 46: 374-88.
Copp BR, Pearce AN. Natural product growth inhibitors of Myco-
bacterium tuberculosis. Nat Prod Rep 2007; 24: 278-97.
Hokuldsson A. PLS regression methods. J Chemometrics 1988; 2:
LP. Multidrug-resistant tuberculosis (MDR-TB):
 Zhaoqi Y, Pinghua S. 3D-QSAR Study of Potent Inhibitors of
Phosphodiesterase-4 Using a CoMFA Approach. Int J Mol Sci
2007; 8(7): 714-22.
Burnouf C, Pruniaux MP. Recent advances in PDE4 inhibitors as
immunoregulators and anti-inflammatory drugs. Curr Pharm Des
2002; 8: 1255-96.
Xu M, Zhang AQ, Han SK, Wang LS. Studies of 3D-quantitative
structure-activity relationships on a set of nitroaromatic com-
pounds: CoMFA, advanced CoMFA and CoMSIA. Chemsphere
2002; 48: 707-15.
Woldemichael GM, Gutierrez-Lugo M-T, Franzblau SG, Wang Y,
Suarez E, Timmermann BN. Mycobacterium tuberculosis growth
inhibition by constituents of Sapium haematospermum. J Nat Prod
2004; 67: 598.
Gu J-Q, Wang Y, Franzblau SG, Montenegro G, Yang D, Tim-
mermann BN. Antitubercular constituents of Valeriana laxiflora.
Planta Med 2004; 70(6): 509-14.
Rodr´ıguez II, Rodr´ıguez AD. Homopseudopteroxazole, a new
antimycobacterial diterpene alkaloid from Pseudopterogorgia elisa-
bethae. J Nat Prod 2003; 66(6): 855-7.
Sayed KAE, Bartyzed P, Shen X, Zjawiony JK, Hamann MT.
Marrine natural products as antituberculosis agents. Tetrahedron
2000; 56: 949-53.
Rukachaisirikul T, Prabpai S, Kongsaeree P, Suksamrarn A. (+)-
Bornyl piperate, a new monoterpene ester from Piper aff. pedicella-
tum roots. Chem Pharm Bull 2004; 52(6): 760-1.
Chumkaew P, Karalai C, Ponglimanont C, Chantrapromma K.
Antimycobacterial activity of phorbol esters from the fruits of
Sapium indicum. J Nat Prod 2003; 66(4): 540-3.
Guchi KI, Saitoh S, Yamata Y. Novel 19-oxygenated Saitoh S,
Yamata Y. Novel 19-oxygenated sterols from the okinawan soft
coral Litophytom Viridis. Chem Pharm Bull 1989; 37: 2553-4.
Boruwa B, Kalita N, Barua JC, et al. Synthesis, Absolute Stereo-
chemistry and Molecular Design of the New Antifungal and Anti-
bacterial Antibiotic Produced by Streptomyces sp. Med Chem Lett
2004; 14: 3571.
Received: December 03, 2010
Revised: March 07, 2011 Accepted: March 10, 2011
© Dwivedi et al.; Licensee Bentham Open.
This is an open access article licensed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons. org/licenses/
by-nc/3. 0/) which permits unrestricted, non-commercial use, distribution and reproduction in any medium, provided the work is properly cited.