4
T. Sreelatha et al. / Bioorg. Med. Chem. Lett. xxx (2014) xxx–xxx
Table 3
rather than that of C-5 of the phenyl group as in case of juglone
and plumbagin. The methyl substituent at C-2 in menadione also
proved to be quite effective. Hence, we can conclude the anticancer
activity of 1,4 naphthoquinone is highest with an oxygenated func-
tionality in the quinone ring. This is further enhanced with a thio-
ether linkage at C-3 position of the quinone. The antimicrobial
studies against Methicillin resistant S. aureus and Fluconazole
resistant C. albicans (Table 3) indicate the derivatives 8g, 10g and
Minimum inhibitory concentration of the compounds against MRSA and FRCA
S. No. Sample code
Minimum inhibitory concentration (MIC) (
l
g/mL)
Gram-positive bacteria
MRSA
Fungal strain
FRCA
1
2
3
4
5
1b
9a
9b
9c
ND
ND
3.9
7.9
7.8
ND
7.8
7.8
11g were most active with MIC values of 15.6 lg/mL. These are
9d
7.8
7.8
4-amino butyric amides of plumbagin, lawsone and menadione.
Based on the zone of inhibition (Table 2) again 8g, 10g and 11g
were most active. Hence we believe introduction of the GABA
group increases the antimicrobial activity of the quinones.
QSAR study: 3D QSAR relating the anticancer activity of the com-
pounds against the two cell lines were developed using VLife QSAR
Plus 1.0 molecular modelling software based on earlier references
with the modelling methodology.28,29 The data were randomly
divided into training and test sets (compounds which did not show
activity were ignored for modelling purposes). The model was
developed using the former and the performance of the model
was tested with the latter set.
The pharmacophore models were generated by aligning all the
molecules on a common template in conjunction with partial least
square method. The alignment of all the molecules on a common
template is shown in (Fig. 6) (A and B). The model with the highest
q2 and pred_r2 is reported below with the corresponding statistical
parameters. Based on the statistical parameters, it could be con-
cluded that the models developed are satisfactory.
6
7
8
9
9e
9f
1c
15.6
15.6
125
15.6
62.5
62.5
7.8
15.6
15.6
250
15.6
62.5
ND
10a
10b
10c
10d
10e
10f
10g
1d
11a
11b
11c
11d
11e
11f
11g
1a
8a
8b
8c
8d
8e
8f
8g
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
62.5
7.8
7.8
15.6
15.6
15.6
15.6
15.6
15.6
15.6
15.6
15.6
15.6
125
125
125
125
7.8
15.6
15.6
15.6
15.6
15.6
15.6
15.6
15.6
15.6
15.6
125
125
125
125
7.8
Model for HeLa
ND
15.6
15.6
ND
7.8
15.6
ND
ND
Activity ¼ ꢀ1:5126 þ 0:0416 ½ESꢁ
ð1Þ
r2 = 0.77, q2 = 0.63, F test = 20.6, pred_r2 = 0.6, n training set = 8, n
Streptomycin 3.9
Fluconazole ND
test = 3.
ND: not determined, as they did not show antimicrobial activity in the well diffu-
sion assay.
Model for SAS
Activity ¼ ꢀ0:13 ꢀ 1:74 ½ESꢁ
ð2Þ
r2 = 0.76, q2 = 0.61, F test = 25.6, pred_r2 = 0.6, n training set = 10, n
test = 4.
of 12 and 14
adione 11f also showed good anticancer activity with an IC50 value
of 16 M (Table 1). It is interesting to note that contrary to our
lM, respectively (Table 1). The valine amide of men-
Both the models reveal that the activity is a function of the
electrostatic nature of the groups near sulphur. For HeLa, as the
electrostatic or electron donating groups near sulphur increases
(like methyl group), activity also increases (since the coefficient
in Eq. 1 is positive). Compound 11f has maximum contribution of
the electrostatic descriptor and it also exhibits very high activity.
Whereas, in the case of SAS cell line, the reverse is true (since
the coefficient in Eq. 2 is negative). So electron donating or with-
drawing groups near the sulphur affect the anticancer activities
of these compounds against these two cell lines. The parity plots
(Figures in Supporting information) which relate the experimental
and predicted activities show the good predictive capability of the
models since all the data points are near the central line.
l
expectations the amides of plumbagin are not bioactive. Since
plumbagin is a potent anticancer compound with an IC50 value of
3 lM against human oral cancer cell lines, it is indeed disappoint-
ing to note that none of its analogues matched its potency. In the
juglone series we have synthesized a more bioactive analogue,
the phenyl alanine amide 9c. Among all the derivatives, those of
lawsone were the most active. This could be attributed to the
chemical structure of lawsone. Lawsone, a hydroxyquinone, can
exist as 1, 2 diketone as well as 1, 4 quinone (Fig. 5).
By structurally modifying the phenolic hydroxyl group, there
should have been a reduction in its activity as reported in previous
studies,27 but we observed an enhancement in the activity. The
MTT assay graphs for selected compounds is shown (Fig. 18) in
Supplementary information. The C-3 substituent, an amino acid
moiety as well as the thioether group could be the reason for ren-
dering the derivatives active. This analogy however fails with the
other quinones. Hence, the factor, which is unique to lawsone, is
the presence of methoxy substituent at C-2 of the quinone ring
O
O
OH
O
O
OH
1, 2-quinone
1, 4-quinone
Figure 5. Keto–enol tautomerism in lawsone.
Figure 6. Alignment of molecules for 3D QSAR for (A) HeLa and (B) SAS cell lines.