Synthesis, characterization and antibacterial activity of cobalt(III) complexes with pyridine-amide ligands

Article abstract of DOI:10.1016/j.ejmech.2007.08.015

The ligands 2-(N-(X-pyridyl)carbamoyl)pyridine (X = 2, 3 or 4 for HL¹-HL³, respectively) and 2,6-bis(N-(Y-pyridyl)carbamoyl)pyridine (Y = 2, 3 or 4 for H₂L⁴-H₂L⁶, respectively) in their mono- and di-deprotonated forms have been used to synthesize kinetically stable cobalt(III) compounds [Co(L^1-3)₃] (1-3) and Na[Co(L^4-6)₂] (4-6), respectively. The Co(III) ion is in octahedral environment and is surrounded by three bidentate ligands in complexes 1-3 and two tridentate ligands in complexes 4-6. Ligands coordinate the cobalt center via amidic-N and pyridine-N centers forming a 5-membered chelate ring. Complexes 1-6 have thoroughly been characterized by the various spectroscopic analyses (¹H NMR, ¹³C NMR, UV-vis, IR, mass), elemental analysis, and conductivity measurement. All complexes have been assayed for in vitro antimicrobial activity against clinically isolated resistant strains of Pseudomonas, Proteus, Escherichia coli and standard strains of Pseudomonas aeruginosa (MTCC 1688), Shigella flexneri (MTCC 1457), Klebsiella planticola (MTCC 2272). All cobalt compounds show mild to moderate activity. However, complexes [Co(L¹)₃] (1) and Na[Co(L⁴)₂] (4) were found to have potent activity against standard and pathogenic resistant bacteria used in the study. Their MIC ranged from 2.7 to 187 μg/ml. In vitro toxicity tests demonstrated that all complexes were less cytotoxic than that of gentamycin on HEK cell lines and the results reveal that these complexes can act as potent antimicrobial agents.

Full text of DOI:10.1016/j.ejmech.2007.08.015

Available online at www.sciencedirect.com
European Journal of Medicinal Chemistry 43 (2008) 2189e2196
http://www.elsevier.com/locate/ejmech
Original article
Synthesis, characterization and antibacterial activity of cobalt(III)
complexes with pyridineeamide ligands
Anurag Mishra a, Nagendra K. Kaushik b, Akhilesh K. Verma b, Rajeev Gupta a,
*
a Department of Chemistry, University of Delhi, Delhi e 110 007, India
b Dr. B.R. Ambedkar Centre for Biomedical Research, University of Delhi, Delhi e 110 007, India
Received 5 May 2007; received in revised form 16 July 2007; accepted 9 August 2007
Available online 14 September 2007
Abstract
The ligands 2-(N-(X-pyridyl)carbamoyl)pyridine (X ¼ 2, 3 or 4 for HL1eHL3, respectively) and 2,6-bis(N-(Y-pyridyl)carbamoyl)pyridine
(Y ¼ 2, 3 or 4 for H2L4eH2L6, respectively) in their mono- and di-deprotonated forms have been used to synthesize kinetically stable cobalt(III)
compounds [Co(L1e3)3] (1e3) and Na[Co(L4e6)2] (4e6), respectively. The Co(III) ion is in octahedral environment and is surrounded by three
bidentate ligands in complexes 1e3 and two tridentate ligands in complexes 4e6. Ligands coordinate the cobalt center via amidic-N and pyr-
idine-N centers forming a 5-membered chelate ring. Complexes 1e6 have thoroughly been characterized by the various spectroscopic analyses
(1H NMR, 13C NMR, UVevis, IR, mass), elemental analysis, and conductivity measurement. All complexes have been assayed for in vitro an-
timicrobial activity against clinically isolated resistant strains of Pseudomonas, Proteus, Escherichia coli and standard strains of Pseudomonas
aeruginosa (MTCC 1688), Shigella flexneri (MTCC 1457), Klebsiella planticola (MTCC 2272). All cobalt compounds show mild to moderate
activity. However, complexes [Co(L1)3] (1) and Na[Co(L4)2] (4) were found to have potent activity against standard and pathogenic resistant
bacteria used in the study. Their MIC ranged from 2.7 to 187 mg/ml. In vitro toxicity tests demonstrated that all complexes were less cytotoxic
than that of gentamycin on HEK cell lines and the results reveal that these complexes can act as potent antimicrobial agents.
Ó 2007 Elsevier Masson SAS. All rights reserved.
Keywords: Cobalt complexes; Pyridineeamide ligands; Antibacterial activity; Cytotoxic activity; MTT assay
1. Introduction
from functional bypassing of that target, or it can be contingent
on impermeability, efflux or enzymatic inactivation of the drug
[4]. In light of these problems, the search for new antibacterial
agents has gained immense popularity in the field of drug
development.
In this race of synthesizing new drugs, cobalt complexes
have attracted a great deal of attention amongst the scientific
community due to their therapeutic uses as tumor imaging
agent [5], antitumor [6], transport protein transferrin (Tf)
[5,6], antimycobacterial [7], antiischaemic [8], antiviral
[9,10], antiparasitic [11], antithrombolytic [11], enzymatic
therapeutics [11], anti-inflammatory activities [12] and as met-
abolic modifier [13]. Herein we describe the synthesis and
characterization of the novel Co(III) complexes as the poten-
tial candidates for antimicrobial activity against standard as
well as clinically isolated resistant bacterial strains with di-
minished cytotoxicity on the HEK cell line.
Emergence of resistance in bacterial strains has become as
one of the prime concerns of the 21st century. There are serious
concerns that untreatable pathogens may develop at an alarm-
ing rate in the near future. Strategies to address this challenge
include the design of improved versions of antibacterial classes
already in clinical use and the use of drug combinations. The
application of these strategies can be quite successful, but
a high risk of rapid resistance development remains [1e3].
Thus, an urgent need for new potent classes of antibiotics
with novel modes of action persists. Resistance in bacteria
can result from the modification of an antibacterial target or
* Corresponding author. Tel.: þ91 11 2766 7624; fax: þ91 11 2766 6605.
E-mail address: rgupta@chemistry.du.ac.in (R. Gupta).
0223-5234/$ - see front matter Ó 2007 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2007.08.015

2190
A. Mishra et al. / European Journal of Medicinal Chemistry 43 (2008) 2189e2196
4'
and/or two pyridine rings per ligand un-coordinated that we
3'
2'
5'
believe have the ability to interact with additional metal ions
[14]. In addition, the central core (cobalt ion) is surrounded
by the hydrophobic ligands, which form an outer lipophilic
sheath. According to our design, such an arrangement may fa-
cilitate the diffusion through bio-membranes, thus enhancing
the antibacterial effectiveness of the complexes. It is to be
noted that similar pyridineeamide-based ligands have also
been used for synthesizing interesting coordination complexes
with novel structural features [15e18].
O
O
Py
N
Co
N
Py
Py
O
N
1. 3NaH, DMF, N2
2. Co(ClO4)2.6H2O
3. Excess O2
N
N
O
3
N
N
NH
Py
N
3
4
[Co(L1)3] (1)
[Co(L2)3] (2)
HL1
6
5
2. Results and discussion
2
4
5
HL2
HL3
2.1. Ligand design and synthesis
N
The bidentate ligands, HL1eHL3, and tridentate ligands,
H2L4eH2L6, have been chosen due to their simple and high
yield method of preparation and their ability to coordinate
metal ions [19]. The ligands HL1eHL3 and H2L4eH2L6 in
their mono- or di-deprotonated form are expected to coordi-
nate transition metal ion via anionic amide-N and pyridine-N
forming a 5-membered chelate ring. This results in the forma-
6
3
5
6
[Co(L3)3] (3)
2
N
Scheme 1. Synthesis of complexes 1e3.
tion of tris-ligated metal complex with [L1e3 ꢀ
tions as a bidentate ligand, while a bis-ligated complex would
result from the tridentate analog, [L4e6 2ꢀ
. The interesting
] where it func-
We have used few pyridineeamide-based bidentate and tri-
dentate ligands in their deprotonated forms to synthesize their
cobalt(III) complexes (Schemes 1 and 2). The bidentate
ligands, HL1eHL3, form the [Co(L1e3)3] complexes whereas
the tridentate ligands, H2L4eH2L6, generate the Na[Co(L4e6)2]
complexes. The ligands coordinate via Namide and Npyridine
donors to the Co(III) ion in an octahedral arrangement.
This coordination mode of the ligands, however, leaves one
]
part is that one pyridine group per ligand of HL1eHL3 and
two pyridine groups per ligand of H2L4eH2L6 remain un-
coordinated due to the geometrical reasons. This arrangement
may orient such un-coordinated pyridine rings to interact with
the cell membrane and/or crucial cations required to maintain
the function of the bacterium cell [14].
10
-
Na+
9
9'
O
O
1. 4NaH, DMF, N2
2. Co(ClO4)2.6H2O
3. Excess O2
N
Co
N
O
O
N
N
N
N
Py
Py
2
Py
Py
HN
NH
N
Py
Py
O
O
5/5'
4/4'
N
N
Na[Co(L4)2] (4)
H2L4
2/2'
6/6'
3/3'
4/4'
3/3'
H2L5
Na[Co(L5)2] (5)
Na[Co(L6)2] (6)
2/2'
N
3/3'
2/2'
5/5'
6/6'
H2L6
Scheme 2. Synthesis of complexes 4e6.

A. Mishra et al. / European Journal of Medicinal Chemistry 43 (2008) 2189e2196
2191
The bidentate ligands, HL1eHL3, and tridentate ligands,
H2L4eH2L6, have been synthesized by the coupling of 2-
pyridinecarboxylic acid (for HL1eHL3) or 2, 6-pyridinedi-
carboxylic acid (for H2L4eH2L6) and respective amine
(2-aminopyridine, 3-aminopyridine or 4-aminopyridine) in
the presence of P(OPh)3 as the coupling reagent. All ligands
give satisfactory microanalytical results and show spectro-
scopic features as reported in the literature [19].
gave satisfactory microanalytical results and authenticated
the purity of the bulk samples.
2.3. NMR studies
The diamagnetic nature of all Co3þ-complexes has helped
1
us in characterizing them by their H (Tables 1 and 2; Figs.
S1 and S2) and 13C NMR spectra (Tables S1 and S2; Figs.
S3 and S4). The NMR spectra of the tris-ligand complexes,
1e3, are quite interesting and deserve special mention. The
mer arrangement of the three deprotonated bidentate ligands
around the cobalt(III) center has caused the asymmetry of all
three coordinated ligands. The effect of this asymmetry is
clearly visible in the form of 3-fold signals in the proton as
well as in the carbon NMR spectra of complexes, 1e3. For ex-
ample, the deprotonated ligand, [L1]ꢀ, [L2]ꢀ, or [L3]ꢀ, is ex-
pected to give eight proton signals due to two chemically
different pyridine rings. Cobalt complexes, 1e3, however,
show 24 signals due to three asymmetric ligands having eight
chemically different protons. Similarly, the 13C NMR spectrum
of the complex 1 clearly shows 33 signals for three asymmetric
ligands each having 11 unique carbon centers (Fig. S3). For ex-
ample, three different signals were observed at 168.88, 168.99
and 169.92 ppm for the C]O quaternary carbon centers in the
case of complex 1. Similarly, the pyridine quaternary carbons
are seen at six different positions in the 13C NMR spectrum.
The 2D {1He13C} correlation spectrum was used for the cor-
rect assignment of the 0 various signals for complex 1
(Fig. S5). For instance, H2 and H6 pyridine protons (observed
as doublets at 7.32, 7.38, 7.38, 8.16, 9.13 and 9.47 ppm) were
2.2. Synthesis and characterization of
cobalt(III) complexes
Complexes 1e6 were synthesized by the reaction of the
respective deprotonated ligand, [Ln]ꢀ/2ꢀ (n ¼ 1e6) with
Co(ClO4)2$6H2O salt in DMF. The in situ generated Co(II)
complex was then oxidized to the final Co(III) complex using
molecular oxygen (Schemes 1 and 2). The red to brown com-
plexes, 1e3, and yellow-green to deep green complexes, 4e6,
were isolated in 60e70% recrystallized yield. Complexes 1e3
are non-electrolyte whereas 4e6 behaved as 1:1 electrolytes in
organic solvents [20]. The FTIR spectra of complexes 1e6 do
not show the NeH stretch and thus clearly indicate the depro-
tonated nature of the amide group. Moreover, a red shift was
observed for the C]Oamide stretch of the Co-complexes com-
pared to the free ligand [21]. These features strongly suggest
the involvement of the anionic Namide in the coordination.
For complexes 1e3, C]Oamide stretches were observed at
three distinct positions indicating the asymmetry of the three
ligands coordinated to the Co(III) ion. For example, three
C]Oamide stretching frequencies were observed at 1625,
1600, and 1580 cmꢀ1 for complex 1. This asymmetry is
most likely the result of the meridional arrangement of the
three ligands around the cobalt center. It is well known from
the literature that the tris-chelate-Co3þ-complexes containing
bidentate ligands can adopt both facial ( fac) and meridional
(mer) geometries. Moreover, both these geometrical isomers
will have D- and L- optical isomers [22,23]. In the mass spec-
tra, the molecular ion peak was clearly observed for the
{[Co(L1e3)3] þ Hþ} and {Na[Co(L4e6)2] þ Hþ} species for
complexes 1e3 and 4e6, respectively, thus asserting the integ-
rity of the complexes in solution. All cobalt compounds also
0
found to be correlated with the six carbon signals (C2 and C6)
found between 147 and 152 ppm for complex 1.
1
The H and 13C NMR spectra of the complexes 4e6 are
quite simple as anticipated for symmetrical bis-ligated co-
balt(III) complex. All individual protons are well separated
and observable in the spectra compared to the free ligand.
Moreover, expected coupling between the protons is also
clearly visible. For example, the proton H10 is clearly seen
0
as a triplet at w8.0 ppm, while the protons H9 and H9 appear
as doublet at w7.0 ppm, for complex 4. The N-pyridine pro-
tons were observed between 6.70 and 7.65 ppm. In line with
the proton NMR spectrum, 13C NMR spectra of complexes
Table 1
1H NMR data for complexes 1e3a
Proton
[Co(L1)] (1)
[Co(L2)3] (2)
[Co(L3)3] (3)
2
e
8.32, 8.19, 7.98 (s, 3H)
9.00 (J ¼ 4.6), 8.77 (J ¼ 5.0), 8.36 (d, 3H)
e
8.30, 8.27, 8.05 (d, 3H)
8.97 (J ¼ 5.3), 8.79 (J ¼ 5.4), 8.34 (d, 3H)
6.94 (J ¼ 5.2), 6.53, 6.21 (d, 3H, J ¼ 5.2)
7.60, 7.53, 7.50 (m, 3H)
e
7.95, 7.92, 7.91 (m, 3H)
6.94 (J ¼ 5.2), 6.53, 6.21 (d, 3H, J ¼ 5.2)
7.90, 7.75, 7.68 (d, 3H)
8.30, 8.27, 8.05 (d, 3H)
e
20
9.47 (J ¼ 5.2), 9.13 (J ¼ 5.2), 8.16 (d, 3H)
7.12 (J ¼ 5.2), 5.78 (J ¼ 7.8), 5.60 (d, 3H, J ¼ 7.9)
7.30, 7.29, 7.29 (m, 3H)
7.26, 7.26, 7.00 (m, 3H)
8.12, 7.71, 7.69 (m, 3H)
6.97, 6.67, 6.56 (m, 3H)
7.91, 7.74, 7.54 (d, 3H)
3
30
4
7.57, 7.54, 7.34 (m, 3H)
7.29, 7.23, 6.98 (d, 3H)
7.91, 7.60, 7.58 (m, 3H)
6.80, 6.77, 6.49 (m, 3H)
8.05, 8.02 (J ¼ 7.6), 7.94 (d, 3H, J ¼ 7.6)
7.92, 7.77, 7.74 (d, 3H)
e
40
5
50
6
7.38 (J ¼ 5.9), 7.38 (J ¼ 5.9), 7.32 (d, 3H)
e
60
a
Spectra recorded in DMSO-d6. J value is in hertz.

2192
A. Mishra et al. / European Journal of Medicinal Chemistry 43 (2008) 2189e2196
Table 2
1H NMR data for complexes 4e6a
the microbroth dilution method (Table 3). Gentamycin was
used as a reference drug for the bacteria. Complexes
[Co(L1)3] (1) and Na[Co(L4)2] (4) showed strong activity
against resistant strains of Pseudomonas, E. coli and standard
strains of Shigella, Klebsiella (MIC ranged from 2.7 to
23.43 mg/ml on different strains). Similarly, complex
Na[Co(L5)2] (5) was found to be effective against Pseudomo-
nas (MIC is 5.5 mg/ml). Other compounds appeared to have
broad spectrum, as they showed mild to moderate activity to-
wards most of the strains. The effectiveness of the complexes
1 and 4 may be due to the position of the un-coordinated pyr-
idine rings that form a cleft suitable to interact strongly with
the cations [14].
Proton Na[Co(L4)2] (4)
Na[Co(L5)2] (5)
Na[Co(L6)2] (6)
2
7.65 (d, 4H, J ¼ 3.8) 7.84 (d, 4H, J ¼ 1.6) 8.10 (d, 4H, J ¼ 5.5)
7.65 (d, 4H, J ¼ 3.8) 7.84 (d, 4H, J ¼ 1.6) 8.10 (d, 4H, J ¼ 5.5)
20
3
6.70 (m, 4H)
6.70 (m, 4H)
7.28 (m, 4H)
7.28 (m, 4H)
7.55 (d, 4H, J ¼ 7.7)
7.55 (d, 4H, J ¼ 7.7)
e
6.77 (m, 4H)
6.77 (m, 4H)
6.72 (m, 4H)
6.72 (m, 4H)
e
6.72 (d, 4H, J ¼ 5.6)
6.72 (d, 4H, J ¼ 5.6)
e
30
4
40
e
5
6.72 (d, 4H, J ¼ 5.6)
6.72 (d, 4H, J ¼ 5.6)
50
6
e
7.63 (d, 4H, J ¼ 1.7) 8.08 (d, 4H)
7.63 (d, 4H, J ¼ 1.7) 8.08 (d, 4H)
60
e
9
7.02 (d, 4H, J ¼ 7.9) 7.44 (d, 4H, J ¼ 7.7) 7.82 (d, 4H, J ¼ 7.7)
7.02 (d, 4H, J ¼ 7.9) 7.44 (d, 4H, J ¼ 7.7) 7.82 (d, 4H, J ¼ 7.7)
90
10
a
8.00 (t, 2H)
7.80 (t, 4H)
8.05 (t, 2H)
Spectra recorded in DMSO-d6. J value is in hertz.
2.6. Cytotoxicity studies
In the present study, the cytotoxic activity of the metal
complexes has also been determined. The study was used to
test the growth inhibition by MTT assay. Data are expressed
in terms of % viability which is directly proportional to the
metabolic active cell number. Cobalt(III) complexes are less
cytotoxic than that of gentamycin (Fig. 1). The cytotoxic activ-
ity results suggest that these cobalt complexes are less cyto-
toxic at moderate concentrations. The importance of such
work lies in the possibility that the new complexes might be
more efficacious drugs against bacterial infections. For this
purpose, a thorough investigation regarding the structureeac-
tivity of the complexes and their stability is required in order
to understand the variation in their biological effects, which
could be helpful in designing more potent antibacterial agents
for the therapeutic use.
4e6 were also simple and all carbon centers were located eas-
ily (Fig. S4).
2.4. Absorption spectra
Absorption spectroscopy of the Co(III) complexes of the
present ligands is the easiest way to characterize them. Com-
plexes 1e3 are red-brown in colour and display the lmax in the
range of 510e540 nm (Fig. S6). Complexes 4e6 are deep
green to yellow-green in colour and show the lmax at
w650 nm (Fig. S7). The difference in the colour as well as
in the position of the lmax for the Co(III) complexes of biden-
tate ligands (HL1eHL3) versus the tridentate ligands (H2L4e
H2L6) is clearly due to the ligand field effect. The high energy
features observed between 350 and 500 nm and those below
350 nm could tentatively be assigned as metal to ligand
charge-transfer and intra-ligand transitions, respectively.
3. Conclusion
2.5. Antibacterial studies
We have shown the preparation of novel cobalt(III) com-
plexes of pyridineeamide-based bidentate and tridentate
ligands. Co(III) complexes have thoroughly been character-
ized by various physical and chemical techniques. The NMR
studies authenticate the octahedral arrangement of the ligands
around the Co(III) ion. The meridional arrangement of the bi-
dentate ligands in the former cobalt complexes, 1e3, has
caused the asymmetry which is clearly reflected in their
In the current study, the synthesized and well characterized
Co(III) complexes have been tested against pathogenic clini-
cally isolated resistant strains of Pseudomonas, Proteus, Es-
cherichia coli and standard strains of Pseudomonas
aeruginosa (MTCC 1688), Shigella flexneri (MTCC 1457),
E. coli (BL21), and Klebsiella planticola (MTCC 2272) using
Table 3
Antibacterial studies on the cobalt complexes
No. Complex
MIC (mg/ml)
Standard bacterial strains
Resistant bacterial strains
Shigella flexneri Klebsiella Planticola Pseudomonas aeruginosa Escherichia coli Proteus vulgaris Pseudomonas aeruginosa
1
2
3
4
5
6
[Co(L1)3]
[Co(L2)3]
[Co(L3)3]
23.43
23.43
46.87
2.7
187
375
375
187
375
375
1.46
5.5
187
187
5.5
375
187
11
187
375
375
187
375
375
187
5.5
187
187
5.5
5.5
187
5.5
187
187
2.7
Na[Co(L4)2] 23.43
Na[Co(L5)2] ea
375
187
1.46
Na[Co(L6)2] 23.43
Gentamycin
1.46
a
No activity.

A. Mishra et al. / European Journal of Medicinal Chemistry 43 (2008) 2189e2196
2193
120
100
80
60
40
20
0
1
3
5
2
4
6
G
1500 750
375 187.5 93.75 46.87 23.48 11.71 5.8
Drug concentration ( g/ml)
2.9
1.46 0.73
Fig. 1. Cytotoxicity of complexes using MTT assay on HEK cells. Cells are grown in the presence of drug at various concentrations up to 48 h and metabolic
activity of these cells was measured using MTT. Results are shown as the % viability. G stands for gentamycin.
NMR spectra. The antibacterial study of the complexes shows
significant activity. Their MIC ranged from 2.7 to 375 mg/ml.
On dilution activity decreases, which shows that the com-
plexes are an effective inhibitor at moderate concentrations.
Complexes 1 and 4 in particular, showed strong activity
against the resistant strains of Pseudomonas, E. coli and stan-
dard strains of Shigella, Klebsiella (MIC ranged from 2.7 to
23.43 mg/ml on different strains). All complexes were also
tested for their cytotoxic activity on the HEK cell lines and
show less cytotoxicity at all tested concentrations when com-
pared with gentamycin. Future studies will be directed to un-
derstand the effect of the un-coordinated nature of the pyridine
ring on the antimicrobial activity and to evaluate the mode of
action of these complexes.
reflux and distillation. Dimethyl sulfoxide (DMSO) was dis-
˚
tilled from 4 A molecular sieves and was stored over sieves.
The ligands, 2-(N-(2-pyridyl)carbamoyl)pyridine (HL1), 2-
(N-(3-pyridyl)carbamoyl)pyridine (HL2), 2-(N-(4-pyridyl)
carbamoyl)pyridine (HL3), 2,6-bis(N-(2-pyridyl)carbamoyl)-
pyridine (H2L4), 2,6-bis(N-(3-pyridyl)carbamoyl)pyridine
(H2L5), and 2,6-bis(N-(4-pyridyl)carbamoyl)pyridine (H2L6),
were synthesized following the procedure for HL1 reported
in the literature [19].
4.2. Physical measurements
The conductivity measurements were done in organic sol-
vents or water using the digital conductivity bridge from Pop-
ular Traders, India (model number: PT-825). Elemental
analysis data were obtained from Elementar Analysensysteme
GmbH Vario EL-III instrument. NMR measurements were
done using an Avance Bruker (300 MHz) instrument. Infra-
red spectra (either as KBr pellet or as a mull in mineral oil)
were recorded using PerkineElmer FTIR 2000 spectrometer.
Absorption spectra were recorded using PerkineElmer
Lambda-25 spectrophotometer. Mass spectra were obtained
from a Jeol SX102/DA-6000 instrument.
4. Experimental
4.1. Materials and methods
All reagents were obtained from the commercial sources
and used as received. FCS, fetal calf serum; PBS,
phosphate-buffered saline; and DMEM were obtained from
Hyclone, USA, whereas DMSO cell culture tested and MTT
were purchased from SRL. N,N0-Dimethyl formamide
4.3. Syntheses
˚
(DMF) was dried and distilled from 4 A molecular sieves
and was stored over sieves. Acetonitrile (MeCN) was dried
by distillation from anhydrous CaH2. Diethyl ether was dried
by refluxing over sodium metal under inert atmosphere. Tetra-
hydrofuran (THF) was dried by refluxing and distilling from
sodium metal and benzophenone. Ethyl alcohol (C2H5OH)
and methyl alcohol (CH3OH) were distilled from magnesium
ethoxide and magnesium methoxide, respectively. Chloroform
(CHCl3) and dichloromethane (CH2Cl2) were purified by
washing with 5% sodium carbonate solution followed by wa-
ter and finally dried over anhydrous CaCl2, before a final
Caution! Metal perchlorate salts are potentially explosive
and should be used in small quantity with great care.
4.3.1. General synthetic procedure for the
cobalt(III) complexes
The respective ligand was dissolved in dinitrogen flushed
DMF and treated with necessary amount of solid NaH. The re-
sulting mixture was then stirred for 30 min at room temperature.
Solid [Co(H2O)6](ClO4)2 was added to the aforementioned

2194
A. Mishra et al. / European Journal of Medicinal Chemistry 43 (2008) 2189e2196
0
mixture. After 30 min of stirring, dry O2 was purged to the solu-
tion for 2 min. The solution was finally stirred for 1 h. The
reaction mixture was filtered followed by the removal of
the solvent under reduced pressure. The crude product was
isolated after washing with diethyl ether. The crude product
thus obtained was dissolved in DMF and subjected to vapour
diffusion of diethyl ether. This results in the highly crystal-
line product within a day. The recrystallized product was
filtered and dried under vacuum.
merged), 126.35, 125.53, 125.04 (C3/C3 for three asymmetric
ligands), 128.93 (two signals are0 merged), 129.62, 134.60
(three signals are merged) (C4/C4 for three asymmetric li-
0
gands), 122.17 (three signals are merged), 122.92 (C5/C5 for
three asymmetric ligand0s), 143.45, 141.67, 140.73, 155.78,
154.88, 154.73 (C6/C6 for0 three asymmetric ligands),
169.91, 169.70, 168.95 (C7/C7 for three asymmetric ligands).
4.3.1.3. [Co(L3)3] (3). Yield: 62%. Anal. Calcd for
C36H33N10O5Co (including one DMF and H2O): C, 58.06;
H, 4.43; N, 18.81. Found: C, 58.26; H, 4.32; N, 18.72. IR
4.3.1.1. [Co(L1)3] (1). Yield: 65%. Anal. Calcd for
C36H33N10O5Co (including one DMF and H2O): C, 58.06;
H, 4.43; N, 18.81. Found: C, 58.36; H, 4.34; N, 18.64. IR
(KBr, n, selected peaks): 1625, 1599, 1582 (C]O) cmꢀ1
.
lmax/nm (DMF, 3, dm3 molꢀ1 cmꢀ1): 534 (210), 412 (sh,
2500), 300 (sh, 18,500). MS (CH3OH) m/z (%): 654.31 (57)
[Co(L3)3 þ Hþ]. Molecular conductivity (w1 mM, DMF,
25 ꢁC): L ¼ 5 S cm2 molꢀ1. dH (300 MHz, DMSO-d6, 25 ꢁC,
TMS): 8.30, 8.27, 8.05 (d, 6H, H2/H6 protons for three asym-
(KBr, n, selected peaks): 1625, 1600, 1580 (C]O) cmꢀ1
.
lmax/nm (CH3CN, 3, dm3 molꢀ1 cmꢀ1): 514 (200), 377 (sh,
3900), 300 (19,000); lmax/nm (DMF, 3, dm3 molꢀ1 cmꢀ1):
514 (270), 375 (sh, 4200), 300 (24,000), 260 (sh, 36,000).
MS (CH3CN) m/z (%): 654.43 (40) [Co(L1)3 þ Hþ].
Molecular conductivity (w1 mM, CH3CN, 25 ꢁC):
L ¼ 8 S cm2 molꢀ1 (the range for 1:1 electrolyte in CH3CN
is 120e160). dH (300 MHz, DMSO-d6, 25 ꢁC, TMS): 9.47,
0
metric ligands), 8.97, 8.79, 8.34 (d, 3H, H2 protons for three
asymmetric ligands), 6.94, 6.53, 6.21 (d, 6H, H3/H5 protons
for three asymmetric ligands), 7.60, 7.53, 7.50 (m, 3H, H3
protons0 for three asymmetric ligands), 7.95, 7.92, 7.91 (m,
3H, H4 protons0 for three asymmetric ligands), 7.90, 7.75,
7.68 (d, 3H, H5 protons for three asymmetric ligands); dC
(300 MHz, DMSO-d6, 25 ꢁC, TMS): 129.83 (three signals
0
0
9.13, 8.16 (d, 3H, H2 protons for three asymmetric0ligands),
7.12, 5.78, 5.60, 7.30, 7.29, 7.29 (d/m, 6H, H3/H3 protons
for three asymmetric 0ligands), 7.26, 7.26, 7.05, 8.12, 7.71,
7.69 (m/m, 6H, H4/H4 protons for three asymmetric ligands),
0
are merged), 154.60, 152.59, 152.29 (C2/C2 for three asym-
0
6.97, 6.67, 6.56, 7.91, 7.74, 7.74 (m/d, 6H, H5/H5 protons for
metric ligands), 122.99 (three signals 0 are merged), 126.60
(two signals are merged), 125.66 (C3/C3 for three asymmetric
ligands), 150.02, 149.53 (two signals are merged), 141.81 (two
three asymmetric ligands), 7.38, 7.38, 7.32 (d, 3H, H6 protons
for three asymmetric ligands); dC (300 MHz, DMSO-d6,
25 ꢁC, TMS): 0 157.02, 156.45, 155.50, 151.07, 150.94,
150.24 (C2/C2 for three asymmetric ligands), 119.97,
119.080, 118.10, 128.65 (two signals are merged), 127.08
(C3/C3 for three asymmetric ligands), 139.85 (two signals
0
signals are merged), 141.09 (C4/C4 for three asymmetric li-
gands), 122.99 (three signals0are merged), 125.23 (two signals
are merged), 123.44 (C5/C5 for three asymmetric ligands),
129.83 (three signals are0 merged), 155.79 (two signals are
merged), 155.06 (C6/C6 for three asymmetric ligands),
168.85 (two signals are merged), 168.44 (C7/C7 for three
asymmetric ligands).
0
are merged), 141.25, 137.40, 136.48, 135.39 (C4/C4 for three
asymmetric ligands), 125.39, 124.48, 123.93, 123.50, 120.67,
0
0
122.50 (C5/C5 for three asymmetric ligands), 148.87, 147.95,
0
147.54, 159.67, 158.24, 160.31 (C6/C6 0 for three asymmetric
ligands), 169.92, 168.88, 168.99 (C7/C7 for three asymmetric
ligands).
4.3.1.4. Na[Co(L4)2] (4). Yield: 52%. Anal. Calcd for
C34H24N10O5CoNa (including one H2O): C, 55.59; H, 3.29;
N, 19.07. Found: C, 55.38; H, 2.97; N, 18.78. IR (KBr, n, se-
lected peaks): 1593, 1578, 1560 (C]O) cmꢀ1. lmax/nm (DMF,
3, dm3 molꢀ1 cmꢀ1): 650 (290), 488 (sh, 1980), 470 (sh, 3040).
MS (CH3OH) m/z (%): 716.89 (100) [Na[Co(L4)2] þ Hþ],
694.93 (20) [[Co(L4)2] þ Hþ]. Molecuꢀla1r conductivity
(w1 mM, DMF, 25 ꢁC): L ¼ 55 S cm2 mol (the range for
1:1 electrolyte in DMF is 60e90). dH (300 MHz0, DMSO-d6,
25 ꢁC, TMS): 7.65 (d, 4H, J ¼ 3.87 Hz, H2/H2 ), 6.70 (m,
4.3.1.2. [Co(L2)3] (2). Yield: 50%. Anal. Calcd for
C36H33N10O5Co (including one DMF and H2O): C, 58.06;
H, 4.43; N, 18.81. Found: C, 58.26; H, 4.32; N, 18.72. IR
(KBr, n, selected peaks): 1622, 1596, 1567 (C]O) cmꢀ1
.
lmax/nm (DMF, 3, dm3 molꢀ1 cmꢀ1): 540 (300), 380 (sh,
7050), 345 (sh, 15,600), 330 (sh, 22,700), 300 (34,000). MS
(CH3OH) m/z (%): 654.18 (60) [Co(L2)3 þ Hþ]. Molecular
conductivity (w1 mM, DMF, 25 ꢁC): L ¼ 7 S cm2 molꢀ1. dH
(300 MHz, DMSO-d6, 25 ꢁC, TMS): 8.32, 8.19, 7.98, 9.00,
0
0
4H, H3/H3 ), 7.28 (m, 4H, H4/H4 ), 7.55 (d, 4H, J ¼ 7.71 Hz,
0
0
H5/H5 ), 7.02 (d, 4H, J ¼ 7.98 Hz, H9/H9 ), 8.0 (t, 2H, H10);
dC (300 MHz, DMSO-d6, 25 ꢁC, TMS): 146.16 (C2), 121.94
(C3), 122.07 (C4), 116.98 (C5), 156.94 (C6), 168.05 (C7),
160.20 (C8), 136.11 (C9), 138.88 (C10).
0
8.77, 8.36 (s/d, 6H, H2/H2 protons0 for three asymmetric li-
gands), 7.57, 7.54, 7.34 (m, 3H, H3 protons for three asym-
metric ligands), 7.29, 7.23, 6.98, 7.91, 7.60, 7.58 (d/m, 6H,
0
H4/H4 protons for three asymmetric ligands), 6.81, 6.77,
0
6.49, 8.05, 8.02, 7.94 (m/d, 6H, H5/H5 protons for three asym-
metric ligands), 7.92, 7.77, 7.74 (d, 3H, H6 protons for three
4.3.1.5. Na[Co(L5)2] (5). Yield: 72%. Anal. Calcd for
C37H33N11O7CoNa (including one DMF and two H2O): C,
53.81; H, 3.99; N, 18.66. Found: C, 53.22; H, 3.90; N,
18.41. IR (KBr, n, selected peaks): 1652, 1593
(C]O) cmꢀ1. lmax/nm (DMF, 3, dm3 molꢀ1 cmꢀ1): 650 (70),
asymmetric ligands); dC (300 MHz, DMSO-d6, 25 ꢁC, TMS):
0
148.35, 145.74, 144.90, 150.03, 149.80, 149.38 (C2/C2 for
three asymmetric ligands), 124.50, 123.53 (two signals are

A. Mishra et al. / European Journal of Medicinal Chemistry 43 (2008) 2189e2196
2195
480 (sh, 910), 385 (sh, 4300), 325 (sh, 13,300). MS (CH3OH)
m/z (%): 716.97 (50) [Na[Co(L5)2] þ Hþ]. Molecular conduc-
tivity (w1 mM, DMF, 25 ꢁC): L ¼ 60 S cm2 molꢀ1. dH
(300 MHz, DMSO-d6, 25 ꢁC, TMS): 7.84 (d, 4H,
cell adhesion. Stock solutions (6 mg/ml) of the compounds
made in water were filter-sterilized and then diluted to 3 mg/ml
in incomplete media. The 3 mg/ml solutions were further di-
luted up to 0.73 mg/ml for treatment against HEK cell lines
to give final concentrations range of 1500e0.73 mg/ml. All as-
says were performed in two independent sets of quadruplicate
tests. Control group containing no drug was run in each assay.
Following 48 h of exposure of cells to drug, each well was
carefully rinsed with 200 ml PBS buffer. Cytotoxicity was as-
sessed using MTT (3-[4,5-dimethylthiazol-2yl]-2,5-diphenylte-
trazolium bromide). MTT solutions of 20 ml (5 mg/ml dd H2O)
along with 200 ml of fresh, complete media were added to each
well and plates were incubated for 4 h. Following incubation,
the medium was removed and the purple formazan
precipitate in each well was sterilized in 200 ml DMSO.
Absorbance was measured using Techan microplate reader
(molecular device) at 570 nm and the results are expressed as
% viability which is directly proportional to metabolic active
cell number. Percentage (%) viability was calculated as:
0
0
J ¼ 1.55 Hz, H2/H2 ), 6.77 (d, 4H, H3/H3 ), 6.72 (m, 4H, H4/
0
0
H4 ), 7.63 (d, 4H, J ¼ 1.74 Hz, H6/H6 ), 7.44 (d, 4H,
0
J ¼ 7.73 Hz, H9/H9 ), 7.80 (m, 2H, H10); dC (300 MHz,
DMSO-d6, 25 ꢁC, TMS): 144.63 (C2), 123.05 (C3), 140.38
(C4), 147.02 (C5), 141.74 (C6), 166.97 (C7), 155.29 (C8),
124.10 (C9), 133.04 (C10).
4.3.1.6. Na[Co(L6)2] (6). Yield: 70%. Anal. Calcd for
C37H33N11O7CoNa (including one DMF and two H2O): C,
53.81; H, 3.99; N, 18.66. Found: C, 53.74; H, 4.38; N,
18.53. IR (KBr, n, selected peaks): 1670, 1626, 1599
(C]O) cmꢀ1. lmax/nm (DMF, 3, dm3 molꢀ1 cmꢀ1): 655 (70),
478 (sh, 890), 380 (sh, 4330), 325 (sh, 12,000). MS
(CH3OH) m/z (%): 716.87 (30) [Na[Co(L2)2] þ Hþ]. Molecu-
lar conductivity (w1 mM, DMF, 25 ꢁC): L ¼ 55 S cm2 molꢀ1
.
dH (300 MHz, DMSO-d6, 25 ꢁC, TMS): 8.10 (d, 4H,
0
% Viability ¼ OD in sample well=OD in control well ꢂ 100:
J ¼ 5.46 Hz, H2/H2 ), 6.72 (d, 4H, J ¼ 5.63 Hz, H3/H5), 6.72
0
0
0
(d, 4H, J ¼ 5.63 Hz, H3 /H5 ), 8.08 (d, 4H, H6/H6 ), 7.82 (d,
0
4H, J ¼ 7.7 Hz, H9/H9 ), 8.05 (m, 2H, H10); dC (300 MHz,
DMSO-d6, 25 ꢁC, TMS): 153.32 (C2), 121.62 (C3), 155.44
(C4), 124.68 (C5), 153.32 (C6), 166.36 (C7), 162.40 (C8),
140.83 (C9), 149.54 (C10).
Acknowledgements
R.G. thanks the generous financial support from the Depart-
ment of Science and Technology (DST), Government of India,
New Delhi.
4.4. In vitro antibacterial activity
All complexes have been screened in vitro against clinically
isolated resistant strains of Pseudomonas, Proteus, E. coli and
standard strains of P. aeruginosa (MTCC 1688), S. flexneri
(MTCC 1457), E. coli (BL21), and Klebsiella planticola
(MTCC 2272). Various methods [24e27] are available for
the evaluation of the antibacterial activity of different types
of drugs. However, the most widely used method [27], which
consists of determining the antibacterial activity of the drug is
to add it in known concentrations to the cultures of the test or-
ganisms, was employed.
Appendix. Supplementary information
Supplementary materials (seven figures and two tables) in-
1
clude H and 13C NMR spectra, absorption spectra and 13C
spectral data for complexes 1e6. Supplementary data associ-
ated with this article can be found in the online version, at
doi:10.1016/j.ejmech.2007.08.015 or from the authors.
References
4.4.1. Microbroth dilution assay
[1] H.S. Gold, R.C. Moellering, N. Engl. J. Med. 335 (1996) 1445.
[2] R. Finch, Clin. Microbiol. Infect. 8 (2002) 21.
[3] I.K. Hosein, D.W. Hill, L.E. Jenkins, J.T. Magee, J. Appl. Microbiol. 92
(2002) 90S.
Different concentrations of the test compounds in 200 ml
of culture medium were prepared in 96-well flat-bottomed
microculture plates (Nunc, Nunclon) by double dilution
method. The wells were prepared in triplicate for each con-
centration. Each well was inoculated with 10.0 ml of bacterial
suspension containing 106 cells/ml. The plates were incubated
at 37 ꢁC for 16 h and the OD was measured at A600 nm of the
suspension to assess the inhibition of cell growth due to treat-
ment with compounds. All the tests were repeated up to three
times.
[4] D.M. Livermore, Clin. Infect. Dis. 36 (2003) S11.
[5] T.A.D. Smith, Bioorg. Med. Chem. 13 (2005) 4576.
[6] F. Liang, P. Wang, X. Zhou, T. Li, Z. Li, H. Lin, D. Gao, C. Zhang,
C. Wu, Bioorg. Med. Chem. Lett. 14 (2004) 1901.
[7] R. Maccari, R. Ottana, B. Bottari, E. Rotondo, M.G. Vigorita, Bioorg.
Med. Chem. Lett. 14 (2004) 5731.
[8] J.F. Unitt, K.L. Boden, A.V. Wallace, A.H. Nall, M.E. Coombs, F. Ince,
Bioorg. Med. Chem. 7 (1999) 1891.
[9] T. Takeuchi, A. Bottcher, C.M. Quezada, T.J. Meade, H.B. Gray, Bioorg.
Med. Chem. 7 (1999) 815.
[10] M.V. Capparelli, M. Goodgame, A.C. Skapski, B. Piggott, Inorg. Chim.
Acta 92 (1984) 15.
4.5. In vitro cell growth inhibition assay
[11] (a) N.D. Rawlings, A.J. Barrett, Methods Enzymol. 244 (1994) 19;
(b) N.D. Rawlings, A.J. Barrett, Methods Enzymol. 244 (1994) 461;
(c) J.H. McKerrow, E. Sun, P.J. Rosenthal, J. Bouvier, Annu. Rev. Micro-
biol. 47 (1993) 821;
Cells were seeded in 96-well plates at a concentration of
2e4 ꢂ 103 cells/well in 200 ml of complete media and incu-
bated for 24 h at 37 ꢁC in 5% CO2 atmosphere to allow for

2196
A. Mishra et al. / European Journal of Medicinal Chemistry 43 (2008) 2189e2196
(d) W.T. Chen, Curr. Opin. Cell Biol. 4 (1992) 802;
(e) J.F. Davies, Z. Hostomska, S.R. Jordon, D.A. Matthews, Science 252
(1991) 88.
[18] (a) A.K. Patra, R. Mukherjee, Inorg. Chem. 38 (1999) 1388;
(b) A.K. Singh, V. Balamurugan, R. Mukherjee, Inorg. Chem. 42 (2003)
6497.
[12] A. Bury, A.E. Underhill, D.R. Kemp, N.J. Oshea, J.P. Smith, P.S. Gomm,
Inorg. Chim. Acta 138 (1987) 85.
[19] (a) S.P. Perlepes, T. Kabanos, V. Hondrellis, J.M. Tsangaris, Inorg. Chim.
Acta 150 (1988) 13;
[13] J.Z. Witczak, D. Boryczewski, Bioorg. Med. Chem. Lett. 8 (1998)
3265.
(b) S. Dutta, P.K. Bhattacharya, E. Horn, E.R.T. Tiekink, Polyhedron 20
(2001) 1815;
[14] Our initial studies show that the complexes 1 and 4 can bind one and two
Zn2þ ions, respectively, in the cleft created by the un-coordinated pyri-
dine rings, Inorg. Chem., submitted for publication.
[15] (a) P.K. Mascharak, Coord. Chem. Rev. 225 (2002) 201;
(b) D.S. Marlin, P.K. Mascharak, Chem. Soc. Rev. 29 (2000) 69;
(c) D.S. Marlin, M.M. Olmstead, P.K. Mascharak, Inorg. Chim. Acta 323
(2001) 1;
(c) N.C. Singha, D.N. Sathyanarayana, J. Mol. Struct. 403 (1997) 123;
(d) K. Gudasi, R. Vadavi, R. Shenoy, M. Patil, S.A. Patil, M. Nethaji,
Inorg. Chim. Acta 358 (2005) 3799.
[20] W.J. Geary, Coord. Chem. Rev. 7 (1971) 81.
[21] K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordina-
tion Compounds, John Wiley & Sons, 1986.
[22] (a) A.W. Jensen, B.A. O’Brien, J. Chem. Educ. 78 (2001) 954;
(b) H.-L. Liu, H. Ishida, Y. Yoshikawa, J. Coord. Chem. 58 (2005) 343;
(c) H.-L. Liu, S. Kasahara, Y. Yoshikawa, J. Coord. Chem. 58 (2005) 1249.
[23] The statistical formation ratio of the mer to the fac isomer should be 3:1
under equilibrium conditions (see Ref. [22b,c]). However, in the present
case no fac isomer was observed. In fact multiple synthetic attempts and
various crystallization solvents only afforded the mer isomer. One possi-
ble explanation could be that the steric crowding of the three un-coordi-
nated or hanging pyridine rings on one side of the molecule in fac isomer
made its isolation energetically least favourable.
(d) F.A. Chavez, C.V. Nguyen, M.M. Olmstead, P.K. Mascharak, Inorg.
Chem. 35 (1996) 6282;
(e) D.S. Marlin, M.M. Olmstead, P.K. Mascharak, Inorg. Chem. 38
(1999) 3258;
(f) F.A. Chavez, M.M. Olmstead, P.K. Mascharak, Inorg. Chem. 36
(1996) 6323.
[16] (a) J.L. Bricks, G. Reck, K. Rurack, B. Schulz, M. Spieless, Supramol.
Chem. 15 (2003) 189;
(b) A.N. Dwyer, M.C. Grossel, P.N. Horton, Supramol. Chem. 16
(2004) 405;
[24] D.S. Blanc, A.J.B. Wenger, J. Clin. Microbiol. 8 (2003) 3499.
[25] D. Saha, J. Pal, Lett. Appl. Microbiol. 34 (2002) 311.
[26] R.K. Tiwari, D. Singh, J. Singh, V. Yadav, A.K. Pathak, R. Dabur,
A.K. Chiller, R. Singh, G.L. Sharma, R. Chandra, A.K. Verma, Bioorg.
Med. Chem. Lett. 16 (2006) 413.
(c) T. Yano, R. Tanaka, T. Nishioka, I. Kinoshita, K. Isobe,
L.J. Wright, T.J. Collins, Chem. Commun. (2002) 1396;
(d) T. Kawamoto, B.S. Hammes, R. Ostrander, A.L. Rheingold,
A.S. Borovik, Inorg. Chem. 37 (1998) 3424.
[17] M. Ray, D. Ghosh, Z. Shirin, R. Mukherjee, Inorg. Chem. 36 (1997)
3568.
[27] I.M.S. Eldeen, E.E. Elgorashi, J.V. Stadan, J. Ethnopharmacol. 102
(2005) 457.