Synthesis, characterization and cytotoxicity studies of Co(III)-flavonolato complexes

Article abstract of DOI:10.1016/j.jinorgbio.2021.111382

Hypoxia activated Co(III) complexes as prodrugs may provide with a selective delivery of cytotoxic or antibacterial compounds. Whithin this field sixteen novel Co(III) ternary complexes with the general formula [Co(4N)(flav)](ClO₄)₂,

Full text of DOI:10.1016/j.jinorgbio.2021.111382

Journal of Inorganic Biochemistry 217 (2021) 111382
Contents lists available at ScienceDirect
Journal of Inorganic Biochemistry
journal homepage: www.elsevier.com/locate/jinorgbio
Synthesis, characterization and cytotoxicity studies of Co
(
III)-flavonolato complexes
a
b
a
c
b
M ´a t ´e Kozsup , XueQuan Zhou , Etelka Farkas , Attila Cs. B ´e nyei , Sylvestre Bonnet ,
d
d
a,*
Tam a´ s Patonay (the late) , Krisztina K o´ nya , P ´e ter Bugly o´
a
Department of Inorganic and Analytical Chemistry, University of Debrecen, H-4032 Debrecen, Egyetem t ´e r 1, Hungary
Leiden Institute of Chemistry, Universiteit Leiden, Einsteinweg 55, 2333 CC Leiden, the Netherlands
Department of Physical Chemistry, University of Debrecen, H-4032 Debrecen, Egyetem t ´e r 1, Hungary
Department of Organic Chemistry, University of Debrecen, H-4032 Debrecen, Egyetem t ´e r 1, Hungary
b
c
d
A R T I C L E I N F O
A B S T R A C T
Keywords:
III
Hypoxia activated Co(III) complexes as prodrugs may provide with a selective delivery of cytotoxic or anti-
bacterial compounds. Whithin this field sixteen novel Co(III) ternary complexes with the general formula [Co
Co complexes
Flavonoles
(
4N)(flav)](ClO
4
)
2
, where 4N = tris(2-aminoethyl)amine (tren) or tris(2-pyridylmethyl)amine (tpa) and flav =
Redox behavior
Anticancer
deprotonated form of differently substituted flavonols have been synthesized, characterized, and their cytotox-
icity assayed under both normoxic and hypoxic conditions. Molecular structures of two free flavonols and seven
complexes are also reported. In all the complexes the bioligands exhibited the expected (O,O) coordination mode
and the complexes showed a slightly distorted octahedral geometry. Cyclic voltammetric studies revealed that
both the substituents of the flavonoles and the type of 4N donor ligands had an impact on the reduction potential
of the complex. The ones containing tren demonstrated significantly higher stability than the tpa analogues,
making these former compounds promising candidates for the development of hypoxia-activated prodrug
complexes. Tpa complexes showed higher activity against both selected human cancer cell lines (A549, A431)
than their free ligand flavonols, indicating that the anticancer activity of the bioligand can be enhanced upon
complexation. However, slight hypoxia-selectivity was found only for a tren complex (11) with moderate
cytotoxicity.
Hypoxia activation
X-ray structures
1
. Introduction
Selective targeting is crucial in cancer therapy. The metal complexes
ligand in the tumor. Previously simple hydroxamates [3–5], an anti-
metastatic matrix metalloproteinase (MMP) inhibitor, marimastat [2],
quinolones [6] were used among others to construct mixed ligand Co
(III) complexes while light activation of Co(III)-curcumin prodrugs was
also studied in an excellent paper by the Hambley group [7].
currently used as anticancer agents are mostly Pt(II) compounds, having
serious side effects due to the lack of selectivity. These problems are
associated with the unwanted interaction of the drug molecules during
their transportation to the cancerous cells [1]. One solution to protect
the compounds from these interactions could be to chaperone them in an
inactive form until they reach the site of action, where they should
become activated [2–4]. The use of such prodrugs therefore is a strongly
appealing option to increase selectivity. As a design principle, significant
difference in kinetic lability between Co(II) and Co(III) complexes makes
this metal suitable to construct prodrugs that can be activated under
hypoxic conditions in the tumor tissues. While Co(III) complexes are
kinetically inert and do not exchange ligands, upon reduction they
become much more labile Co(II) complexes that can release a bioactive
Flavonoids belong to the group of polyphenolic compounds that are
mainly found in tea, apple, onion and fungi [8–10]. They have a wide
range of biological activities, like antibacterial, antiviral, neuro-
protective, cardioprotective, antioxidant and antitumor effects [9].
Anticancer activity of flavonoids is based on their ability to inhibit
several enzymes, like protein kinases and topoisomerases that are
necessary for the life cycles of cancer cells [11–16]. Furthermore, they
inhibit the secreation of MMPs connected to tumor metastasis [17–19].
Having these properties flavonoid type compounds can inhibit cell
growth and tumor cell invasion. Up to now four flavonoids, flavopiridol,
silibinin, quercetin and a water-soluble glycine N-monosubstituted
*
Corresponding author.
E-mail address: buglyo@science.unideb.hu (P. Bugly o´ ).
https://doi.org/10.1016/j.jinorgbio.2021.111382
Received 2 November 2020; Received in revised form 27 January 2021; Accepted 27 January 2021
Available online 3 February 2021
0
162-0134/© 2021 The Authors.
Published by Elsevier Inc.
This is an open access article under the CC BY-NC-ND license
(
http://creativecommons.org/licenses/by-nc-nd/4.0/).

M. Kozsup et al.
Journal of Inorganic Biochemistry 217 (2021) 111382
′
carbamate ester prodrug of quercetin, 3 -(N-carboxymethyl)carbamoyl-
Table 1
′
III
3
,4 ,5,7-tetrahydroxyflavone, (QC12), reached the phase of clinical in-
Numbering of the Co complexes used in this study.
vestigations as anticancer agents, but none of them have been intro-
duced into market. The main problem is their poor bioavailability
resulting from susceptibility to glucourinidation and methylation by
intestinal and liver metabolism [9]. Due to the presence of a hydroxyl
group and an oxo group in chelating position in the structure of flavonols
Complex
Abbreviation
[
Co(tren)(flav)](ClO
Co(tpa)(flav)](ClO
4
)
2
5
[
4
)
2
6
[Co(tren)(Brflav)](ClO4)2
[Co(tpa)(Brflav)](ClO
7
4
)
2
8
[
[
[
[
Co(tren)(Meflav)](ClO
Co(tpa)(Meflav)](ClO
Co(tren)(NO Fflav)](Cl)
Co(tpa)(NO Fflav)](ClO
4
)
2
9
(
a subgroup of flavonoids), they are potential chelating ligands by
4
)
2
10
11
12
13
14
15
16
17
18
19
20
coordinating to metal ions via these two oxygen donor atoms.
Complexation of flavonoids was shown to increase their pharmacoki-
netic and pharmacological properties [20,21].
2
2
2
4
)
2
[Co(tren)(NO
2
flav)](Cl)(ClO
flav)](ClO
Co(tren)(chrys)](Cl)(ClO
Co(tpa)(chrys)](ClO
Co(tren)(ClOMeflav)](ClO
4
)
[
[
[
[
Co(tpa)(NO
2
4 2
)
In the present work the synthesis and characterization of redox-
active cobalt(III) complexes of differently substituted flavonols are re-
ported. Tris(2-aminoethyl)amine (tren) or tris(2-methylpyridyl)amine
4
)
4 2
)
4
)
2
(
tpa) tripodal amines were used as tetradentate ligands to synthesize
[Co(tren)(iPrflav)](ClO4)2
[Co(tpa)(iPrflav)](ClO
[Co(tren)(nar)](Cl)(ClO
the kinetically inert cobalt(III) complexes. A series of bioactive ligands
4 2
)
′
4
)
with anticancer potential, i.e. 3-hydroxyflavone (flavH), 6-F-3 -NO
2
-3-
′
′
2
hydroxyflavone (NO FflavH), 3 -NO
2
-3-hydroxyflavone (NO
2
flavH), 4 -
′
Me-3-hydroxyflavone (MeflavH), 6-Cl-4 -OMe-3-hydroxyflavone (ClO-
hydroxyacetophenone, flavH, CoCl
ingenin, NaClO
SigmaAldrich, TCI Chemicals, Acros Organics, Scharlau or Reanal and
used as received. Tpa [22], [Co(tren)(NO ]Cl, [Co(tren)Cl ]Cl, [Co
tpa)(NO ]Cl, [Co(tpa)Cl ]Cl [23], MeflavH [24], ClOMeflavH [25],
2
∙6H
2
O, NaNO
2
, tren, chrysin, nar-
′
MeflavH), 6-Br-3-hydroxyflavone (BrflavH), 4 -iPr-3-hydroxyflavone
4
, and methanol were commercial products from Merck,
(
iPrflavH), chrysin (chrysH) and naringenin (narH), were used to com-
III
plete the coordination sphere of the Co ion. The formulae and abbre-
viations of all ligands are summarized in Fig. 1 while numbering of the
complexes in Table 1.
2
)
2
2
(
2
)
2
2
BrflavH [26], iPrflavH [27] were synthesized and purified according to
literature procedures.
2
. Experimental
.1. Materials and reagents
-nitrobenzaldehyde,
2
′
′ ′
5 -F-2 -
3
2 -hydroxyacetophenone,
H N
NH2
2
N
N
N
N
N
H N
2
tren
tpa
O
O
O
OH
O
O
OH
Cl
OH
O
HO
OMe
MeflavH
O
ClOMeflavH
chrysH
O
OH
O
OH
F
OH
HO
O
O
O
OH
NO2
NO
2
narH
NO
2flavH
NO FflavH
2
O
O
O
O
OH
Br
OH
OH
O
O
flavH
BrflavH
iPrflavH
Fig. 1. The formulae and abbreviations of the ligands used in this study.
2

M. Kozsup et al.
Journal of Inorganic Biochemistry 217 (2021) 111382
2
2
.2. Syntheses
2.2.5.1. [Co(tren)(flav)](ClO
mmol), NaOH (12.80 mg, 0.32 mmol) and [Co(tren)Cl
0.32 mmol). Brown crystalline solid, yield: 86.20 mg (42%). H NMR
tren), 3.01 (m, 3H,
tren), 3.16 (m, 3H, -CH tren), 3.52 (m,
4
)
2
(5). Using flavH (76.43 mg, 0.32
2
]Cl (100.00 mg,
′
2
.2.1. 5-NO -2 -hydroxychalcone (1)
1
3
-Nitrobenzaldehyde (1.16 g, 7.0 mmol) was dissolved in 10 ml
(400 MHz, DMSO): δ/ppm = 2.78–2.90 (m, 3H, -CH
2
′
methanol and 2 -hydroxyacetophenone (0.83 g, 6.0 mmol) was added.
Solution of 0.9 g NaOH in 0.9 ml water was added dropwise and the
reaction mixture was stirred at room temperature overnight. The
precipitated brown solid was treated with 20 ml 10% HCl and the
mixture was filtered. The collected crude product was washed with
water and recrystallized from methanol. The yellow crystalline solid was
-CH
0.6H, -CH
1.4H, -NH
4H, -NH tren), 7.59 (m, 4H, Ar
0.3H, Ar-H-isomer A) 8.40 (d, 0.7H, Ar-H-isomer B), 8.65 (m, 2H,
2
tren), 3.10 (m, 2H, -CH
2
2
2
tren-isomer B) 3.66 (m, 1.4H, -CH
2
tren-isomer A), 5.08 (m,
2
tren-isomer A) 5.22 (m, 0.6H, -NH tren-isomer B), 5.39 (m,
2
H), 7.91–8.04 (m, 2H, Ar H), 8.10 (d,
– –
2
ꢀ
1
Ar
–
21
H). IR (KBr)/cm : 3447, 1491, 1099, 755, 625. Anal Required for
27Cl CoN 11∙0.5H O: C, 38.79, H, 4.34, N, 8.62%. Found: C,
38.85, H, 4.49 N, 8.55%. MS (ESI positive ion): m/z: 221.070 ([Co(tren)
1
washed with water and dried in vacuo. Yield: 1.57 g (97%). H NMR
C
H
2
4
O
2
(
400 MHz, DMSO): δ/ppm = 7.01–7.05 (m, 2H, Ar
–
H), 7.57–7.61 (m,
H), 7.92–7.96 (d, 1H, -CH), 8.22–8.26 (d,
H, -CH), 8.27–8.36 (m, 3H, Ar H), 8.80 (m, 1H, Ar H), 12.35 (s, 1H,
1
1
H, Ar
–
H), 7.76 (t, 1H, Ar
–
(flav)]2 ), 541.095 ([Co(tren)(flav)](ClO
+ +
4
) ).
–
–
ꢀ
1
-
OH). IR (KBr)/cm : 3432, 1646, 1590, 1525, 1359, 1100. Anal
2.2.5.2. [Co(tpa)(flav)](ClO4)2 (6). Using flavH (76.43 mg, 0.32
Required for C15
H
11NO
4
: C, 66.91, H, 4.12, N, 5.20%. Found: C, 66.75,
mmol), NaOH (12.80 mg, 0.32 mmol) and [Co(tpa)Cl ]Cl (145.81 mg,
2
1
H, 4.04 N, 5.32%.
0.32 mmol). Pale brown solid, yield: 115.60 mg (46%). H NMR (400
MHz, DMSO): δ/ppm = 5.22 (d, 1.3H, -CH
2
tpa-isomer A), 5.29 (d, 0.7H,
′
′
2
.2.2. 5 -F-5-NO
2
-2 -hydroxychalcone (2)
-CH tpa-isomer B) 5.32–5.34 (m, 2H, -CH tpa), 5.58 (d, 1.3H, -CH tpa-
2
2
2
′
′
The synthesis was similar to that of 1, using 5 -F-2 -hydrox-
isomer A), 5.86 (d, 0.7H, -CH tpa-isomer B) 7.44–7.50 (m, 1H, Ar
–
H),
2
yacetophenone (0.92 g, 6.0 mmol). The product was isolated as a yellow
crystalline solid. Yield: 1.48 g (86%). 1H NMR (400 MHz, DMSO):
7.57 (m, 4H, Ar
–
H), 7.79–7.89 (m, 5H, Ar
–
H), 7.97 (d, 0.3H, Ar-H-
isomer B) 8.04–8.14 (m, 4H, Ar
–
H), 8.36–8.44 (m, 3H, Ar
–
H), 8.51
δ/ppm = 7.03–7.06 (m, 1H, Ar
–
H), 7.44–7.49 (m, 1H, Ar
H), 7.92–7.96 (d, 1H, -CH), 8.11–8.14 (m, 1H, Ar
.15–8.19 (d, 1H, -CH), 8.28–8.30 (m, 1H, Ar H), 8.35 (d, 1H, Ar
.80 (m, 1H, Ar
593, 1525, 1353, 1174. Anal Required for C15
–
H), 7.76 (t,
(d, 0.7H, Ar-H-isomer A), 9.03 (t, 1H, Ar H), 9.17 (d, 0.3H, Ar-H-
–
ꢀ 1
1
8
8
1
3
H, Ar
–
–
H),
isomer B), 9.60 (m, 0.7H, Ar-H-isomer A). IR (KBr)/cm : 3436, 1612,
–
–
H),
1491, 1436, 1094, 766, 623. Anal Required for C33H27Cl Co-
2
ꢀ
1
–
H), 12.05 (s, 1H, -OH). IR (KBr)/cm : 3434, 1649,
N O ∙0.5H O: C, 49.89, H, 3.55, N, 7.05%. Found: C, 49.89, H, 3.53 N,
4
11
2
2
+
H
10FNO : C, 62.72, H,
4
7.05%. MS (ESI positive ion): m/z: 293.073 ([Co(tpa)(flav)] ).
.51, N, 4.88%. Found: C, 62.76, H, 3.52 N, 5.02%.
2
.2.5.3. [Co(tren)(Brflav)](ClO
4
)
2
(7). Using BrflavH (49.25 mg 0.16
]Cl (50.00 mg, 0.16
mmol). Brown solid, yield: 41.15 mg (36%). H NMR (400 MHz, DMSO):
tren), 2.88–3.05 (m, 4H, -CH tren),
′
2
.2.3. 3 -NO
2
-3-hydroxyflavone (3)
mmol), NaOH (6.4 mg, 0.16 mmol) and [Co(tren)Cl
2
1
1
(1.00 g, 3.71 mmol) was dissolved in 21 ml methanol and 5 ml 8%
NaOH. H
2
O
2
(3.4 ml, 30%) was added dropwise to the well stirred re-
δ/ppm = 2.74–2.86 (m, 3H, -CH
tren), 3.50 (m, 0.6H, -CH
tren-isomer A), 5.09 (m, 1.4H, -NH
tren-isomer B), 5.40 (m, 4H, -NH
2
2
◦
action mixture for 15 min at 0 C. The mixture was stirred for 1 h and
poured into 80 g cracked ice. The pH of the mixture was adjusted to 1.0
using 20% HCl. The precipitated solid was filtered, washed with water
and recrystallized from acetic acid. The pale brown crystalline solid was
collected and dried in vacuo. Yield: 0.36 g (35%). 1H NMR (400 MHz,
3
.08–3.19 (m, 3H, -CH
m, 1.4H, -CH
m, 0.6H, -NH
2
2
tren-isomer B), 3.76
tren-isomer A), 5.24
tren), 7.54–7.68 (m,
(
2
2
(
2
2
3H, Ar–
(
H), 7.98–8.17 (m, 2H, Ar H), 8.63–8.67 (m, 3H, Ar H). IR
–
–
ꢀ 1
KBr)/cm
:
3457, 1489, 1162, 769, 625. Anal Required for
DMSO): δ/ppm = 7.49 (m, 1H, Ar
d, 1H, Ar H), 8.33 (d, 1H, Ar H), 8.61 (d, 1H, Ar
Ar H), 10.22 (s, 1H, -OH). IR (KBr)/cm : 3250, 1608, 1570, 1527,
347, 1137. Anal Required for C15 NO : C, 63.61, H, 3.20, N, 4.95%.
–
H), 7.81–7.90 (m, 3H, Ar
–
H), 8.14
C
21
H26BrCl
2
CoN 11∙0.5H O: C, 34.59, H, 3.73, N, 7.68%. Found: C,
4
O
2
(
–
–
–
H), 9.07 (m, 1H,
3
4.52, H, 3.79 N, 7.64%. MS (ESI positive ion): m/z: 260.062 ([Co(tren)
ꢀ 1
–
Brflav)]2 ).
+
(
1
H
9
5
Found: C, 63.49, H, 3.24 N, 4.88%.
2
.2.5.4. [Co(tpa)(Brflav)](ClO
mmol), NaOH (5.6 mg, 0.14 mmol) and [Co(tpa)Cl
mmol). Light brown solid, yield: 53.17 mg (44%). 1H NMR (400 MHz,
DMSO): δ/ppm = 5.19 (d, 0.7H, -CH tpa-isomer B), 5.21 (d, 1.3H, -CH
tpa-isomer A), 5.32–5.33 (m, 2H, -CH tpa), 5.59 (d, 0.7H, -CH tpa-
isomer B), 5.95 (d, 1.3H, -CH tpa-isomer A), 7.45 (m, 1H, Ar H),
.54–7.60 (m, 3H, Ar H), 7.72–7.89 (m, 5H, Ar H), 7.95–8.19 (m, 5H,
Ar H), 8.65 (d, 0.7H, Ar-H-isomer A), 9.00
m, 1H, Ar
4
)
2
(8). Using BrflavH (45.50 mg, 0.14
2
]Cl (65.51 mg, 0.14
′
2
.2.4. 6-F-3 -NO
2
-3-hydroxyflavone (4)
The synthesis was similar to that of 3, using 2 (1.20 g, 4.2 mmol). The
product was isolated as a pale brown crystalline solid. Yield: 0.41 g
2
2
2
2
1
(
32%). H NMR (400 MHz, DMSO): δ/ppm = 7.72–7.80 (m, 2H, Ar
–
H),
–
2
7
8
.88 (t, 1H, Ar
.62 (d, 1H, Ar
–
H), 7.92–7.96 (m, 1H, Ar
H), 8.35 (m, 1H, Ar
– –
H),
7
–
–
–
H), 9.07 (m, 1H, Ar
–
H), 10.35 (s, 1H, -OH). IR (KBr)/
–
H), 8.43–8.49 (m, 3H, Ar
–
ꢀ
1
cm : 3343, 1626, 1581, 1522, 1349, 1083. Anal Required for
FNO ∙0.6H O: C, 57.74, H, 2.97, N, 4.49%. Found: C, 57.46, H,
.89 N, 4.30%.
(
–
H) 9.14 (d, 0.7H, Ar-H-isomer A) 9.25 (d, 0.3H, Ar-H-
C
15
H
8
5
2
ꢀ 1
isomer B) 9.64 (d, 0.3H, Ar-H-isomer B). IR (KBr)/cm : 3425, 1493,
2
1
2 4 2
442, 1094, 769, 624. Anal Required for C33H26BrCl CoN O11∙1.2H O:
C, 44.74, H, 3.23, N, 6.32%. Found: C, 44.79, H, 3.15 N, 6.24%. MS (ESI
2
.2.5. General procedure for the synthesis of the Co(III)-flavonolato/4N
2+
positive ion): m/z: 332.026 ([Co(tpa)(Brflav)] ).
donor ligand ternary complexes
CAUTION: Although we never have experienced any problems,
perchlorate salts are potentially explosive.
2
.2.5.5. [Co(tren)(Meflav)](ClO
4
)
2
(9). Using MeflavH (50.00 mg 0.20
]Cl (61.78 mg, 0.20
mmol). Brown solid, yield: 64.32 mg (49%). H NMR (400 MHz, DMSO):
δ/ppm = 2.43 (m, 3H, -CH ), 2.76–2.92 (m, 3H, -CH tren), 2.99–3.02
m, 3H, -CH
tren), 3.06–3.11 (m, 2H, -CH
CH tren), 3.50 (m, 0.7H, -CH
isomer A), 5.09 (m, 1.3H, -NH
isomer B), 5.35–5.49 (m, 4H, -NH
.47 (d, 0.7H, Ar-H-isomer B), 7.60 (m, 1H, Ar
Ar H), 8.10 (m, 0.3H, Ar-H-isomer B), 8.38
H), 7.98–8.03 (m, 1H, Ar
mmol), NaOH (8.0 mg, 0.20 mmol) and [Co(tren)Cl
2
The flavonol was dissolved in 15 ml methanol and NaOH (1 equiv.)
1
was added. [Co(tren)Cl
2
]Cl or [Co(tpa)Cl
2
]Cl (1 equiv.) was added and
3
2
◦
the deep brownish reaction mixture was stirred at 60 C overnight. After
cooling, the mixture was filtered, using a cotton wool, and the solution
was brought to dryness in rotavap. The residue was dissolved in 5 mL
(
2
2
tren), 3.14–3.18 (m, 2H,
-
2
2
tren-isomer B), 3.64 (m, 1.3H, -CH
tren-isomer A), 5.23 (m, 0.7H, -NH
2
tren-
tren-
2
2
4
water and filtered. Solid NaClO (2 equiv.) was added to the solution. On
2
tren), 7.40 (d, 1.3H, Ar-H-isomer A),
H), 7.92 (m, 1H,
standing at room temperature crystalline solid appeared after few days.
The solid was filtered, washed with water and dried in vacuo. In some
cases the crude products were recrystallized from water or methanol.
7
–
–
–
3

M. Kozsup et al.
Journal of Inorganic Biochemistry 217 (2021) 111382
ꢀ
(
m, 0.7H, Ar-H-isomer A), 8.56 (m, 2H, Ar
–
H). IR (KBr)/cm 1: 3443,
611, 1537, 1491, 1430, 1104, 753, 625. Anal Required for
29Cl CoN 11: C, 40.26, H, 4.61, N, 8.54%. Found: C, 40.09, H,
2.2.5.10. [Co(tpa)(NO
0.17 mmol), NaOH (6.8 mg, 0.17 mmol) and [Co(tpa)Cl
0.17 mmol). Brown solid, yield: 59.14 mg (42%). 1H NMR (400 MHz,
DMSO): δ/ppm = 5.27–5.39 (m, 4H, -CH tpa), 5.54 (d, 0.6H, -CH tpa-
isomer B), 5.88 (d, 1.4H, -CH tpa-isomer A), 7.47 (m, 1H, Ar H),
H), 7.78–7.97 (m, 5H, Ar H), 8.05–8.15 (m, 5H,
H), 8.35–8.44 (m, 2H, Ar), 8.55 (d, 1H, Ar H), 8.59 (m, 0.6H, Ar-
2 4 2 2
flav)](ClO ) (14). Using NO
flavH (48.15 mg
1
2
]Cl (77.46 mg,
C
22
H
2
4
O
4
.53 N, 8.45%. MS (ESI positive ion): m/z: 228.579 ([Co(tren)
2
2
–
Meflav)]2 ).
+
(
2
7
.54–7.59 (m, 2H, Ar
–
–
–
Ar
–
2
0
0
.2.5.6. [Co(tpa)(Meflav)](ClO
4 2
)
(10). Using MeflavH (43.26 mg,
]Cl (78.06 mg,
.17 mmol), NaOH (6.8 mg, 0.17 mmol) and [Co(tpa)Cl
2
H-isomer A), 8.72 (m, 0.4H, Ar-H-isomer B) 9.06 (d, 0.4H, Ar-H-isomer
B), 9.15 (d, 0.6H, Ar-H-isomer A), 9.29 (d, 0.6H, Ar-H-isomer A), 9.39 (s,
0.4H, Ar-H-isomer B), 9.59 (d, 0.4H, Ar-H-isomer B), 10.25 (s, 0.6H, Ar-
1
.17 mmol). Light brown solid, yield: 60.05 mg (44%). H NMR (400
-isomer B), 2.56 (s, 2H, -CH
), 5.57 (d, 0.6H, -CH tpa-isomer B),
tpa-isomer A), 7.39–7.60 (m, 4H, Ar H), 7.68 (d,
H), 7.79–7.90 (m, 4H, Ar H), 7.98 (m, 0.7H, Ar-H-isomer A),
.02–8.15 (m, 4H, Ar H), 8.34 (d, 1H, Ar H), 8.39 (d, 0.6H, Ar-H-
H) 9.00 (m,
.3H, Ar-H-isomer B), 9.15 (d, 0.7H, Ar-H-isomer A), 9.58 (d, 0.3H, Ar-
MHz, DMSO): δ/ppm = 2.39 (s, 1H, -CH
3
3
-
ꢀ 1
isomer A), 5.20–5.34 (m, 4H, -CH
2
2
H-isomer A). IR (KBr)/cm : 3436, 1541, 1495, 1349, 1093, 768, 624.
Anal Required for C33 26Cl CoN 13: C, 47.73, H, 3.16, N, 8.43%.
5
1
8
.86 (d, 1.4H, -CH
2
–
H
2
5
O
H, Ar
–
–
Found: C, 47.41, H, 3.05 N, 8.29%. MS (ESI positive ion): m/z: 315.564
+
flav) + OMe]+).
–
–
([Co(tpa)(NO
2
flav)]2 ), 662.145 ([Co(tpa)(NO
2
isomer B), 8.48 (d, 1.4H, Ar-H-isomer A) 8.91 (d, 1H, Ar
–
0
2.2.5.11. [Co(tren)(chrys)](Cl)(ClO
0.32 mmol), NaOH (12.8 mg, 0.32 mmol) and [Co(tren)Cl
0.32 mmol). Light brown solid, yield: 72.17 mg (38%). 1H NMR (400
4
) (15). Using chrysH (81.36 mg
ꢀ 1
H-isomer B). IR (KBr)/cm : 3431, 1610, 1492, 1431, 1093, 770, 624.
Anal Required for C34 29Cl CoN 11∙0.8H O: C, 50.11, H, 3.91, N,
.88%. Found: C, 50.09, H, 3.72 N, 6.84%. MS (ESI positive ion): m/z:
2
]Cl (100 mg,
H
2
4
O
2
6
3
MHz, DMSO): δ/ppm = 2.75–2.77 (m, 4H, -CH
tren), 3.18 (m, 3H, -CH tren), 3.48 (m, 2H, -CH
NH tren), 5.50 (m, 2H, -NH tren), 5.70 (m, 2H, -NH
2H, Ar–H), 7.50 (s, 1H, -CH), 7.63 (m, 3H, Ar
2
tren), 2.87 (m, 3H, -CH
2
00.079 ([Co(tpa)(Meflav)]2 ).
+
2
2
tren), 5.31 (m, 2H,
-
2
2
2
tren), 6.34 (m,
–
H), 8.10 (m, 2H, Ar
IR (KBr)/cm : 3420, 1634, 1531, 1430, 1169, 1107. Anal Required for
27Cl CoN ∙1H O: C, 41.26, H, 4.78, N, 9.17%. Found: C, 41.13,
H, 4.78 N, 9.22%. MS (ESI positive ion): m/z: 229.069 ([Co(tren)
–
H).
2
0
0
.2.5.7. [Co(tren)(NO
.17 mmol), NaOH (6.8 mg, 0.17 mmol) and [Co(tren)Cl
.17 mmol). The solid was obtained without adding the perchlorate
2
Fflav)](Cl)
2
(11). Using NO
2
FflavH (50.00 mg
ꢀ 1
2
]Cl (53.00 mg,
C
21
H
2
4
O
8
2
counter ion, isolated as a reddish crystalline solid. The pure product was
obtained after recrystallization from MeOH. Yield: 37 mg (38%). 1
NMR (400 MHz, DMSO): δ/ppm = 2.87 (m, 3H, -CH tren), 2.94 (m, 2H,
CH tren), 3.56. (m, 3H, -CH tren), 5.53
tren), 6.05 (m, 2H, -NH tren), 6.29 (m, 2H, -NH tren),
.81 (m, 1H, Ar H), 8.19 (m, 1H, Ar H),
H), 7.89–7.98 (m, 2H, Ar
.38 (d, 1H, Ar H), 8.86 (d, 1H, Ar H), 9.76 (s, 1H, Ar H). IR (KBr)/
25Cl CoF-
∙MeOH: C, 43.44, H, 4.81, N, 11.51%. Found: C, 43.12, H, 4.73 N,
2+ +
H
(chrys)] ), 457.129 ([Co(tren)(chrys)-H] ).
2
-
2
tren), 3.04–3.16 (m, 4H, -CH
m, 2H, -NH
2
2
2.2.5.12. [Co(tpa)(chrys)](ClO
mmol), NaOH (12.8 mg, 0.32 mmol) and [Co(tpa)Cl
0.32 mmol). Brown solid, yield: 79.9 mg (28%). 1H NMR (400 MHz,
DMSO): δ/ppm = 4.97 (d, 1H, -CH tpa-isomer A), 5.12 (d, 1H, -CH tpa-
isomer B), 5.30 (d, 2H, -CH tpa), 5.59 (d, 1H, Ar-H-isomer A), 5.70 (d,
1H, Ar-H-isomer B), 6.43 (d, 1H, Ar H), 6.56 (s, 0.5H, Ar-H-isomer A),
7.27 (s, 0.5H, Ar-H-isomer B), 7.41–7.46 (m, 1H, Ar H), 7.51–7.62 (m,
3H, Ar H), 7.92 (d, 1H, Ar H), 8.00–8.06
H), 7.72–7.84 (m, 4H, Ar
m, 2H, Ar H), 8.11 (m, 2H, Ar H), 8.23 (m, 2H, Ar H), 8.28 (m, 2H,
4
)
2
(16). Using chrysH (81.36 mg 0.32
(
2
2
2
2
]Cl (145.81 mg,
7
8
–
–
–
–
–
–
2
2
ꢀ
1
cm : 3430, 1505, 1346, 1242, 797. Anal Required for C21
H
2
2
N
5
O
5
–
+
1
1.31%. MS (ESI positive ion): m/z: 252.559 ([Co(tren)(NO
2
Fflav)]2 ),
–
+
5
04.110 ([Co(tren)(NO
2
Fflav)-H] ).
–
–
–
(
–
–
–
Ar–H), 9.26 (d, 0.5H, Ar-H-isomer A), 9.44 (d, 0.5H, Ar-H-isomer B),
11.05 (s, 0.5H, -OH-isomer A), 11.38 (s, 0.5H, -OH-isomer B). IR (KBr)/
2
.2.5.8. [Co(tpa)(NO
2
Fflav)](ClO
4
)
2
(12). Using NO
2
FflavH (50.00
]Cl (75.64
mg, 0.17 mmol). Light brown solid, yield: 67.08 mg (46%). 1H NMR
tpa), 5.53 (d, 0.7H,
tpa-isomer A), 7.47 (m, 1H,
H), 7.77–7.90 (m, 4H, Ar H), 8.01–8.29 (m,
H), 8.36 (m, 0.4H, Ar-H-isomer B), 8.44 (m, 0.6H, Ar-H-isomer
A), 8.54 (d, 1H, Ar H), 8.59 (m, 0.6H, Ar-H-isomer A) 8.73 (d, 0.4H, Ar-
mg, 0.17 mmol), NaOH (6.8 mg, 0.17 mmol) and [Co(tpa)Cl
2
ꢀ 1
cm
:
3434, 1633, 1525, 1101, 771, 623. Anal Required for
(
400 MHz, DMSO): δ/ppm = 5.38–5.43 (m, 4H, -CH
CH tpa-isomer B), 5.92 (d, 1.3H, -CH
H), 7.57 (m, 2H, Ar
–
2
C
33
H
27Cl CoN 12∙1H O: C, 48.37, H, 3.57, N, 6.84%. Found: C, 48.09,
2
4
O
2
-
2
2
H, 3.61 N, 7.09%. MS (ESI positive ion): m/z: 301.069 ([Co(tpa)
+
+
Ar
–
H, Ar
–
(chrys)]2 ), 601.129 ([Co(tpa)(chrys)-H] ), 701.085 ([Co(tpa)(chrys)
+
6
–
+ ClO
4
] ).
–
H-isomer B), 8.88 (m, 0.4H, Ar-H-isomer B), 9.17 (d, 0.6H, Ar-H-isomer
2.2.5.13. [Co(tren)(ClOMeflav)](ClO
4
)
2
(17). Using
ClOMeflavH
]Cl
(51.44 mg, 0.17 mmol). Brown solid, yield: 67.65 mg (56%). H NMR
(400 MHz, DMSO): δ/ppm = 2.76–2.85 (m, 2H, -CH tren), 2.88–3.03
(m, 4H, -CH tren), 3.14–3.17 (m, 2H,
tren), 3.07–3.12 (m, 2H, -CH
-CH tren), 3.51 (m, 0.7H, -CH tren-isomer B), 3.73 (m, 1.3H, -CH tren-
isomer A), 3.89 (m, 3H, -CH ), 5.07 (m, 1.3H, NH tren-isomer A), 5.21
m, 0.7H, NH tren), 7.14 (d,
tren-isomer B), 5.36–5.43 (m, 4H, NH
A), 9.28 (d, 0.6H, Ar-H-isomer A), 9.37 (s, 0.4H, Ar-H-isomer B), 9.65 (d,
(53.00 mg 0.17 mmol), NaOH (6.8 mg, 0.17 mmol) and [Co(tren)Cl
2
ꢀ 1
1
0
.4H, Ar-H-isomer B), 10.23 (s, 0.6H, Ar-H-isomer A). IR (KBr)/cm
432, 1501, 1448, 1347, 1094, 624. Anal Required for C33 25Cl CoF-
13∙H O: C, 45.75, H, 3.14, N, 8.08%. Found: C, 45.63, H, 3.12 N,
:
3
H
2
2
N
5
O
2
2
2
7
.96%. MS (ESI positive ion): m/z: 680.134 ([Co(tpa)(NO
2
Fflav) +
2
2
2
+
OMe] ).
3
2
(
2
2
2
.2.5.9. [Co(tren)(NO
2
flav)](Cl)(ClO
4
) (13). Using NO
2
flavH (48.15
]Cl (53.00
1.3H, Ar-H-isomer A), 7.23 (d, 0.7H, Ar-H-isomer B), 7.91 (m, 1H,
Ar–H), 7.99 (d, 0.4H, Ar-H-isomer B), 8.04 (d, 0.6H, Ar-H-isomer A),
mg 0.17 mmol), NaOH (6.8 mg, 0.17 mmol) and [Co(tren)Cl
2
mg, 0.17 mmol). The pure product was obtained after recrystallization
8.10 (d, 0.4H, Ar-H-isomer B), 8.47 (d, 0.6H, Ar-H-isomer A), 8.64–8.67
1
ꢀ 1
(m, 2H, Ar
–
H). IR (KBr)/cm : 3463, 3259, 1603, 1492, 1433, 1108,
from MeOH. Orange solid, yield: 68.15 mg (52%). H NMR (400 MHz,
DMSO): δ/ppm = 2.85 (m, 3H, -CH
2
tren), 2.94 (m, 2H, -CH
tren), 3.55. (m, 3H, -CH
tren), 5.49–5.57 (m,
tren), 5.82 (m, 2H, -NH tren), 7.63 (t, 1H, Ar
H), 7.94–8.01
H), 8.08 (d, 1H, Ar H), 8.13 (d, 1H, Ar H), 8.38 (d, 1H,
H), 8.84 (d, 1H, Ar H), 9.77 (s, 1H, Ar H). IR (KBr)/cm : 3439,
542, 1496, 1347, 1098, 624. Anal Required for C21 26Cl CoN ∙-
MeOH: C, 40.38, H, 4.62, N, 10.70%. Found: C, 40.06, H, 4.47, N,
2
tren),
3 4 2
626. Anal Required for C22H28Cl CoN O12∙H O: C, 36.51, H, 4.18, N,
3
4
.08–3.15 (m, 4H, -CH
H, -NH
2
2
7.74%. Found: C, 36.28, H, 3.92, N, 7.52%. MS (ESI positive ion): m/z:
2
+
–
253.057 ([Co(tren)(ClOMeflav)] ), 505.107 ([Co(tren)(ClOMeflav)-
2
2
+
+
(
m, 2H, Ar
–
–
–
–
H] ), 605.063 ([Co(tren)(ClOMeflav) + ClO
4
] ).
ꢀ
1
Ar
–
–
1
H
2
5
O
9
2.2.5.14. [Co(tren)(iPrflav)](ClO
0.18 mmol), NaOH (7.2 mg, 0.18 mmol) and [Co(tren)Cl
0.18 mmol). Brown solid, yield: 55.23 mg (45%). 1H NMR (400 MHz,
DMSO): δ/ppm = 1.27 (d, 6H, -CH ), 2.78 (m, 3H, -CH tren), 2.94–3.03
4
)
2
(18). Using iPrflavH (50.00 mg
2
]Cl (55.54 mg,
+
1
0.42%. MS (ESI positive ion): m/z: 243.564 ([Co(tren)(NO
2
flav)]2 ).
3
2
4

M. Kozsup et al.
Journal of Inorganic Biochemistry 217 (2021) 111382
2
(
m, 5H, -CH
tren), 5.39 (m, 2H, -NH
NH tren), 7.45 (d, 2H, Ar
.97 (d, 1H, Ar H), 8.36 (d, 1H, Ar
cm : 3512, 1615, 1603, 1493, 1090, 764, 623. Anal Required for
33Cl CoN 11∙1.5H O: C, 40.58, H, 5.11, N, 7.89%. Found: C,
0.69, H, 5.08 N, 8.03%. MS (ESI positive ion): m/z: 242.094 ([Co(tren)
2
tren, -CH), 3.16 (m, 3H, -CH
tren), 5.53 (m, 2H, -NH
H), 7.59 (t, 1H, Ar H), 7.91 (t, 1H, Ar
H), 8.57 (d, 2H, Ar H). IR (KBr)/
2
tren), 3.63 (m, 2H, -CH
tren), 5.89 (m, 2H,
H),
2
and refined on F using SHELXL [28] program incorporated into WINGX
[29] suite. Tables were extracted from the CIF files with PublCIF pro-
gram [30] and PLATON [31] was used for the crystallographic calcu-
lations. Details of data collection and refinement with the data of the
crystals were given in Supplementary Table S1. Deposition numbers for
structures published in this paper are CCDC 2,041,489–2,041,497 for
2
2
-
2
–
–
–
7
–
–
–
ꢀ
1
C
24
H
2
4
O
2
4
2
complexes 5, 8, 9, 11, 13, 15, 18 as well as for NO flavH (3) and
iPrflav)]2 ), 483.180 ([Co(tren)(iPrflav)-H] ).
+
+
MeflavH, respectively.
(
2
0
0
.2.5.15. [Co(tpa)(iPrflav)](ClO
.18 mmol), NaOH (7.2 mg, 0.18 mmol) and [Co(tpa)Cl
.18 mmol). The pure product was obtained after recrystallization from
4
)
2
(19). Using iPrflavH (50.00 mg
2.5. Cyclic voltammetric (CV) studies
2
]Cl (55.54 mg,
Cyclic voltammetric measurements were performed within the
voltage range 500 to ꢀ 1200 mV, at room temperature in water:meth-
anol = 1:1, using a BASi Epsilon Eclipse instrument equipped with a
1
MeOH. Brown solid, yield: 34.15 mg (22%). H NMR (400 MHz, DMSO):
δ/ppm = 1.23 (d, 2H, -CH -isomer B), 1.39 (d, 4H, -CH -isomer A), 2.96
m, 0.3H, -CH-isomer B), 3.14 (m, 0.7H, -CH-isomer A), 5.20–5.29 (m,
3
3
(
3
three-electrode system, that consists of Ag/AgCl/3 M KNO reference
2
H, -CH
2
), 5.32 (s, 0.6H, -CH
2
-isomer B), 5.35 (s, 1.4H, -CH
2
-isomer A),
-isomer A), 7.79–7.89
H),
H), 8.34 (d, 0.7H, Ar-H-isomer A), 8.38–8.40 (m, 1H,
H), 8.49 (d, 2H, Ar H), 8.99 (d, 1H, Ar H), 9.03 (d, 0.3H, Ar-H-
isomer B), 9.18 (d, 0.7H, Ar-H-isomer A), 9.60 (d, 0.3H, Ar-H-isomer
electrode (E1/2 = +209 mV vs. NHE), a platinum wire auxiliary elec-
5
.58 (d, 0.6H, -CH
2
-isomer B), 5.85 (d, 1.4H, -CH
2
trode (ALS Co. Japan), and a glassy carbon (CHI104) working electrode.
(
m, 4H, Ar
–
H), 7.96–7.98 (m, 1H, Ar
–
H), 8.04–8.08 (m, 2H, Ar
–
3 6
Aqueous solution of K [Fe(CN) ] was used to calibrate the system (E1/2
8
.12 (t, 2H, Ar
–
–
= +458 mV vs. NHE in 0.50 M KCl) [32]. The samples were degassed
Ar
–
–
before the measurements using argon gas. The concentration of the
ꢀ
1
samples was 1 mM and the potentional sweep rates were 200 mVs
ꢀ 1
B). IR (KBr)/cm : 3435, 1492, 1431, 1094, 765, 623. Anal Required
for C36 33Cl CoN 11: C, 52.25, H, 4.02, N, 6.77%. Found: C, 52.58, H,
.15, N, 6.69%. MS (ESI positive ion): m/z: 314.096 ([Co(tpa)
during the determination of the redox potentials. 0.20 M KNO
as supporting electrolyte.
3
was used
H
2
4
O
4
iPrflav)]2 ).
+
(
2.6. Biological studies
2
.2.5.16. [Co(tren)(nar)](Cl)(ClO
4
) (20). Using narH (87.12 mg 0.32
]Cl (100.00 mg,
.32 mmol). The pure product was obtained after recrystallization from
2.6.1. Cytotoxicity studies
mmol), NaOH (12.8 mg, 0.32 mmol) and [Co(tren)Cl
2
Cell Culturing. Cells were thawed and at least passaged twice under
normoxia (21% O ) before starting cytotoxicity experiments. Each cell
0
2
1
MeOH. Brown solid, yield: 89.19 mg (46%). H NMR (400 MHz, DMSO):
δ/ppm = 2.70–2.89 (m, 4H, -CH ), 2.91–2.99 (m, 2H, -CH ), 3.00–3.15
m, 3H, -CH ) 2.78 (m, 3H, -CH tren),
), 3.18–3.37 (m, 5H, -CH
.22–5.55 (m, 6.6H, -NH tren, -CH-isomer A), 5.75 (d, 0.6H, Ar-H-
line was cultured in Dulbecco’s Modified Eagle Medium with phenol red,
supplemented with 8.0% v/v fetal calf serum (FCS), 0.2% v/v penicillin/
streptomycin (P/S), and 0.9% v/v Glutamine-S (GM) (DMEM complete).
2
2
(
2
2
2
2
2
5
2
Cells were cultured in either 25 cm or 75 cm flasks and split at 70–80%
2
isomer A), 5.78 (d, 0.4H, Ar-H-isomer B), 5.98 (m, 0.4H, -CH-isomer
B), 6.05 (d, 0.6H, Ar-H-isomer A), 6.15 (d, 0.4H, Ar-H-isomer B),
confluence (three times per week for 25 cm flasks, once per week for 75
2
cm flasks). The normoxic cells were incubated in the dark at 21% O
2
,
ꢀ
1
◦
6
.88–6.95 (m, 2H, Ar
–
H), 7.37–7.42 (m, 2H, Ar
–
H). IR (KBr)/cm
29Cl CoN
:
37 C and 7.0% CO
2
. The hypoxic cells were incubated in the dark using
concentration which was lower (1%,
incubator, MCO-170M). The medium was
3
421, 1610, 1100, 836, 624. Anal Required for C21
H
2
4
O
9
∙-
similar conditions except the O
using a PHCbi O /CO
2
MeOH: C, 41.07, H, 5.17, N, 8.71%. Found: C, 40.81, H, 4.99, N, 8.54%.
2
2
2
+
MS (ESI positive ion): m/z: 238.075 ([Co(tren)(nar)] ), 475.141 ([Co
refreshed three times a week. Cells used in all biological experiments
were cultured for a maximum of eight weeks.
+
+
(
tren)(nar)-H] ), 575.098 ([Co(tren)(nar) + ClO
4
] ).
Cytotoxicity assay. Human cancer cell lines (A431, human epider-
moid carcinoma; A549, human lung carcinoma) were distributed by the
European Collection of Cell Cultures (ECACC), and purchased through
Sigma Aldrich. Dulbecco’s Minimal Essential Medium (DMEM, with and
without phenol red, without glutamine), 200 mM Glutamine-S (GM),
trichloroacetic acid (TCA), glacial acetic acid, sulforhodamine B (SRB),
tris(hydroxylmethyl)aminomethane (tris base), 4,5,6,7-tetrachloro-
2
.3. NMR, IR and Electrospray Ionization Mass Spectrometric (ESI-MS)
measurements
NMR measurements were carried out using a Bruker Avance 400
6
NMR spectrometer at room temperature on samples prepared in d -
DMSO. Calibration was performed using the residual solvent peak of
DMSO (2.50 ppm). IR spectra as KBr pellets were recorded on a Perkin
Elmer FTIR Paragon 1000 PC instrument at the Department of Organic
Chemistry, University of Debrecen. ESI-TOF MS measurements in the
positive mode were carried out on a Bruker MaXis II. uhr ESI-TOF MS
instrument at the Department of Inorganic and Analytical Chemistry,
′
′ ′ ′
2
,4 ,5 ,7 -tetraiodo-fluorescein disodium salt (rose bengal), and cis-
diamineplatinum(II) dichloride (cisplatin), were purchased from
Sigma Aldrich. (2R,3R,4R,5R)-hexan-1,2,3,4,5,6-hexol (D-mannitol) was
purchased from Santa Cruz Biotechnology via Bio-Connect. Fetal calf
serum (FCS) was purchased from Hyclone. Penicillin and streptomycin
were purchased from Duchefa and were diluted to a 100 mg/ml peni-
cillin/streptomycin solution (P/S). Trypsin and Opti-MEM® (without
phenol red) were purchased from Gibco® Life Technologies. Trypan
blue (0.4% in 0.81% sodium chloride and 0.06% potassium phosphate
dibasic solution) was purchased from BioRad. Plastic disposable flasks
and 96-well plates were from Sarstedt. Cells were counted using a Bio-
Rad TC10 automated cell counter with BioRad Cell Counting Slides.
UV–vis measurements for analysis of 96-well plates were performed on a
M1000 Tecan Reader. Cells were inspected with an Olympus IX81
microscope.
University of Debrecen. The concentration of the samples was 10 g/ml
μ
and the solvent was methanol. The instrument was equipped with an
electrospray ion source, where the voltage was 4.5 kV. The drying gas
was N
2
. The flow rate was 4 l/min and the drying temperature was
◦
2
00 C. Na-formate was injected after each separation enabling internal
m/z calibration. The spectra were evaluated with the Bruker Compass
Data Analysis 4.4. software.
2
.4. Crystal structure analysis
The cytotoxicity of the compounds to two human cancer cell lines
(A431, A549) was assessed using an assay described by Hopkins et al.
[33]. In short, cell cultures with a maximum confluence of 70–80% were
trypsinized and centrifuged (1.5 min, 1.2 relative centrifugal force),
X-ray data were collected on a Bruker D8 Venture type diffractom-
eter with Mo (λ = 0.71073 Å) and Cu (λ = 1.54184 Å) radiation. Mo-
lecular structures of the compounds were solved using direct methods
5

M. Kozsup et al.
Journal of Inorganic Biochemistry 217 (2021) 111382
trypsin and DMEM complete were removed, and the cells were re-
The UV–Vis spectra were recorded in the interval 200–500 nm using
a Perkin Elmer Lambda 25 UV/Vis type spectrophotometer.
suspended using Opti-MEM supplemented with 2.4% v/v FCS, 0.2% v/
v P/S, and 1.0% v/v GM, (hereafter called Opti- MEM complete). 10
μ
L
of cell suspension and 10 L of trypan blue were mixed and pipetted into
μ
3. Results and discussion
a cell counting slide, and cells were counted using a BioRad TC10
automated cell counter. The cell suspension was diluted to the appro-
3.1. Synthesis and characterization
3
3
priate seeding density (A431, 8 × 10 ; A549, 5 × 10 cells/well) and
seeded in the wells of a 96-well plate. The aqueous cisplatin positive
control solution was prepared from a stock solution based on clinical
formulation (3.3 mM cisplatin, 55 mM mannitol, 154 mM NaCl). Ster-
ilized dimethylsulfoxide (DMSO) was used to dissolve the cobalt com-
plexes or ligands, in such amounts that the maximum % (v/v) of DMSO
per well did not exceed 0.5% (v/v).
The chalcone derivatives 1 and 2 were synthesized by Claisen
condensation of the appropriately substituated acetophenone and 3-
nitrobenzaldehyde in methanol (Scheme 1). Crude products were puri-
fied by recrystallization from methanol leading to the final compounds
with high yield. The obtained chalcones were converted into the cor-
responding flavonols 3 and 4 in Algar-Flynn-Oyamada reaction using
NaOH and 30% H O (Scheme 1). The final products were obtained after
Both for normoxic and hypoxic conditions the complete cytotoxicity
experiment lasted 96 h: cells were seeded at t = 0 h, treated at t = 24 h,
and the cells were fixed at t = 96 h. 96-well plates were seeded with the
2
2
recrystallization of the crude products from acetic acid [35–37].
Stoichiometric amount of [Co(4N)Cl ]Cl (4N = tren or tpa) precursor
2
correct amount of cells in 100
μ
L Opti-MEM complete per well. Border
and one of the different types of flavone/flavonol/flavanone ligands
◦
wells B12-G12 were seeded with cells for qualitative positive controls
were reacted in MeOH in the presence of one equivalent base at 60 C
such as cisplatin. In the remaining outer wells 200
complete was pipetted.
μ
L of Opti-MEM
overnight. After the solvent was changed to water, mostly brownish
coloured crystalline solids appeared in few days at room temperature by
ꢀ
For the cytotoxicity assay under hypoxia, the following modification
were applied following Lameijer et al. [34]: the cells were cultured at
least for two weeks under hypoxia before starting the cytotoxic assay.
For the assay itself, after seeding the cells in air, the plates were incu-
bated under hypoxic conditions for 24 h. Treatment was realized in air at
t = 24 h. The cells were put back in the hypoxic incubator and incubated
for 72 h. The cells were finally fixed at t = 96 h for the end-point SRB
assay.
adding 2 eq. bulky ClO4 anion to the reaction mixture. For 11, the
ꢀ
complex was obtained in the absence of the ClO4 ion. For 3, 5, 8, 9, 11,
13, 15, 18 and MeflavH slow evaporation of the solution at room tem-
perature afforded single crystals of the compounds suitable for X-ray
diffraction studies. All of the novel complexes are air stable crystalline
solids soluble in polar solvents, (MeOH, DMSO) and, in low concentra-
tion, in water too.
1H NMR spectra of the complexes showed the expected resonance
signals and indicated the appearance of isomers in many cases. In the
spectra of the tren complexes the signals of the –CH2 groups of tren
appear in the range of 3.6–2.6 ppm. Therefore, for the samples prepared
Both for normoxic and hypoxic cytotoxicity assay, the end-point (t =
9
6 h) sulforhodamine B
(
SRB) assay was performed following Vichai et al. In short, cells were
6
fixed at t = 96 h by adding cold trichloroacetic acid (10% (v/v) 100
μL)
in d -DMSO, signals of the –NH groups also appeared at around 5.5
2
◦
in each well. The plates were stored at 4 C for 14 h, then the TCA
medium mixture was removed, the cells were rinsed with demineralized
water three times, and air dried. Then, each well was stained with 100
ppm. For these products the assignation of the signals of the different
groups of the two ligands was simple, because of the aliphatic character
of tren, unlike the flavH ligands with mostly aromatic protons. For 5, 7,
9, 17, 20 the 1H NMR spectra of the compounds clearly indicated the
presence of two geometric isomers. As it is illustrated in Fig. 2 these are
likely the cis- and trans-isomers depending on the position of the coor-
dinated deprotonated hydroxyl group to the tertiary amine of the
tripodal tren or tpa.
μ
L SRB solution (0.6% w/v SRB in 1% v/v acetic acid) for 30 min, the
SRB solution was removed, and washed with acetic acid (1% v/v) until
no SRB came off, normally requiring 3–5 times. Once the plates were air-
dried overnight, 200
μ
L of tris base (tromethamine, 10 mM) was
pipetted to each well. To determine the cell viability the absorbance at
10 nm was measured using a M1000 Tecan Reader. To make sure all
5
The isomerisation was investigated on the signal of a –CH group at
2
the SRB was dissolved, this measurement was performed at least 30 min
after addition of tris base. The SRB absorbance data per compound per
concentration were averaged over three identical wells (technical rep-
licates, nt = 3) in Excel and made suitable for use in GraphPad Prism.
Relative cell populations were derived from the average of the untreated
around 3.5 ppm, the signals of the –NH groups both belonging to the
2
tren ligand and the signals of some aromatic or methyl protons of the
bioligand, therefore these peaks were separated well enough to identify
which isomer they belong to. Based on the integral values of these sig-
nals, the ratio of the two isomers was ~2:1 in all cases. Complete
assignation of the aromatic peaks was achievable for 9 and 17. As a
representative example, the aromatic region of the spectrum of 9 with
the assignation is presented in Fig. 3.
controls (n
t
= 6). The data from three independent biological replica-
tions were used to obtain the dose-response curves and EC50 values using
non-linear regression of Hills-slope equation with a fixed Y maximum
′
′
(
100%) and minimum (0%) relative cell population values.
In Fig. 3 3 and 5 protons have one signal in each isomers. The peak
with higher intensity at 7.40 ppm belongs to the main A isomer while the
lower intensity signal next to it, at 7.47 ppm relates to the same protons
2
.6.2. Lipophilicity measurements
′
′
Distribution coefficients (D7.40) of compounds 5, 6, 9, 10, 15, 16
in isomer B. The corresponding signals of protons 2 and 6 of the isomers
overlap at 8.5–8.6 ppm, but the difference between the intensity of the
peaks of the two isomers is still detectable. For 6 (7.60 ppm) and 7 (7.92
ppm), the multiplicity of the peaks indicates the presence of the two
isomers. Signals of 8 and 5 protons of the isomers are separated more in
the spectrum. The highest difference in chemical shifts is observed be-
tween the peaks corresponding to 5 proton of the two isomers most
likely due to the vicinity of this proton to the Co(4N) core. Assignation of
the signals to the corresponding protons is further supported by COSY
measurements (Supplementary Fig. S1). Notably, for 11, 13, 15, 18 the
isolated complexes contanied only one isomer. A representative aro-
matic region of spectrum of 18 is presented in Supplementary Fig. S2.
Assignation of the NMR signals was more difficult for the analogous
tpa complexes due to the presence of the pyridyl signals making
were determined by shake-flask method in n-octanol/buffered aqueous
◦
solution at pH = 7.40 phosphate bufferered saline (PBS) at 25.0 C. The
complexes were dissolved in n-octanol pre-saturated buffered aqueous
phase with the concentration of ~100 M. Aqueous buffered phases and
μ
water pre-saturated n-octanolic phases were mixed in 1:3 volume ratios
for 16 h. After phase separation, the UV–Vis spectrum of the complex in
the aqueous phase was compared to that of the original stock solution
and the corresponding D7.40 values were calculated according to the
following equation:
[
]
A (stock sol.)
V (aqueous phase)
D
7.40
=
ꢀ 1
x
A (aqueous phase after separation)
V (n ꢀ octanol phase)
(
1)
6

M. Kozsup et al.
Journal of Inorganic Biochemistry 217 (2021) 111382
Scheme 1. Synthesis of the chalcones and flavonols.
2+
2+
N
Co
N
N
Co
N
O
O
O
O
N
N
O
O
N
N
cis-
trans-
2
+
Fig. 2. The two isomers of the studied complexes using the example of [Co(tpa)(flav)] (6).
1
6
Fig. 3. Aromatic region of the H NMR spectrum of 9 in d -DMSO with the assignation of the signals. A and B indicate the signals belonging to the two different
isomers of the complex.
7

M. Kozsup et al.
Journal of Inorganic Biochemistry 217 (2021) 111382
1
6
Fig. 4. Aliphatic region of the H NMR spectrum of 16 and 8 in d -DMSO with the assignation of the signals of the isomers. For 16 the ratio of the two isomers is ~1:1
while for 8– 2:1.
crowded the low field part of the spectra. As it is illustrated in Fig. 4 for
6 and 8, in the aliphatic region the –CH groups of coordinating tpa
showed two sets of signals belonging to two isomers (A and B): two
doublets and one singlet. This reveals that protons of two –CH groups in
axial position in the complexes are non-equivalent and coupling of them
results in doublets in the spectrum. On the contrary, the singulet for the
third methylene group supports the equivalence of its two protons due to
equatorial position of this arm in the complexes [5,6]. For all these
1
2
1
2
complexes (6, 8, 10, 12, 14, 16, 19) H NMR spectra showed the mixture
of the two isomers. Based on the relative intensity of the –CH
2
tpa
Fig. 5. Ortep view of [Co(tren)(flav)](ClO
Cl) ∙4MeOH (11), [Co(tren)(NO flav)](ClO
4
)
2
∙H
2
O (5), [Co(tpa)(Brflav)](ClO
4
)
2
∙0.25H
(Cl)
2
O (8), [Co(tren)(Meflav)](ClO
4
)(Cl)∙H
2
O (9), [Co(tren)(NO
2
Fflav)]
(
2
2
4
)
2
∙2MeOH (13), [Co(tren)(chrys)]
2
(ClO
4
)
2
2
∙3H O (15) and [Co(tren)(iPrflav)](ClO
2
4
)(Cl)∙MeOH (18). Thermal
ellipsoids shown at 40% probability level with partial numbering scheme. The counter ions and solvent molecules are omitted for clarity.
8

M. Kozsup et al.
Journal of Inorganic Biochemistry 217 (2021) 111382
protons, the ratio of the isomers for 6, 8, 10, 12, 14, 19 was found to be
complexes. Electrochemical measurements were carried out in the +600
to ꢀ 1000 mV voltage range under the conditions detailed in the
~
2:1, the same like in the case of tren complexes (vide supra). The only
exception was 16, where the intensity of the signals of the two isomers
were almost identical indicating ~1:1 ratio of the isomers (Fig. 4).
ESI-MS analysis in the positive mode provided further proof for the
identity of the complexes. In the mass spectra of all products, the peaks,
Experimental part. All the determined cathodic (Epc) or anodic (Epa)
peak potentials measured against Ag/AgCl were converted to NHE. The
obtained Epc and Epa values of the complexes are summarized in Table 2.
Results of the electrochemical studies revealed major differences
between the tren and tpa complexes in agreement with former literature
findings [5,6,10,41]. In particular, the recorded cyclic voltammograms
of the tren complexes showed irreversible reduction while those of the
tpa analogues exhibited reversible redox processes of the complexes.
Although typically the ΔE = Epa ꢀ Epc or the E1/2 = (Epa ꢀ Epc)/2 values
are used for the comparison of reversible redox processes inasmuch as
this is not the case here for the tren complexes, Table 2 contains the Epa
and Epc values instead. The difference in reversibility indicates that upon
reduction the tren complexes dissociate practically completely due to
the low stability of the corresponding Co(II) complex. On the contrary,
the reversible reduction for the tpa complexes can be explained by the
higher thermodynamic stability of the Co(II) complex formed and
belonging to the [Co(4N)(flav)]2 or [Co(4N)(flav)-H] ions, appeared.
+
+
ꢀ
For 5, 16, 17, 20, under ESI-MS conditions, the ClO
4
counter ions did
not dissociate completely either, in agreement with former literature
findings [6]. All the mass spectra displayed the correct isotopic pattern
(
Supplementary Fig. S3).
For all the new ligands and complexes the microanalytical results
also proved their integrity and purity as the experimental data are in
good agreement with the calculated ones. For 13, 15 and 20 the CHN
data showed that the complex crystallized with mixed counter ions, one
ꢀ ꢀ
ClO and one Cl . The results also indicated that most of the complexes
4
also have water molecules in their crystal structure and this was also
detected during the X-ray diffraction studies.
Molecular structure of the novel compounds 5, 8, 9, 11, 13, 15, 18 as
well as 3 and MeflavH were investigated by single crystal X-ray structure
determination. For the complexes the determined structures are shown
in Fig. 5 while for 3 and MeflavH in Supplementary Fig. S4.
decreased rate of dissociation of the ligands, which are due the -back-
π
bonding capability of tpa. This feature of tpa also results in significantly
more positive reduction peak potential values for the tpa complexes
compared to the corresponding tren analogues; as a representative
example this is demonstrated in Fig. 6 where voltammograms registered
for 9 and 10 are presented. These results are in accordance with the
findings on the redox properties of previously reported similar type Co
complexes [6,40].
X-ray studies confirmed the expected structure and octahedral ge-
ometry of the complexes in all cases. The structures consist of one 4N
donor ligand and one flavonolato ligand coordinating to the Co(III)
centre. For all complexes the flavonols are bound to the metal ion via the
carbonyl and the hydroxyl O donor atoms. Based on the obtained mo-
lecular structures complexes 3, 5, 8, 9, 13, 15, 18 were crystallized as
the trans-isomer since the tertiary amine of the 4N ligand was observed
to be trans to the deprotonated hydroxyl group of the flavonol ligand.
Only complex 11 was found to be crystallized as the cis-isomer. It should
be emphasized, however, that the crystals were selected by chance and
in each case only one of them was analyzed. As the NMR pointed out
some of the isolated solid complexes are mixture of isomers (vide supra).
In the structure of 15 two cobalt complex molecules were found in the
The reduction potential values of the complexes are also influenced
by the different substituents of the flavonolato ligands. Voltammograms
′
′
of 3 -NO -3-hydroxyflavone, 3, and 6-F-3 -NO -3-hydroxyflavone 4,
2
2
complexes showed higher reduction potential values than the analogous
complexes of the unsubstituted flavonol indicating that the strongly
2
deactivating -NO group with electron withdrawing character makes the
complex more reducible by decreasing the electron density at the cobalt
ion. Flavonolato complexes having activating substituents with electron
donating property found to have more negative reduction potentials
than the complexes of the unsubstituted flavonol. This can be explained
by the increased electron density at the metal ion leading to a less
reducible complex. Data in Table 2 also reveal that the chrysin and
naringenin complexes exhibit the most negative Epc values among the
studied ones. This is most likely due to the size of the (O,O) chelate ring
formed with the metal ion being six with these ligands. The low Epc
values indicate that these complexes with six membered chelate rings
are harder to be reduced than the ones containing five membered
chelate ring. Moreover, both chrysin and naringenin contain activating
ꢀ
ꢀ
asymmetric unit with two Cl and two ClO
possibility that these type of complexes can be crystallized with mixed
anions. The Co O bond lengths were found to be shorter than the Co
lengths, which is in accordance with the data of similar type of Co(III)
complexes [38]. The length of the two different Co O bonds were
4
counter ions proving the
–
–
N
–
slightly different for all investigated complexes, that was obsereved in
the case of previously published Ru(II) complexes of flavonols [36,39].
The key bond length and bond angle values are shown in Supplementary
Table S2.
–
OH substituent(s) that can also cause shift of the Epc values of the
complexes to the more negative region. Registered CV curves of all
3
.2. Electrochemical studies
Cyclic voltammetry was used to explore the redox properties and to
determine the peak potential values of the synthesized Co(III)
Table 2
Cathodic (Epc) and anodic (Epa) peak potential values (mV vs. NHE) of the [Co
2
+
(
4N)(flav)] (4N = tren or tpa, see Table 1) complexes using glassy carbon
ꢀ 1
electrode, in water: MeOH 1:1, I = 0.20 M KNO
3
, at a sweep rate of 200 mVs .
Flavonolate ligand in [Co(4N)
4N = tren
4N = tpa
2
+
(
flav)]
#
E
pc
#
E
pc
E
pa
(
mV)
(mV)
(mV)
flav
5
ꢀ 206
ꢀ 285
ꢀ 187
ꢀ 194
ꢀ 175
ꢀ 331
ꢀ 264
ꢀ 291
ꢀ 532
6
51
69
57
71
80
ꢀ 3
–
197
241
185
309
252
94
Brflav
7
8
Meflav
9
10
12
14
16
–
NO
NO
2
Fflav
flav
11
13
15
17
18
20
2
chrys
Fig. 6. Cyclic voltammetry curves registered for [Co(tren)(MeFlav)](ClO
and [Co(tpa)(MeFlav)](ClO (10) in H
O:MeOH = 1:1, referenced to Ag/AgCl
electrode at a potential sweep rate of 200 mV/s, I = 0.20 M KNO and c =
.0 mM.
4 2
) (9)
ClOMeflav
iPrflav
nar
–
187
–
4
)
2
2
19
–
29
–
3
1
9

M. Kozsup et al.
Journal of Inorganic Biochemistry 217 (2021) 111382
complexes are presented in Supplementary Fig. S5.
2 2
normoxic (21% O ) and hypoxic (1% O ) conditions. The determined
Since the reduction potential range of cellular reductases is known to
fall within the range of ꢀ 200 to ꢀ 400 mV vs. NHE [6,42–46], if the Epc
values of the complexes are considered in Table 2 the tren complexes
might be suitable for selective reduction in the reductive environment of
the tumor tissues. On the contrary, the Epc values of the tpa analogues in
general might be too high for the selective reduction of the complex in
the cancer cells. In order to obtain more information regarding this
assumption 10 fold excess of ascorbic acid was added to an aqueous
solution of complex 7 and the UV–Vis spectrum in the range 350–800
nm was monitored for 48 h. No measureable change was observed
revealing no reduction of the complex under these conditions.This in-
dicates that the Epc value alone is not enough to predict the mechanism
of a reduction process occuring in a biological system.
cell growth inhibition effective concentrations (EC50 in
cated in Table 3.
μ
M) are indi-
In accordance with previous literature results none of the precursor
[Co(4N)Cl ]Cl complexes were found to be active [6,40,41]. The A549
2
cell line was significantly more sensitive to the tested compounds than
A431, with the exception of 9 for which the EC50 values were found
lower for the A431 cells than for A549. Out of the free ligands MeflavH
and chrysH showed no cytotoxicity against either cell types while flavH
was slightly active against the A549 cell line under normoxic conditions
2 2
only. Ligands NO flavH, NO FflavH and BrflavH were active only
against A549 either but their cytotoxic effects were found to be signif-
icant in this cell line. These results show the remarkable effects of the
different substituents of the biological flavonoid ligands on their cyto-
toxicity, in agreement with literature data [47].
Among the cobalt(III) compounds the tpa analogues were found to be
significantly more active comparing to the corresponding tren com-
plexes. This tendency was again in good agreement with the cytotoxicity
results obtained for the tren and tpa cobalt complexes of the 1-methyl-3-
3
.3. Biological studies
Stability of the complexes in aqueous solution was monitored with
1
the aid of H NMR before the biological studies. The measurements were
(
2-naphtyl)propane-1,3-dione ligand (naacH), investigated by the
performed in a D
2
O: DMSO = 5: 1 (v/v) solvent mixture. No changes
Hambley group [40]. Low IC50 values were found for complexes 6, 8, 12
and 14, indicating their high antitumor activity against the A549 cell
line, even under hypoxia. Since the corresponding free ligands showed
lower cytotoxicity than their cobalt complexes, it appears that the ob-
tained antitumor activity was clearly enhanced upon complexation.
However, only compound 11 showed a statistically significantly lower
EC50 value under hypoxia compared with normoxia; compound 7
showed a slightly lower EC50 value under hypoxia, but that was not
1
were observed in the H NMR spectra of the complexes after 48 h at
room temperature, indicating that they are stable under these conditions
and suitable for further biological studies. As an illustration 1H NMR
spectra of 9 recorded at 0 h and in 48 h are shown in Supplementary
Fig. S6.
The cytotoxic activity of most of the complexes and of the free
flavone ligands was tested against A549 (lung carcinoma) and A431
(
skin epidermoid carcinoma) human cancer cell lines under both
Table 3
Cytotoxicity of the compounds against A549 (lung carcinoma) and A431 (skin epidermoid carcinoma) human cancer cell lines under normoxic (21% O
1% O ) conditions determined after 72 h incubation by an end-point SRB assay. Cell-growing inhibition effective concentrations (EC50) are reported in
2
) and hypoxic
M with 95%
(
2
μ
confidence interval (CI) in
μ
М. Data is the mean over three independent experiments.
Compound
EC50
/μM
A431
A549
Normoxia
CI
Hypoxia
CI
Normoxia
CI
Hypoxia
CI
[
Co(tren)Cl
Co(tpa)Cl
2
]Cl
]Cl
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
36
>100
>100
>100
>100
[
2
chrysH
flavH
+7
–4
MeflavH
BrflavH
>100
>100
>100
>100
>100
84
+29
7.0
+1.0
ꢀ 1.0
+3
12
+7
ꢀ 4
+3
ꢀ 3
+3
ꢀ 2
ꢀ
19
NO
2
2
flavH
FflavH
64
+7
>100
>100
15
16
17
15
ꢀ
6
ꢀ 2
NO
>100
+2
ꢀ
2
[
Co(tren)(flav)](ClO
Co(tpa)(flav)](ClO
4
)
2
(5)
(6)
>100
>100
>100
>100
[
4
)
2
45
+4
47
+7
8.3
+1.2
21
+3
ꢀ 4
ꢀ
3
ꢀ 6
1.1
[
[
[
[
Co(tren)(Meflav)](ClO
)
4 2
(9)
44
17
55
13
+5
88
26
91
17
+15
ꢀ 12
+4
>100
14
>100
17
ꢀ
4
Co(tpa)(Meflav)](ClO
4
)
2
2
(10)
(7)
+3
+1
+11
ꢀ 7
ꢀ
2
ꢀ 4
ꢀ 1
Co(tren)(Brflav)](ClO
4
)
+11
+39
ꢀ 23
+2
89
+12
ꢀ 12
+1.0
ꢀ 0.8
73
+20
ꢀ 16
+3.6
ꢀ 2.9
ꢀ
9
Co(tpa)(Brflav)](ClO
4
)
2
(8)
+2
6.8
8.9
ꢀ
2
ꢀ 2
[
[
Co(tren)(chrys)](ClO
4
)
2
(15)
(16)
>100
>100
>100
>100
Co(tpa)(chrys)](ClO
4
)
2
68
+17
13
64
+15
ꢀ 12
34
+4
33
+5
ꢀ
ꢀ 3
ꢀ 4
[
[
[
Co(tren)(NO
2
Fflav)](ClO
Fflav)](ClO
flav)](ClO
4
)
2
(11)
>100
15
>100
22
72
+15
48
+8
ꢀ
12
ꢀ 7
Co(tpa)(NO
Co(tpa)(NO
2
2
4
)
2
(12)
+3
+4
5.2
6.3
1.8
+0.7
ꢀ 0.6
+1.0
ꢀ 0.9
+0.2
ꢀ 0.2
6.9
7.3
2.4
+1.2
ꢀ 1.1
+1.3
ꢀ 1.4
+0.7
ꢀ 0.6
ꢀ
2
ꢀ 3
4
)
2
(14)
14
+4
23
+5
ꢀ
3
ꢀ 4
Cisplatin
3.2
+0.7
0.6
5.6
+1.3
ꢀ 1.1
ꢀ
1
0

M. Kozsup et al.
Journal of Inorganic Biochemistry 217 (2021) 111382
statistically significantly different from the normoxia value, while all
other complexes showed either identical or slightly lower toxicity under
hypoxia, compared to normoxia.
Table 4
Partition coefficient log D7.40 values determined for a selection of cobalt(III)
◦
complexes at pH = 7.40 (PBS buffer), t = 25 C.
Flavonolate ligand in [Co(4N)(flav)]
2
+
While for the tren compounds only a mild toxicity or no toxicity at all
was found, one should note that making cytotoxic agents that do not lose
their cytotoxicity under hypoxia, is an interesting result. Indeed, many
cytotoxic agents lose at least part of their toxicity due to increased
resistance of hypoxic cells.
4N = tren
4N = tpa
#
log D7.40
#
log D7.40
flav
5
ꢀ 0.141
ꢀ 0.524
ꢀ 0.349
6
0.756
Meflav
chrys
9
10
16
0.337
15
ꢀ 0.277
In order to explore the reason of the differences between the bio-
logical activity of tren and tpa complexes the partition coefficient log
tren cobalt complexes exhibited reduction in the good potential scale,
they were not activated by hypoxia or very cytotoxic to the cancer cells,
probably due to their too high hydrophilicity that is detrimental to
cellular uptake. Complex 11 stands out as it showed a slight activation
under hypoxia. Overall, the design of cancer-selective, hypoxia-acti-
vated cobalt complexes requires a careful balance among various pa-
rameters including the redox potential of the complex and its
lipophilicity. Efforts are currently undergoing to obtain tren derivatives
with higher lipophilicities and higher cellular uptake.
D
7.40 of a selection of six cobalt complexes were measured to assess on
their lipophilicity and ability to cross biological membranes by passive
diffusion. The log D7.40 values of the studied complexes are shown in
Table 4. In all cases the log D7.40 values were higher for the tpa com-
plexes, compared to the tren analogues, indicating the more lipophilic
character of the former ones, while the latter ones were found all rela-
tively water-soluble. This finding can be explained by the three aromatic
pyridyl groups in the tpa ligand, which increase the lipopilicity of the
complexes. These results also correlate with the cytotoxicity studies, as
all complexes of the tpa series were found more cytotoxic than their tren
analogue. Therefore, the higher lipophilic character of the tpa com-
plexes is probably one reason for their higher anticancer activity.
Based on these results, we hypothesize that although the reduction
potentials of the tren series of complexes are negative enough for them
to be reduced in hypoxic cells, their anticancer activity was too low to
measure significant activation under hypoxia because of their more
hydrophilic character and low cellular uptake. Mild hypoxia activation
was observed for the tren complex 11 which bears one of the most active
Author statement
M a´ t ´e Kozsup: Synthesis, Investigation, Draft peraration; XueQuan
Zhou: Investigation; Etelka Farkas: Conceptualization, Draft prepara-
tion; Attila Cs. B ´e nyei: Investigation; Sylvestre Bonnet: Supervision,
Methology, Draft preparation; Tam a´ s Patonay (the late) and Krisztina
K o´ nya; Providing some of the ligands; P ´e ter Bugly o´ : Conceptualization,
Methodology, Supervision, Writing and finalizing the manuscript.
2
cytotoxic flavonoid ligands of the series, i.e. NO Fflav. The reduction
potential values of tpa complexes are probably too positive for triggering
activation under hypoxia, although these complexes show significant
cytotoxicity in A549 lung cancer cells due to their higher lipophilicity
and probably better cellular uptake. Overall, an ideal hypoxia-activated
cobalt(III) complex would be either a tpa derivative with a more nega-
tive Epc value, or a more hydrophobic tren complex. It should also be
noted that the high anticancer activity of most tpa-based cobalt(III)
compound can probably not be related to the release of the active
flavonol ligand, since their anticancer efficiency was found much higher
than that of the corresponding free flavonol ligands. The activated co-
balt(II) complex itself, reduced both under normoxia and hypoxia in the
tpa series, might either have off-side cellular targeted that have not been
identified yet, or generate reactive oxygen species for example via a
Fenton-like reaction.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Acknowledgements
The research was supported by the EU and co-financed by the Eu-
ropean Regional Development Fund under the project GINOP-2.3.2-15-
2
016-00008.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.jinorgbio.2021.111382.
4
. Conclusions
Nine substituted flavonols (flavH, MeflavH, BrflavH, ClOMeflavH,
References
2 2
iPrflavH, NO flavH, NO FflavH, narH and chrysH) and two tetradentate
nitrogen ligands (tren or tpa) were used to synthesize a series of sixteen
[
1] L. Amable, Pharmacol. Res. 106 (2016) 27–36.
Co(4N)(flav)]2 complexes that were studied as potential hypoxia-
+
[2] T.W. Failes, C. Cullinane, C.I. Diakos, N. Yamamoto, J.G. Lyons, T.W. Hambley,
[
Chem. Eur. J. 13 (2007) 2974–2982.
activated anticancer agents. All analytical data, together with a series
of X-ray crystal structures, are consistent with an (O,O) coordination
mode of the flavonol ligand to the metal ion and with an octahedral
geometry. Cyclic voltammetric studies of the complexes revealed that
both the 4N ligand and the substituents on the flavonol bioligand in-
fluence the redox properties of the complexes. The tren-containing
compounds were found to be reduced at significantly lower potential
than their tpa analogues, while electron-donating substituents on the
flavonol also shifted the reduction potential of the cobalt complex to the
more negative region. All prepared tpa-containing complexes were
found to have higher anticancer activity towards A549 and A431 human
cancer cell lines than the corresponding free flavonol ligands, indicating
that the anticancer efficiency of the bioligands can be enhanced upon
complexation to the metal. However, no hypoxia activation was
observed in this series of compounds because their reduction potentials
are high enough to see activation already under normoxia. Although the
[
[
3] T.W. Failes, T.W. Hambley, Dalton Trans. (2006) 1895–1901.
4] P.J. Kim, R. Hocking, J.K. Clegg, P. Turner, S.M. Neville, T.W. Hambley, Dalton
Trans. 41 (2012) 11293–11304.
5] P. Bugly o´ , I. Kacsir, M. Kozsup, I. Nagy, S. Nagy, A.C. B ´e nyei, E. Kovats, E. Farkas,
Inorg. Chim. Acta 472 (2018) 234–242.
´
[
[
6] M. Kozsup, E. Farkas, A.C. B ´e nyei, J. Kasparkova, H. Crlikova, V. Brabec, P. Bugly o´ ,
J. Inorg. Biochem. 193 (2019) 94–105.
[7] A.K. Renfrew, N.S. Bryce, T.W. Hambley, Chem. Eur. J. 21 (2015) 15224–15234.
[
8] M.G.L. Hertog, P.C.H. Hollman, M.B. Katan, J. Agric, Food Chem. 40 (1992)
2
379–2383.
[
9] M. Khater, D. Ravishankar, F. Greco, H.M. Osborn, Future Med. Chem. 11 (21)
(2019) 2845–2867.
[
[
10] D. Raffa, B. Maggio, M.V. Raimondi, F. Plescia, G. Daidone, Eur. J. Med. Chem. 142
(
2017) 213–228.
11] C. Kanadaswami, L.T. Lee, P.H. Lee, J.J. Hwang, F.C. Ke, Y.T. Huang, M.T. Lee, In
Vivo 19 (2005) 895–910.
[
[
[
12] J.J. Champoux, Annu. Rev. Biochem. 70 (2001) 369–413.
13] K. Chahar, N. Sharma, M.P. Dobhal, Y.C. Joshi, Pharmacogn. Rev. 5 (9) (2011).
14] I.K. Wang, S.Y. Lin-Shian, J.K. Lin, Eur. J. Cancer 35 (1999) 1517–1525.
[15] C. Bailly, Curr. Med. Chem. 7 (2000) 39–58.
1
1

M. Kozsup et al.
Journal of Inorganic Biochemistry 217 (2021) 111382
[
16] Sukardiman, M. Tanjung, A. Darwanto, M.O. Darmadi, Clin. Hermorheol.
Microcirc. 23 (2000) 185–190.
[35] C.J. Mills, N.N. Mateeva, K.K. Redda, J. Heterocyclic Chem. 43 (2006) 59.
[36] A. Kurzwernhart, W. Kandioller, S. B a¨ hler, C. Bartel, S. Martic, M. Buczkowska,
G. Mühlgassner, M.A. Jakupec, H.B. Kraatz, P.J. Bednarski, V.B. Arion, D. Marko,
B.K. Keppler, C.G. Hartinger, J. Med. Chem. 55 (2012) 10512–10522.
[37] C.X. Quin, X. Chen, R.A. Hughes, S.J. Williams, O.L. Woodman, J. Med. Chem. 51
(2008) 1874–1884.
[
[
[
[
17] L.M. Matrisian, Trends Genet. 6 (1990) 121–125.
18] L.A. Liotta, Med. Sci. Am. 266 (1992) 54–59 (62–63).
19] J.S. Rhee, L.M. Coussens, Trends Cell Biol. 12 (5) (1992) 209–211.
20] M. Samsonowicz, E. Regulska, M. Kalinowska, Chem. Biol. Interact. 273 (2017)
2
45–256.
[38] A.K. Renfrew, N.S. Bryce, T.W. Hambley, Chem. Sci. 4 (2013) 3731.
[39] A. Kurzwernhart, W. Kandioller, C. Bartel, S. B a¨ hler, R. Trondl, G. Mühlgassner, M.
A. Jakupec, V.B. Arion, D. Marko, B.K. Keppler, C.G. Hartinger, Chem. Commun.
48 (2012) 4839–4841.
[
[
21] M.M. Kasprzak, L. Szmigiero, E. Zyner, J. Ochocki, J. Inorg. Biochem. 105 (4)
(
2011) 518–528.
22] Z. Tyekl ´a r, R.R. Jacobson, N. Wei, N.N. Murthy, J. Zubieta, K.D. Karlin, J. Am.
Chem. Soc. 115 (1993) 2611–2689.
[40] P.D. Bonnitcha, B.J. Kim, R.K. Hocking, J.K. Clegg, P. Turner, S.M. Neville, T.
W. Hambley, Dalton Trans. 41 (2012) 11293.
[
[
23] E. Kimura, S. Young, J.P. Collman, Inorg. Chem. 9 (5) (1970) 1183–1191.
24] W. Adam, D. Golsch, L. Hadjiarapoglu, T. Patonay, Tetrahedron Lett. 32 (8) (1991)
[41] M. Kozsup, O. D o¨ m o¨ t o¨ r, S. Nagy, E. Farkas, E.A. Enyedy, P. Buglyo, J. Inorg.
Biochem. 204 (2020) 110963.
´
´
1
041–1044.
[
[
[
25] R. Khanna, R. Kumar, A. Dalal, R.C. Kamboj, J. Fluoresc. 25 (2015) 1159–1163.
26] D.A. Akrawi, T. Patonay, K. K o´ nya, P. Langer, Synlett. 24 (2013) 860–864.
27] S. Das, I. Mitra, S. Batuta, M.N. Alam, K. Roy, N.A. Begum, Bioinorg. Med. Chem.
Lett. 24 (21) (2014) 5050–5054.
[42] B.P. Green, A.K. Renfrew, A. Glenister, P. Turner, T.W. Hambley, Dalton Trans. 46
(2017) 15897–15907.
[43] M.D. Hall, T.W. Failes, N. Yamamoto, T.W. Hambley, Dalton Trans. (2007)
3983–3990.
[
[
[
[
[
28] G.M. Seldrick, Acta Cryst A64 (2008) 112–122.
[44] V. Mallikarjun, D.J. Clarke, C.J. Cambell, Free Radic. Biol. Med. 53 (2012)
280–288.
29] L.G. Farrugia, J. Appl. Crystallogr. 32 (1999) 837–838.
30] S.P. Westrip, J. Appl. Crystallogr. 43 (2010) 920–925.
[45] D.C. Ware, B.D. Palmer, W.R. Wilson, W.A. Denny, J. Med. Chem. 36 (1993)
1839–1846.
31] A.L. Spek, J. Appl. Crystallogr. 36 (2003) 7–13.
32] E. Farkas, P. Bugly o´ , E.A. Enyedy, M.A. Santos, Inorg. Chim. Acta 357 (2004)
´
[46] J. Jiang, C. Auchinvole, K. Fischer, C.J. Cambell, Nanoscale 6 (2014)
12104–12110.
2
451–2461.
[
[
33] S.L. Hopkins, B. Siewert, S.H.C. Askes, P. Veldhuizen, R. Zwier, M. Heger,
S. Bonnet, Photochem. Photobiol. Sci. 15 (2016) 644–653.
[47] A. Kurzwernhart, S. Mokesch, E. Klapproth, M.S. Adib-Ravazi, M.A. Jakupec, C.
G. Hartinger, W. Kandioller, B.K. Keppler, Eur. J. Inorg. Chem. (2016) 240–246.
34] L.N. Lameijer, D. Ernst, S.L. Hopkins, M.S. Meijer, S.H.C. Askes, S.E. Le D ´e v ´e dec,
S. Bonnet, Angew. Chem. Int. Ed. 56 (2017) 11549–11553.
1
2