Correlation between the SOD-like activity of hexacoordinate iron(II) complexes and their Fe3+/Fe2+ redox potentials

Article abstract of DOI:10.1016/j.inoche.2010.10.023

Hexacoordinate, high-spin iron(II) complexes with different isoindoline-derivative ligands have been prepared and characterized. The Fe ³⁺/Fe²⁺ redox transition is reversible in each case, but the E_1/2 values vary in a ~ 400 mV range depending on the ligand. SOD-like activity of the complexes was determined by indirect methods with cytochrome c, and nitroblue tetrazolium indicator at pH 7.6. The measured activities correlate with the redox potentials for the Fe³⁺/Fe ²⁺ couples. The results indicate that the superoxide dismutation takes place via an inner-sphere mechanism at the iron site.

Full text of DOI:10.1016/j.inoche.2010.10.023

Inorganic Chemistry Communications 14 (2011) 205–209
Contents lists available at ScienceDirect
Inorganic Chemistry Communications
journal homepage: www.elsevier.com/locate/inoche
Correlation between the SOD-like activity of hexacoordinate iron(II) complexes and
their Fe3+/Fe2+ redox potentials
Balázs Kripli a, Gábor Baráth a, É. Balogh-Hergovich a, Michel Giorgi b, A. Jalila Simaan b, László Párkányi c,
a
a
József S. Pap a, , József Kaizer , Gábor Speier
⁎
a
Department of Chemistry, University of Pannonia, 8201 Veszprém, Hungary
Aix-Marseille Université, FR1739, Spectropole, Campus St. Jérôme, Avenue Escadrille Normandie-Niemen, 13397 Marseille cedex 20, France
Chemical Research Center, Hungarian Academy of Sciences, 1525, Budapest, Hungary
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
Hexacoordinate, high-spin iron(II) complexes with different isoindoline-derivative ligands have been
prepared and characterized. The Fe3+/Fe2+ redox transition is reversible in each case, but the E1/2 values vary
in a ~400 mV range depending on the ligand. SOD-like activity of the complexes was determined by indirect
methods with cytochrome c, and nitroblue tetrazolium indicator at pH 7.6. The measured activities correlate
with the redox potentials for the Fe3+/Fe2+ couples. The results indicate that the superoxide dismutation
takes place via an inner-sphere mechanism at the iron site.
Received 31 May 2010
Accepted 22 October 2010
Available online 30 October 2010
Keywords:
Iron
SOD mimics
Isoindoline ligand
Redox chemistry
© 2010 Elsevier B.V. All rights reserved.
Superoxide dismutase (SOD) enzymes [1] act as a primary defense
system against oxidative damage in living species by accelerating the
dismutation of superoxide during the catalytic cycle shown in Eq. (1).
iminoisoindoline-derivative 3N-donor ligands (Scheme 1) and their
SOD-like activity with respect to their redox chemistry. The ligands in
Scheme 1 were chosen because of their strong chelating ability (see
discussion later) and uncomplicated synthesis. To be noted, that these
isoindoline-derivatives show eye-catching similarities with the very
popular porphyrins, too: the 3N-donor set is planar, resembling the 3/
4 part of a full porphyrin ring, they exhibit informative electronic
spectra that are rich in bands, and the extended π-delocalisation
within the molecule makes them relatively resistant to chemical
changes.
The known 1,3-bis(2′-pyridylimino)isoindoline [8] (HL2,
Scheme 1) has been widely used in enzyme models with copper [9],
manganese [10], cobalt [11] and iron [12]. We synthesized derivatives
with various side arms on the imine moiety (HL1, HL3–HL5) [13]. For
HL2, two ways of metal ion chelation have been demonstrated. It may
act as a tridentate ligand in its neutral [14], or in its deprotonated,
anionic form [12,15], due to tautomerism according to Scheme 1. It
may be presumed as a general feature for each derivative presented
here. In the free ligands, however, the tautomeric equilibrium is
shifted toward the isomer, in which the dissociable proton is located
on the isoindoline nitrogen. Indeed, as it is seen on the structure of HL4
(Fig. 1, Table 1) [16], strong interactions between N1, N3, N6 and H1
stabilise the symmetric diimine isomer, thus the molecule adopts
planar geometry in the crystalline form.
Life forms exhibiting SOD activity contain this structurally versatile
enzyme with various metal cofactors: Fe [2], Cu/Zn [3], Mn [4] and
NiSODs [5] are known. Permanent malfunction of such enzymes may
contribute to chronic diseases such as Alzheimer's, Parkinson's, or
inflammations, due to overexposing the cells to superoxide and to
secondary reactive oxygen species derived from superoxide [6]. It is
widely acknowledged, that the redox potential of the metal cofactor is
an important determinant of the SOD activity. However, only some
studies focus on systematically modified ligand series which are
applied to fine-tune the redox properties of the chelated metal ion
while providing unchanged stability for the complexes [7]. Here we
report the characterization of some iron(II) complexes with a series of
The bond distances between the carbon and nitrogen atoms are
typical for a conjugated π-system. The double bonds are found
between the C1–N2 and C8–N5 atom pairs (~1.29 Å) in the diimine-
isoindoline moiety, and between the C9–N3 and C17–N6 atom pairs
⁎
Corresponding author. Tel.: +36 88624720.
E-mail address: jpap@almos.vein.hu (J.S. Pap).
1387-7003/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.inoche.2010.10.023

206
B. Kripli et al. / Inorganic Chemistry Communications 14 (2011) 205–209
coordinated ligands, whereas lower energy bands between 440 and
500 nm, are tentatively rendered to charge transfer transitions from
the iron(II) ion to the ligand. In the infrared region the free isoindoline
ligands exhibit intense bands in the 1600–1660 cm−1 region that can
be assigned as coupled vC=N vibrations. Complexes containing the
deprotonated ligands on the other hand show weak band(s) above
1600 cm−1 and stronger ones between 1600 and 1500 cm−1
.
The one-electron oxidized analogue of 1, [FeIII(L1)2](CF3SO3).
CH3CN (1ox·CH3CN) was synthesized under dioxygen atmosphere
[18] in order to compare the SOD-like activity of the iron(II) and iron
(III) forms (vide infra). Besides the spectroscopic characterization, X-
ray structural analysis of the compound was also accomplished. The
structure of 1ox (Fig. 2a) shows an almost distortion-free octahedral
geometry of the six nitrogen donor atoms around the iron(III) center.
The two Nind atoms (N1 and N1A) are situated in the trans position to
each other, closer to the iron center than the thiazolyl nitrogens by
~0.05 Å. The planes of the coordinated ligand moieties are nearly
perpendicular to each other. The Fe–N bond distances in 1ox (~1.95
and 2.00 Å) are shorter than in the analogous iron(II) complexes (2.07
and 2.27 Å in avg. for 2) [12], and (2.05 and 2.15 Å in avg. for 4).
The iron(II) complex, 4 was also structurally characterized
(Fig. 2b). The two Fe–Nind bond distances (Fe1–N1, 2.0568(17) and
Fe1–N1A, 2.0529(17)) are close to the corresponding values for 2
(avg. 2.07 Å). The hexacoordinate iron(II) resides in the center of a
strongly distorted octahedron in contrast with that of 2, or 1ox. In the
structure of 4, a helical twist of the two ligand planes is seen, causing a
significant distortion of the octahedral geometry. This occurs, because
the shorter Fe–Nbim distances compared to the Fe–Npy distances in 2
(the difference is ~0.12 Å) bring the aromatic hydrogens in the 5′-
position of the benzimidazolyl groups very close (~2.5 Å) to the
indoline-plane of the other ligand, and the steric hinderance impedes
the settle of an undistorted geometry.
Scheme 1. Tautomerism of the ligands and the possible metal chelating modes.
(~1.32 Å) in the benzimidazole heterocycles. Due to the extended
conjugation of the π-bonds within the molecule, the UV–vis
absorption spectra of the ligands exhibit high intensity, and relatively
low energy, multiple π–π* bands in the 360–500 nm region (Table 2).
The bis-chelate complexes [FeII(L1)2] (1), [FeII(L3)2] (3), [FeII(L4)2]
(4), [FeII(L5)2] (5) reported here, bearing octahedral geometry were
prepared according to Eq. (2), analogously to a reported procedure for
[FeII(L2)2] (2) [12,18].
Electrochemical properties of the complex series were investigat-
ed with cyclic voltammetry in DMF solution. All compounds show one
reversible, one-electron redox wave in the plotted potential range
(Fig. 3). This is attributed to the Fe3+/Fe2+ redox transition, and spans
a 423 mV potential range from 1 to 5 (Table 3). To make a compound
thermodynamically competent in the SOD-like reaction, the redox
potential of the Fe3+/Fe2+ couple should be, as encountered in SODs,
between the potential of the two couples O2/O−2 and O2−/H2O2 that is,
at pH=7, −0.40 V and +0.65 V vs. SCE, respectively [7c]. All E1/2
values for the iron(II) complexes fall into this range. The process is
reversible with very similar Ea and Ec peak separations for each
complex (~100 mV), similarly to the ferrocene internal standard.
Although a possible explanation for this relatively wide potential
range could be the change of the spin-state of the iron(II) ion
depending on the coordinating ligand, based on the available
Mössbauer data for the two structurally characterized complexes, 2
[12] and 4 (zero field isomer shifts are 1.018 and 1.12 mms−1, and
quadrupole splittings are 1.975 and 1.32 mms−1 at 80 K, respective-
ly), these iron(II) complexes are high-spin. Assuming that one of the
two eg electrons is removed during the oxidation step, the relative
energy of these electrons throughout the series can be the governing
factor for the changes in the E1/2. The relative energy of the eg set
should follow the tendencies in the ligand-field strength, as a result
the observed increase in the E1/2 from 1 to 5 allows us to suppose a
decrease in the ligand-field. In other words, the stronger the ligand-
field effect, the easier it is to oxidize the iron(II) center. The notable
difference between the bond distances measured for 2 and 4 is in
accordance with this explanation.
The UV–vis absorption maxima of 1–5 (Table 2) that are found
between 350 and 460 nm are attributed to the π–π* transitions of the
Reactivity of 1–5 and 1ox with superoxide anion was investigated in
aqueous HEPES buffer following the modified McCord–Fridovich assay [19]
in the presence of catalase enzyme [20]. Complexes 1–5 and 1ox inhibit the
reduction of NBT [21]. Since the IC50 concentrations are proportional to the
concentration and type of indicator, we calculated the apparent kcat rate
constants as kcat =kI[I]/IC50 (where kI is 5.94×104 M−1 s−1, when I=NBT
Fig. 1. Molecular structure of HL4. Ellipsoids are plotted at the 50% probability. Selected
bond distances (Å) and angles (°): C1–N1 1.3843(18), C8–N1 1.3800(19), C1–N2
1.2906(19), C8–N5 1.2939(19), C9–N2 1.3767(19), C17–N5 1.3799(18), C17–N6 1.3235
(18), C9–N3 1.3264(18), N3–H1 2.106, N6–H1 2.112, N1–C1–N2 173.68(6), C1–N1–C8
85.80(6), N1–C8–N5 90.77(7), C1–N2–C9 169.62(6), C8–N5–C17 98.60(7), N1–H1–N6
122.20, and N1–H1–N3 122.29.

B. Kripli et al. / Inorganic Chemistry Communications 14 (2011) 205–209
207
Table 1
Crystal structure details for 1ox·CH3CN, 4 and HL4.
1
ox·CH3CN
4
HL4
Chemical formula
Formula weight
Space group
a, Å
b, Å
c, Å
C13H19F3FeN11O3S5 C48H35F3FeN14 C24H19N3
866.77
863.75
405.46
Manoclinic, P 21/c
14.0476(14)
20.5817(17)
12.2870(10)
90
103.000(2)
90
3461.4(5)
4
Triclinic, P-1
13.0522(19)
14.261(3)
14.691(2)
63.983(6)
74.430(6)
70.724(6)
2294.9(7)
2
Monoclinic, P 21/n
8.4949(2)
16.9348(3)
13.946(3)
90
95.758(1)
90
1996.24(7)
4
α, deg
β, deg
γ, deg
V, Å3
Z
Dcalc, mg m−3
Temperature, K
Unique reflections
1.663
1.250
1.349
296(2)
293(2)
203(2)
7926
10,065
5085
dataN 2σ parameters/ 4999/507/0
7964/572/0
3818/282/0
restraints
a
R1 [F2N2σ (F2)],
0.0478, 0. 1060
1.030
0.0495, 0.1377 0.0496, 0.1125
1.043 1.083
b
wR2 (F2)
Goodness of fit
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
a
R
=
ð
∑ jF j ꢀ jF j
Þ
=
ð
∑ F
Þ
.
ꢀ
ꢀ
1
o
c
o
h
i
wR
=
∑w F ꢀ F
∑w F
1=2,w=1/σ2(Fo2)+(AP)2+BP, where P[Fo2+2
ꢁ
ꢂ
ꢁ
ꢂ
2
2
b
2
2
2
2
o
c
o
(Fo2+2(Fc2)]/3, A=0.0562, 0.0978 and 0.0532, B=0, 0 and 0.7956 for 1ox.CH3CN, 4and HL4,
respectively.
[22] and 2.0×105 M−1 s−1 when I=cytochrome c [23]) for comparisons
with the literature data (Table 3). The [FeII(Ln)2] complexes show activities
comparable to those of iron(II) species containing TPEN-derivatives [7a].
Although positive indirect tests applied here do not provide conclusive
evidence on the real catalytic activity of putative SOD mimics, testing both
the oxidized and reduced forms is a good indication that the complexes can
indeed act as catalysts in the dismutation of superoxide anion. With 1ox
,
which is the iron(III) analogue of 1 similar activity was measured with that
of 1 (Table 3). Therefore we propose, that these complexes may be
considered as catalysts of superoxide anion dismutation.
The determined kcat (also the IC50 values) for 1–5 correlate with the
3+ 2+
observed EFe
potentials [24] (Fig. 4). Similar correlations with a
/Fe
positive slope have been described previously for manganese(III)- and
iron(III)-porphyrins [7]. Increase in the SOD-like activity with the redox
potential becomes explainable if one assumes that the reduction of the
Table 2
UV-vis absorption data of the free ligands and the complexes 1–5 in DMF.
Ligand
π–π* CT bands of the free
ligands λmax/nm (log ε)
[Fe(L)2] λmax/nm (log ε)
Ref.
Fig. 2. Molecular structures of a) 1ox·CH3CN and b) 4. Selected bond distances (Å) and
angles (°) for 1ox·CH3CN: Fe1–N1 1.946(2), Fe1–N1A 1.952(2), Fe1–N2 2.003(2), Fe1–N2A
2.005(2), Fe1–N3 2.003(2), Fe1–N3A 1.997(2), N1–Fe1–N1A 178.70(10), N2–Fe1–N3
175.99(10), N3A–Fe1–N2A 177.98(10), N1–Fe1–N2 89.23(9), N2–Fe1–N2A 92.06(9), N1–
Fe1–N2A 89.62(9); and for 4: Fe1–N1 2.0568(17), Fe1–N1A 2.0529(17), Fe1–N4 2.1356
(17), Fe1–N4A 2.1471(17), Fe1–N6 2.1296(17), Fe1–N6A 2.1318(17), N1–Fe1–N1A 173.68
(6), N1–Fe1–N4 85.80(6), N1–Fe1–N6 90.77(7), N6–Fe1–N6A 169.62(6), N1A–Fe1–N4
98.60(7), N4–Fe1–N4A 168.40(7). Solvent molecules and hydrogen atoms are removed for
clarity. The CF3SO−3 counterion and some ligands are plotted as spheres for viewing
purposes. Ellipsoids are plotted at the 50% probability.
HL1
485 (3.18)
448 (3.86)
420 (4.05)
396 (4.01)
374 (3.94)
410 (3.99)
386 (4.23)
366 (4.19)
408 (4.11)
385 (4.29)
366 (4.24)
478 (4.11)
447 (4.35)
420 (4.33)
393 (4.38)
373 (4.39)
470 (4.06)
440 (4.30)
413 (4.31)
388 (4.35)
369 (4.37)
788 (4.01)
463 (4.43)
435 (4.50)
404 (4.53)
346 (4.30)
440 (4.58)
410 (4.81)
390 (4.75)
439 (4.50)
411 (4.65)
392 (4.57)
487 (4.45)
454 (4.62)
433 (4.60)
349 (4.55)
This work
HL2
HL3
HL4
14a
12b
8a
This workb
iron(III) form is the rate-limiting step. On the basis of our results,
discerning between an inner-sphere, or an outer-sphere redox
mechanism is uncertain. However, in the case of the latter, in which
superoxide anion does not coordinate to the metal center, one would
expect that those complexes will show higher activity, which have
redox potentials close to the midway value between the two potential
values for the superoxide anion as a result of an equal thermodynamic
driving force for both the reduction and the oxidation steps [7f]. In
addition, it was shown for monohydroxoiron(III) and aquomanganese
(III) porphyrins that in case of an outer-sphere mechanism, a 120 mV
increase in E1/2 imparts a roughly 10-fold increase in kcat, in accordance
This work
HL5
493 (4.28)
453 (4.49)
389 (4.40)
367 (4.45)
10a
This workb
a
Ligand.
Complex.
b

208
B. Kripli et al. / Inorganic Chemistry Communications 14 (2011) 205–209
°′
Fig. 4. Dependence of the measured SOD-like activity of 1–5 on the E
values.
Fe3+=Fe2+
(Potentials are plotted vs. SCE. EFc+/Fc vs. SCE was determined experimentally to be + 500
′
10 mV in our system and was added to the experimentally determined E°Fe3+
values to
Fig. 3. CVs of compounds 1–5 (0.5–1.0 mM) taken at a scan rate of 100 mVs−1 using a
GCE working electrode, Pt-wire auxiliary electrode, and a Ag/AgCl reference electrode
in DMF. Potentials are referenced to the ferrocene/ferrocenium couple.
=Fe2+
gain the above plot.)
Acknowledgements
with the Marcus equation for outer-sphere electron-transfer reactions
[7b]. With our complexes the observed change in kcat is only 30% per
~120 mV. Based on these considerations, we think the inner-sphere
mechanism is more plausible for the redox steps according to Eq. (1) for
1–5. To further elucidate this question, thiocyanate was added to a
solution of compound 2 in 5:1 ratio, because this small ligand was
investigated in heptacoordinate iron(II) complexes earlier [25]. In the
UV–vis absorption spectrum of the mixture a new band occurs at
460 nm [26] besides the unchanged bands of 2 indicating the presence
of a heptacoordinate species. This emphasizes that an additional small,
anionic ligand like SCN− may add to the iron(II) center.
Financial support of the Hungarian National Research Fund (OTKA
K67871, OTKA K75783 and OTKA PD75360), TÉT and COST is
gratefully acknowledged.
Appendix A. Supplementary material
Files contain the supplementary crystallographic data for this
paper. These data can be obtained free of charge from The Cambridge
Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/
cif.
Supplementary data to this article can be found online at
doi:10.1016/j.inoche.2010.10.023.
Cross-reactions (that could modify the observed IC50 values)
between the redox indicator NBT and the tested high-spin iron
complexes could be excluded based on our experiments. Therefore
the observed linear correlation between the redox potential of the
complexes and their SOD-like activity indicates that the rate
determining step in the superoxide dismutation is the reduction of
the metal center. Formation of a heptacoordinate transition state by
the addition of superoxide to the octahedral complexes is a plausible
References
[1] J.M. McCord, I. Fridovich, J. Biol. Chem. 244 (1969) 6049.
[2] (a) W.C. Stallings , T.B. Powers , K.A. Pattridge , J.A. Fee , M.L. Ludwig , Proc. Natl
Acad. Sci. USA: Biol. Sci. 80 (1983) 3884;
mechanism based on our tests with SCN−
.
(b) D. Ringe , G.A. Petsko , F. Yamakura , K. Suzuki , D. Ohmori , Proc. Natl Acad.
Sci. USA: Biol. Sci. 80 (1983) 3879.
[3] J.A. Tainer, E.D. Getzoff, J.S. Richardson, D.C. Richardson, Nature 306 (1983) 284.
[4] (a) W.C. Stallings , K.A. Pattridge , R.K. Strong , M.L. Ludwig , J. Biol. Chem. 260
(1985) 6424;
(b) G.E. Borgstahl , H.E. Parge , M.J. Hickey , W.F. Hellewell Jr., R.A. Beyer , J.A.
Tainer , Cell 71 (1992) 107.
[5] a) B. Palenik, et al., Nature 424 (2003) 1037;
Table 3
Apparent rate constants and redox potentials of iron complexes exhibiting SOD-like
activity.
Comp1exa
kcat (106M−1s−1
)
E°′cox/red (V)
Ref.
b
1
1ox
5.00 0.35 (4.62 0.32)
0.555
This work
b) D.P. Barondeau, C.J. Kassmann, C.K. Burns, J.A. Tainer, E.D. Getzoff, Biochemical
43 (2004) 8038.
5.28 0.30
2
3
4
5
3.83 0.20
3.62 0.20
3.06 0.15 (3.28 0.15)
2.50 0.22 (2.26 0.24)
0.21 0.05
0.041
2.15
3.5
14.0
0.319
0.265
0.205
0.123
This work
This work
This work
This work
This work
[7c]
[24]
[7a]
[7a]
[7a]
[6] D.P. Riley, Chem. Rev. 99 (1999) 2573.
[7] a) M. Tamura, Y. Urano, K. Kikuchi, T. Higuchi, M. Hirobe, T. Nagano, Chem.
Pharm. Bull. 48 (2000) 1514;
b) I. Batinic-Haberle, I. Spasojevic, P. Hambright, L. Benov, A.L. Crumbliss, I.
Fridovich, Inorg. Chem. 38 (1999) 4011;
c) S. Durot, C. Policar, F. Cisnetti, F. Lambert, J.-P. Renault, G. Pelosi, G. Blain, H.
Korri-Youssoufi, J.-P. Mahy, Eur. J. Inorg. Chem. (2005) 3513;
d) M. Sjödin, J. Gatjens, L.C. Tabares, P. Thuéry, V.L. Pecoraro, S. Un, Inorg. Chem.
47 (2008) 2897;
e) G.-F. Liu, K. Dürr, R. Puchta, F.W. Heinemann, R. van Eldik, I. Ivanovic-
Burmazovic, Dalton Trans. (2009) 6292;
f) I. Batinic-Haberle, I. Spasojevic, R.D. Stevens, P. Hambright, P. Neta, A. Okado-
Matsumoto, I. Fridovich, Dalton Trans. (2004) 1696.
FeCl2
[FeIIIBIG)Cl2]
[FeII(TPAA]
[FeII(6MeTPEN)]2+
[FeII(TPEN]2+
[FeII((4Me)4TPEN)]2+
[FeII((4MeO)4TPEN]2+
Fe-SOD
0.640
0.590
0.510
0.425
0.020
0.160
35.0
70.0
[7a]
[27]
[6]
[28]
~220d
~2000d
700
Cu,Zn-SOD
cob(II)alamin
[8] W.O. Siegl, J. Org. Chem. 42 (1977) 1872.
[9] É. Balogh-Hergovich, J. Kaizer, G. Speier, G. Huttner, A. Jacobi, Inorg. Chem. 39
(2000) 4224.
[10] a) J. Kaizer, B. Kripli, G. Speier, L. Párkányi, Polyhedron 28 (2009) 933;
b) J. Kaizer, T. Csay, P. Kövári, G. Speier, L. Párkányi, J. Mol. Catal. A: Chem. 280
(2008) 203.
[11] P.T. Selvi, H. Stoeckli-Evans, M. Palaniandavar, J. Inorg. Biochem. 99 (2005) 2110.
[12] É. Balogh-Hergovich, G. Speier, M. Reglier, M. Giorgi, E. Kuzmann, A. Vértes, Inorg.
Chem. Commun. 8 (2005) 457.
aLigand abbreviations: BIG, N,N-bis[1-methyl-2-imidazolyl)methyl] glycinate TPAA. tris{2-
[N-(2-pyridylmethyl)amino]ethyl}amine; 6MeTPEN, N-(6-methyl-2-pyridy1emrthyl)-N,N′,
N′-(2-pyridylmethyl)ethylenediamine, TPEN, N,N,N′,N′-(2-pyridylmethyl)
ethylenediamine; (4Me)4TPEN, N,N,N′,N′-(4-methoxy-2-pyridylmethyl)ethylenediamine;
(4MeO)4TPEN, N,N,N′,N′-(4-methoxy-2-pyridylmethyl)ethylene-diamine. bValues in
parenthesis are from the cytochrome assays.cvs. SCE. dkcat/Km values.

B. Kripli et al. / Inorganic Chemistry Communications 14 (2011) 205–209
209
[13] 1, 3-bis(2′-pyridylimino)isoindoline (HL2) [8], 1, 3-bis(4′-methyl-2′-pyridyli-
mino) isoindoline (HL3) [8] and 1, 3-bis(2′-benzimdazolylimino)isoindoline
(HL5) [10] were prepared according to published procedures. 1,3-bis(2′-
thiazolylimino)isoindoline (HL1). Phthalonitrile (3.84 g, 30 mmol) and 6.00 g
(60 mmol) 2-aminothiazole were melted at 180 °C and stirred vigorously for
9 hours, until ammonia evolution ceased. The product was cooled to room
temperature and diethyl ether was added (100 mL). The dark brown crystals of
the product were filtered and vacuum dried. The crude ligand was recrystallized
from a minimal amount of boiling N, N-dimethylformamide. Yield: 7.44 g (80%).
M.p.: 258–260 °C. Anal. Calcd for C14H9N5S2: C, 54.00; H, 2.91; N, 22.49. Found: C,
54.30; H, 2.82; N, 22.32. FT-IR bands (KBr pellet, cm−1) 3207(w), 3109(w), 3076
(w), 1622(s), 1478(m), 1376(w), 1299(m), 1213(s), 1131(s) 1050(m), 873(w),
776(m), 706(s), 596(w). UV–vis (DMF) [λmax, nm (logε)] 485 (sh, 3.18), 448
(3.86), 420 (4.05), 396 (4.01), 374 (3.94), 288 (3.54). 1H-NMR (CDCl3), δ/ppm:
8.00 (dd, 2H, J1 =5.8 Hz, J2 =3.2 Hz), 7.77 (d, 2H, J=3.3 Hz), 7.63 (dd, 2H,
J1 =5.8 Hz, J2 =3.3 Hz), 7.18 (d, 2H, J=3.3 Hz). 13C-NMR (CDCl3), δ/ppm: 117.6,
122.9, 132, 135, 141, 152.7, 171.8. 1,3-bis(1′-methyl-2′-benzimidazolylimino)
isoindoline (HL4). This ligand was synthesized as it was reported for HL5 earlier
[9]. Yield: 85%. M.p.: 197–199 °C. Anal. Calcd for C24H19N7: C, 71.09; H, 4.72; N,
24.18. Found: C, 70.92; H, 4.83; N, 24.32. FT-IR bands (KBr pellet, cm−1) 3111(w),
3048(w), 2957(w), 2924(w), 1621(s), 1478(m), 1434(m), 1321(m), 1266(m),
1209(s), 1095(s) 1050(m), 824(w), 739 (s), 694(w). UV–vis (DMF) [λmax, nm
(logε)] 478 (4.11), 447 (4.35), 420 (4.33), 393 (4.38), 373 (4.39), 348 (4.33), 316
(4.19), 289 (4.37). 1H-NMR (CDCl3), δ/ppm: 8.23 (m, 4H), 7.90 (m, 4H), 7.62 (m,
4H), 4.06 (s, 6H).
(w), 1572(s), 1516(s), 1466(s), 1394(m), 1346(w), 1288(m), 1184(m), 1066(s),
996(m), 931(m), 805(m), 706(m). UV–vis (DMF) [λmax, nm (logε)] 439 (4.50),
411 (4.65), 392 (4.57), 347 (4.19), 331 (4.32), 307 (4.43). 4. Yield: 61%. Anal.
Calcd for C48H36FeN14: C, 66.67; H, 4.20; N, 22.68. Found: C, 66.51; H, 4.13; N,
22.72. FT-IR bands (KBr pellet, cm−1) 3056(w), 2933(w), 1622(s), 1552(s), 1483
(w), 1471(m), 1434(w), 1386(m), 1325(m), 1287(w), 1186(m) 1099(m), 1074
(m), 739(m). UV–vis (DMF) [λmax, nm (logε)] 487 (sh, 4.45), 454 (4.62), 433 (sh,
4.60), 349 (4.55). Crystals suitable for X-ray structural determination were
obtained from DMF. 5. Yield: 72%. Anal. Calcd for C44H28FeN14: C, 65.35; H, 3.49; N,
24.25. Found: C, 65.19; H, 3.37; N, 24.08. FT-IR bands (KBr pellet, cm−1) 3419(m),
3060(w), 1616(m), 1533(s), 1471(m), 1446(m), 1416(m), 1374(w), 1315(m),
1271(m), 1191(m) 1115(m), 1029(w), 748(m). UV–vis (DMF) [λmax, nm (logε)]
493 (4.28), 453 (4.49), 389 (4.40), 367 (4.45), 319 (3.29).
[19] a) J.M. McCord, I. Fridovich, J. Biol. Chem. 244 (1969) 6049;
b) C. Beauchamp, I. Fridovich, Anal. Biochem. 44 (1971) 276.
[20] The SOD-like activity of the iron complexes was studied at 298 K by using indirect
method of nitroblue tetrazolium (NBT), or cytochrome
c reduction. The
superoxide radical anion was generated in situ by the xanthine/xanthine oxidase
reaction, and detected spectrophotometrically by monitoring the formation of
diformazan from NBT at 560 nm, or alternatively, the reduced form of cytochrome
c
at 550 nm. The test reactions were carried out in 0.01 M HEPES (N-2-
hydroxyethylpiperazine-N′-2-ethanesulfonic acid) buffer at pH=7.6. The reac-
tion was initiated by adding an appropriate amount of xanthine oxidase to a
stirred solution of NBT (1.7×10−4 M), or cytochrome (5.2×10−5 M), xanthine
(10−4 M) and catalase enzyme (175 U/3 mL) to generate ΔA560 550 =0.025–
or
[14] a) R.R. Gagné, D.N. Marks, Inorg. Chem. 23 (1984) 65;
0.028 min−1. The NBT reduction rate was measured either in the absence or
presence of putative SOD mimics (0–20×10−6 M). The possible inhibition of the
xanthine oxidase enzyme by the SOD mimics was excluded by monitoring the
urate formation at 296 nm in separate test reactions. The SOD-like activity was
then characterized by determining the IC50 values (μM, the concentration where
inhibition of NBT reduction is 50%) for each compound. The IC50 concentrations
can be obtained from ΔA0/ΔAX −1 vs. complex concentration plots as follows:
complex concentration, X [μM]=IC50 where ΔA0/ΔAX −1=1 (ΔA0 is the change
in absorbance at 560 nm min−1 in the absence of complex; ΔAX is the change in
b) D.M. Baird, K.Y. Shih, J.H. Welch, R.D. Bereman, Polyhedron 8 (1989) 2365.
[15] a) J. Kaizer, J. Pap, G. Speier, M. Reglier, M. Giorgi, Trans. Met. Chem. 29 (2004)
630;
b) J. Kaizer, J. Pap, G. Speier, L. Párkányi, Z. Kristallogr. NCS 219 (2004) 141;
c) R.J. Letcher, W. Zhang, C. Bensimon, R.J. Crutchley, Inorg. Chim. Acta 210
(1993) 183.
[16] The crystal evaluations and intensity data collections were performed either on a
Rigaku R-AXIS Rapid single-crystal diffractometer (1ox and 4), or a Bruker–Nonius
Kappa CCD single-crystal diffractometer (HL4) using Mo Kα radiation
(λ=0.71070 Å) at 296, 293, and 203 K, respectively. Crystallographic data and
details of the structure determination are given in Table 1, whereas selected bond
lengths and angles are listed in the corresponding figure captions. SHELX-97[17]
was used for structure solution and full matrix least squares refinement on F2. CIF
files are available in the CCDC database: 1ox·CH3CN (CCDC 756934), 4 (CCDC
756933) and HL4 (CCDC 756935).
absorbance at 560 nm min−1 in the presence of
X μM of the complex).
Alternatively, IC50 can be determined when 100(ΔA0 −ΔAX)/ΔA0 is plotted
against the complex concentration. The data can be fitted with a first order
exponential curve, and the 50% inhibition on the curve shows the IC50. The two
methods are equally suitable for the determination of the IC50 values.
[21] In separate test reactions the inhibition of the xanthine/xanthine oxidase system
by the iron(II) complexes was excluded. The reduction of NBT by superoxide
[17] G.M. Sheldrick, SHELX97, Program for the Refinement of Crystal Structures,
University of Göttingen, Germany, 1997.
anion takes place in a one-electron process giving tetrazolinyl radical, that
disproportionates to formazan, which gives indicator colour and NBT. Under assay
conditions with 1 mM/min superoxide anion production, where the concentra-
tion of the complex and NBT is higher than that of the substrate, a mixture of the
reduced and oxidised form of the complex is present simultaneously. This way,
there is a possibility of cross reaction between the oxidised form of the putative
SOD mimic complex, and the tetrazolinyl radical that leads to overestimation of
the SOD activity. To elucidate this problem, the SOD-like activity was tested in
three cases (1, 2 and 5, Table 4) with cytochrome c indicator, too. No significant
difference in the kcat values was found using cytochrome c compared to the
results from the NBT assays, therefore the above depicted cross reaction does not
interfere with the superoxide scavenging reactions.
[18] [FeII(L1)2] (1). To a stirred suspension of 1.00 g (3.22 mmol) of HL1 and 0.41 g
(1.61 mmol) of FeII(ClO4)2·xH2O in methanol (20 mL) 450 μL (3.22 mmol)
triethylamine was added dropwise under an argon blanketing atmosphere. The
mixture was stirred for overnight at room temperature and then filtered. The dark
green precipitate was washed with 20 mL methanol and 20 mL diethylether and
vacuum dried. Yield: 0.70 g (64%). Anal. Calcd for C28H16FeN10S4: C, 49.70; H,
2.38; N, 20.70. Found: C, 49.55; H, 2.52; N, 20.52. FT-IR bands (KBr pellet, cm−1
)
3110(w), 3090(w), 1520(s), 1471(w), 1290(m), 1230(m), 1187(s), 1102(s) 1072
(m), 921(m), 874(m), 812(m), 710(s), 643(w), 523(w). UV–vis (DMF) [λmax, nm
(logε)] 788 (4.02), 463 (4.43), 435 (4.50), 404 (4.53), 346 (4.30), 288 (4.56). [FeIII
(L1)2](CF3SO3) (1ox). To a stirred suspension of 0.32 g (1.02 mmol) L1H and 0.18 g
(0.51 mmol) FeII(CF3SO3)2 in acetonitrile (10 mL) 140 μL (1.01 mmol) triethyla-
mine was added dropwise under an argon blanketing atmosphere. The solution
then was put under dioxygen atmosphere and stirred for overnight. The green
precipitate was filtered off, and the clear mother liquor was slowly evaporated to
give 1ox in a crystalline form. Crystals suitable for X-ray structural analysis were
selected from the bulk product. Yield: 0.20 g (41%). Anal. Calcd for C29H16F3-
FeN10O4S5·CH3CN: C, 42.96; H, 2.21; N, 17.78. Found: C, 42.92; H, 2.16; N, 17.82.
FT-IR bands (KBr pellet, cm−1) 3400(m), 3105(w), 3083(w), 1593(m), 1517(m),
1491(w), 1298(s), 1251(s), 1235(s), 1213(m), 1162(m) 1101(m), 1032(s), 879
(w), 746(m), 714(w), 651(m), 634(w), 515(w). UV–vis (DMF) [λmax, nm (logε)]
480 (4.40), 440 (4.50), 414 (4.43), 345 (4.31), 290 (4.52). [FeII(L3)2] (3), [FeII
(L4)2] (4), [FeII(L5)2] (5). All complexes were synthesized analogously to 1. 3.
Yield: 43%. Anal. Calcd for C40H32FeN10: C, 67.80; H, 4.55; N, 19.77. Found: C,
68.04; H, 4.62; N, 19.85. FT-IR bands (KBr pellet, cm−1) 3056(w), 2921(w), 2847
[22] Z.-R. Liao, X.-F. Zheng, B.-S. Luo, L.-R. Shen, D.-F. Li, H.-L. Liu, W. Zhao, Polyhedron
20 (2001) 2813.
[23] C. Auclair , E. Voisin , in: R.A. Greenwald (Ed.), CRC Handbook of Methods for
Oxygen Radical Research, CRC, Boca Raton, FL, 1985, p. 123.
[24] R.H. Weiss, A.G. Flickinger, W.J. Rivers, M.M. Hardy, K.W. Aston, U.S. Ryan, D.P.
Riley, J. Biol. Chem. 268 (1993) 23049.
[25] a) S. Zhang, K.S. Snyder, R.E. Shepherd, Inorg. Chim. Acta 201 (1992) 223;
b) M.S. Ward, R.E. Shepherd, Inorg. Chim. Acta 286 (1999) 197.
[26] Originates from the thiocyanate to iron(II) charge transfer band. This ligand was
chosen because there is no overlap between the 460 nm band and other
absorption bands of 2.
[27] W.B. Greenleaf, D.N. Silverman, J. Biol. Chem. 277 (2002) 49282.
[28] E. Suarez-Moreira, J. Yun, C.S. Birch, J.H.H. Williams, A. McCaddon, N.E. Brasch, J.
Am. Chem. Soc. 131 (2009) 15078.