Chemical speciation of copper(II) diaminediamide derivative of pentacycloundecane - A potential anti-inflammatory agent

Article abstract of DOI:10.1039/b614878f

Formation constants of copper(ii), zinc(ii) and calcium(ii) with 3,5-diaminodiamido-4-oxahexacyclododecane (cageL) has been studied by glass electrode potentiometry at 25 °C and an ionic strength of 0.15 mol dm ^-3. Copper(ii) forms more stable complexes with cageL than zinc(ii) and calcium(ii). Metal ion complexation promotes deprotonation and coordination of the amide nitrogens resulting in overall tetragonal coordination of Cu ²⁺ suggested by the UV-visible electronic spectra. Speciation calculations using a blood plasma model suggest that zinc(ii) and calcium(ii) are good competitors of copper(ii) in vivo. Bio-distribution experiments using ⁶⁴Cu-labelled Cu(ii)-cageL show that about 50% dose of the complex is retained in the body after 24 h. This journal is The Royal Society of Chemistry.

Full text of DOI:10.1039/b614878f

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Chemical speciation of copper(II) diaminediamide derivative of
pentacycloundecane—a potential anti-inflammatory agent
Sebusi Odisitse,a Graham E. Jackson,a Thavendran Govender,b Hendrik G. Krugerb and Amith Singhb
Received 12th October 2006, Accepted 26th January 2007
First published as an Advance Article on the web 8th February 2007
DOI: 10.1039/b614878f
Formation constants of copper(II), zinc(II) and calcium(II) with 3,5-diaminodiamido-4-
oxahexacyclododecane (cageL) has been studied by glass electrode potentiometry at 25 ◦C and an ionic
strength of 0.15 mol dm−3. Copper(II) forms more stable complexes with cageL than zinc(II) and
calcium(II). Metal ion complexation promotes deprotonation and coordination of the amide nitrogens
resulting in overall tetragonal coordination of Cu2+ suggested by the UV-visible electronic spectra.
Speciation calculations using a blood plasma model suggest that zinc(II) and calcium(II) are good
competitors of copper(II) in vivo. Bio-distribution experiments using 64Cu-labelled Cu(II)-cageL show
that about 50% dose of the complex is retained in the body after 24 h.
they were still hydrophilic and hence were cleared from the
Introduction
body via the kidneys. In order to improve the lipophilicity
Rheumatoid arthritis (RA) is a debilitating disease affecting
of the complex, mixed amino/amido ligands were designed in
some 5% of the Western World.1 There is no cure, however,
which coordination to the metal would lead to deprotonation
of the amide nitrogen and a neutral complex.10 Partition co-
efficient measurements showed that these complexes were in-
deed more lipophilic with 5% of the complex going into the
organic layer. In order to further improve the lipophilicity,
the pentacyclo-undecane derived ligand 3,5-diaminodiamido-4-
oxahexacyclo[5.4.1.02,6.03,10.05,9.08,11]dodecane (cageL) shown in
Fig. 1 was designed. This ligand has the amino/amido coordi-
nating structure used before but now attached to a pentacyclo-
undecane derivative.
The chemistry of pentacyclo-undecane cage derivatives has been
extensively studied.15–17 A number of amino cage compounds have
promising potential medicinal and pharmaceutical activities.18–23
The advantages of incorporating a rigid cage structure into bio-
active compounds has been extensively reported.24–30 Due to the
large lipophilic nature of the cage moiety it was able to cross
various membranes quite efficiently. An added advantage, is that
the rigidity of the cage should increase the stability of the metal
complex by forcing the ligand into an ideal conformation for
complexation.
the symptoms and progression of the disease can be controlled
using immunosuppressive and anti-inflammatory drugs.2,3 Copper
complexes were observed to be effective in the treatment of RA and
other degenerative connective tissue diseases as early as 1941.4,5
Indeed, the copper bangle has been used for centuries.7 More
recently, we and others have shown that copper complexes are
able to alleviate the inflammation associated with RA.7–13 Serum
copper levels are elevated during RA and it has been postulated
that endogenous copper might have a protective function in
chronic inflammatory conditions.4–7 With this in mind, we have
embarked on a programme of ligand design with the objective of
producing a ligand which will complex copper and increase its
bioavailability and at the same time not disrupt the homeostasis
of other endogenous metal ions. To this end, the ligand has to be
a strong chelator of copper, although it should not be so strong
that the complex is excreted intact before the copper can exert
its therapeutic potential. The complex should also be primarily a
nitrogen donor ligand so that the selectivity for copper is increased.
The concentrations of calcium(II) and zinc(II) in blood plasma
are far higher than that of copper(II) and the ligand would have
to overcome this concentration differential in order to bind to
copper(II) in vivo. The complex should also be kinetically labile
to be able to release the copper at the active site and should be
lipophilic in order to be absorbed dermally.
Results and discussion
Ligand synthesis
The above criteria led to the design of the polyamine
ligands 3,6,9,12-tetra-azatetradecanedioic acid (L6) and 3,6,9-
triazatetradecanedioic acid (L7).8,14 In animal tests, these lig-
ands proved to be powerful chelators of copper, so much so
that they were rapidly excreted, unchanged, in urine. Although
these complexes were formally neutral, being dicarboxylic acids,
The diaminodiamido cage ligand was synthesized from
pentacyclo-undecane dione as illustrated in Scheme 1.
The dione is photocyclized from the Diels–Alder adduct be-
tween p-benzoquinone and cyclopentadiene.31 Treatment of the
dione with allylmagnesium bromide produced the endo–endo diol
2.32–33 The subsequent reactions to synthesize the intermediates
up to the acid chloride 5 have been described previously.34–36 The
novel ester 6 was obtained through reaction of the acid chloride
with excess ethanol. The required diaminodiamido cage ligand 7
was obtained by treatment of the ester 6 with ethylenediamine (see
Scheme 2).
aDepartment of Chemistry, University of Cape Town, Private Bag Ronde-
bosch, 7701, South Africa
bSchool of Chemistry, University of KwaZulu-Natal, Durban, 4041, South
Africa. E-mail: jackson@science.uct.ac.za
1140 | Dalton Trans., 2007, 1140–1149
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
Fig. 1 Schematic diagrams of cageL and related ligands.
Scheme 1
the presence of the amide N–H proton. The multiplet at 3.13 ppm
can be attributed to methylene protons H-15. The relative flat
proton signal at 1.9 ppm was assigned to the amino protons.
Potentiometry
The protonation constants for the two terminal amine groups of
cageL are given in Table 1. The first protonation constant is lower
than that of methylamine because of the electronic withdrawal
effect of the amide. Similarly, pKa1 is 1.23 log units greater than
pKa2 because of the electrostatic repulsion of the first proton.
We have used the complex formation (ZM-bar)37,38 (eqn (1))
and deprotonation functions (Q-bar)38,39 (eqn (2)) as criteria in
speciation model selection where the former measures the number
of ligands bound per metal ion while the latter indicates the
number of protons released upon metal ion complexation. The
two functions are derived from the free and total concentrations of
the participating components as well as the evaluated protonation
constants. The classical ZM-bar function is strictly only defined
for simple mononuclear complex formation. However, deviations
from ideal behaviour are indicative of the different speciation
Scheme 2
The 13C NMR spectrum of the diaminodiamido ligand 7
shows the amide carbonyls at 170.05 ppm, quaternary carbons at
93.6 ppm and methylene carbons C-12, C-14 and C-15 at 39.3 ppm,
41.6 ppm and 41.0 ppm, respectively. The “cage” carbons are
1
reported in the Experimental section. The H NMR spectrum
shows two doublets (doublet of doublets) that are due to protons
H-4a (1.40 ppm) and H-4s (1.75 ppm) coupling with protons at the
two non-equivalent sides of the cage. The peak at 7.19 ppm shows
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 1140–1149 | 1141
©

View Article Online
Table 1 Logarithms of overall stability constant, bpqr, of copper(II),
zinc(II) and calcium(II) complexes of cageL at 25 ◦C and ionic strength
0.15 mol dm−3 NaCl. RH is the Hamilton R-factor, Rlim is the minimum
R-factor based on the estimated errors in the analytical data. The standard
deviation in the log b is given in parentheses. The general formula of the
complex MpLqHr is denoted by the coefficients pqr
Metal
H+
p
q
r
Log bpqr
9.52(1)
17.81(2)
15.50(3)
9.96(2)
RH
Rlim
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
0
0.01
0.02
0.02
0.02
Cu(II)
−1
−2
1
2.71(3)
−7.05(4)
13.81(4)
5.55(3)
−2.93(4)
−11.74(3)
13.19(3)
3.80(3)
Zn(II)
Ca(II)
0.01
0.01
0.03
0.02
0
−1
−2
1
0
−1
−7.57(6)
occurring. Thus, if the curves at different metal : ligand ratios are
not superimposable, protonated or polynuclear species formation
is indicated, while if the curves fan back hydroxyl species formation
is indicated.
ZM-bar and Q-bar are defined by:
ZM-bar = (TL − [L])/TM
(1)
(2)
Q-bar = (TH* − TH)/TM
Where TL, TM and TH refer to the total ligand, metal and proton
concentration, respectively, [L] is the free ligand concentration and
T*H is the calculated total concentration of protons that would be
necessary to obtain the same pH if no complexation took place.
Fig. 2 shows the ZM-bar function for the Cu(II)–cageL system
plotted against pL (−log[L]). These curves at different metal-to-
ligand ratios were not superimposable, indicating the presence of
polynuclear or protonated species. At low pL or high pH the curves
fanned back indicating the formation of hydroxy species. The same
speciation is shown by the Q-bar curves. Again these curves are
not superimposable and cut the n-bar curve around pH 6.0. Since
the n-bar curve is a plot of the number of protons that would be
bound to the ligand at a particular pH in the absence of the metal
ion, if the Q-bar is greater than n-bar, it means that more protons
have been lost from the ligand than the ligand had to lose. That is
an hydroxy species has formed.
Fig. 2 Experimental and theoretical (a) Z-bar and (b) Q-bar curves for
the Cu(II)-cageL system at 25 ◦C and an ionic strength of 0.15 M (Cl−). M :
L ratios 1 : 2 (ꢀ), 1 : 3 (᭜), 1 : 4 (ꢁ) are displayed. The theoretical line was
calculated using the model given in Table 1.
The potentiometric data were analysed using the ESTA suite of
computer programs39 which yielded the results given in Table 1.
The low standard deviation in the log b’s, the low Hamilton factors
and the agreement between the observed and calculated data lend
confidence to the results. In fact RH ≈ Rlim, means that it is not
statistically possible to improve the model. It is interesting to note
that the stability of the Cu(II)–cageL complexes is greater than
its non-cage analogue. It was anticipated that this would be the
case because in cageL, the ligand is pre-formed in the correct
conformation for metal-ion coordination. A species distribution
diagram, calculated using the data in Table 1 is shown in Fig. 3.
From this we see that complexation commences at about pH 3
when the protonated MLH species is formed. At pH 6, the ML
species predominates while at pH 8.5 most of the copper is in the
MLH−1 species.
Fig. 3 Speciation distribution curves for a Cu(II)–cageL solution (with
M : L of 1 : 1 and [cageL] = 0.0034 mol dm−3) plotted as a function of pH.
Zn(II) and Ca(II) were also found to form complexes with
cageL albeit less stable complexes (Table 1). The order of stability
Cu(II) > Zn(II) > Ca(II) is as expected.40 These two metal ions were
studied because they are potential competitors of Cu(II) in vivo.
Even though they are weaker chelators than Cu(II), their in vivo
concentration is far higher and so they could potentially compete.
From the potentiometric data it is not possible to tell the struc-
ture of the different complexes formed. However, by comparison
1142 | Dalton Trans., 2007, 1140–1149
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
Table 2 Wavelengths corresponding to maximum absorption coefficients
Structure 4(a) has the metal bidentately coordinated to one
terminal amine and a carbonyl oxygen. The other terminal
amine is still protonated giving the correct stoichiometry. Billo43
and subsequently Sigel and Martin44 have proposed a simple
empirical method of estimating the electronic energy of a proposed
configuration. If one amine is coordinated to the copper a kmax of
743 nm is expected and with two amines a kmax of 661 nm is
expected. The kmax of the MLH species is 720 nm which is much
closer to the value for single amine coordination. Since we believe
the ligand is bidentate the carbonyl oxygen must be coordinated.
This would not change kmax. Further support for this conclusion
comes from the demonstration that with copper(II)–glycinamide
(pKa 7.93)42 and b-alaninamide (pKa 9.19)42 the metal coordinates
through the amine and the carboxyl group with log b of 5.3 and 5.1,
respectively.42 These values are very similar to the log K of 6.0 for
our system where the pK1 = 9.52. It is possible for both carbonyl
groups to be coordinated as depicted in Fig 4(b). The UV-visible
spectrum of this structure would be similar to that of structure 4(a)
as would their formation constants. Hence we cannot distinguish
between these two structures.
Loss of a proton from the MLH species gives rise to the ML
complex. If structure 4(b) is the correct structure for MLH then
ML could form by coordination of the second terminal amine
to give structure 4(c). Alternatively, one of the amides could
deprotonate to give structure 4(d) where the second carbonyl may
or may not be coordinated. The pKa for this process (log b110 − log
b111) is 5.0. This can be compared to the same process occurring
with glycinamide (pKa 6.8) where it is known that coordination
swaps from the amide oxygen to the amide nitrogen.44 kmax for the
ML complex of cageL is 660 nm. Structure 4(c) and 4(d) have
calculated kmax values of 661 and 646 nm, respectively. Hence we
conclude that the ML species has structure 4(c).
of the various Cu(II) species formed in solution with cageL
kmax/nm, e/dm3mol−1 cm−1
M
MLH
ML
MLH−1
790
13.2
720
27.4
660
44.2
590
95.5
with known systems, it is possible to make some inference as to
the site of coordination. Ligands L1 and L2 in Fig. 1 are similar to
cageL but do not have the cage moiety.
Solution structures
During the pH titration of the Cu(II)–cageL system, the colour of
the solution changed from blue to blue–violet. Electronic spectra
were recorded for this system as a function of pH. A single
broad absorption band was observed which envelopes the expected
2
2
2
2
2
2
spin allowed A1g← B1g, B2g← B1g and Eg← B1g transitions of
a tetragonally distorted copper complex.41 Analysis of the data
yielded the kmax and molar extinction coefficients listed in Table 2.
Electronic spectra are useful because the energy of the transition
is influenced by the coordination sphere of the metal ion. Thus
kmax affords a measure of the solution structure of the complex.
Computer optimized structures for the MLH species are shown
in Fig. 4. The log K for the equilibrium M + LH ↔ MLH is
6.0. Given the basicity of the amine, this value is too high for
monodentate coordination (cf. methylamine, pKa = 10.6; b110
=
4.1).42
Desseyn et al.45 have proposed a structure in which only the
two amines are coordinated to the metal. This structure also has
a calculated kmax of 661 nm. We believe, however, that the size
of the chelate ring argues against this structure. Comparing 1,2-
ethylenediamine (pK1 = 9.89; pK2 = 7.08),42 1,3-propylenediamine
(pK1 = 10.56; pK2 = 8.76)42 and 1,4-butylenediamine (pK1
=
10.72; pK2 = 9.44),42 which form 5,6 and 7 membered chelate rings,
the copper binding constants are 10.5, 9.7 and 8.6, respectively.42
The increasing basicity of the amines should increase the stability
of the complex but this is countered by the decreased chelate effect.
If only the two terminal amines of CageL were to coordinate there
would be no chelate stabilisation. However, its log b110 is 9.96 while
pK1 is 9.52 and pK2 is 8.289.
Structures 4(e)–4(h) are all possible for the MLH−1 species.
From the kmax value for this complex of 590 nm structure 4(f)
can be eliminated. The other possible structures have calculated
kmax values between 573 and 584. Structure 4(e) has been proposed
by Zuberbu¨ler and Kaden46 but we do not favour this structure
because of the large chelate ring.
Structure 4(g) requires that both amides are deprotonated while
the terminal amine remains protonated. The pKa for the loss of
a proton from ML to form MLH−1 is 7.2. This value is below
pKa2 (8.3) but one may argue that electronic repulsion from the
charged metal center would decrease the basicity of the amine.
Also, this complex exists in the pH range 7–10. At this high pH,
it is unlikely that the terminal amine would remain protonated
while the amides deprotonated. If the amide were to coordinate
Fig. 4 Optimised structures for the different possible isomers of
Cu(II)–cageL, For clarity axial waters and hydrogen atoms bonded to
carbon atoms have been omitted.
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 1140–1149 | 1143
©

View Article Online
to the metal ion then it would be easy for the amine to coordinate
giving the MLH−2 stoichiometry. Similarly structure 4(f), where
the carbonyl oxygen is coordinated and the unprotonated amine
is not, is unlikely. Hence we favour structure 4(h) for the MLH−1
species.
MLH−2 forms from the deprotonation of MLH−1 with a pKa of
9.7. This proton could come from a coordinated water molecule or
from the uncoordinated amide nitrogen. The hydrolysis constant
of [Cu(H2O)6]2+ is 7.87, but this is for an equatorially coordinated
water molecule. Because of Jahn–Teller distortion, hydrolysis of
axial water will be much lower (i.e. the pKa would be much
higher). Hence it is possible that the proton comes from water.
However, we have shown that coordination of OH− in the axial
position causes a 20 nm red shift in kmax. While we were unable
to obtain the electronic spectrum of MLH−2, spectra for related
ligands L1–L5 all show a substantial blue shift.9,10 Thus we propose
that deprotonation of the amide occurs and the ligand rearranges
to give structure 4(i). This structure is supported by several other
studies47 including the crystal structure of Ni–L3.48
Nuclear magnetic resonance studies
Fig. 6 Proton NMR spectra for complexation of Cu(II) (0.007 M) with
cageL (0.015 M) as a function of pH. The protons are labeled according
to Fig. 1.
Fig. 5 shows a plot of chemical shift as a function of pH where (a),
(b) and (c) refer to the chemical shifts of the methylene protons
adjacent to the carbonyl, amine and amide groups (in Fig. 1),
respectively. There is no significant change in the chemical shift of
the (a) protons while (b) and (c) shift substantially. This is a clear
indication that the terminal amines are undergoing deprotonation.
Since the molecule is symmetrical and the protons undergo rapid
chemical exchange, only a single set of resonances is seen. The
inflection point in the chemical shift curve can be used to estimate
the average pKa of the two protonation sites. The value of 9.5 is
in very good agreement with the average value of 9.52 obtained
potentiometrically.
broadened substantially and by pH 7.5 has virtually disappeared
into the baseline. This is a clear indication that complexation starts
at the terminal amine. The fact that the (c) proton signals do not
broaden at the same rate indicates that they are not at the same
distance from the metal ion. If the amide was also coordinated one
would expect signals (b) and (c) to broaden at the same rate. Above
pH 8.3 the (c) protons do broaden while the (a) protons are still
relatively sharp. This is consistent with coordination switching
from the carbonyl group to the amide nitrogen. These results
support the structures postulated above for the different species.
Molecular mechanics
Molecular mechanics has been used extensively to study the
preferred conformation of organic molecules. However, its use in
inorganic chemistry or coordination chemistry is not as extensive.
The main reason for this is the difficulty of obtaining a reliable
force field which will accurately describe the bonding of metals.
A second problem is the large number of different coordination
geometries available to metal ions. Notwithstanding these difficul-
ties, Accelrys have developed the extensible systematic force field
(esff).49 This force field employs semi-empirical rules to translate
atomic-based parameters to parameters typically associated with
a covalent force field. The force field has been applied to molecular
simulations of a wide variety of systems including transition metal
complexes.50
Fig. 5 Change in proton chemical shift (ppm) as a function of pH of
0.015 mol dm−3 cageL at 25 ◦C. The protons are labeled according to
Fig. 1.
Fig. 4 shows the optimised structures for the different possible
isomers of Cu(II)–cageL. It is not possible to compare the total
potential energy because of the different bonding and overall
charge of the different structures. However, the internal energy
represents the strain energy introduced into the ligand when it is
forced to adopt a particular conformation upon metal binding.
It is, therefore, one of the components of the total potential
energy, the others being the ligand field stabilization energy,
electrostatic interaction energy and van der Waal’s energy. Care
must be exercised in comparing these energies to the measured
Fig. 6 shows the effect of Cu(II) on the spectrum of cageL. Cu(II)
is paramagnetic which can affect both the chemical shift and
relaxation time of the protons. This manifests itself in a broadening
and shifting of the NMR signals. The broadening effect of the
metal is attenuated by 1/r6 where r is the internuclear distance
between the Cu(II) and the observed proton. Copper(II) exchange
is very rapid and so only an average spectrum is seen. Inspection of
Fig. 6 shows that at pH 4 no complexation has taken place as none
of the signals are broadened. At pH 6.2, the (b) proton signal has
1144 | Dalton Trans., 2007, 1140–1149
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
Table 3 Bonding potential energies differences (kcal mol−1) calculated
for ligand conformations found in the four possible structures of MLH−1
(Fig. 4) relative to the ligand in its lowest energy conformation (cvff force
field) together with the total internal potential energy of the metal complex
(esff force field)49
has been made of the ECCLES model51 of blood plasma to which
the constants determined in this study have been added. Fig. 7(a)
gives the p.m.i. curves for cageL and some related ligands. These
curves show that cageL mobilizes Cu(II) more than L1, however, it
is much poorer than L6. The reason for this is the relatively high
affinity of cageL for Zn(II) and Ca(II). Fig. 7(b) shows that cageL
causes substantial mobilization of Ca(II) and Zn(II) and so there
is little of the ligand left to complex Cu(II). These results indicate
that cageL could find application as a carrier of copper in dermal
absorption but that it would release any bound copper once it
entered the circulatory system.
Structures
Energy
e
f
g
h
Total
Bond
Angle
Torsion
Total for complex
21.5
46.9
5.8
18.0
20.1
30.1
42.2
−3.3
11.0
34.3
102.4
−2.3
−2.2
8.4
5.5
14.8
107.8
26.0
102.2
109.8
thermodynamic equilibrium constants because no account has
been taken of possible entropy changes. Notwithstanding these
limitations, the internal potential energy can be useful in assessing
the viability of a proposed structure.
Since it is difficult to compare the potential energy of complexes
with differing modes of coordination, we have calculated the
change in internal potential energy of the ligand in going from its
lowest energy conformation to the conformation depicted in the
optimised structures given in Fig. 4. Thus this change in potential
energy represents the ligand conformational penalty which must
be paid in order for the ligand to coordinate to the metal ion.
The internal potential energy of the lowest energy conformation
of the ligand is 102.4 kcal mol−1. This high potential energy comes
mainly from the strain within the cage structure, in particular the
angle bending and torsion angle energies. By forcing the ligand to
adopt a particular conformation so that it can coordinate to the
metal ion, the internal potential energy increases. Table 3 lists the
potential energy increase of 4 possible structures for the MLH−1
complex. In the previous section, it was argued that structure 4(e)
was unlikely because of the large chelate ring. This structure has
the highest internal strain energy. Also, a change in coordination
from amide carbonyl to amide nitrogen 4(f) to 4(g) increases the
torsion angle strain by 12 kcal mol−1. Most of this strain energy
comes from distortion of the amide bond planarity as shown in
Fig. 4. The most likely structure was postulated to be 4(h). This
structure has the 2nd highest strain energy which comes from bond
angle and torsion angle distortion. However, in this structure the
ligand is also quadradentate, while in the other 3 isomers it is
tridentate. In order to form then, the bond strength of the metal
amine bond must more than compensate for the increase in strain
energy. Using the esff force field it is possible to calculate the
internal energy of the complex. When the metal is included in the
calculation then structure 4(h) does indeed have the lowest internal
energy.
Fig. 7 P.m.i. curves for (a) Cu(II) with L1, L6 and cageL, and (b) cageL
with Cu(II), Zn(II) and Ca(II).
Superoxide dismutase mimetic activity studies
The beneficial role of copper in minimizing inflammation has
been attributed to its redox activity, particularly the ability
of copper in such enzymes as superoxide dismutase (SOD) to
remove the highly pro-inflammatory superoxide anion radical.52,53
Along with hydrogen peroxide and hydroxyl radical, superoxide
has been implicated in oxidative damage phenomena related to
aging, inflammation and post-ischaemic injury via reperfusion.53,54
Copper-zinc superoxide (CuZn-SOD) is an enzyme that catalyses
the dismutation of the superoxide anion radical to molecular
oxygen, water and/or hydrogen peroxide. A number of copper(II)
complexes including those of polypeptides have been reported to
exhibit SOD-mimic activity and thus are viewed as alternative
human therapeutics to remove pro-inflammatory superoxide an-
ion radical in vivo.52,55 We investigated the SOD activity of the
Blood plasma simulation
The design of copper based anti-inflammatory drugs is based,
amongst other things, upon the ability of the ligand to mobilize
copper in vivo as reflected by a high plasma mobilizing index
(p.m.i.). P.m.i. of a ligand for a metal ion is defined as the
percentage increase in the total concentration of low molecular
mass complexes of the metal ion caused by the ligand. The
calculation of p.m.i. takes into account competition between the
ligand and all the endogenous metal ions and low molecular mass
ligands present in blood plasma. In calculating the p.m.i curves use
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 1140–1149 | 1145
©

View Article Online
[Cu(II)-cageLH−1] complex using the NBT assay52,56,57 by mea-
suring the concentration of the complex required to reduce
diformazan formation by 50% (also termed as the IC50 value).
An IC50 value of 233 lmol dm−3 for the [Cu(II)-cageLH−1] species
was determined. This value is much higher than the IC50 of
0.011 lmol dm−3 for the native CuZn-SOD.58 The high IC50 value
exhibited by the [Cu(II)-cageLH−1] complex is indicative of its very
low SOD mimetic activity. This may be due to the unavailability
of the binding sites for the superoxide anion radical to coordinate
to the metal ion.
IC50 values of Cu(II) complexes of amino acids such as tyrosine
(45 lmol dm−3) and lysine (86 lmol dm−3) and, non-steroidal
anti-inflammatory drugs (NSAIDs) such as salicylates (2–28
lmol dm−3) have been reported.59,60 These values are also lower
than that obtained in this study. However, we have previously re-
ported IC50 values of 58 and 110 lmol dm−3 for related ligands.12,13
to the ML species on the basis of the distribution curves (Fig. 3)
for the Cu(II)–cageL system. The log Poct/aq at the physiological
pH 7.4 is −1.65, can be approximated as the value for MLH−1
species. Positive log Poct/aq values in the range 0.618 to 4.128 of
some anti-inflammatory drugs have been reported.63 This suggests
that for a drug (complex) to be reasonably lipophilic, the log Poct/aq
value must be at least 0.618. However, several workers64–66 studied
64Cu-labeled complexes of log Poct/aq values in the range −1.60 to
−3.02 and therefore, our results fall within this region.
Although the results for the partition coefficient studies suggest
that the copper(II) complexes of cageL are largely hydrophilic,
at least 2.9% of the Cu(II)–cageL complexes is extracted into
the organic phase of the octanol–water mixture at physiological
pH. The contributing factors to the hydrophilicity of these
complexes may be the presence of coordinated water molecules,
hydrogen bonding and the overall charge of the MLH, ML and
MLH−1 complexes.
1-octanol–water partition coefficient studies
Bio-distribution experiments on mice
One important aspect upon which biological activity depends
is the ability of the drug to reach the target area. For dermal
absorption transport across the skin and bio-membranes in
general is a process of limited diffusion governed by the drug’s
chemical and physical properties such as lipophilicity and protein
binding.61,62 For this reason, this study also investigated the
octanol/water partition coefficients of copper complexes with a
view of establishing whether these species can be administered
transdermally. The octanol/water mixture was used as a bio-phase
model of the membrane.
Based on the aforementioned in vitro results, it was deemed
necessary to perform bio-distribution experiments. Since the body
is a dynamic system, the bio-distribution was measured as a
function of time. Table 4 shows the bio-distribution of 64Cu-
labelled Cu(II)–cageL and CuCl2 at 6 and 24 h post-injection.
The results revealed an initial rapid clearance of [64Cu]Cu(II)–
cageL complex from the blood and rapid uptake by the liver.
The high uptake by the liver is due to the fact that copper storage
and metabolism occurs in this organ. The Cu(II)–cageL complex
has a much longer biological half-life than the copper complexes
of L1, L6 and L7. The [64Cu]CuCl2 used as a control, was rapidly
excreted via hepatobiliary and renal routes. The activity retention
in the body for the [64Cu]Cu(II)-cageL system is about 50% dose as
compared to 20% dose for the control. Such activity accumulation
and retention over 24 h in the body is encouraging and merits
evaluation of this copper chelating agent for possible use in
Fig. 8 shows the logarithm of the partition coefficients (log
P
oct/aq) of Cu(II)–cageL complexes plotted as a function of
pH. Observed in Fig. 8 is the fact the all log Poct/aq values
are negative, thus indicating that these complexes are largely
hydrophilic. The low log Poct/aq values in the acidic pH range 3–4
are due to the high concentration of hydrated [Cu(OH2)6]2+ in this
region. Complexation of Cu(II) with cageL begins at pH 2.5 with
formation of MLH species predominating at pH 5.1 thus giving a
log Poct/aq value of −2.44. A log Poct/aq value of −1.82 is assigned
Table 4 % Dose per organ and per gram tissue (in brackets) for bio-
distribution of [64Cu]Cu(II)–cageL species (mean std. dev., n = 3)
[64Cu]Cu(II)–cageL
[64Cu]CuCl2
Organ
Blood
Carcass
Head
6 h 24 h
6 h
24 h
0.86 0.24
1.07 0.17
1.86 0.14
0.80 0.23
(0.87 0.10) (1.49 0.25) (0.62 0.17) (0.71 0.22)
6.94 0.79 13.25 2.87 7.85 1.87 11.35 1.77
(0.61 0.05) (1.16 0.35) (0.63 0.17) (0.95 0.20)
2.00 0.26 3.72 0.21 1.94 0.04 7.56 2.74
(0.63 0.08) (1.13 0.17) (0.59 0.02) (2.60 1.09)
0.28 0.04 0.44 0.04 0.19 0.02 0.21 0.07
(1.78 0.19) (3.55 0.38) (1.56 0.15) (1.73 0.43)
Heart
Intestines 16.80 4.97 25.34 5.61 22.17 2.87 11.61 6.94
(5.84 2.00) (9.09 4.09) (8.14 1.12) (3.95 1.86)
Kidney
1.15 0.09
1.94 0.01
1.04 0.15
0.96 0.42
(3.18 0.17) (5.10 0.20) (3.09 0.41) (3.03 1.03)
67.89 5.42 33.57 7.56 63.69 5.06 11.39 7.11
(56.36 1.65) (30.69 7.56) (56.20 0.45) (11.32 0.64)
Liver
Lung
0.61 0.07
(3.33 0.41) (5.74 0.81) (1.85 0.25) (2.27 0.34)
0.07 0.01 0.15 0.09 0.06 0.04 0.04 0.03
(0.44 0.10) (0.90 0.18) (0.42 0.13) (0.25 0.10)
0.28 0.05 0.44 0.11 0.23 0.04 0.12 0.02
(2.30 0.13) (3.86 0.38) (1.73 0.13) (1.32 0.11)
2.89 1.08 18.34 2.52 1.50 1.29 55.44 8.72
0.96 0.04
0.53 0.08
0.44 0.18
Muscle
Spleen
Urine
Fig. 8 Cu(II)–cageL partition coefficients measured at 25 ◦C and an ionic
strength of 0.15 mol dm−3 (Cl−) as a function of pH.
1146 | Dalton Trans., 2007, 1140–1149
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
Table 5 % Dose per organ and per gram tissue for the bio-distribution
of the dermally absorbed [64Cu]Cu(II)–cageL complexes (mean std. dev.,
n = 4) 24 h post-dosing
in deionised water and extracted with ethyl acetate. The organic
extract was concentrated in vacuo to yield the diester (0.550 g) as
a light yellow oil. 1H NMR [CDCl3, 400 MHz]: dH 2.72 (m, 2H),
2.61 (m, 2H), 2.38 (m, 2H), 1.47 (d, H4a, J 10.44 Hz), 1.83 (d, H4s,
J 10.44 Hz), 2.64 (m, 2H), 2.79 (s, 2H), 4.10 (q, 2H, H16, H21, J
7.142), 1.22 (t, 2H); 13C NMR [CDCl3, 100 MHz]: dc 48.4 (d), 41.8
(d), 44.5 (d), 43.3 (t), 92.8 (s), 59.1 (d), 92.8 (s), 38.4 (t), 170.4 (s),
60.4 (t), 14.1 (q), and elemental analysis calculated for C19H24O5:
C, 68.66; H, 7.28%: found C, 68.76, H, 7.31%.
[64Cu]Cu(II)-cageL
[64Cu]CuCl2
Organ
Per organ
Per gram
Per organ
Per gram
Blood
Carcass
Head
0.21 0.03 0.19 0.05
10.50 2.88 0.96 0.27 14.20 10.2
8.63 2.66 3.08 0.91
0.04 0.02 0.48 0.26
2.25 0.08
1.50 0.23
0.89 0.59
1.02 0.20
2.10 0.43
6.27 0.70
3.61 0.63
3.34 0.67
0.37 0.07
Heart
Intestines 10.83 6.80 4.00 2.59 20.92 2.53
Kidney
Liver
Synthesis of 3,5-diaminodiamido-4-
0.22 0.17 0.85 0.60
1.57 1.00 1.52 1.05 18.48 3.16 13.80 2.52
1.74 0.27
oxahexacyclo[5.4.1.02,6.03,10.05.9.08,11]dodecane 7
Lung
0.08 0.06 0.81 0.50
0.04 0.02 0.52 0.53
0.06 0.03 0.79 0.42
67.82 9.39
0.67 0.12
0.13 0.03
0.19 0.02
37.80 13.4
3.04 0.52
0.58 0.27
2.11 0.35
Muscle
Spleen
Urine
Freshly distilled ethylenediamine (∼50 ml) was added to the diester
6 (0.500 g, 1.50 × 10−3 mol) and the reaction mixture was refluxed
for 20 h under nitrogen. The excess ethylenediamine was removed
in vacuo. The product 7 was extracted with tetrahydrofuran and
concentrated in vacuo to give a light brown oil (0.500 g, 92%). 1H
NMR [CDCl3, 400 MHz]: dH 2.53 (m, 2H), 2.56 (m, 2H), 2.33
(m), 1.40 (d, H4a, J 10.44 Hz), 1.75 (d, H4s, J 10.44 Hz), 2.50
(m, 2H,), 2.58 (s, 2H), 7.19 (2H), 3.13 (t, 2H), 2.64 (t, 2H), 1.90
(2H), 13C NMR [CDCl3, 100 MHz]: dc 47.9 (d), 41.2 (d), 43.9
(d), 43.2 (t), 93.6 (s), 58.4 (d), 39.3 (t), 170.05 (s), 41.64 (t), 40.95
(t), and elemental analysis calculated for C19H28N4O3: C, 63.33;
H, 7.77, N, 15.55%: found C, 63.76, H, 7.77, N, 15.96%. The
NMR spectra were recorded on a Varian Unity Inova 400 MHz
spectrometer. Elemental analyses were obtained from a Leco
CHNS 932 instrument. The purity of cageL was confirmed by
acid–base titration and was found to be greater than 95%.
chemotherapy and diagnosis. However, this complex is unlikely
to penetrate the blood–brain barrier in view of its octanol–water
partition coefficient.
Anderson et al.67,68 have investigated thermodynamically and
kinetically stable 64Cu-labelled macrocyclic complexes with differ-
ent formal charges as biofunctional chelators. All the complexes
were observed to be rapidly cleared from circulation and positively
charged complexes had higher uptake and slow clearance. Other
reporters63–65 have shown that positively charged 64Cu complexes of
the macrocyclic ligands exhibited rapid uptake in the liver and kid-
neys with slow clearance, whereas the negatively charged and neu-
tral complexes cleared rapidly from all tissues by the renal route.
Since the preferred route of administration is as a cream
via the skin, bio-distribution studies were also performed using
dermal absorption. Results of cageL and CuCl2 control are given
in Table 5. Of note here is that the same bio-distribution as
intravenous injection is not obtained. With dermal absorption
less activity is excreted in the urine and more activity is retained
in the body.
Potentiometric measurements
All solutions for potentiometry were prepared in glass distilled
water which had been boiled to remove dissolved carbon dioxide.
Recrystallized NaCl was used as background electrolyte at an ionic
strength of 0.15 mol dm−3 (Cl−). Other reagents, copper(II) chloride
dihydrate, zinc(II) chloride dihydrate, HCl, NaOH and EDTA
(Merck) were commercially available and of analytical grade.
These were used without purification. The 0.1 mol dm−3 solutions
of NaOH and HCl were prepared from Merck Titrisol ampoules
and standardized by titrating with potassium hydrogen phthalate
(KHP) and sodium tetraborate decahydrate (Borax), respectively,
using standard methods of Vogel. The NaOH solutions prepared
were further standardized against standard HCl solution. Po-
tentiometric data were collected, under an inert atmosphere of
purified nitrogen, at 25 ◦C, using an automatic titration procedure
described previously.10,12 The data were analysed using the ESTA
suite of computer programs.39 The slope of the glass electrode
(Metrohm 6.0222.100) was determined from a buffer line and the
final E◦ determined in situ. Carbonate contamination of the titrant
solutions was checked by titration using the Gran method.69 Under
these conditions the water ionization constant was determined to
be 13.73 log units.
Conclusions
In this study we have shown that incorporation of the pentacyclo-
undecane cage derivative into a diaminodiamide ligand does not
affect the ability of the ligand to complex copper(II). Indeed,
the rigidity of the cage appears to increase the stability of the
complexes. This is presumably because the ligand is pre-formed in
the correct conformation for metal binding. The bio-distribution
and dermal absorption studies show that Cu(II) complexes of
cageL survive in vivo.
Experimental
Synthesis of the novel oxahexacyclododecane diester 6
To a suspension of the diacid 4 (0.552 g, 2.00 × 10−3 mol) in dry
dichloromethane (10 ml) was added oxalyl chloride (2 ml, 2.30 ×
10−2 mol) drop-wise with stirring under nitrogen. The solution was
stirred for a further 5 h after the reaction became homogeneous
and concentrated under vacuo to yield the diacyl choride 5 as
a colourless oil. Excess ethanol ∼50 ml was then added to the
diacyl chloride and the mixture was stirred under nitrogen for
18 h. The reaction mixture was concentrated in vacuo, dissolved
UV-visible and nuclear magnetic resonance spectroscopy
UV-visible spectra were recorded at 25 ◦C on a Hewlett Packard
8425A Diode Array in the wavelength region between 340 and
800 nm. Aqueous solutions containing 1 : 2 ratio Cu(II) : cageL
were taken over pH range 2–11. Small amounts of 0.1 mol dm−3
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 1140–1149 | 1147
©

View Article Online
NaOH and HCl were used to adjust the pH during the titration.
NMR samples, 0.015 mol dm−3 of cageL solution and a solution
of 1 : 2 molar ratio of the Cu(II) to cageL were prepared in D2O.
1H NMR spectra were recorded on a Varian Unity Plus 400 MHz
instrument. The pH of the solutions was adjusted using NaOD or
DCl. The pH values were not corrected for the isotope effect.
an area of approximately 2 × 2 cm on the dorsal side of all test
animals 15 h prior to dosing. The animals were anaesthetized
with saline solution of ketamine/xylazine following clipping and
a small tubing ring (1.2 cm diameter and 0.5 cm height) was fixed to
their backs using cyanoacrylate adhesive. At 24 h post-dosing time
point, groups of mice (four at a time) were anaesthetized by carbon
monoxide inhalation and the procedure for removing samples
followed that of bio-distribution experiments. The percentages
of radio-activity dose per organ and/or per gram (g) tissue were
computed from the mean organ weights after correction for radio-
activity decay.
The bio-distribution and dermal absorption studies on mice
were approved by the Research Animals Ethics Committee of the
University of Cape Town (permission number 005/039). The au-
thority to possess and use radioactive nuclide 64Cu was granted by
the university’s radiation protection and health safety committee
in conjunction with the Department of Health (authority number
33/01/0327).
Molecular mechanics
Molecular mechanics calculations were performed using the Dis-
cover 3 module of the Accelrys life sciences molecular modelling
software, InsightII.49 For the free ligand, the cvff force field was
used while for the metal complex the esff force field was used.
Blood plasma simulation, superoxide dismutase mimic activity and
octanol–water partition coefficient studies
Blood plasma modelling was carried out by incorporation of the
formation constants determined in this study into a blood plasma
model70,71 consisting of data for 10 metal ions and more than 5000
ligands. This enlarged database was efficiently and conveniently
interrogated by the Evaluation of Constituent Concentrations in
Large Equilibrium Systems (ECCLES) computer program to yield
results pertaining to the influence of the chelating agents on the
equilibria in terms of the plasma mobilizing index (p.m.i) of each
agent. P.m.i, defined as the ability to move metal from a protein
bound form to a low-molecular weight form, can be represented
by the following expression;
Caution: 64Cu decays by positrons emissions, which annihilate
to produce tissue penetrating high energy c-radiation. Therefore,
it is advisable to use much less activity of 64Cu and perform all
experiments behind lead bricks inside a fume cupboard.
Acknowledgements
The authors thank the Botswana Government and the Coun-
cils of the Universities of Cape Town and KwaZulu-Natal for
financial support and a scholarship to S. O. We are also grateful
to Dr J. R. Zeevaart and D. Jansen (South African Nuclear
Energy Cooperation (Ltd) for supplying us with 64CuCl2 and,
J. Boniaszczuk (Nuclear Medicine Dept), J. Visser (Animal Unit),
Dr J. N. Zvimba and T. Jackson for technical assistance with
animal experiments.
(total concentration of low molecular weight metal complex species in the presence of drug)
-
-
p.m.i =
(total concentration of low molecular weight metal complex species in normal plasma)
-
-
SOD mimetic activity of the Cu(II)–cageL system was determined
12
using nitroblue tetrazolium (NBT) assay as described earlier.
Partition coefficients were measured at 25 ◦C and an ionic strength
of 0.15 mol dm−3 (Cl−) using the shake flask method reported
12
previously.
Bio-distribution and dermal absorption studies
References
64Cu was prepared by neutron irradiation of 1.7–2.5 mg copper(II)
oxide (99.9999%, Aldrich Chem. Co., WI, USA) for 24 h at a
thermal neutron flux of 1.0 × 1014 neutrons cm−2 s−1 in Safari-1
Research Reactor. The targets were dissolved in 1.0 ml 1.0M HCl
and diluted to 3.0 ml with 1.0 M HCl. Typical specific activities of
∼600 MBq mg−1 target material were obtained.
1 D. J. MacCarthy, Arthritis and Allied Conditions, A textbook of
Rheumatology, 11th edn, Lea and Febiger, Philadelphia, 1989, p. 507.
2 A. Aeschlinmann, B. R. Simmen and B. A. Michel, in; Rheumatoid
Arthritis, ed. H. Baumgartner, J. Dvorak, D. Grob, U. Munzinger and
B. R. Simmen, Thieme Medical Publishers, Inc., New York, 1985.
3 R. F. Wilkens and S. L. Dahl, Therapeutic Controversies in the
Rheumatic Disease, 1987, Grune and Stratton, Inc., Orlando.
4 J. R. J. Sorenson, Prog. Med. Chem., 1976, 19, 135.
5 J. R. J. Sorenson, Prog. Med. Chem., 1978, 15, 211.
6 J. R. J. Sorenson, Prog. Med. Chem., 1989, 26, 437.
7 J. R. J. Sorenson, in Metal ions in biological Systems, Inorganic Drugs
in Deficiency and Disease, ed. H. Sigel, Marcel Decker, Inc., 1982, vol.
14, p. 77.
64Cu-labelled Cu(II)–cageL complex was prepared by spiking
20 cm3 of Cu2+ (0.001 mol dm−3)–cageL (0.003 mol dm−3) solution
with 10 mCi 64CuCl2(aq) (Radiochemical purity >98%). Bio-
distribution studies were carried out by injecting 6–8 weeks old
Balb/c mice with 0.2 cm3 of 5 lCi [64Cu]Cu(II)–cageL solution and
64CuCl2(aq) in control experiments. At 6 and 24 h post-injection
time points, groups of mice (three at a time) were anaesthetized
by carbon monoxide inhalation and aliquots of blood were taken
from the heart. Various samples were removed, rinsed with NaCl
(0.9%), dried and then weighed. The radioactivity in these samples
as well as in 10 cm3 aliquots of urine extract were counted using a
Minaxi Auto-gamma 5000 Series counter with the window set at
340–540 KeV. The dermal absorption experiments were performed
by applying 0.2 cm3 of 5 lCi [64Cu]Cu(II)–cageL and 64CuCl2(aq)
dosing solutions on the enclosed skin of the anterior dorsal side
of the test animals. The application site was prepared by clipping
8 G. E. Jackson and M. J. Kelly, J. Chem. Soc., Dalton Trans., 1989, 2429.
9 G. E. Jackson, P. W. Linder and A. Voye, J. Chem. Soc., Dalton Trans.,
1996, 4605.
10 G. E. Jackson, L. Gama-Mkhonta, A. Voye and M. J. Kelly, J. Inorg.
Biochem., 2000, 79, 147.
11 S. Odisitse, MSc Thesis, University of Cape Town, 2003.
12 T. E. Nomkoko, G. E. Jackson, B. S. Nakani, K. A. Louw and J. R.
Zeevaart, Dalton Trans., 2004, 741.
13 T. E. Nomkoko, G. E. Jackson and B. S. Nakani, Dalton Trans., 2004,
1432.
14 G. E. Jackson, in Handbook of Metal–Ligand Interactions in Biological
Fluids, Bioinorganic Chemistry, ed. G. Berthon, Marcel Dekker, Inc.,
New York, 1995, vol. 2, p. 1228.
15 A. P. Marchand, Chem. Rev., 1989, 89, 1011.
16 G. W. Griffin and A. P. Marchand, Chem. Rev., 1989, 89, 997.
1148 | Dalton Trans., 2007, 1140–1149
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
17 A. P. Marchand, in Advances in Theoretically Interesting Molecules, ed.
R. P. Thummel, JAI Press, Greenwich, CT, vol. 1, 1989, 357.
18 D. W. Oliver, D. G. Dekker, F. O. Snykers and T. G. Fourie, J. Med.
Chem., 1991, 34, 851.
45 F. J. Quaeyhaegens, H. O. Desseyn, S. P. Perlepes, J. C. Plakatouras, B.
Bracke and A. T. H. Lenstra, Transition Met. Chem., 1991, 16, 92.
46 A. Zuberbuhler and Th. Kaden, Helv. Chim. Acta, 1968, 51, 1805.
47 H. Ojima and K. Nonoyama, Coord. Chem. Rev., 1988, 921, 857.
48 R. M. Lewis, G. H. Nancollas and P. Coppens, Inorg. Chem., 1972, 11,
1371.
19 D. W. Oliver, D. G. Dekker and F. O. Snyckers, Eur. J. Med. Chem.,
1991, 26, 375.
20 D. W. Oliver, D. G. Dekker and F. O. Snyckers, Drug Res., 1991, 41,
49 InsightII, System Guide, issued by Accelrys, Cambridge, UK, July
2005.
545.
21 Y. Inamoto, K. Aiyami, N. Tasaishi and Y. J. Fujikura, J. Med. Chem.,
1967, 19, 536.
22 Y. Inamoto, K. Aiyami, T. Kadono, H. Makayama, A. Takatsuki and
G. Tumura, J. Med. Chem., 1977, 20, 1371.
50 S. Shia, L. Yan, Y. Yang, J. Fisher-Shaulsky and T. Thacher, J. Comput.
Chem., 2003, 24(9), 1059.
51 G. E. Jackson, P. M. May and D. R. Williams, J. Inorg. Nucl. Chem.,
1978, 40, 1227.
23 J. C. A. Boeyens, L. M. Cook, G. N. Nelson and T. G. Fourie,
S. Afr. J. Chem., 1994, 47(2), 72.
52 M. Gonzalez-Alvarez, G. Alzuet, J. Borras, L. C. Agudo and J. M.
Montego-Bernado, J. Biol. Inorg. Chem., 2003, 8, 112.
53 I. Fridovich, J. Biol. Chem., 1989, 264(14), 7761.
54 V. Pelmenschikov and P. E. M. Siegbahn, Inorg. Chem., 2005, 44(9),
3311.
55 R. N. Patel and K. B. Pandega, J. Inorg. Biochem., 1998, 72, 109.
56 C. Auclair and E. Voisin, in CRC Handbook of Methods for Oxygen
Radical research, ed. R. A. Greenwald, 1985, CRC Press, Boca Raton,
FL.
57 I. Fridovich, in CRC Handbook of Methods for Oxygen Radical
research, ed. R. A. Greenwald, 1985, CRC Press, Boca Raton, FL.
58 H.-L. Zhu, L.-M. Zheng, D.-G. Fu, X.-Y. Huang, M.-F. Wu and W.-X.
Tang, J. Inorg. Biochem., 1998, 70, 211.
59 C. T. Dillion, T. W. Hambley, B. J. Kennedy, P. A. Lay, Q. Zhou, N. M.
Davies, J. R. Biffin and H. L. Regtop, Chem. Res. Toxicol., 2003, 16,
28.
24 R. S. Schab, A. C. England, D. C. Poskanzer and R. R. Young, J. Am.
Med. Assoc., 1969, 208, 1168.
25 J. L. Neumeyer, in Principles of Medicinal Chemistry, ed. W. O. Foye,
Lea and Febiger, London, 1989, p. 223.
26 K. Gerzon and D. Kou, J. Med. Chem., 1967, 10, 189.
27 K. Gerzon, R. J. Kraay and R. T. Rapala, J. Med. Chem., 1965, 8, 580.
28 A. N. Voldeng, C. A. Bradley, R. D. Kee, E. L. Knight and F. L. Melder,
J. Pharm. Sci., 1968, 57, 1053.
29 K. B. Brookes, P. W. Hickmott, K. K. Jutle and C. A. Schreyer,
S. Afr. J. Chem., 1992, 45, 8.
30 W. J. Geldenhuys, S. F. Malan, J. R. Bloomquist, A. P. Marchand and
C. J. Van der Schyf, Med. Res. Rev., 2005, 25(1), 21.
31 R. C. Cookson, E. Grundwell, R. R. Hill and J. Hudec, J. Chem. Soc.,
1964, 3062.
32 A. P. Marchand, Z. Huang, Z. Chen, H. K. Hariprakasha, I. N. N.
Namboothiri, J. S. Brodbelt and M. L. Reyzer, J. Heterocycl. Chem.,
2001, 38, 1361.
33 T. Govender, H. K. Hariprakasha, H. G. Kruger and A. P. Marchand,
Tetrahedron: Asymmetry, 2003, 14, 1553.
34 A. P. Marchand, K. A. Kumar, A. S. McKim, K. M. Majerski and G.
Kragol, Tetrahedron, 1997, 10, 3467.
60 J. E. Weder, C. T. Dillion, T. W. Hambley, B. J. Kennedy, P. A. Lay,
Q. Zhou, N. M. Davies, J. R. Biffin, H. L. Regtop and N. M. Davies,
Coord. Chem. Rev., 2002, 232, 95.
61 S. K. Poole and C. F. Poole, J. Chromatogr., B, 2003, 797, 3.
62 G. Klopman and H. Zhu, Mini-Rev. Med. Chem., 2005, 5, 127.
63 F. Pehourcq, M. Matoga, C. Jarry and B. Bannwarth, J. Liq. Chro-
matogr. Relat. Technol., 2001, 24(14), 2177.
35 G. A. Boyle, T. Govender, H. G. Kruger and G. E. M. Maguire,
Tetrahedron: Asymmetry, 2004, 5(17), 2661.
36 T. Govender, H. K. Hariprakasha, H. G. Kruger and A. P. Marchand,
Tetrahedron: Asymmetry, 2003, 14(11), 1553.
64 X. Sun, M. Wuest, Z. Kovacs, A. D. Sherry, R. Motekaitis, Z. Wang,
A. E. Martell, M. J. Welch and C. J. Anderson, J. Biol. Inorg. Chem.,
2003, 8, 217.
65 X. Sun, M. Wuest, G. R. Weisman, E. H. Wong, D. P. Reed, C. A.
Boswell, R. J. Motekaitis, A. E. Martell, M. J. Welch and C. J. Anderson,
J. Med. Chem., 2002, 45, 469.
66 A. B. Packard, J. F. Kronauge, P. J. Day and S. T. Treves, Nucl. Med.
Biol., 1998, 25, 531.
67 T. M. Jones-Wilson, K. A. Deal, C. J. Anderson, D. W. McCarthy, Z.
Kovac, R. J. Motekaitis, A. D. Sherry, A. E. Martell and M. J. Welch,
Nucl. Med. Biol., 1998, 25, 523.
68 C. S. Cutler, M. Wuest, C. J. Anderson, D. E. Reichert, Y. Sun, A. E.
Martell and M. J. Welch, Nucl. Med. Biol., 2000, 27, 375.
69 G. Gran, Analyst, 1952, 77, 309.
70 P. M. May, P. W. Linder and D. R. Williams, J. Chem. Soc., Dalton
Trans., 1977, 548.
71 G. E. Jackson and M. J. Kelly, Inorg. Chim. Acta, 1988, 152, 215–217.
37 J. C. Rossotti and H. Rossotti, J. Chem. Educ., 1965, 42, 375.
38 G. E. Jackson and M. J. Kelly, J. Chem. Soc., Dalton Trans., 1989, 2429.
39 K. Murray and P. M. May, ESTA (Equilibrium Simulation for Titration
Analysis) manual, Version. 3, UWIST, Cardiff, Wales, 1989.
40 D. R. Williams, The metals of Life, van Nonstrand Reinhold Company,
London, 1971.
41 H. H. Jaffe and M. Orchin, Theory and Application of Ultraviolet
Spectroscopy, John Wiley & Sons, New York, 1962.
42 A. E. Martel, R. M. Smith and R. J. Motekaitis, NIST Critical Stability
Constants of Metal Complexes Database, Texas A&M University
College Station, TX, 1993.
43 E. J. Billo, Inorg. Nucl. Chem. Lett., 1974, 10, 613.
44 H. Sigel and A. E. Martell, Chem. Rev., 1982, 82, 385.
This journal is
The Royal Society of Chemistry 2007
Dalton Trans., 2007, 1140–1149 | 1149
©