The interaction of silver(II) complexes with biological macromolecules and antioxidants

Article abstract of DOI:10.1007/s10534-019-00198-0

Silver is widely used for its antimicrobial properties, but microbial resistance to heavy metals is increasing. Silver(II) compounds are more oxidizing and therefore have the potential to overcome resistance via extensive attack on cellular components, but have traditionally been hard to stabilize for biological applications. Here, the high oxidation state cation was stabilised using pyridinecarboxylate ligands, of which the 2,6-dicarboxypyridine Ag(II) complex (Ag2,6P) was found to have the best tractability. This complex was found to be more stable in phosphate buffer than DMSO, allowing studies of its interaction with water soluble antioxidants and biological macromolecules, with the aim of demonstrating its potential to oxidize them, as well as determining the reaction products. Spectrophotometric analysis showed that Ag2,6P was rapidly reduced by the antioxidants glutathione, ascorbic acid and vitamin E; the unsaturated lipids arachidonic and linoleic acids, model carbohydrate β-cyclodextrin, and protein cytochrome c also reacted readily. Analysis of the reaction with glutathione by NMR and electrospray mass spectrometry confirmed that the glutathione was oxidized to the disulfide form. Mass spectrometry also clearly showed the addition of multiple oxygen atoms to the unsaturated fatty acids, suggesting a radical mechanism, and cross-linking of linoleic acid was observed. The seven hydroxyl groups of β-cyclodextrin were found to be completely oxidized to the corresponding carboxylates. Treatment of cytochrome c with Ag2,6P led to protein aggregation and fragmentation, and dose-dependent oxidative damage was demonstrated by oxyblotting. Thus Ag2,6P was found to be highly oxidizing to a wide variety of polar and nonpolar biological molecules.

Full text of DOI:10.1007/s10534-019-00198-0

Biometals
https://doi.org/10.1007/s10534-019-00198-0(0123456789(). -, volV
(
)
0
1
2
3
4
5
8
9
(
)
.
-
v
o
l
V
)
The interaction of silver(II) complexes with biological
macromolecules and antioxidants
.
Katherine D. Trotter Olawale Owojaiye Stuart P. Meredith
.
.
.
Pat E. Keating Mark D. Spicer John Reglinski Corinne M. Spickett
.
.
Received: 15 February 2019 / Accepted: 30 April 2019
Ó The Author(s) 2019
Abstract Silver is widely used for its antimicrobial
properties, but microbial resistance to heavy metals is
increasing. Silver(II) compounds are more oxidizing
and therefore have the potential to overcome resis-
tance via extensive attack on cellular components, but
have traditionally been hard to stabilize for biological
applications. Here, the high oxidation state cation was
stabilised using pyridinecarboxylate ligands, of which
the 2,6-dicarboxypyridine Ag(II) complex (Ag2,6P)
was found to have the best tractability. This complex
was found to be more stable in phosphate buffer than
DMSO, allowing studies of its interaction with water
soluble antioxidants and biological macromolecules,
with the aim of demonstrating its potential to oxidize
them, as well as determining the reaction products.
Spectrophotometric analysis showed that Ag2,6P was
rapidly reduced by the antioxidants glutathione,
ascorbic acid and vitamin E; the unsaturated lipids
arachidonic and linoleic acids, model carbohydrate b-
cyclodextrin, and protein cytochrome c also reacted
readily. Analysis of the reaction with glutathione by
NMR and electrospray mass spectrometry confirmed
that the glutathione was oxidized to the disulfide form.
Mass spectrometry also clearly showed the addition of
multiple oxygen atoms to the unsaturated fatty acids,
suggesting a radical mechanism, and cross-linking of
linoleic acid was observed. The seven hydroxyl groups
of b-cyclodextrin were found to be completely
oxidized to the corresponding carboxylates. Treatment
of cytochrome c with Ag2,6P led to protein aggrega-
tion and fragmentation, and dose-dependent oxidative
damage was demonstrated by oxyblotting. Thus
Ag2,6P was found to be highly oxidizing to a wide
variety of polar and nonpolar biological molecules.
Electronic supplementary material The online version of
this article (https://doi.org/10.1007/s10534-019-00198-0) con-
tains supplementary material, which is available to authorized
users.
K. D. Trotter Á O. Owojaiye Á P. E. Keating Á
M. D. Spicer Á J. Reglinski
Keywords Ag(II) 2,6-dicarboxypyridine Á
Antimicrobial metal Á Glutathione Á Lipid
peroxidation Á Oxidative stress
Department of Pure & Applied Chemistry, Strathclyde
University, 295 Cathedral Street, Glasgow G1 1XL, UK
S. P. Meredith Á C. M. Spickett
School of Life and Health Sciences, Aston University,
Aston Triangle, Birmingham B4 7ET, UK
C. M. Spickett (&)
Strathclyde Institute of Pharmacy and Biomedical
Sciences, Strathclyde University, 161 Cathedral Street,
Glasgow G4 0NR, UK
e-mail: c.m.spickett@aston.ac.uk
123

Biometals
Introduction
Heiman et al. 2015; Vindigni and Surawicz 2015),
development of more powerful formulations of silver
with higher oxidizing potential would be desirable,
especially for external use in disinfection and
cleansing.
The antiseptic and oligodynamic behaviour of silver
has been appreciated for more than 200 years (Marx
and Barillo 2014; Mijnendonckx et al. 2013). How-
ever, even in recent years its application as a biocide
has been limited to the use of a few relatively simple
In addition to Ag0 and Ag?, silver also has two
higher oxidation states that are potentially extremely
powerful oxidants: Ag2? (Ag2? ? e- ? Ag?;
E0 = 2.0 V) and Ag3? (Ag3? ? e- ? Ag2?; Eo
unknown) (Weast 1979). Biocidal silver compounds
such as SSD are all based on compounds of silver in its
less reactive, lower oxidation states (Ag0/Ag?). This
choice, in part, was due to an inability to stabilize and
control silver in its higher oxidation states in early
synthetic studies. However, many of these issues have
been resolved and routes are available for the produc-
tion of a wide range of silver(II) compounds and a
more limited range of silver(III) species (Levason and
Spicer 1987). This opens the door for development of
novel, high oxidation state silver compounds for
antimicrobial disinfection.
´
preparations (Azocar et al. 2014). Many medical items
such as catheters and prostheses are coated with the
elemental form of silver, which prevents colonization
of their surface by pathogens (Hetrick and Schoenfisch
2006). The release of silver cations through slow
dissolution of silver halide is an effective method of
disinfecting water for domestic consumption (Sil-
vestry-Rodriguez et al. 2007). Since the late 1960s,
silver sulfadiazine (SSD, or FlamazineÒ) has been
common in anti-bacterial preparations for the treat-
ment of burns (Atiyeh et al. 2007). Even this
compound is a simple preparation comprising an
antimicrobial cation and anion formulated together as
a weakly bonded complex (Cook and Turner 1975).
Although there has been robust interest in developing
new silver-based anti-infective agents such as silver
nanoparticles (Ag/Ag2O), novel silver complexes and
silver impregnated fabrics (Ag0/Ag?) (Fromm 2013;
Konop et al. 2016; Le Ouay and Stellacci 2015;
Simoncic and Klemencic 2016; Singh et al. 2015),
these are mainly reformulations of silver in its
commonest Ag0/Ag? forms.
Increasing the redox potential of the silver agent is
an effective method of enhancing biocidal activity, as
it limits the effectiveness of antioxidant defence.
Powerful oxidants such as silver(II) can be expected to
irreversibly chemically oxidize a wide range of
functional (sulfhydryl, vicinal diols) and structural
components (unsaturated lipid, proteins, carbohy-
drates) on the surface and inside the microbial cell.
However, a change in oxidation state not only
increases the redox potential, it also changes the
preferred shape of the metal complex. Silver(I) has a
marked preference for tetrahedral geometry, whereas
d9 silver(II) predominantly adopts square planar
geometry. It is known that the geometry a metal
complex adopts can affect its biological activity; for
example, the ability of platinum compounds to interact
with DNA (Rosenberg et al. 1969) and the antimicro-
bial and anticancer activities of various metal com-
plexes (Malik et al. 2018). Metal complexes can be
transported across membranes by passive and active
mechanisms (Martinho et al. 2018), and it has been
reported that specific coordination structures may
occur during active transport; for example, in N-MBD
Cu?-ATPases Cu? adopts a trigonal planar form
(Arguello et al. 2012). Thus, biocides based on
silver(II) could allow an enhanced oxidative attack
and, depending on their geometry, might exert diverse
effects on biological systems. There is a wide range of
The mechanisms of antimicrobial action of these
silver formulations are still not completely under-
stood, but the consensus is that they can act on a
variety of targets, including interactions with bacterial
cell membranes, binding to and inhibition of thiol-
containing proteins, and release of reactive oxygen
species through processes still to be fully elucidated
(Konop et al. 2016). As Ag? is a moderate oxidizing
agent (Ag? ? e- ? Ag; E0 = 0.80 V), only the most
sensitive redox sites are affected by it. Although it has
sometimes been suggested that the multimodal mech-
anism of Ag0/Ag? formulations make resistance less
likely to develop, resistance to heavy metals in general
is well known (Pal et al. 2014), and reports of
resistance to silver compounds are on the increase
(Hanczvikkel et al. 2018; Panacek et al. 2018; Percival
et al. 2005). In view of the fact that antimicrobial
resistance is a major problem worldwide and aggres-
sive microbial species such as E. coli 0157, MRSA,
and C. difficile are on the increase (Brandt et al. 2014;
123

Biometals
simple ligands which can be used to stabilize
silver(II), but the pyridinecarboxylates are an excel-
lent initial choice, as they have been synthesized
previously and mostly form planar complexes (Drew
et al. 1970; Drew et al. 1971; Fowles et al. 1968),
although the silver(II) complex with 2,6-dicar-
boxypyridine has been reported to be octahedral
(Drew et al. 1969). However, their ability to react
with biological molecules and cause oxidative damage
has not been studied previously. Consequently, sil-
ver(II) complexes were prepared using pyridinecar-
boxylates as ligands, and the stability of these
complexes was investigated. The aim of the study
was to determine the effectiveness of the silver(II)
complexes for oxidizing biological antioxidants, lipids
and proteins.
solvent. The resulting solution was subjected imme-
diately to spectrophotometric analysis and the con-
centration of Ag2,6P in solution calculated
retrospectively using molar extinction coefficient
obtained from the reference sample.
Thus although all the experiments were carried out
in duplicate or triplicate, the difficulty of producing
completely dry complex meant that it was impossible
to generate solutions containing exactly the same
concentration of reagents. Consequently, the data
shown are derived from representative experiments.
The stability of bis-(2,6-
dicarboxypyridyl)silver(II) in solution
A solution of bis-(2,6-dicarboxypyridyl)silver(II) was
prepared either in phosphate buffer (0.1 M KH2PO4,
pH 7.0) at 4.47 mM or in DMSO at 18.4 mM. Aliquots
of solutions were transferred immediately to a cuvette
and the absorbance (400–1000 nm) was monitored
over a 2 h period.
Experimental
All reagents were obtained commercially. UV–Vis
spectra were recorded on an Agilent Technologies
Cary 60 UV–Vis spectrophotometer. NMR analysis
was carried out on a Bruker AMX 400 operating at
400 MHz for 1H. Solid reflectance spectra
(400–900 nm) were recorded on a Photonics CCD
array UV–Vis spectrophotometer. Silver(II) com-
plexes of 2-carboxypyridine, 2,3-dicarboxypyridine,
2,4-dicarboxypyridine, 2,5-dicarboxypyridine and
2,6-dicarboxypyridine (Ag2,6P) as were prepared
using literature methods (Drew et al. 1970; Drew
et al. 1971; Fowles et al. 1968).
Reaction of bis-(2,6-dicarboxypyridyl)silver(II)
with glutathione
An 8.5 mM solution of Ag2,6P in phosphate buffer
was prepared and 2.25 mL was transferred to a cuvette
and the visible spectrum (400–900 nm) recorded.
Aliquots (20 lL) of reduced glutathione (GSH) solu-
tion (84.3 mM) in phosphate buffer were added to the
cuvette and the spectrum re-recorded after each
addition until the band (kmax 570 nm) attributed to
silver(II) disappeared (* 120 lL).
The protocol for handling Ag2,6P in solution
To investigate the products of the reaction using
NMR, three solutions of reduced glutathione (6.1 mg
in 1 mL of D2O, 20 mmol) were treated with 8.2 mg
(18.5 mmol), 16.6 mg (36 mmol) or 24.4 mg
(53 mmol) of Ag2,6P respectively. The solutions
were allowed to react overnight and then filtered into
A reference sample of Ag2,6P was prepared using
published methods (Fowles et al. 1968). The sample
was subjected to elemental analysis (found: C 32.77, H
2.68, N 5.85%: expected for Ag2,6PÁ4H2O: C 32.97, H
2.77, N 5.49%), which confirmed the hydration state.
This reference sample was used to calculate the molar
extinction coefficient of Ag2,6P in water (e570, 252/M/
cm; e890, 207/M/cm) and DMSO (e600, 88.6/M/cm).
Ag2,6P was found to decompose slowly with the
natural green/black colour giving way to a white
product. Hence small batches of Ag2,6P were pre-
pared immediately prior to use to avoid problems
associated with degradation, and given amounts of
Ag2,6P were quickly dissolved in a given amount of
1
a 5 mm NMR tube. H NMR spectra were obtained
using a Bruker AVANCE 3 spectrometer operating at
400.12 MHz. Samples were maintained at 300 K
during spectral acquisition. The NMR spectra were
collected using a standard pulse sequence. The free
induction decay was generated by a 3.13 ls pulse
width corresponding to a 30o pulse. Each data set (4 k
scans; no water suppression) was collected in 32 k of
memory. A 1 Hz line broadening function was applied
123

Biometals
Reaction of bis-(2,6-dicarboxypyridyl)silver(II)
before Fourier transformation to reduce the effect of
the baseline noise.
with b-cyclodextrin
The reaction with b-cyclodextrin (25 mmol/L) with
Ag2,6P was tested essentially as for ascorbic acid
described above, except that the reaction was carried
out in a sample bottle and 200 lL aliquots of b-
cyclodextrin were added sequentially until the band at
570 nm had been extinguished (after addition of
2.0 mL).
Reaction of bis-(2,6-dicarboxypyridyl)silver(II)
with ascorbic acid
The reaction of bis-(2,6-dicarboxypyridyl)silver(II)
(7.6 mM) with ascorbic acid (144 mM) was investi-
gated essentially as described for glutathione. Aliquots
(20 lL) of the ascorbate solution were added to a
cuvette containing 2.25 mL of Ag2,6P and the spec-
trum re-recorded until the band (kmax 570 nm) disap-
peared, which corresponded to the addition of 80 lL.
For analysis by mass spectrometry, aqueous solu-
tions (2 mL) of b-cyclodextrin containing 100 mg
(88 lmoles) were treated with Ag2,6P (200 mg,
0.44 mmol). The solutions were allowed to react
overnight and then filtered before analysis as
described below.
Reaction of bis-(2,6-dicarboxypyridyl)silver(II)
with vitamin E (a-tocopherol)
Electrospray mass spectrometric analysis of small
molecules
The visible spectrum (400–900 nm) of 2.25 mL of
9.1 mM
bis-(2,6-dicarboxypyridyl)silver(II)
in
DMSO was recorded before and after addition of
10 lL aliquots of 54.3 mM a-tocopherol in DMSO,
with loss of absorbance at 620 nm after addition of
70 lL. Due to the competition between the concurrent
reactions of Ag2,6P with a-tocopherol and Ag2,6P
with DMSO, a definite end point cannot be given for
the reaction of Ag2,6P with a-tocopherol (vide infra).
Electrospray mass spectra were recorded using an
Agilent 6130 (dual source). Samples of glutathione,
vitamin E, linoleic acid, and arachidonic acid prepared
as described above were diluted in methanol, while b-
cyclodextrin was diluted in 50:50 acetonitrile con-
taining 0.1% formic acid:water and introduced into the
instrument with an infusion rate of 0.2 mL/min using
methanol. Spectra were acquired with the following
parameters: ionization mode, MM-ES ? APCI, -ve
ionization; source temperature, 300 °C; Voltage,
4000 V; Curtain gas flow rate, 12 L/min; m/z range
50–2000. Spectra were typically acquired for 30 s and
averaged. The MS data were analysed using Agilent
Chemstation.
Reaction of bis-(2,6-dicarboxypyridyl)silver(II)
with linoleic and arachidonic acid
A 2.25 mL aliquot of 19.5 mM bis-(2,6-dicar-
boxypyridyl)silver(II) in DMSO in a cuvette was
reacted sequentially with 10 lL aliquots of 91.1 mM
sodium linoleate or 98.7 mM sodium arachidonate,
both prepared in DMSO. The reaction was monitored
by recording the visible spectrum until the band (kmax
620 nm) had disappeared. No definite end point could
be given due to the competing reaction with DMSO
occurring.
Reaction of bis-(2,6-dicarboxypyridyl)silver(II)
with cytochrome-c
Aqueous solutions (2 mL) of cytochrome c (16.5 mg,
54 lmoles) were incubated with 20 mg (44 lmoles),
40 mg (88 lmoles) or 60 mg (131 lmoles) of Ag2,6P.
The solutions were allowed to react overnight and then
filtered before analysis as described below.
For analysis by mass spectrometry, 2 mL of
aqueous suspensions of the sodium salts of the fatty
acids (sodium linoleate; 16.5 mg, 54 lmoles or
arachidonic acid; 16.5 mg, 50 lmoles) were reacted
with 20 mg (45 lmoles), 40 mg (91 lmoles) or 60 mg
(136 lmoles) of Ag2,6P. The solutions were allowed
to react overnight and then filtered before analysis as
described below.
Analysis of protein oxidation by oxyblotting
for DNP-carbonyl adducts
Aliquots of the samples (10 lL; * 75 lg protein)
were resolved by SDS-PAGE with a 12% resolving gel
123

Biometals
(Sambrook and Russell 2006) and then either stained
with InstantBlue stain (Sigma-Aldrich, UK) or trans-
ferred onto PVDF membrane for oxyblotting as
described previously (Shacter 2000). After washing,
the membrane was acidified with 2 N HCl and labelled
with 10 mM dinitrophenylhydrazine (DNPH) (Sigma-
Aldrich, UK) for 5 min. After further washing and
blocking the membrane was incubated in blocking
buffer containing monoclonal primary antibodies
cell membrane transport, in the case of Ag(II) their
potential may be limited by their lack of solubility.
In view of the high redox potential of silver(II)
(equation 2; E0 = 2.0 V NHE (Weast 1979)), the
silver (II) compounds were expected to be quite
reactive even to the extent of potentially oxidizing
water. Therefore, the first step was to investigate the
lifetime of Ag2,6P in aqueous buffer and DMSO, by
monitoring the visible absorbance spectrum (Suppl.
Fig. 1a). In aqueous solution a small but manageable
degradation (* 10%) of Ag2,6P was observed over a
2 h period. The profile of the degradation process was
linear within the lifetime of the experiment, suggest-
ing that decomposition does not occur via an SN2
displacement of the axial carboxylates by water or
phosphate. In contrast, the stability of Ag2,6P in
DMSO was poor, having a half life of only 25 min
(Suppl. Fig. 1b). DMSO can be oxidized to dimethyl-
sulphone (Me2SO2; E0 = 1.54 V vs. NHE) and the
silver(II) complex studied here is therefore theoreti-
cally capable of driving this reaction (Krtil et al.
1996). Solid reflectance spectrophotometry indicated
that the solid and DMSO solution phase structures of
Ag2,6P are similar (kmax 600 nm), suggesting that
decomposition occurs via electron transfer rather than
ligand exchange. The reaction of the silver complex
with DMSO limits the interpretation of reactions with
other compounds carried out in this solvent, but in
some cases there was no feasible alternative. To
obviate problems with the slow decomposition of
Ag2,6P in solution, fresh solutions were prepared
immediately before the start of each experiment.
rabbit
anti-dinitrophenylhydrazone
(anti-DNP)
(D9656, Sigma-Aldrich, UK) at a working dilution
of 1:1000 overnight at 4 °C. The secondary antibody
was HRP-linked goat anti-mouse (6154, Sigma-
Aldrich, UK) antibody (working dilution 1:1000) for
2 h at room temperature. The membrane was washed
again as described above and HRP-linked anti-mouse
was detected using enhanced chemiluminescence
(ECL kit 34078, Thermo Fisher Scientific, Hemel
Hempstead, UK) according to the manufacturer’s
instructions. The membrane was scanned using a
G:BOX system (Syngene, Cambridge, UK) running
the GeneSys software (Syngene, Cambridge, UK).
Results and discussion
Solubility and stability of silver(II) complexes
Silver(II) complexes of 2-carboxypyridine (Ag2P),
2,3-dicarboxypyridine
boxypyridine (Ag2,4P),
(Ag2,3P),
2,4-dicar-
2,5-dicarboxypyridine
(Ag2,5P) and 2,6-dicarboxypyridine (Ag2,6P) were
prepared using literature methods (Drew et al. 1970;
Drew et al. 1971; Fowles et al. 1968). Ag2P, Ag2,3P,
Ag2,4P and Ag2,5P had limited solubility in both
water and DMSO, and were not studied further owing
to the limited relevance to biological environments. In
contrast, Ag2,6P was observed to be reasonably
soluble in both water (* 10 mM) and DMSO
(* 20 mM). The solubility profile of the compounds
most likely arises from their solid state structures:
Ag2P and Ag2,3P are planar species and prone to p-
stacking in the solid state (Fowles et al. 1968), which is
known to affect solubility detrimentally. In contrast,
Ag2,6P adopts an octahedral geometry in the solid
state and is unable to p-stack, which lowers the lattice
energy and promotes its solubility in polar solvents
(Drew et al. 1970). The structures are shown in Fig. 1.
Thus, despite the desirability of planar compounds for
Reaction with antioxidants
Biological systems utilize a number of species as co-
factors and reducing agents (e.g. glutathione, ascorbic
acid, a-tocopherol), and depletion or oxidation of
antioxidants and structural biological molecules is an
early stage in the stress leading to the toxic effects of
oxidizing compounds (Halliwell and Gutteridge
1998). Consequently, the reactions of Ag2,6P with
these three antioxidants was investigated. The reaction
with an antioxidant can readily be inferred by
spectrophotometric titrations in which the stepwise
reduction of the coloured Ag2,6P to its colourless
silver(I) product is observed. The reactions of glu-
tathione and ascorbate were carried out in aqueous
solution, whereas the reaction of a-tocopherol was
123

Biometals
Fig. 1 Structures of the silver (II) complexes synthesized. 2,3-dicarboxypyridine (Ag2,3P), 2,4-dicarboxypyridine (Ag2,4P), 2,5-
dicarboxypyridine (Ag2,5P) and 2,6-dicarboxypyridine (Ag2,6P)
carried out in DMSO. Using the diminution of the
band at 570 nm it was possible to titrate Ag2,6P with
glutathione (Fig. 2), and the molar ratio at the end
point was calculated to be slightly greater than 2:1
Ag2,6P:GSH. This would be generally consistent with
the 2-electron oxidation of GSH to GSSG, assuming
that the Ag2,6P undergoes a 1-electron reduction to
Ag?, as the appearance of metallic silver was not
observed.
Fig. 2), which clearly showed that treatment of GSH
(m/z 306.1) with Ag2,6P resulted in conversion to
GSSG (m/z 611.1).
The reaction of Ag2,6P with ascorbic acid was
monitored spectrophotometrically in the same way as
that of glutathione, and it could clearly be seen that
addition of ascorbic acid resulted in loss of absorbance
at 570 nm (Suppl. Fig. 3). The molar ratio at the end
point was calculated to be approx. 2:1 Ag2,6P:ascor-
bate, consistent with the 2-electron oxidation of
ascorbate to dehydroascorbate. However, this reaction
was not investigated further due to the lack of stability
of dehydroascorbate in the presence of redox metals;
metal-mediated ascorbic acid oxidation and redox
cycling is a facile process involving low as well as
high valent metals (Halliwell and Gutteridge 1998;
Skov and Vonderschmitt 1975).
To investigate the nature of the oxidation in more
1
depth, the reaction of GSH was monitored using H-
NMR. Figure 3 clearly demonstrates that Ag2,6P
oxidized GSH to the disulfide form (GSSG), with
increasing Ag2,6P amounts correlating with increased
loss of the GSH triplet signals at * 2.95 ppm and
appearance of the GSSG pairs of doublets at * 3.05
and 3.3 ppm. This finding was supported by negative
ion electrospray mass spectrometry analysis (Suppl
Fig. 2 Spectrophotometric
analysis of the reaction
between Ag2,6P and
glutathione. The titration of
8.5 mM Ag2,6P
(18.7 lmoles in 2.2 mL) in
0.1 M KH2PO4, pH 7.0 with
glutathione (GSH). The
glutathione (84.3 mM) was
added in 20 uL aliquots (a
total of 6) and the
corresponding lmoles of
GSH are indicated on the
right-hand side of the traces
123

Biometals
Fig. 3 1H-NMR analysis of
the reaction between Ag2,6P
and glutathione. 27 lmols
of glutathione (GSH) was
reacted with increasing
amounts of Ag2,6P as
indicated on the spectra. The
spectrum of commercial
glutathione disulfide is
shown at the bottom for
comparison
The reaction of a-tocopherol with Ag2,6P in
DMSO was also discernable using spectrophotometry
(Suppl. Fig. 4). However, due to the competing
interaction of Ag2,6P with DMSO discussed above,
it was not possible to obtain an accurate end point for
the titration.
ESI-mass spectrometry was used to monitor the
reaction of Ag2,6P with the unsaturated fatty acid
linoleic acid (Fig. 4). A small amount of adventitious
oxidation of the control sample is apparent in Fig. 4a,
but the major signal is the native fatty acid at m/z 279
([M–H]-). There was also a strong signal at m/z 325,
which was identified as the formate adduct of linoleic
acid ([M ? CHO2]-). At the lower treatment concen-
tration (Fig. 4b) the signal of the native lipid was
greatly reduced and the major signal was at m/z 311,
corresponding to the addition of O2 (? 32 Da). There
was also evidence of addition of a single oxygen atom
at m/z 295, and another signal at m/z 293, which was
most probably due to loss of water from the species at
m/z 311, suggesting that it may be a bis-hydroxide
rather than a hydroperoxide (Spickett and Pitt 2015).
Interestingly, a peak was observed at m/z 557, which
was consistent with the formation of a cross-linked
dimer of linoleic acid (loss of 3H but singly charged)
and there were also dimers containing 2, 3 and 4
additional oxygens. Crosslinking of oxidized fatty
acyl chains under highly oxidizing conditions has been
ESI–MS analysis of the reaction with fatty acids
The unsaturated fatty acids linoleic acid and arachi-
donic acid were used as models to investigate the
ability of Ag2,6P to oxidize lipid. Preliminary studies
by spectrophotometric analysis indicated that reac-
tions occurred, but the reactions were so fast that
kinetic analysis was not possible; moreover the
experiments were carried out in DMSO as the fatty
acid salts were sparingly soluble in aqueous solution,
and therefore were limited by the issues with this
solvent mentioned above. The focus of these studies
was therefore the analysis of oxidation products of the
biomolecules.
123

Biometals
Fig. 4 The reaction of
linoleic acid with Ag2,6P
studied by ESI-mass
spectrometry in negative ion
mode. a Untreated linoleic
acid. b 29 mM Linoleic acid
treated with 37 mM Ag2,6P
and c 29 mM linoleic acid
treated with 110 mM
Ag2,6P
reported previously (Muizebelt and Nielen 1996;
Schroter et al. 2016; Tolvanen et al. 2008). At the
higher treatment concentration the native lipid was
further depleted, probably resulting from degradation
of the lipid to small, non-ionized breakdown products,
but otherwise the oxidation pattern was comparable.
Treatment of arachidonic acid with Ag2,6P also
clearly showed the occurrence of oxidation (Fig. 5).
As with linoleic acid, there was some adventitious
123

Biometals
Fig. 5 The reaction of
arachidonic acid with
Ag2,6P studied by negative
ion ESI-mass spectrometry.
a ESI-mass spectra of
untreated arachidonic acid.
b arachidonic acid (27 mM)
treated with Ag26P
(22.5 mM) and
c arachidonic acid (27 mM)
treated with Ag26P
(68 mM). All samples were
diluted equivalently in
methanol prior to infusion
into the instrument. The
signal at m/z 349 is probably
a formate adduct, while the
one at 333 appears to be a
contaminant
123

Biometals
oxidation in the untreated sample, reflecting the
susceptibility of this polyunsaturated fatty acid to
autoxidation (Fig. 5a), but even at the low treatment
concentration (Fig. 5b) the incorporation of oxygen
was increased, with signals at m/z 319 (1 O), m/z 335
(2 O), m/z 351 (3 O) and m/z 367 (4 O). A strong signal
at 349.2 was also observed, corresponding to
? 46 Da; in view of its appearance only in treated
samples, this is likely to be a dehydration product
following the addition of 4 oxygens such as an
epoxyisoprostane, which are known as relatively
stable products of arachidonic acid (Spickett and Pitt
2015). At the higher treatment concentration the
native signal at m/z/303 was almost abolished
(Fig. 5c), and the fatty acid was more highly oxidized
with the strongest signal at m/z 335.1 and all other
products at higher masses and levels of oxidation. In
contrast, there was no evidence of dimers of arachi-
donic acid analogous to those observed with linoleic
acid, which should have occurred at m/z 605 (data not
shown), although the ions that appeared between m/z
515–553 were not identified. Comparing the reactions
of linoleic and arachidonic acids with Ag2,6P, it was
clear that higher concentrations of the silver were
required to deplete the more unsaturated fatty acid,
reflecting its greater capacity for oxidative
modification.
previously (Fraschini and Vignon 2000). Spectropho-
tometric titrations in aqueous solution (400–900 nm)
clearly showed evidence of a reaction of Ag2,6P with
b-cyclodextrin (Fig. 6). An end point was obtained at
approximately 3:1 Ag2,6P: b-CD, which suggests that
multiple oxidations might be occurring. Analysis by
mass spectrometry also suggested that oxidation had
occurred (Fig. 7). Figure 7a shows untreated b-cy-
clodextrin, which was the major species in the sample
at m/z 1133.1 and therefore 100% relative intensity. As
the samples were prepared in solvent containing
formate, there was also a significant formate adduct
at 1179.0. Treatment of the b-cyclodextrin with a
5-fold excess of Ag2,6-P led to free 2,6-picolinate
presenting the strongest signal (100%) at m/z 166.1
(data not shown), while the b-cyclodextrin signal was
substantially depleted and a signal at m/z 1231.1
appeared, corresponding to the oxidation of all of the
hydroxyl groups into carboxylic acid to form a hepta-
carboxylate b-cyclodextrin (plus 7 9 14 Da). A 2,6-
dicarboxypyridine adduct of cyclodextrin at m/z
1300.3 was also observed, probably reflecting the high
level of the free ligand present in the sample after
reduction of Ag2,6-P by the carbohydrate. In contrast
to the treatment of b-cyclodextrin with HOCl reported
previously [20], there was no clear evidence of
intermediate oxidation products, such as tris and
hexakis carboxylate species, presumably owing to the
highly oxidizing nature of the Ag2? complex.
Reaction with the carbohydrate beta-cyclodextrin
To investigate the effects of Ag2,6P on carbohydrates,
b-cyclodextrin was used as a model, as it can readily be
observed by mass spectrometry ([M-H]- at m/z 1133),
and moreover its oxidation by HOCl has been studied
Reaction with cytochrome c
Many adverse effects of oxidizing antimicrobial
agents are mediated by protein oxidation and
Fig. 6 Spectrophotometric
analysis of the reaction of b-
cyclodextrin with Ag2,6P.
7.6 mM Ag2,6P (17.2 lmol
in 2.2 mL) in 0.1 M
KH2PO4, pH 7.0 was
titrated against b-
cyclodextrin (b-CD). The b-
cyclodextrin (25.6 mM) was
12.0
10.0
8.0
6.0
4.0
added in 200 lL aliquots
(9 10). The mole ratio at the
end point was calculated to
2.0
be 1:3 Ag2,6P: b-CD
mol:mol
0.0
0.0
2.0
4.0
6.0
8.0
10.0
12.0
[cyclodextrin] mmol/L
123

Biometals
Fig. 7 Electrospray mass
spectrometry analysis of the
reaction of Ag2,6P with b-
cyclodextrin a the negative
ion ESI-mass spectra of b-
cyclodextrin, also showing a
formate adduct at ? 46 Da;
b b-cyclodextrin (100 mg,
88 lmoles) treated with a
5-fold excess of Ag2,6P
(200 mg, 0.44 mmol); and
c structure of b-cyclodextrin
together with its formula.
Note that the y-axis scale is
relative signal intensity,
where in a the native b-
cyclodextrin is the major
species and therefore 100%,
whereas in b 2,6-picolinate
at m/z 166.1 was the major
species at 100%
disruption of cell signaling, so the effect of Ag2,6P on
cytochrome C as a model ubiquitous protein was
studied. Preliminary studies by spectrophotometry
showed that Ag2,6P was consumed by relatively small
amounts of protein (data not shown), which is
consistent with the presence of multiple oxidation
sites on the polypeptides. To confirm the oxidative
action of Ag2,6P on the protein, the formation of
protein carbonyls was investigated, as these are well-
established oxidation products (Domingues et al.
2013; Shacter 2000). Carbonyl formation may occur
by oxidative deamination of lysines, or radical attack
and fragmentation of various other residues (Davies
2016). The samples were first separated on denaturing
polyacrylamide gels and visualized by staining with
coomassie blue; then labeling of carbonyl groups with
DNPH followed by western blotting with antibody to
the DNP-adduct (commonly known as oxy-blotting)
was carried out. Figure 8 shows that even low levels of
Ag2,6P resulted in loss of the cytochrome c band at
12.3 kDa and appearance of high molecular weight
aggregates that were retained at the top of the
resolving gel, while higher concentrations led to more
aggregates and additionally some degradation prod-
ucts observed at the bottom of the gel (Fig. 8a).
Oxyblotting for oxidative damage to the protein
showed the presence of increased carbonyls with
increasing severity of the Ag2,6P treatment, initially
in the cytochrome c band but this was lost at higher
Ag2,6P concentrations and staining of the high
123

Biometals
Fig. 8 The effect of Ag2,6P treatment on Cytochrome c.
Protein (8.25 mg/mL) was treated with 0, 63, 126, 189 or
252 lM Ag2,6P overnight. Molecular weights markers (MWM)
are shown on the lefthandside with numbers in kDa; the
molecular weight of horse heart Cytc is 12,384 Da. a Coomassie
blue-stained 12% SDS-PAGE reducing gel. b Oxyblot of a
comparable gel using an anti-DNPH primary antibody to
determine the formation of carbonyl groups on cytochrome c
as a marker of oxidative damage
molecular weight aggregates predominated (Fig. 8b).
This clearly confirmed that extensive protein oxida-
tion occurred following Ag2,6P treatment of the
cytochrome c.
intermediate action via oxidation of the solvents H2O
or DMSO, leading to production of hydroxyl radicals
(HOÁ) or other radical species.
The reactivity of Ag2,6P with a range of different
biomolecules suggests that it is likely to have signif-
icant toxicity to bacterial cells and therefore good
biocidal potential. Its solubility in the organic solvent
DMSO implies that it may be able to penetrate lipid
membranes and thus gain access to the intracellular
environment, as lipophilicity is known to be a factor in
biological transport of metal complexes (Oldfield et al.
2007). The reaction with beta-cyclodextrin is impor-
tant and similar reactions could contribute to microbial
toxicity by altering carbohydrates in the cell wall or on
the surface of microbial cells.
Concluding remarks
Although planar metal complexes have been reported
to be useful in terms of their ability to penetrate cell
membranes, it is clear that for Ag(II) dicar-
boxypyridine complexes, the planar ones have lower
aqueous solubility compared to one with octahedral
geometry, namely Ag(II) 2,6-dicarboxypyridine
(Ag2,6P). This compound has satisfactory stability
in aqueous solution and was found to oxidize a variety
of biomolecules. It was shown to react readily with the
thiol-containing antioxidant glutathione, as well as
with ascorbic acid (vitamin C) and a-tocopherol
(vitamin E). It caused extensive oxidation of polyun-
saturated fatty acids as well as cross-linking of the
chains, and induced carbonyl formation, a common
marker of oxidative damage, on the small model
protein cytochrome c. Together, these effects suggest
a one-electron or free radical mechanism of action,
although it is not clear whether it involves direct
reaction of the metal ion with the biomolecules, or
However, Ag(II) complexes remain challenging to
work with, owing to their limited stability in relevant
solvents or even in pure form at room temperature.
Although the study has shown that it should be
possible to design improved biocides based on silver
in a ?2 formal oxidation state, further effort is
required to design new ligands that support and target
this moiety better.
Acknowledgements SM and CMS acknowledge BSSRC and
Mologic for the Industrial CASE Award BB/J012939/1. CMS
and JR acknowledge support from the UK Engineering and
123

Biometals
Physical Sciences Research Council (EPSRC) funded ‘‘Aston
Multidisciplinary Research for Antimicrobial Resistance: The
AMR4AMR project’’ Grant No. EP/M02735X/1.
molecular structure of bis(pyridine-2,3-dicarboxylato)sil-
ver(II) dihydrate. J Chem Soc A. https://doi.org/10.1039/
j19710002959
Fowles GWA, Matthews RW, Walton RA (1968) Studies on co-
ordination compounds of silver(II). Part I. Magnetic and
spectral properties of complexes with pyridine carboxylic
acids. J Chem Soc A 25:124. https://doi.org/10.1039/
j19680001108
Fraschini C, Vignon MR (2000) Selective oxidation of primary
alcohol groups of beta-cyclodextrin mediated by 2,2,6,6-
tetramethylpiperidine-1-oxyl radical (TEMPO). Carbo-
hydr Res 328:585–589. https://doi.org/10.1016/s0008-
6215(00)00129-4
Open Access This article is distributed under the terms of the
Creative Commons Attribution 4.0 International License (http://
creativecommons.org/licenses/by/4.0/), which permits unre-
stricted use, distribution, and reproduction in any medium,
provided you give appropriate credit to the original
author(s) and the source, provide a link to the Creative Com-
mons license, and indicate if changes were made.
Fromm KM (2013) Silver coordination compounds with
antimicrobial properties. Appl Organomet Chem
27:683–687. https://doi.org/10.1002/aoc.3047
Data availability The original data associated with this
project are available at https://doi.org/10.17036/researchdata.
aston.ac.uk.00000418.
Halliwell B, Gutteridge JMC (1998) Free radicals in biology and
medicine, 5th edn. Oxford University Press, Oxford
Hanczvikkel A, Fuzi M, Ungvari E, Toth A (2018) Transmis-
sible silver resistance readily evolves in high-risk clone
isolates of Klebsiella Pneumoniae. Acta Microbiol Imm H
65:387–403. https://doi.org/10.1556/030.65.2018.031
Heiman KE, Mody RK, Johnson SD, Griffin PM, Gould LH
(2015) Escherichia coli O157 outbreaks in the United
States, 2003-2012. Emerg Infect Dis 21:1293–1301.
https://doi.org/10.3201/eid2108.141364
References
Arguello JM, Raimunda D, Gonzalez-Guerrero M (2012) Metal
transport across biomembranes: emerging models for a
distinct chemistry. J Biol Chem 287:13510–13517. https://
doi.org/10.1074/jbc.R111.319343
Atiyeh BS, Costagliola M, Hayek SN, Dibo SA (2007) Effect of
silver on burn wound infection control and healing: review
of the literature. Burns 33:139–148. https://doi.org/10.
1016/j.burns.2006.06.010
Hetrick EM, Schoenfisch MH (2006) Reducing implant-related
infections: active release strategies. Chem Soc Rev
35:780–789. https://doi.org/10.1039/b515219b
Konop M, Damps T, Misicka A, Rudnicka L (2016) Certain
aspects of silver and silver nanoparticles in wound care: a
minireview. J Nanomater. https://doi.org/10.1155/2016/
7614753
´
´
´
Azocar MI, Gomez G, Levın P, Paez M, Mun˜oz H, Dinamarca N
(2014) Review: antibacterial behavior of carboxylate sil-
ver(I) complexes. J Coord Chem 67:3840–3853. https://
doi.org/10.1080/00958972.2014.974582
´
´
Krtil P, Kavan L, Hoskovcova I, Kratochvilova K (1996)
Anodic oxidation of dimethyl sulfoxide based electrolyte
solutions: an in situ FTIR study. J Appl Electrochem
26:523–527. https://doi.org/10.1007/bf01021976
Brandt C, Makarewicz O, Fischer T, Stein C, Pfeifer Y, Werner
G, Pletz MW (2014) The bigger picture: the history of
antibiotics and antimicrobial resistance displayed by
scientometric data. Int J Antimicrob Agents 44:424–430.
https://doi.org/10.1016/j.ijantimicag.2014.08.001
Cook DS, Turner MF (1975) Crystal and molecular structure of
silver sulphadiazine (N1-pyrimidin-2-ylsulphanilamide).
J Chem Soc, Perkin Trans 2:1021–1025. https://doi.org/10.
1039/P29750001021
Davies MJ (2016) Protein oxidation and peroxidation. Biochem
J 473:805–825. https://doi.org/10.1042/BJ20151227
Domingues RM, Domingues P, Melo T, Perez-Sala D, Reis A,
Spickett CM (2013) Lipoxidation adducts with peptides
and proteins: deleterious modifications or signaling
mechanisms? J Proteom 92:110–131. https://doi.org/10.
1016/j.jprot.2013.06.004
Le Ouay B, Stellacci F (2015) Antibacterial activity of silver
nanoparticles: a surface science insight. Nano Today
10:339–354. https://doi.org/10.1016/j.nantod.2015.04.002
Levason W, Spicer MD (1987) The chemistry of copper and silver
intheirhigheroxidation-states. CoordChemRev76:45–120.
https://doi.org/10.1016/0010-8545(87)85002-6
Malik MA, Dar OA, Gull P, Wani MY, Hashmi AA (2018)
Heterocyclic Schiff base transition metal complexes in
antimicrobial and anticancer chemotherapy. MedChem-
Comm 9:409–436. https://doi.org/10.1039/c7md00526a
Martinho N, Santos TCB, Florindo HF, Silva LC (2018) Cis-
platin-membrane interactions and their influence on plat-
inum complexes activity and toxicity. Front Physiol
9:1898. https://doi.org/10.3389/fphys.2018.01898
Marx DE, Barillo DJ (2014) Silver in medicine: the basic sci-
ence. Burns 40(Suppl 1):S9–S18. https://doi.org/10.1016/j.
burns.2014.09.010
Drew MGB, Fowles GWA, Matthews RW, Walton RA (1969)
Silver(II) Bis(pyridine-2,6-dicarboxylate) monohydrate a
novel six-coordinate structure.
J Am Chem Soc
91:7769–7771. https://doi.org/10.1021/ja50001a056
Drew MGB, Matthews RW, Walton RA (1970) Studies on
coordination compounds of silver 3. Crystal structure of
silver(2)-Bis(Pyridine-2,3-dicarboxylate) dihydrate. Inorg
Nucl Chem Lett 6:277. https://doi.org/10.1016/0020-
1650(70)80231-8
Mijnendonckx K, Leys N, Mahillon J, Silver S, Van Houdt R
(2013) Antimicrobial silver: uses, toxicity and potential for
resistance. Biometals 26:609–621. https://doi.org/10.1007/
s10534-013-9645-z
Muizebelt WJ, Nielen MWF (1996) Oxidative crosslinking of
unsaturated fatty acids studied with mass spectrometry.
J Mass Spectrom 31:545–554
Drew MGB, Matthews RW, Walton RA (1971) Studies on co-
ordination complexes of silver(II). Part VI. Crystal and
123

Biometals
Oldfield SP, Hall MD, Platts JA (2007) Calculation of
lipophilicity of a large, diverse dataset of anticancer plat-
inum complexes and the relation to cellular uptake. J Med
Chem 50:5227–5237. https://doi.org/10.1021/jm0708275
Pal C, Bengtsson-Palme J, Rensing C, Kristiansson E, Larsson
DGJ (2014) BacMet: antibacterial biocide and metal
resistance genes database. Nucleic Acids Res 42:D737–
D743. https://doi.org/10.1093/nar/gkt1252
Simoncic B, Klemencic D (2016) Preparation and performance
of silver as an antimicrobial agent for textiles: a review.
Text Res
0040517515586157
J
86:210–223. https://doi.org/10.1177/
Singh R, Shedbalkar UU, Wadhwani SA, Chopade BA (2015)
Bacteriagenic silver nanoparticles: synthesis, mechanism,
and applications. Appl Microbiol. Biot 99:4579–4593.
https://doi.org/10.1007/s00253-015-6622-1
Panacek A et al (2018) Bacterial resistance to silver nanopar-
ticles and how to overcome it. Nat Nanotechnol 13:65.
https://doi.org/10.1038/s41565-017-0013-y
Percival SL, Bowler PG, Russell D (2005) Bacterial resistance
to silver in wound care. J Hosp Infect 60:1–7. https://doi.
org/10.1016/j.jhin.2004.11.014
Rosenberg B, VanCamp L, Trosko JE, Mansour VH (1969)
Platinum compounds: a new class of potent antitumour
agents. Nature 222:385–386
Sambrook J, Russell D (2006) SDS-polyacrylamide gel elec-
trophoresis of proteins. Cold Spring Harb Protoc 3:18–47
Schroter J et al (2016) Unexpected products of the hypochlorous
acid-induced oxidation of oleic acid: a study using high
performance thin-layer chromatography-electrospray ion-
ization mass spectrometry. J Chromatogr A 1439:89–96.
https://doi.org/10.1016/j.chroma.2015.11.059
Shacter E (2000) Quantification and significance of protein
oxidation in biological samples. Drug Metab Rev
32:307–326. https://doi.org/10.1081/Dmr-100102336
Silvestry-Rodriguez N, Sicairos-Ruelas EE, Gerba CP, Bright
KR (2007) Silver as a disinfectant. Reviews of environ-
mental contamination and toxicology. Springer, New
York, NY, pp 23–45. https://doi.org/10.1007/978-0-387-
69163-3_2
Skov KA, Vonderschmitt DJ (1975) Kinetics of iron and copper
catalysis of ascorbate oxidation. Bioinorg Chem
4:199–213.
3061(00)80103-9
https://doi.org/10.1016/S0006-
Spickett CM, Pitt AR (2015) Oxidative lipidomics coming of
age: advances in analysis of oxidized phospholipids in
physiology and pathology. Antioxid Redox Sign
22:1646–1666. https://doi.org/10.1089/ars.2014.6098
Tolvanen P, Maki-Arvela P, Eranen K, Warna J, Holmbom B,
Salmi T, Murzin DY (2008) Thermal polymerisation and
autoxidation of technical grade linoleic acid. J Am Oil
Chem Soc 85:567–572. https://doi.org/10.1007/s11746-
008-1229-7
Vindigni SM, Surawicz CM (2015) C. difficile infection:
changing epidemiology and management paradigms. Clin
Transl Gastroen 25:124. https://doi.org/10.1038/ctg.2015.
24
Weast RC (1979) Handbook of chemistry and physics, 60th edn.
CRC Press, Florida
Publisher’s Note Springer Nature remains neutral with
regard to jurisdictional claims in published maps and
institutional affiliations.
123