Nanobio Silver: Its Interactions with Peptides and Bacteria, and Its
Uses in Medicine
Sonja Eckhardt,* Priscilla S. Brunetto, Jacinthe Gagnon, Magdalena Priebe, Bernd Giese,
and Katharina M. Fromm*
Department of Chemistry, University of Fribourg, Chemin du Muse
e 9, 1700 Fribourg, Switzerland
1. INTRODUCTION
Silver, known in metallic form since antiquity, has very early
been recognized by mankind for its antimicrobial properties, a
phenomenon observed, for example, in the context of drinking
water (a silver coin in a well), food (silver cutlery, water storage
recipients), and medicine (silver skull plates, teeth).
1
Silver
compounds were also shown to be useful. For example, dilute
solutions of silver nitrate served long, and still do in some
countries, as antimicrobial ointment to be instilled into
Published in "&KHPLFDO5HYLHZV±"
which should be cited to refer to this work.
http://doc.rero.ch
newborn babies’ eyes to prevent contraction of gonorrhea from
the mother.
1
Used frequently until the middle of the 20th century, silver
and its compounds were then replaced by the newly discovered
antibiotics and nearly forgotten for almost 50 years. However,
one rare example for the use of silver in this time period is silver
sulfadiazine. Upon increasing bacterial resistance against
current antibiotics, and with the steady development of
nanoscale science, research on silver and its compounds has
regained in interest,
2
however, facing the challenge to catch up
the time during which research activities in that area were low.
Today, silver presents itself as bulk material, as nanoparticles,
clusters, and a large variety of silver compounds that are
available for numerous applications, one of them being in
medicine.
For ca. 20 years, the number of publications dealing with
silver in medical applications is increasing steadily; for example,
the number of papers corresponding to a search on “silver in
medicine” on the Web of Science database in April 2012 is
given as below five in 1993, and above 60 in 2011, with ca. 1900
citations in 2011. These publications involve scientists from all
areas: physicists, biochemists, chemists, materials scientists,
microbiologists, and medical doctors, all tackling the problem
from their respective points of view and with their own
methods, respectively, highlighting applications from materials,
analytics, detection, and diagnosis to treatment or drug delivery.
This is, for example, reflected in a large number of different
assays described to study the antimicrobial properties, the
biocompatibility, and the toxicity in vitro and in vivo, based on
many different bacterial strains and cell lines and leading to
sometimes opposite results. If chemists, physicists, and material
scientists tend to know the respective silver compounds used in
their study very well, they usually limit their biological studies
to in vitro antimicrobial tests.
3−7
On the other hand, medically
oriented reports are less clear about the exact materials used,
and present more advanced in vitro as well as in vivo results,
the latter of which are sometimes contradictory to in vitro
data.
8,9
Besides, the research at different scales, for example, at
the macro- and microscopic level (phenomenological observa-
tions), on the one hand, and the molecular studies, on the other
hand, is not always congruent.
Recent reviews and books covered so far individual aspects
of, for example, silver in medicine,
10−12
silver coordination
compounds,
13
and biomaterials containing silver.
14
Other
reports on nanotechnology and its use in medicine contain
also chapters dedicated to silver.
15−17
We thus felt it timely to attempt to bring together these
different aspects, to collect the data out in the literature, and to
link and present these results of the current state of the art
together with the remaining challenges. This Review addresses
thus many readers: the chemist will learn about the molecular
effects of silver in biological environments; the materials
scientist who is interested in making nanodevices may just as
well find interesting examples herein as the microbiologist or
medical doctor who wants to find out about the material
composition and its side effects. We focus on four main aspects:
(i) the interaction of silver at a molecular level, on the one
hand, (ii) at a cellular level, on the other hand, (iii) literature
dealing with nanomaterials from which we can learn and get
clues on the functioning of silver at all levels, and (iv) studies
on the biocompatibility of silver. These parts are followed by a
critical discussion of the state of knowledge.
For silver and its compounds, the following sections will
elucidate on the molecular interactions of biomolecules with
silver ions and silver nanoparticles (AgNPs), focusing on
peptides as main reaction partners in cells, then on bacterial
interactions with silver, on the use of silver in biomaterials, and
on the biocompatibility aspects of this metal and its
compounds. We will thus shine light on the different aspects
of silver in medicine, exploring the different levels of activity,
from the molecular to the cellular and micrometer dimensions.
Silver presents itself in different forms; among them ionic silver
and AgNPs are the most frequently discussed in the literature.
We will also present bulk silver surfaces as well as the so-called
nanosilver, a recently described, apparently new class of silver.
For the latter, we will discuss it wherever we find the, still rare,
literature.
From a chemical point of view, it is convenient to first
highlight interactions of silver at the molecular level because it
is the basis for the interactions at higher levels. Thus, in the
next section, we will start the journey with the examination of
the silver binding to amino acids and peptides.
2. PEPTIDES AND SILVER
2.1. Silver Binding to Amino Acids and Peptides
The interaction of metal ions with biomolecular targets, for
example, amino acids, peptides, or proteins, is known to play a
fundamental role in many biological processes such as electron
transfer reactions, oxygen transport, as well as metal transport
and storage. Further on, the inhibition of enzymes by metal
complexes with labile ligands is well-known. It functions by
ligand exchange reactions, where the labile ligand present in the
administered drug is replaced by the targeted enzyme, a
principle used in many metal-based drugs.
18
It is therefore one
of the most important processes in bioinorganic chemistry.
19
Although silver ions do not seem to be involved in natural
systems, their medical use makes them an important target.
Ag(I) can be classified as a soft cation according to Pearson,
20
and therefore it prefers to bind to polarizable, so-called soft,
ligands. In addition, it is able to form strong σ−π-bonds
through back-donation of the electrons located in its d-orbitals
to the π*-orbitals of the ligand. In terms of electronic structure,
Ag(I) is closely related to Cu(I) because both ions have the
same outer electronic configuration with a closed-shell structure
(d
10
). Copper as trace element plays a fundamental role as
cofactor in all organisms.
21
Because the presence of larger
amounts of free copper ions is toxic for the cells, many peptides
and proteins are involved in its homeostasis.
21
Because of the
close chemical relation between Ag(I) and Cu(I), it is not
surprising that silver has been also found to bind to peptides
and proteins. Because amino acids are the small subunits used
to construct peptides and proteins, it is reasonable to first have
a closer look at their interaction with silver ions to answer the
question on how Ag(I) binds to their polymeric form.
2.1.1. Silver Binding to Amino Acids − Theory. The
silver complexes of all common 20 naturally occurring α-amino
acids have been calculated using, for example, the hybrid
density functional theory (DFT)
22,23
or the quantitative
structure−property relationship methodology.
24
A comparison
of these data is not easy, due to the different levels of
calculation, varying basis sets, or the nonidentical treatment of
the basis set superposition error. Nevertheless, these theoretical
studies reveal a common trend, with the three basic amino acids
arginine, lysine, and histidine being the strongest silver ion
http://doc.rero.ch
binders, while the nonpolar aliphatic amino acids with the
exception of methionine have the weakest silver ion affinity
among the naturally occurring amino acids, whose side chains
are shown in Figure 1. This affinity is defined as the binding
energy at 298 K. The calculated silver ion affinity of all 20
natural amino acids on the basis of a 1:1 stoichiometry can be
found in Table 1.
23
Most interesting for a classical coordination
chemist, cysteine is not among the calculated main candidates
for silver binding even though the sulfur atom is considered as
the softest group present in proteins. Cysteine is therefore,
together with methionine, often considered as preferential
binding site for silver ions in proteins,
25
an assumption that is
not confirmed by theory.
The total binding energy can be decomposed into two
energy terms, one for the preorganization of the amino acid and
the second for the interaction between amino acid and cation.
Because this latter interaction is attractive, this term is always
negative and may compensate for the preorganization energy,
which is positive due to the higher energy of the amino acid in
the complexed form (less degrees of freedom) as compared to
their free state. The order in theoretical binding energies for the
metal ion−amino acid complex formation is very similar for
Ag(I) and Cu(I), the main exceptions being proline and serine,
both having a higher affinity toward silver as compared to
copper with respect to the other amino acids. Jover et al.
23
also
compared the affinity toward silver with the binding affinities of
the amino acids for other monocharged cations, identifying the
order Cu(I) > Li(I) > Ag(I) > Na(I) > K(I) for most of the
amino acids. The calculated binding affinities also take into
account possible coordination modes between Ag(I) and the
amino acid. While some metal ions are able to adopt different
modes, others tend to prefer a single coordination geometry.
These preferences in terms of coordination are thought to be
one important factor on how nature realizes the metal ion
selectivity of metal ion-containing enzymes, which is indis-
pensable for their biological function. The toxicity of some
metals, which are usually not present in biological systems, is
often based on their ability to occupy these (under biological
conditions) metal-speci fic binding sites through effective
competition with the biologically relevant metal ion and strong
coordination to its binding site.
26
In principle, there can be up
to three different coordination sites present in an amino acid:
(i) the amino nitrogen-donor at the N-terminus, (ii) the oxygen
atoms of the carboxylic group of the C-terminus, and if present
(iii) the heteroatom-containing side chains. A selection of
possible coordination modes for the Ag(I) complexes of amino
acids is given in Figure 2.
24
The binding can either occur
without or with the help of the side chain. In the first case, the
amino acid acts as bidentate ligand either via both termini (1
and 2) or via both oxygen atoms of the carboxylic group of the
C-terminus (3 and 4). If the silver binding occurs with the help
of the side chain, the amino acid can either act as tridentate (5)
or bidentate ligand (6 and 7).
In theory, monodentate coordination via only one
heteroatom is possible as well, leading to a 2:1 complex
between the amino acid and silver with respect to the preferred
metal ion coordination number of two of the metal ion.
However, because there is no stabilizing contribution due to
chelation for the monodentate forms, they are expected to be
higher in energy. Hopkinson divided the structures of silver−
amino acid complexes obtained by theoretical studies into three
major categories.
27
The first one represents five-membered
cyclic structures with the silver ion dicoordinated by the two
termini (1−3). Members of the second category coordinate
Figure 1. The 20 common natural α-amino acids classified according
to their side chain R.
Table 1. Calculated Enthalpies ΔH° and Free Energies ΔG°
(kJ mol
−1
) for the Amino Acid−Ag(I) Complexes
α-amino acid ΔH°
a
ΔG°
a
glycine (Gly) 206.1 170.1
alanine (Ala) 212.8 176.8
valine (Val) 216.1 181.0
leucine (Leu) 219.5 185.6
isoleucine (Ile) 221.1 188.9
serine (Ser) 224.5 190.2
cysteine (Cys) 230.2 194.4
threonine (Thr) 233.2 199.8
aspartic acid (Asp) 232.4 199.0
proline (Pro) 234.5 199.8
phenylalanine (Phe) 236.2 198.6
glutamic acid (Glu) 239.9 203.1
tyrosine (Tyr) 239.9 202.3
asparagine (Asn) 250.8 217.4
tryptophan (Trp) 260.0 221.5
methionine (Met) 262.1 219.9
glutamine (Gln) 264.2 225.5
histidine (His) 284.2 249.1
lysine (Lys) 296.8 260.8
arginine (Arg) 336.5 279.8
a
Estimated from ref 23.
http://doc.rero.ch
Ag(I) by three or four atoms, yielding structures with multiple
rings (5), while the members of the last group bind Ag(I) via
the carboxylate of their zwitterionic form in a bidentate fashion
(4). Further on, the members of the first two categories have
been described as charge-solvated because the silver ion is well
solvated by the amino acid in the gas phase, while the last group
forms salt-bridge structures.
22
Structure 1 was found to be the
predominant coordination mode of unpolar aliphatic amino
acids with hydrogen or alkyl groups in their side chain. The
Ag(I) is symmetrically coordinated by the two termini with
almost identical Ag−N and Ag−O distances, for example, 2.418
and 2.390 Å in case of the Ag(I)−glycine complex.
26
The
Ag(I)−proline complex is the only case where the salt bridge
structure 4 is favored over the charged solvated forms.
23
This is
probably, on the one hand, due to the constraints of the five-
membered ring that do not allow the optimum orientation of
the lone pair on the nitrogen atoms, and, on the other hand,
due to the larger basicity of the secondary amine. All of the
remaining amino acids are predicted to coordinate in a
tridentate fashion with side chain participation (5). In these
cases, the silver-coordination bonds with the termini are
elongated as compared to the bidentate structures of the
unpolar, aliphatic amino acids (by ca. 0.050 Å) due to the
participation of the third coordination site. For the four amino
acids possessing the highest proton affinities (Arg, Lys, His, and
Gln), the additional Ag− X bond is calculated to be shorter than
those with the termini, indicating a stronger binding of the side
chain.
After having discussed the theoretical work on coordination
of amino acids toward silver, we will in the next subsection
focus on experimental efforts undertaken in this research area.
2.1.2. Silver Binding to Amino Acids − Experiment. To
our current knowledge, the only experimental data for the
Ag(I) affinity of all 20 naturally occurring amino acids (with the
exception of Cys) have been provided by Siu and co-workers,
28
who also determined the binding energies of the silver ion to
simple alcohols and amides.
27
They determined the relative
silver ion affinities of the amino acids using relative ΔG° values
obtained by the kinetic method developed by Cooks and co-
workers.
29
The basis of this method is the competitive
decomposition of a set of ion-bound heterodimers, whose
general structure in case of the determination of silver binding
affinities can be described as [Aaa
x
AgAaa
y
]
+
(Aaa
x
and Aaa
y
representing two different amino acids). The so-obtained
experimental data
28
are in good agreement with the previously
described DFT calculations
23
(Table 1).
Considering the number of amino acids, possible coordina-
tion modes (see section 2.1.1) and crystallization conditions,
the structural information for silver−amino acid complexes
based on X-ray structures is still rare. This is at first glance quite
surprising because a detailed analysis of the coordination
chemistry of amino acids is a prerequisite for the understanding
of bioinorganic reaction mechanisms since they serve as model
compounds for peptide−silver interactions.
30
However, the
relatively little amount of published crystal structures
summarized in Table 2
31−41
can be explained by the fact that
the resulting compounds are difficult to synthesize because they
are light sensitive in solution and poorly soluble in common
solvents.
39
Nomiya and co-workers recently showed that for
alanine and asparagine the light stability of the silver complex
can be enhanced with additional triphenylphosphine ligands
38
or by acetylation of the N-terminus
40
of the amino acid. To
further classify the binding modes of Ag(I) in these structures
(with the exception of sulfur-containing compounds), the
following classification was proposed:
39
the silver ion is
involved (i) only in Ag−O bonds (bond length <2.6 Å),
42
(ii) only in Ag−N bonds (bond length <2.5 Å),
42
or (iii) in
both Ag−O and Ag−N bonds.
The summary in Table 2 shows that the latter class is the
most frequent one. Further on, it reveals that in the published
structures the coordination geometry around the silver ion is, as
expected, preferentially linear, leading either to 1D coordina-
tion polymers or to dimeric structures (e.g., for a silver to
amino acid ratio of 1:2). For some of the presented structures,
the Ag−Ag distance is so short that the interaction of both d
10
ions becomes discussable
43,44
because the bonds are shorter
than the van der Waals diameter of 3.44 Å
45
and in the range of
the distance of 2.88 Å
46
found in metallic silver. It is surprising
that among the few published structures there is a reasonable
amount of structures with amino acids, which were calculated
to be weak silver binders because they allow no side chain
participation.
2.1.3. Silver Binding to Peptides. Going from amino
acids to peptides, the number of coordination sites per
molecule increases and the backbone amide can function as
an additional coordination site as well. Shoeib and c o-
workers
26,47
investigated the structure and free energy of di-
and tripeptides by means of DFT calculations. For the
investigated silver−peptide complexes, they found structures
at low energy minima with coordination numbers between two
and four. Tricoordinated systems were found in the case of
dipeptides, while tri- and tetracoordination were observed for
tripeptides. As one might expect, experimental data based on X-
ray crystallography are even less abundant for silver−peptide
complexes than for amino acids. To our current knowledge, the
only published structure is the one published by Acland and
Freeman
31
for Gly−Gly, the simplest of all dipeptides. At pH =
6, this dipeptide forms a dimeric complex 8 in which two silver
Figure 2. Selection of possible coordination modes between Ag(I) and
amino acids. R represents the side chain of the amino acid in general, if
it does not participate in the silver binding, while X represents the
coordinating heteroatom present in the side chain.
http://doc.rero.ch
Table 2. Overview of Single-Crystalline Silver − Amino Acid Complexes Analyzed by X-ray Diffraction
amino acid
space
group structure important distances [Å]
coordination geometry around
Ag(I) formula ref
L-Ala P2
1
1D-polymer → 3D-network via CH−π interactions O−Ag 2.111(5) linear (O−Ag−N) [Ag(C
3
H
6
NO
2
)] 34
N
i
−Ag 2.125(7)
DL-Ala (+PPh
3
) P
1
centrosymmetric dimer → 1D polymer through H-bonding O1−Ag 2.4697(16) distorted tetrahedron {[Ag[DL-Ala)(PPh
3
)]2H
2
O} 38
N1−Ag 2.243(2)
O2
i
−Ag 2.6079(18)
P−Ag 2.3330(8)
β-Ala P2
1
/n centrosymmetric dimer Ag−Ag 2.855(4) linear (O−Ag−O) [Ag(C
3
H
7
NO
2
)](NO
3
) 32, 33
O1−Ag 2.198(19)
O2
i
−Ag 2.210(19)
L-Asn C222
1
stair-like chiral polymer O1−Ag 2.103(2) linear (O−Ag−N) [Ag(C
4
H
7
N
2
O
3
)] 36
N
i
−Ag 2.131(2)
Ag−Ag 3.437(3)
D-Asn C222
1
stair-like chiral polymer O1−Ag 2.106(2) linear (O−Ag−N) [Ag(C
4
H
7
N
2
O
3
)] 36
N
i
−Ag 2.131(2)
Ag−Ag 3.437(3)
DL-Asn (+PPh
3
) P
1
centrosymmetric dimer → 2D polymer through H-bonding O1−Ag 2.4574(18) distorted tetrahedron {[Ag[DL-Ala)(PPh
3
)]2H
2
O} 38
N1−Ag 2.203(2)
O2
i
−Ag 2.6849(18)
P−Ag 2.3196(7)
DL-Asp P2
1
/c centrosymmetric dimer → self-assembly into stair-like polymer Ag−Ag 2.781(1) linear (O−Ag−O) [Ag
2
(C
8
H
12
N
2
O
8
)]1.5H
2
O36
O1−Ag 2.234(3)
O2
i
−Ag 2.223(3)
O−3
ii
−Ag 2.403(4)
L-Cys(Me) P2
1
2
1
2 1D-polymer S−Ag 2.8436(7) distorted tetrahedron [Ag(L-Cys(Me)] 40
O−Ag 2.590(2)
O
i
−Ag 2.188(2)
N−Ag 2.212(2)
Gly P
1
1D-polymer O−Ag 2.11 linear (O−Ag−N) [Ag(Gly)] 31
N
i
−Ag 2.14
Gly P2
1
/n 1D-polymer O−Ag1 2.12 linear (O−Ag−O; N−Ag−N) [Ag(Gly)]0.5H
2
O31
O
i
−Ag1 2.13
N
(mean)
−Ag2 2.15
Gly P2
1
/n helical polymer Ag−Ag 3.241 linear (O−Ag−O; N−Ag−N) [Ag(Gly)]
2
H
2
O36
O2−Ag2 2.134(4)
O3−Ag2 2.115(4)
N1−Ag1 2.152(4)
N2−Ag1 2.147(4)
L-His P2
1
left-handed helical polymer N
amino
i
−Ag 2.125(4) linear (N−Ag−N) [Ag(C
6
H
8
N
3
O
2
)]0.2EtOH 35
N
π
−Ag 2.097(4)
D-His P2
1
right-handed helical polymer N
amino
i
−Ag 2.131(3) linear (N−Ag−N) [Ag(C
6
H
8
N
3
O
2
)] 39
N
π
−Ag 2.092(3)
DL-His C2/c dimeric structure → coordination polymer through Ag−Ag interactions Ag1−Ag1 3.0385(4) trigonal bipyramidal [Ag(C
6
H
8
N
3
O
2
)]
2
6H
2
O39
http://doc.rero.ch
