ARTICLE
Auteur(s) : Dave Van Den Plas, Kris
De Smet, Dimitri Lens, Philippe Sollie
Research Department, Flen Pharma nv, Drie Eikenstraat 661,
B-2650 Edegem, Belgium
accepté le 27 Février 2008
The ultimate goal of wound management and therapy is fast
healing and re-epithelialisation with minimal complications. In
this way the risk of reduced functionality and aesthetic discomfort
greatly diminishes. In order to realize this objective, a great
deal of attention is paid to preventing infection of the wound.
Provided that the patient’s immune status can handle the bacterial
burden, the presence of a certain number of micro-organisms
generally does not impair wound healing. As a rule of thumb,
infection could negatively influence wound healing once a microbial
burden of 105 micro-organisms/g of tissue is reached [1,
2]. Signs of infection are most often swelling, redness, an
increased amount of wound exudates and increased wound pain caused
largely by the activation of immune cells, resulting in a delay in
wound healing and general discomfort to the patient.
Micro-organisms impede wound healing by competing for
nutritional compounds, by secreting toxins, disrupting the
inflammatory cytokine balance in favour of pro-inflammatory
mediators and degrading newly formed tissue, and in general give an
unpleasant odour to the wound [3, 4]. One strategy gaining renewed
attention for fighting the threat of microbial infection and
preventing wound sepsis is the use of silver. Traditionally, two
main products have been used for this purpose. Silver-nitrate is
active against a variety of micro-organisms and is used in a
concentration of 0.5% on patients with extensive burns [5]. Silver
sulfadiazine combines the inhibitory action of the silver with the
antibacterial effect of sulfadiazine [6].
In the years following the introduction of these products,
however, more concerns arose about their safety and disadvantages.
The main complication occurring during treatment with
AgNO3 was a drop in serum sodium and chlorine due to ion
exchange between Ag+ and Cl−,
HCO3−, CO3− and protein
anions, leading to the production of very slightly soluble or
insoluble salt solutions. Secondly, it was seen that during
AgNO3 treatment all objects which came into contact with
the AgNO3 coloured black on exposure to light [7].
Thirdly, Bader [8] found elevated silver levels in the kidneys,
spleen, liver and muscles of two patients on post-mortem
examination.
To resolve these problems, wound-care companies searched for
improved products combining the strength of silver with
technological advances in wound dressings. The new products offer
the opportunity of a slow release of silver enabling less frequent
changes of dressings. The result is a myriad of Ag-dressings on the
market (for review see [3]). Although the cytotoxicity of these
dressings is reported to be less pronounced, some authors do
describe reduced viability of cells after contact with such silver
compounds [9-11].
In order to assess the ambiguity surrounding the subject, we
decided to check the cytotoxicity of various commercially available
Ag-dressings, all profoundly different in both composition and type
of silver. Comfeel®-Ag (Coloplast) is a sticky
hydrocolloid plate containing a silver complex which, on contact
with wound exudates, releases the silver. Seasorb®-Ag
(Coloplast) is an alginate dressing containing calcium alginate
(fibres), sodium carboxymethylcellulose (CMC) and an ionic silver
complex which, in the presence of wound exudates, releases silver
ions for a period of 7 days. Acticoat® 7
(Smith&Nephew) is a silver dressing consisting of 5 layers (2
internal absorbing layers and 3 silver-containing polyethylene
layers) in which the silver is present in the form of
Nanocrystalline™ silver. The dressing is reported to be active for
7 days. Biatain®-Ag (Coloplast) is a polyurethane foam
dressing with an antibacterial silver complex homogeneously
dispersed in its structure.
Materials and methods
Materials
Silver dressings were purchased from the respective manufacturers.
Comfeel®-Ag, Seasorb®-Ag and
Biatain®-Ag were bought from Coloplast and
Acticoat® 7 from Smith&Nephew. DMEM, FBS,
Penicillin/Streptomycin, L-Glutamine were all purchased from Sigma
(Bornem, Belgium). The growth medium in which eukaryotic cells were
grown consisted of DMEM supplemented with 10% heat inactivated calf
serum, 4 mM L-glutamine, 100 U/mL streptomycin and
100 μg/mL penicillin. Bacteria were routinely grown on Tryptic
Soy Agar (TSA) and yeasts on Sabouraud Dextrose Agar (SDA) (Becton
Dickinson (BD), Erembodegem, Belgium) plates. Tryptic Soy (TS)
broth was used for liquid cultures. MTT
[3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] was
bought from Sigma (Bornem, Belgium) and a stock solution was made
(5 mg/mL). The final concentration in medium for cytotoxicity
tests was 0.5 mg/mL. HaCaT keratinocytes were kindly provided
by Prof. Dr J. Merregaert (Lab Molecular Biotechnology, University
of Antwerp). Mouse fibroblast NCTC clone 929 cells were purchased
from ECACC (Salisbury, UK) Polycarbonate cell culture transwell
inserts for 6-well plates (0.4 μ) Millicell®-PCF
were purchased from Millipore (Brussels, Belgium). Staphylococcus
aureus (ATCC6538), Escherichia coli (ATCC8739) and Candida albicans
(ATCC10231) were purchased from the Belgian Coordinated Collection
of Micro-organisms (Brussels, Belgium).
Cell death analysis in the presence of silver dressings
Cells were seeded subconfluently into 6 well plates and were grown
overnight at 37 °C/5% CO2 until 90%-100% confluency
was reached. Silver dressings were divided into 1.5 × 1.5 cm
dimensions and were weighed:
- – Acticoat® 7: (0.044 ± 0.005) gram;
- – Biatain®-Ag: (0.146 ± 0.010) gram;
- – Comfeel®-Ag: (0.471 ± 0.040) gram;
- – Seasorb®-Ag: (0.044 ± 0.005) gram.
The dressings were placed on the transwell inserts and overlaid
with 1.2 mL of growth medium. Cells grown in the presence of
transwell inserts alone were used as a control. The experimental
set-up ensured no direct contact between cells and dressings.
Because of this, consequent cytotoxicity can only be the result of
diffusion from the dressings under investigation (figure 1A).
Cells with dressings were incubated at 37 °C/5%
CO2 for 4 hours, after which the cells were analysed
visually. For cell death analysis the cells were incubated for 2
hours in the same experimental set-up as above. The type of cell
death programme was analysed using the Apoptosis Detection Kit
(Sigma, Bornem, Belgium) according to the manufacturer’s
instructions. In brief, cells were detached from the wells by a
short trypsin/EDTA incubation and pooled with the cells from the
supernatants. Next, cells were washed twice in PBS and the cell
pellet was re-suspended in AnnexinV binding buffer and propidium
iodide. Labelled cells were analysed by flow cytometry on a FACScan
(Becton Dickinson) at gates FL1/FL3. As a positive control, cells
were incubated in the presence of 5 mM hydrogen peroxide.
Bioactivity of silver from different dressings
Gram-positive S. aureus and Gram-negative E. coli cells were
dispersed in liquid broth (TS) to obtain an optical density of 600
nm (OD600) of 0.1; C. albicans cells were re-suspended
to obtain an OD600 of 0.5.
Dressings were dispersed in liquid TS medium at 40 mg/mL
for 24 h at 37 °C/230 rpm. Afterwards, 1/2 serial
dilutions of the ‘extracts’ were made in microbial growth media.
One hundred microlitres of the freshly made dilutions were poured
into 96 well plates and 2 μL of the microbial suspensions were
added to the mixture. Wells with only liquid growth medium served
as a positive control for growth. Micro-organisms were allowed to
grow for 24 h at 37 °C. Growth was measured
spectrophotometrically at 600 nm (MRX II, Dynex Technologies,
US) and MIC50 values were calculated.
Zone of inhibition
S aureus, E coli and C albicans were susupended in sodium
chloride-peptone (OD600 = 1) and plated on agar plates.
Dressings were divided into pieces of 1 cm × 1 cm and
placed directly on top of the agar plates. Plates were incubated
for 24 hours at 37 °C. After incubation the zone of inhibition
surrounding the dressings was measured.
Statistical analysis
The results are the mean and standard deviation of at least 3
independent experiments. The data were analysed by the Student
t-test.
Results
Influence of active silver on keratinocytes
To verify the influence of active silver from dressings on cells in
vitro, we visually analysed the cells after 2 h incubation in
the presence of various dressings. Incubation was as described in
Materials and Methods (figure 1A). After 4h of
incubation, a short time relative to the indented contact time on
wounds, visual inspection of the cells clearly showed that for all
but one dressing (Biatain®-Ag) cells were rounding up or
deforming, indicative of a cell being in a stress situation (figure 1B). Although
the Biatain®-Ag dressing least influenced cell
morphology, it should be noted that this dressing strongly absorbed
the growth medium throughout the experiment. MTT analysis confirmed
our visual observations in that the silver dressings significantly
influence cell survival compared to control cultures
(p < 0.01). Cell loss in Biatain®-Ag
treated cultures was significantly less severe
(p < 0.03) compared to the other dressings (figure 1C).
Both types of analysis demonstrate that all tested silver
dressings release their active silver complexes which then diffuse
into the surrounding medium thus influencing cell morphology and
survival.
Analysis of cell death programme induced by the various
dressings
Because of the fast onset of cell death observed, we decided to
investigate in more detail the cell death process induced by the
different dressings. As programmed cell death is typically induced
within a few hours of the challenge [12], we decided to perform
flow cytometric analysis after 2 h incubation of cells in an
experimental set-up analogous to figure 1A. After
incubation, cells were double stained with AnnexinV-FITC, a marker
for early apoptosis (phosphatidylserine exposure to the outer cell
surface) and propidium iodide, a marker for necrosis (cell membrane
damage). In the cytometric analysis plot (figure 2), the four
quadrants are indicative of different types of cell death: 1 for
necrosis, 2 for late apoptosis/necrosis, 3 for living cells and 4
for early apoptosis.
A marked difference in response was seen for the two cell lines
(table 1).
For the keratinocyte HaCaT cells (figure 2A), a statistical
difference in cell survival was seen for all dressings.
Seasorb®-Ag roughly showed equal amounts of cells in
early apoptosis and late apoptosis/necrosis, whereas in that same
time frame Acticoat® 7 and Biatain®-Ag
cultures were mostly in early phase apoptosis.
Comfeel®-Ag challenged cells were in late phase
apoptosis or necrosis. Incubation with the pro-oxidant hydrogen
peroxide (5 mM) resulted in a high number of cells in early
apoptosis, although a significant number of cells were also in late
apoptosis (quadrant 2) and necrosis (quadrant 1 and 2). In this
experimental set-up, Biatain®-Ag was the least cytotoxic
when compared with the other products. More keratinocytes were
dying in the control set-up compared with fibroblasts.
Fibroblast 142BR cells overall (figure 2B, table 1) showed higher survival rates than HaCaT
cells. Cell survival was significantly reduced by the silver
dressings compared with control cells, with most surviving cells
found in Biatain®-Ag treated cultures and least in
Comfeel®-Ag treated cultures. A small but significant
number of cells challenged with Biatain®-Ag or hydrogen
peroxide were in early apoptosis. All silver dressing treated
cultures showed significant numbers of cells in necrosis and late
apoptosis or necrosis (quadrant 1 and quadrant 2 respectively).
Table 1 Flow cytometric analysis of HaCaT keratinocytes
and Fibroblast 142BR. Values are means of three experiments
expressed as percentage; numbers between brackets represent
standard deviations. * p < 0.05; ** p < 0.03
|
Necrotic
|
Late apoptotic or necrotic
|
Living
|
Early apoptotic
|
|
HaCaT
|
142BR
|
HaCaT
|
142BR
|
HaCaT
|
142BR
|
HaCaT
|
142BR
|
|
Control
|
0.40 (0.15)
|
1.62 (0.37)
|
4.86 (1.00)
|
0.02 (0.02)
|
83.37 (6.47)
|
98.33 (0.38)
|
11.36 (7.50)
|
0.05 (0.07)
|
|
H2O2
|
1.62 (0.33)**
|
9.41 (1.28)**
|
27.27 (6.96)*
|
42.22 (1.67)**
|
26.00 (10.62)**
|
43.63 (5.92)**
|
45.13 (4.50)*
|
1.39 (0.48)*
|
|
Acticoat® 7
|
0.75 (0.51)
|
16.58 (1.33)**
|
11.71 (1.14)**
|
19.07 (4.92)**
|
39.97 (8.78)**
|
63.73 (4.63)**
|
47.57 (8.22)*
|
0.61 (0.52)
|
|
Biatain®-Ag
|
0.38 (0.29)
|
4.47 (0.37)**
|
5.6 (0.89)
|
3.61 (0.16)**
|
58.47 (5.47)**
|
89.27 (0.67)**
|
35.53 (5.67)**
|
1.99 (0.73)*
|
|
Comfeel®-Ag
|
1.23 (0.82)
|
36.87 (4.78)**
|
58.50 (10.92)*
|
34.04 (13.18)**
|
16.42 (9.17)**
|
32.27 (4.82)**
|
23.97 (4.18)
|
0.28 (0.24)
|
|
Seasorb®-Ag
|
0.08 (0.07)
|
28.99 (3.52)**
|
51.27 (6.81)**
|
13.71 (0.26)**
|
2.99 (1.33)**
|
56.80 (3.16)**
|
45.67 (7.53)*
|
0.50 (0.33)
|
Anti-microbial activity of the silver dressings on agar plates:
zone of inhibition
In order to compare the anti-microbial activity of the four silver
dressings by diffusion, S. aureus, E. coli and C. albicans
suspensions on agar plates were challenged with the different
silver dressings on plate. Figure 3 and table 2 clearly show that the susceptibility of the
micro-organisms to the dressings is very different. Whereas E. coli
was susceptible to Biatain®-Ag, S. aureus and C.
albicans were to a much lesser extent. E. coli was also most
susceptible to Comfeel®-Ag, but was not towards
Acticoat® 7. The latter dressing was active against all
three micro-organisms.
Table 2 Zone of inhibition measurements in mm. Results
are expressed as mean + SD of three independent
experiments
|
Staphylococcus aureus
|
Candida albicans
|
Escherichia coli
|
|
Acticoat® 7
|
2.10 (0.10)
|
3.67 (0.29)
|
1.83 (0.76)
|
|
Biatain®-Ag
|
0.05 (0.07)
|
0.05 (0.07)
|
2.25 (0.35)
|
|
Comfeel®-Ag
|
2.83 (1.04)
|
3.90 (0.10)
|
6.00 (0.50)
|
|
Seasorb®-Ag
|
0 (0.0)
|
0 (0.0)
|
0.83 (0.58)
|
Biological effect by dilution
As every dressing has its own density, this results in varying
amounts of silver being released per surface area and so we decided
to suspend a fixed amount of the silver dressings in growth medium
for 24 hours and make serial dilutions of the “extract”. Next, a
fixed number of micro-organisms were challenged with the dilutions
for 24h and growth was assessed spectophotometrically.
From table 3 it is clear that for the
three micro-organisms tested, the smallest MIC50 values
were obtained for the Acticoat® 7 dressing compared with
the other dressings. These differences between Acticoat®
7 and the other dressings were statistically different for all
organisms (p < 0.05) other than Seasorb®-Ag on E.
coli, where the statistical difference was p = 0.08. For S. aureus,
a statistical difference was seen for Biatain®-Ag and
Seasorb®-Ag (p < 0.01) but not for
Biatain®-Ag versus Comfeel®-Ag or
Comfeel®-Ag versus Seasorb®-Ag. For C.
albicans, a statistical difference was seen between
Comfeel®-Ag and Seasorb®-Ag (p < 0.02) but
not for Biatain®-Ag versus Comfeel®-Ag or
Biatain®-Ag and Seasorb®-Ag. For E. coli, no
statistical difference was noticed between Biatain®-Ag,
Comfeel®-Ag and Seasorb®-Ag.
Table 3 MIC50 values for dressings extracted
for 24 h in bacterial growth media (see materials and
methods). Values are means (± SD) of at least three independent
experiments expressed as mg/mL in bacterial growth medium
|
Staphylococcus aureus
|
Candida albicans
|
Escherichia coli
|
|
Acticoat® 7
|
7.6 (1.8)
|
7.9 (1.6)
|
8.1 (1.3)
|
|
Biatain®-Ag
|
29.9 (1.0)
|
28.5 (6.4)
|
26.7 (1.9)
|
|
Comfeel®-Ag
|
23.7 (7.8)
|
37.3 (0.8)
|
25.9 (2.2)
|
|
Seasorb®-Ag
|
21.9 (0.2)
|
20.8 (3.1)
|
23.7 (8.6)
|
Conclusion
As reported by Atiey et al. [3], no two silver dressings are alike.
This is also true of the form in which the silver is incorporated
in the various dressings. Silver can be present in a variety of
forms: It can be either metallic (Ag0) or ionic
(Ag+). In the latter case, many silver salts are
insoluble and hence precipitate from water solutions.
Nanocrystalline silver consists of metallic silver (Ag0)
with altered grain boundaries and it has been suggested that this
is a third form of silver [13]. These different forms or states of
silver therefore make it difficult to compare the
cytotoxicity/bactericidicity of the silver dressings on the market.
In order to circumvent the problem, it seemed justified that we
compare the relative cytotoxicity/bactericidicity of the different
dressings in a well-defined water solution (growth medium),
irrespective of the absolute amount of silver in the dressings. We
further hypothesized that this would mirror more closely the
situation in wound conditions as compared with absolute amounts of
the metal.
In a first experimental set-up, we placed silver dressings on a
Transwell insert with cells growing beneath it. In such a set-up,
emerging cytotoxicity can only be the result of diffusion from
silver from the dressings. It was observed that both keratinocyte
cells (HaCaT) as well as fibroblast cells (142BR) were sensitive to
all silver dressings. However, more surviving fibroblasts were
observed compared to keratinocytes. This difference in sensitivity
is probably the result of the different metabolisms and oxidative
sensitivities of the two cell types. Interestingly,
Biatain®-Ag had the least influence on both cell types,
although this could be attributed to the high absorption of the
dressing resulting in a delayed equilibrium concentration of silver
in the growth medium. Concurrent with this hypothesis, we clearly
observed cell stress when leaving the cells for 24 h in
contact with this dressing (data not shown).
We hypothesize that the amount of cells in early apoptosis
compared to late apoptosis or necrosis is dependent on the amount
of biologically available silver in the medium. Indeed we observed
for Biatain®-Ag, where high absorption of growth medium
was noticed, the highest ratio of cells in early apoptosis after
2 h incubation. Comfeel®-Ag on the other hand,
demonstrating little absorption of growth medium, resulted in the
highest number of cells in late apoptosis or necrosis after
2 h incubation. Interestingly, this dressing was the heaviest
per surface unit, which could well result in a high silver release
at equilibrium state. This dressing also showed the largest zones
of inhibition on agar plates.
A similar observation was made by Burd et al. [14] when they
demonstrated cytotoxicity for different silver dressings. This
group, as well as Brett [15] and Walker and colleagues [16],
clearly established that silver release is strongly dependent on
the incubation conditions of the dressings. This could explain the
great discrepancy in our observations for Seasorb® Ag in
the high mortality of HaCaT cells, yet absence of zones of
inhibition, but MIC50 values comparable to other
dressings.
In our two experimental set-ups for the antimicrobial activity
of the dressings, no clear correlation could be found. Some
dressings showed poor zones of inhibition, but clearly had
anti-microbial effects in the 24 h extraction procedure. This,
however, should not be a surprise. In the extraction procedure,
complete equilibrium between the silver in the dressing and the
surrounding medium should have been established after 24 h, as
for Burd et al. [14] reported a pre-incubation of only 10 minutes
in different media. In the case of diffusion on agar plates,
however, only at the place in direct contact between the plates and
the silver dressing, was there a release of silver. Concurrent with
this hypothesis, Comfeel®-Ag, on average 5-10 times
heavier per square unit, was the most potent of all dressings in
this experimental set-up. An interesting observation was made for
Acticoat® 7. It performed well in both anti-microbial
set-ups. This could be related to its reported nanocrystalline
structure [13, 17, 18] which might be less susceptible to
inhibition by organic materials present in the bacterial growth
media.
In conclusion, our experiments clearly show that the
antimicrobial activity of silver dressings is accompanied by
cellular cytotoxicity. This is in line with the conclusion reported
by Atiyah and colleagues [3] who stated that silver-based products
cannot discriminate between healthy cells and pathogenic bacteria.
As our experimental data were performed on time scales which are
relatively short compared with in vivo situations, it therefore
seems that silver dressings should be used only on critically
contaminated wounds rather than used de facto. A delay in
reepithelialisation could result from cytotoxicty of the newly
formed, sensitive, keratinocytes, as has been shown by Burd et al.
[14] in in-vivo situations. The search for antimicrobial dressings
with good antimicrobial activity but with minimal toxicity towards
eukaryotic cells should therefore continue.
Acknowledgement
We wish to thank Dr P. Ponsaerts of the University of Antwerp for
his assistance in the flow cytometry analysis. Financial support:
none. Conflict of interest: none.
References
1 Robson MC. Wound infection. A failure of wound healing
caused by an imbalance of bacteria. Surg Clin North Am 1997; 77(3):
637-50.
2 Robson MC. Infection in the surgical patient: an
imbalance in the normal equilibrium. Clin Plast Surg 1979; 6(4):
493-503.
3 Atiyeh BS, Costagliola M, Hayek SN,
Dibo SA. Effect of silver on burn wound infection control and
healing: review of the literature. Burns 2007; 33(2): 139-48.
4 Warriner R, Burrell R. Infection and the chronic
wound: a focus on silver. Adv Skin Wound Care 2005; 18(Suppl 1):
2-12.
5 Moyer CA, Brentano L, Gravens DL,
Margraf HW, Monafo WW. Treatment of large human burns
with 0.5% silver nitrate solution. Arch Surg 1965; 90: 812-67.
6 Stanford W, Rappole BW, Fox Jr. CL.
Clinical experience with silver sulfadiazine, a new topical agent
for control of pseudomonas infections in burns. J Trauma 1969;
9(5): 377-88.
7 Klasen HJ. A historical review of the use of silver in
the treatment of burns. II. Renewed interest for silver. Burns
2000; 26(2): 131-8.
8 Bader KF. Organ deposition of silver following silver
nitrate therapy of burns. Plast Reconstr Surg 1966; 37(6):
550-1.
9 Hussain SM, Hess KL, Gearhart JM,
Geiss KT, Schlager JJ. In vitro toxicity of nanoparticles
in BRL 3A rat liver cells. Toxicol In Vitro 2005; 19(7):
975-83.
10 Poon VK, Burd A. In vitro cytotoxity of silver:
implication for clinical wound care. Burns 2004; 30(2): 140-7.
11 Lam PK, Chan ES, Ho WS, Liew CT. In vitro
cytotoxicity testing of a nanocrystalline silver dressing
(Acticoat) on cultured keratinocytes. Br J Biomed Sci 2004; 61(3):
125-7.
12 Teraki Y, Shiohara T. Apoptosis and the skin. Eur J
Dermatol 1999; 9(5): 413-25.
13 Birringer R. Nanocrystalline materials. Mater Sci Eng
1989; A117: 33-43.
14 Burd A, Kwok CH, Hung SC, Chan HS,
Gu H, Lam WK, Huang L. A comparative study of the
cytotoxicity of silver-based dressings in monolayer cell, tissue
explant, and animal models. Wound Repair Regen 2007; 15(1):
94-104.
15 Brett DW. A discussion of silver as an antimicrobial
agent: alleviating the confusion. Ostomy Wound Manage 2006; 52(1):
34-41.
16 Walker M, Cochrane CA, Bowler PG,
Parsons D, Bradshaw P. Silver deposition and tissue
staining associated with wound dressings containing silver. Ostomy
Wound Manage 2006; 52(1): 46-50.
17 Djoki SS, Burrell RE, Field DJ. An
electrochemical analysis of thin silver films produced by reactive
sputtering. J Electrochem Soc 2001; 148(3): 191-6.
18 Taylor PL, Omotoso P, Wiskel JB,
Mitlin D, Burrell RE. Impact of heat on nanocrystalline
silver dressings. Part II. Physical properties. Biomaterials 2005;
26(35): 7230-40.
|
Printable version
Version PDF
