IntroductionThe skin acts as a primal barrier against loss of water and endogenous substances from the human body; furthermore, the skin prevents the ingress of exogenous pathogens and pollutants. The outermost layer of the skin, the stratum corneum (SC) represents the main barrier of the skin and prevents entrance of potentially harmful substances into the viable skin layers (underlying epidermis and dermis). While a large number of studies have investigated penetration of organic and inorganic solutes, little is known about percutaneous penetration of particles. At present there is a growing interest in this area because of the ascent of nanomedicine and drug delivery by nano-carriers (Jensen et al., 2011; Labouta et al., 2011; Lademann et al., 2007; Lieb et al., 1992; Schaefer and Lademann, 2001). As the SC is regarded as a primary barrier for skin absorption, the amount of a substance measured at different depths of the SC may be used to investigate percutaneous penetration. Tape stripping is a technique that allows for sequential removal of SC cell layers using an adhesive tape and has become a basic method to study the penetration and the reservoir behavior of topically applied substances (Bronaugh & Stewart, 1985; Jakasa et al., 2006; Lademann et al., 2009).
The role of silver as topical agent to reduce the colonization of bacteria is well known and is generally believed to be related to the release of silver ions from material. In the last decades the use of silver nanoparticles (AgNPs) as source of silver ions is strongly increased in various medical and consumer products (Ansari et al., 2011; Martinez-Castanon et al., 2008) and a number of studies address their toxicity (Arbabi et al., 2013; Korani et al., 2011, 2015). Since even a thin layer of metallic silver on the material could impart a nanostructured surface (Oriňáková et al., 2013), the characterization of silver in these materials is needed. Often, the presence of AgNPs in commercially available products is not stated by the manufacturer. Therefore, the aim of this study was to investigate the presence of AgNPs in a garment (i.e. textile sleeve covering forearm skin) and subsequently whether AgNPs are present in the skin after dermal exposure. As medical devices containing AgNPs are increasingly being used to combat frequent bacterial infections in atopic dermatitis (AD) (Jain et al., 2009; Keck & Schwabe, 2009;Keck et al., 2014; Kim et al., 2008), in the present study we included healthy subjects and AD patients. The European Scientific Committee on Cosmetics and Non-Food Products (SCCNP) nanotechnology report declares that the data on skin penetration of nanoparticles through diseased skin are limited. An impaired skin barrier will not only increase the amount of a compound that penetrate across the skin but will possibly also allow for penetration of particles that normally would not be able to penetrate due to their large size (Kezic & Nielsen, 2009). This study is a part of a larger previously published investigation (Pluut et al., 2015) in which percutaneous penetration of total silver (i.e. ionic and particulate form) has been investigated in 20 healthy volunteers and 20 AD patients. From the cohort of patients described in Pluut et al. (2015), tape strips from seven AD and seven healthy subjects after the fifth day of exposure were randomly chosen and further investigated for the presence of silver particles in the skin by using more advanced techniques, atomic force microscopy (AFM) and scanning electron microscopy with energy dispersive X-ray analysis (SEM-EDX), than were used in Pluut et al. (2015).
MethodsDermal exposureThe subjects wore on their forearm a silver-containing sleeve for 8 h during the night on five consecutive days. As a control, on the other forearm the participants wore a placebo (silver free) sleeve. The silver garment (Medical device Class I) consisted of 79% Micromodal fiber, 13% Polyamide Ag and 7% Lycra. The area of the sleeve was on average 600 cm2. To determine the concentration of silver in the material, samples of 1 cm2 from three sleeves were digested for 24 h in 5 mL of nitric acid 69% v/v. After 24 h, the resulting solutions were diluted to 50 mL with Milli-Q water (Bedford, MA) and analyzed for silver content by inductively coupled plasma-atomic emission spectroscopy (ICP-AES). The average concentration of silver was 3.6 (±0.1 SD) (w/w).
Study populationThe characterization of silver particles has been performed in the SC tape collected from 14 Caucasian subjects (seven healthy subjects and seven subjects with AD), 10 female and 4 male, with a mean age of 31.6 years (range 21–68 years). The participants gave a written informed consent before starting the study, which was approved by the Medical Ethics Committee of the Academic Medical Center, University of Amsterdam (METC 2013_031).
Tape stripping procedureAfter the fifth day of exposure, the layers of the SC were sequentially removed with circular adhesive tapes (3.8 cm2, standard D-Squame, Monaderm, Monaco, France). The tapes were consecutively attached to the volunteer’s underarm and pressed for 10 s with a constant pressure (225 g/cm2) using a D-Squame pressure instrument D500 (CuDerm, Dallas, TX, USA). An average number of 20 tapes was needed to remove the whole SC. The SC depth of a particular tape could be deduced from the amount of protein measured on each tape as described elsewhere (Jakasa et al., 2007; Mohammed et al., 2012). The tapes were stored separately in cryo vials at 4 °C prior to analysis by SEM and AFM. The concentration of silver on the tapes was determined as describe previously in Pluut et al. (2015). Briefly, 1 mL MilliQ water and 0.5 mL of ultrapure HNO3 (69%) was added to the tape. Samples were sonicated for 15 min and diluted to a final volume of 6.5 mL with MilliQ water. Only the first four tapes were analysed individually, other tapes were pooled (three or four tapes). The total silver concentration in tape strips was determined using inductively coupled plasma mass spectrometry (ICP-MS) (7500CE, Agilent Technologies, Santa Clara, CA, USA). The limit of detection of the method, LoD was 0.05 mg/L, RSD<5%).
Characterization of silver particles by SEM and AFMSEM with EDX analysis of the silver-containing and placebo textiles and SC tapes was performed blindly by three different laboratories. AFM analysis was performed by a single laboratory. For each sample (i.e. tape or material) at least three randomly selected areas were analysed.
FE-SEM-EDX – laboratory I (National Institute for Occupational Safety and Health, USA)Textile samples were first visually observed under a light microscope. Both the silver textile and placebo had the same construction – an outer surface composed only of white threads and an internal surface (side intended to contact skin) composed of white threads inter-woven with bundles of golden colored threads. Individual golden threads were pulled from the bulk textile pieces using clean tweezers and placed on double-sided conductive carbon tape mounted on an aluminum electron microscopy stub. All microscopy stub samples were sputter coated with gold/palladium prior to analysis to enhance conductivity. Each sample was visually inspected using a field emission scanning electron microscope (FE-SEM, Model S-4800, Hitachi, Tokyo, Japan) at varying magnifications. The elemental composition of particles was evaluated using EDX analysis (Quantax, Bruker Instruments, Berlin, Germany). Identification of elements was made by matching spectra to the EDX reference library.
SEM-EDX – laboratory II (Serend-ip GmbH, Germany)The samples were placed upon conductive carbon tape. To enhance the electrical conductivity, the samples were coated with a layer of gold approximately 5 nm thick. The images were taken in standard operation mode (15 kV) with a secondary electron detector by a LEO 1530 VP (Zeiss, Oberkochen, Germany) equipped with an EDX detector Oxford INCA energy 200 (Oxford Instruments Analytical, Wycombe, UK).
SEM-EDX – laboratory III (Department of Chemical and Pharmaceutical Sciences, University of Trieste, Italy)Analysis was performed by a Hitachi model TM-3000 – equipped with Oxford Instruments X-ray Microanalysis (Hitachi Ltd., Tokyo, Japan). The tapes were fixed rigidly onto a circular metallic sample holder by means of an adhesive tape (silver free). All the analyses were run at 15 kV. A prior analysis at 50× was performed to have an overview of the whole sample stub. SEM analysis has been done at magnification of 2000× to identify the presence of silver aggregates for 500 fields (corresponding to an area of about 4 mm2). Any aggregate was then characterized at magnifications up to 8000× and the EDX spectrum was recorded. The starting field of analysis was randomly chosen at the upper left corner of the sample stub, and then the resting fields were selected at grid points formed by evenly spaced horizontal and vertical lines, with the distance between fields being sufficient to prevent field overlap.
AFM – laboratory IV (Laboratory for Bioelectrochemistry and Surface Imaging, Ruđer Bošković Institute, Croatia)The AFM imaging was performed using a Multimode AFM with Nanoscope IIIa controller (Bruker, Billerica, MA, USA) with a vertical engagement (JV) 125 μm scanner using both, contact and tapping modes. Contact mode imaging was performed using silicon-nitride tips (NP-20, Veeco, nom. frequency 56 kHz, nom. spring constant of 0.32 N/m). The tapping mode was applied using silicon tips (RTESP, Veeco, resonance frequency 289–335 kHz, spring constant in the range 20–80 N/m). The maximal resolution of 512 samples per line was used if not stated otherwise. The processing and analysis of images was done by NanoScope™ software (Digital Instruments, version V614r1). All the measurements were performed under ambient conditions i.e. at room temperature and a relative air humidity of 50–60%. Contact mode was used to analyze the silver-containing garment following previously developed protocol (Ercegović et al., 2011). Samples of silver-containing garment (1 cm2) were fixed onto the sample holder (stainless steel, diameter 1.5 cm) immediately before imaging using adhesive tape. The sample surface areas investigated ranged from 25 μm × 25 μm down to 2 μm × 2 μm, so as to ensure that the high-resolution images (scan areas of 2 μm × 2 μm) capture the features of interest.
Tape samples with precut surface area of 0.7 × 0.7 cm2 were fixed onto the sample holder (stainless steel, diameter 1 cm) immediately before imaging using adhesive tape.
The AFM imaging was performed using tapping mode. The linear scanning rate was optimized between 1.0 and 2.0 Hz with scan resolution of 512 samples per line, except for large scans (15 μm × 15 μm) were resolution was 256 samples per line. Scanned areas were 225 μm2, 25 μm2 and 4 μm2 corresponding to 15 μm × 15 μm, 5 μm × 5 μm and 2 μm × 2 μm scan size, respectively. For each SC sample, we have recorded at least three randomly selected large areas. The large-area images were used to select the areas of interest that were imaged at high resolution. The high-resolution images were analyzed for presence of particles that were suspected to be silver particles, which were measured using section analysis (height profiles along indicated lines).
ResultsSilver particles in the garmentAFM images showed a grain-like morphology of the fiber surface with scattered particles of spherical or tetragonal shape (Figure 1). In placebo textile, no such particles have been found.
The height of particles was measured taking into account the curvature of the fiber (Figure 1(B)). We measured the height of 117 individual particles and performed statistical analysis using “R” software environment for statistical computing and graphics (R Core Team, 2013). Histogram of particles height distribution is given in Figure 1(C), and the summary of statistical analysis in Supplementary Table 1. Particles display Gumbel distribution with parameters a = 36.2 and b = 15.9.
Figure 2(A–C) shows the SEM images fiber surfaces of the silver-containing textile at different magnification before use. Generally, fiber surfaces on the textile before use had a uniform coating though there were occasionally uncoated spots on the fibers (Figure 2(A,B)). Closer inspection of the coating revealed the presence of nanoscale particles (Figure 2(C)); the size of individual particles that comprised the coating were not measured individually to obtain a size distribution, though qualitatively they are on the order of 10s to hundreds of nanometers in diameter. EDX analysis of coating on a fiber (Figure 2(D)) confirmed that the nanoscale particles that comprised the covering were made of silver (other elements identified were carbon and oxygen). EDX analysis did not identify silver particles on the uncoated spots of the garment textile or on the placebo textiles (data not shown). The EDX spectrum of the coating on the garment textile after use also contained only silver, carbon and oxygen signals (data not shown). were uniformly coated with particles, though as seen from Figure 2(A) and (B), there were occasionally uncoated spots on the fibers. SEM-EDX analysis of silver-containing fiber (Figure 2(D)) confirmed that the particles composing the covering were made of silver while no AgNP were identified on the placebo textiles.
No differences in the EDX spectra of the textile before and after its use were found. There also did not appear to be any difference in the morphology of the AgNPs on the fibers before and after use; however, qualitatively there seemed to be more areas on the “after” fibers that were missing the silver coating compared to “before” fibers though this observation is based on visual inspection of only a few fields of view (images not shown).
Silver particles on the SC tapesTable 1 summarizes the information regarding SEM-EDX analysis of inorganic aggregates found in healthy and AD subjects exposed to the silver-containing garment. Furthermore, for each investigated tape, the tape number and the corresponding depth of the SC from which the tape originates are given. In all tapes, which were collected from investigated skin site, also the silver concentration has been determined by ICP-MS. As the tape in which SEM or AFM imaging has been done could not be analyzed for silver, we show in Supplementary Figure 1 the profile of silver concentration (expressed as ng Ag/cm2) vs. tape number which is obtained from other tapes. The silver concentrations on all investigated tapes (up to number 20) were above the limit of the detection of the ICP-MS method (0.05 μg/L). Numerous particles, ranging in size from a few hundred nanometers to 2 μm were detected almost in all SC samples collected from the silver exposed forearm in both healthy and AD subjects.
The electron density of silver resulted in these particles appearing as bright spots (Figures 3 and 4); moreover, the shape of silver aggregates was very irregular often revealing the single components borders. An elemental mapping for silver has been done in a number of samples consistently showing that the bright spots in the SEM image correspond with high density of silver in the elemental mapping, which is further confirmed by EDX spectra (Figure 3). Consistently, the EDX of the sites without bright spots (Figure 4, spot 3) showed absence of silver. Furthermore, in the control samples (placebo exposed skin) no silver was detected, although on the first tape several dull gray particles were detected on skin cell containing K–Ca–S (not shown).
Although the number of silver particles in the deeper SC layers was clearly lower than in the uppermost layers Figure 5, they could be detected even in the last collected tape (no. 20) (Table 1, Figure 3).
In the uppermost SC layer, (tape no.1) aggregates were found on the cells as well between the cells, whereas in the deeper layers it seems that they were located mainly on the cell surface.
Most of aggregates observed in the SEM images showed in the EDX spectra the presence of silver as major component although some aggregates, in particular those found in the deeper layers showed next to silver also chlorine, calcium, potassium, sulfur and sodium (Table 1, Figure 5(A) and (B)). Example on the fifth tape shown in Figures 5(A) and (B), one particle contained silver (spectrum 1), three particles showed presence of silver, chlorine and sodium (spectrum 2, 3 and 4) while one particle showed no presence of silver but only potassium, chlorine and sodium (spectrum 5). Particles that do not contain silver differed in morphology from silver aggregates as they showed more regular and compact shapes.
No striking differences were found between healthy and AD subjects regarding the morphology and number of the silver aggregates revealed.
As the elemental composition cannot be verified by AFM, the identification of silver particles was based on the particle shape (sharp edges, tetragonal or spherical shape, clusters) and by the comparison between placebo and silver exposed samples. In addition, the recorded phase images often provide a clear distinction between “organic” and “inorganic” particles. Sometimes, large objects were encountered but the edges were not sharp. In such cases, we changed the imaging parameters by applying a larger force while imaging. This way we succeeded in peeling-off the organic layer (presumably lipid) uncovering sharp objects. This finding might indicate the pathway of particle penetration being lipid bilayers between the cells. Table 2 summarizes the results of AFM analysis performed on tapes of healthy and AD subjects exposed to the silver garment. According to AFM analysis, inorganic nanoparticles (30–100 nm), with a tetrahedral morphology characteristic for AgNPs (Pletikapić et al., 2012) were identified. Majority of the particles suspected to be silver were situated on the cells, fewer in the space between the cells (Table 2). Consistently with the SEM-EDX analysis, also AFM revealed presence of clusters and aggregates (e.g. Figures 6 and 7). In contrast to silver-containing garment, in all but one subject no particles that were suspected to be silver have been identified by AFM (Table 2). Consistently with SEM-EDX analysis, no significant differences were revealed between healthy and AD subjects regarding the number and morphology of nanoparticles and aggregates or agglomerates found.
DiscussionThis study was part of a larger study aiming to investigate dermal uptake of total silver (i.e. soluble and insoluble) in 20 AD patients and 20 healthy subjects (Pluut et al., 2015). In a number of subjects, next to dermal uptake we characterized the silver particles in the SC, which is described in this paper. The presence of Ag particles and their aggregates in different matrices was detected using AFM and SEM-EDX. The main advantage of AFM is high vertical resolution (1–2 nm), but on the other side absolute identification and elemental analysis is not possible. The first selection by using AFM was made according to the distinct morphology of the AgNPs and in comparison with the placebo samples. Next, textile and skin samples were analyzed by SEM-EDX. Both techniques used in this study require only minimal sample handling before analysis; therefore, the possibility of artifacts is strongly reduced. Moreover, SEM-EDX analysis were performed by three distinct laboratories with different instrumentations showing consistent results.
The garment characterization by AFM and SEM-EDX before and after dermal exposure revealed the presence of silver particles on the textile surface of nano- and micron-size. After five days exposure, the presence of silver particles of different size and composition was shown across the whole SC in both atopic and healthy skin, while these particles were not present in the placebo samples (with exception of one tape sample analysed by AFM). Interestingly, most of the Ag-particles on the tape 1 (approximately second–third layer of the SC) appeared between the cells, whereas the preferred localization on the deeper tapes was rather on cell surfaces. In some cases, the EDX spectra revealed also sulfur and chlorine as components of silver aggregates found in SC. The presence of chlorine in the aggregates could be ascribed to the reaction of silver ions released from the garment with chlorine from biological media, as no chlorine has been detected in the garment before exposure. The formation of silver chloride after immersion of silver-containing textiles in biological media was previously reported (Impellitteri et al., 2009; Stefaniak et al., 2014). Stefaniak et al. (2014) suggested that silver chloride binding on nanoparticles surface is more likely than silver chloride precipitation. The presence of sulfur in silver aggregates might reflect silver binding to cysteine residues of skin proteins as, for example, keratin which is highly abundant in the SC (van der Zande et al., 2012). Our findings are consistent with previous studies, which reported the presence of silver compounds (Ag–S and Ag–Se) in different organs after oral exposure to CH3CO2Ag and AgNO3 (Loeschner et al., 2011; Liu et al., 2010). Furthermore, the formation of Ag–S and Ag–Se in the skin was also reported in argyria patients (Massi & Santucci, 1998). Presence of silver agglomerates in the SC is consistent with our parallel in vitro study with the same textile in which silver and silver chloride aggregates of sizes of up to 1 μm were identified both in the epidermis and dermis (Bianco et al., 2015).
To date, there are very few in vivo studies investigating penetration of AgNPs in human skin after topical application (George et al., 2014; Prow et al., 2011; Zhu et al., 2015) and some papers reported the penetration of AgNPs in burned skin (Rigo et al., 2013; Trop et al., 2006; Vlachou et al., 2007; Wang et al., 2009). Consistently with our results, George et al. (2014) showed the presence of silver clusters up to 750 nm in size below the epidermis after the five-day dermal exposure of 16 healthy subjects to nanocrystaline silver dressing. Prow et al. (2011) reported by using reflectance confocal microscopy the presence of AgNPs aggregates in furrows and hair follicles after treatment with an AgNPs spray. In a recent study, Zhu et al. (2015) showed by using a sensitive surface-enhanced Raman scattering that the penetration depth of coated AgNPs (70 ± 20 nm) can exceed the SC. As in that study AgNPs were coated thus preventing oxidation to silver ions, it confirms that next to Ag ions also AgNPs can penetrate through the SC. Though, as stated by the authors it was not clear whether the presence of AgNPs is caused by the penetration of AgNPs through the SC or by the measurements inside the hair follicle. Some authors state that the skin penetration of nanoparticles larger than several nm through the intercellular space is not possible due to the small dimensions of the lipid bilayers which are the principal penetration route for most of chemicals (Watkinson et al., 2013). In the present study, we did not observe localization of the aggregates along the hair follicles. Furthermore, especially in the deeper SC, the aggregates were present on the cell surface.
The aggregation of silver in the SC might explain the low penetration of silver that we previously found both in our parallel in vitro and in vivo studies with the same textile (Bianco et al., 2015; Pluut et al., 2015). This is also consistent with a study of George et al. (2014), which reported low serum and urinary levels of silver after topical application of a nanosilver textile (10–40 nm) although silver NPs have been found below epidermis (George et al., 2014).
Considering main application of silver garments i.e. combating microorganisms on the skin surface, the retention of silver in the SC due to aggregation might be beneficial. Formed silver aggregates in the SC can be seen as reservoir for silver ions that could prolong the antibacterial effect in the skin. As the aggregates are of μm size, they will not be able to penetrate further and will eventually be removed from the SC by desquamation. In that way, the formation of aggregates can be seen as a detoxification pathway as less silver will reach the systemic circulation. On the other side, the risk of local effects (e.g. pro-inflammatory effect) of silver in the skin will theoretically be increased. However, in our accompanied in vivo study we did not find the changes in the inflammatory IL-1 cytokines in the SC (Pluut et al., 2015).
Previously, Larese et al. (2009) showed significant differences between intact and damaged skin permeation of Ag NPs dispersed in synthetic sweat. Thus, we expected a higher penetration of Ag NPs through atopic skin since it is known that the SC of AD patients has a reduced barrier function. However, in the present study, there was no striking difference between the atopics and non-atopics concerning the number and size of the particles and their abundance at different depths. This was also consistent with our parallel study on silver uptake, which did not show a significant difference in silver percutaneous absorption between atopic and healthy skin. It has to be noted that the number of subjects in this study was limited and the number and size of the particles at certain depths has not been assessed entirely in a systematic manner. Though, as a rule minimal three measurements on randomly chosen locations have been done per tape so we assume that the results are representative.