In in vitro-experiments to study retention of allergenic metals in skin, conditions mimicking intense hand hygiene practices using water, soap and hand-disinfectant, were obtained by simultaneous exposure to nickel alone or in combination with cobalt and chromium, and the exposure solvents Milli-Q water, 0.5% sodium lauryl sulphate (SLS) and ethanol, respectively. In addition, skin with impaired barrier properties, to further resemble damage from intensive hand hygiene, was created via pre-treatment with SLS. In practical terms, the study was divided into three different experimental parts; I - in which conditions causing an impaired skin barrier were tested and evaluated, II - in which treated and untreated skin were exposed to nickel in Milli-Q water, 0.5% SLS and ethanol, and III - in which treated and untreated skin were co-exposed to nickel, cobalt and chromium in Milli-Q water, 0.5% SLS and ethanol, Fig. 1. (Flowchart of the chronological order of the study can be found in Supplementary Material Figure S1). All material used in experiments were acid washed (soaked for 24 h in 10% HNO3, rinsed three times with ultrapure water and dried in ambient laboratory air) or cleaned with ethanol, to avoid any possible metal contamination.

Schematic illustration of the three experimental parts (I, II, III) of the study

Full-thickness skin of stillborn piglets from commercial breeders was used in the present study. As the animals were not bred for research purposes, the use is exempt from the Swedish Agency for Agriculture’s requirements for ethical vetting of research involving animals. Although the OCED TG 428 does not specify the use of pig skin, the GD 156 [16] state the fact that pig skin is considered an appropriate alternative to human skin, which is also in line with the results from a review of in vitro penetration studies by Barbero et Frasch [17]. Pig skin is routinely used in skin permeation assays as it has been shown to have similar permeability characteristic to human skin [17,18,19,20]. Stillborn piglet skin has also been reported to have comparable permeability to human skin for organic compounds [21, 22]. No data is available regarding its metal permeability, but it has been used in other studies of metal retention, for nickel, cobalt, chromium and lead [23, 24].

At arrival to the laboratory, stillborn piglets were rinsed with lukewarm water, after which skin integrity was checked by measuring the transepidermal water loss (TEWL, Dermalab, Cortex Technology, Hadsund, Denmark).

To simulate experimental exposure conditions affected by intense hand hygiene practices, the skin of stillborn piglets was washed with water and soap for 5 min (DAX Mildtvål Oparfymerad, KiiltoClean, Hyllie Stationstorg 2, Malmö, Sweden) or repeatedly treated 25 times with hand disinfectant (DES 75 vol%, LIV by Clemondo, Helsingborg, Sweden) in situ. Based on the results from TEWL measurements following each step in the procedure, the approach was concluded to not efficiently impair the skin barrier and hence, were not used for the experiments (for more information on TEWL values and the procedure see Supplementary Material TableS1).

Full thickness skin (mean thickness 0.86 ± 0.22 mm) was collected from the back and flank of the stillborn piglets (< 24 h post-mortem) and the effect of excision on the skin was checked measuring TEWL at four different locations in each skin piece. Skin thickness was measured with a digital micrometre (model number 293-666-20 Mitutoyo, Kawasaki, Japan). An average TEWL ≥ 11 g⋅m− 2⋅h− 1 was used as a cut-off for inclusion [25, 26]. However, no skin pieces had to be discarded. The average TEWL for the skin was 7.25 ± 1.22 g⋅m− 2⋅h− 1. Next, the skin was wrapped in polyethene film and aluminium foil and stored at − 20ºC until later use within 3 months.

On the day of experiments, using a sterile scalpel (Kiato, Sylak AB, Askim, Sweden) 3 × 3 cm skin pieces were cut from each frozen skin and placed in a petri dish to thaw for 30 min at room temperature. Thereafter, the barrier integrity of each skin piece was controlled. The measured TEWL of all skin samples were < 11 g⋅m− 2⋅h− 1.

Pre-treatment of skin to alter the barrier integrity can be performed by physical means [27], but for the purposes of this study, a pre-treatment with aqueous SLS was elaborated based on the OECD TG 439 for in vitro skin irritation [28].

After the thawing of skin, 500 µl PBS (PBS, pH = 7.4, Gibco Life Technologies, Thermo Fisher Scientific, Waltham, MA, USA) was put in the petri dish underneath the 3 × 3 cm skin piece to prevent dehydration. The skin surface was exposed to 200 µl of SLS-solutions (diluted from 20% SLS in H2O, Sigma-Aldrich, Schnelldorf, Germany) at different concentrations; 0.5, 1, 2, 5, and 10% in Milli-Q water (18.2 MΩ ⋅ cm− 1, Merck Millipore, Darmstadt, Germany) for 1 h, covered by the petri dish lid [29]. The concentrations were chosen based on available literature where 0.5-2% SLS concentrations has been used to irritate human skin [11, 30, 31]. OECD test guidelines for in vitro skin irritation using reconstructed human epidermis suggest the use of 5% aqueous SLS as positive control [28]. Due to the different skin models used in the available literature, and to ensure that we select the SLS concentration with the highest effect on the skin barrier, we decided to test also 10%, a concentration above those reported in the previous literature. The SLS was removed by rinsing with 4 ml (2 ml per side) of deionized water (dH2O, 16.8 MΩ ⋅ cm− 1). In total, four replicate samples were produced for each concentration tested with skin originating from four different piglets. The experiments for the two highest concentrations 5 and 10% respectively, were repeated and the results are thus based on 8 replicate samples. The TEWL values for each skin sample was recorded 20 min after removing the treatment.

A series of experiments were conducted to evaluate the ability of the skin barrier to retain metals given conditions without and with SLS pre-treatment, to alter the skin barrier, and the simultaneous exposure to Milli-Q water, 0.5% SLS and ethanol, to mimic intensive hand hygiene practices using water, soap and hand sanitizer, respectively. The OECD TG 428 for skin absorption [15] and GD 156 [16] constituted the starting point for experiments with a focus on the study of the skin barrier as boundary for exposure.

Six jacketed Franz cells (orifice diameter 11.28 mm, corresponding to an exposure area of 0.95 cm2, receptor volume 3 ml, Permegear, Bethlehem, PA, USA) were mounted on an adapted magnetic stirrer plate (HP 6 Variomag, H + P Labortechnik, Munick, Germany) and by means of circulating water from a thermostat water bath (AT 110, Heto, Alleod, Denmark) the diffusion cells were tempered at 32ºC. PBS was used as receptor fluid and was kept stirred using Teflon coated magnetic stirring bars. Skin pieces were mounted onto the Franz cells 15 min before the start of metal exposures.

This study comprises twenty-four different exposure scenarios each tested on both treated and untreated skin (Table 1), with a dose range of relevance for occupational settings and exposure time that mimic real-life work periods (short exposure and full day work shift) [32,33,34]. In the experimental part II, skin was exposed to nickel (1.36 µmol corresponding to a dose of 80 µg Ni/cm2) dissolved in Milli-Q water, 0.5% SLS and ethanol (≥ 96%, v/v, TechniSolv®, France) for 2 and 8 h. In part III, skin was similarly co-exposed to equimolar amounts of nickel, cobalt and chromium (4.09 µmol corresponding to a dose of 80 µg Ni + 80 µg Co + 71 µg Cr/cm2) in the three exposure solvents. The donor solutions were prepared using two metal reference materials: a standard nickel stock solution (10 000 µg Ni/ml in 2.5% HNO3, Spectrascan, Teknolab, Ski, Norway) and a special, equimolar high concentration reference material of nickel + cobalt + chromium (Ni + Co + Cr 200 mmol/l in 10% HNO3, Spectrascan, Teknolab, Ski, Norway).

Once the skin exposures to metal were initiated, the donor compartment and sampling port were occluded with parafilm (PARAFILM®, American National Can™). Blank (Milli-Q) exposures were carried out in parallel to enable control for any metal baseline quantities found in the skin (Supplementary Material Figure S2).

Post exposure, the skin surface was rinsed with 2 ml dH2O per side (4 ml in total). Biopsy punches (Kai medical, 8 mm diameter) were taken from the exposed area and placed in polypropylene-plastic tubes (12 ml, Sarstedt, Nümbrecht, Germany) with 1 ml of 67% HNO3 for 48 h (until fully digested). Prior to metal analysis, 50 µl of digested skin was diluted with 4.95 ml of dH20 and spiked with 20 µl of indium (1.255 µg In/ml, diluted from stock solution of 999 ± 5 µg In/ml in 2% HNO3, Spectrascan, Teknolab, Ski, Norway).

Quantitative analyses of Ni, Co and Cr were performed using Inductively Coupled Plasma-Mass Spectrometry (ICP-MS iCAP Q Thermo Fisher Scientific, Qtegra version 2.10). Concentrations of 58Ni, 60Ni, 59Co, and 52Cr, were analysed in kinetic energy discrimination (KED) measurement mode using helium gas to reduce any polyatomic interference and argon as nebulizer gas, cool gas, and auxiliary gas.

Matrix-matched standards for calibration with the concentrations of 0, 0.1, 1, 5, 10, 50, 100 and 500 µg/l Ni, Co, Cr and Pb in 2% HNO3 (67–69% HNO3, VWR, Normatom, Leuven, Belgium) were diluted from single metal reference materials (Ni: 1001 ± 4 µg/ml in 2% HNO3 (v/v); Co: 1000 ± 3 µg/ml in 3% HNO3 (v/v); Cr: 1002 ± 4 µg/ml in 2% HNO3 (v/v); Pb: 998 ± 4 µg/ml in 0.5% HNO3 (v/v), Spectrascan, Teknolab, Ski, Norway).

To ensure statistical certainty, each sample was analysed three to five times. The limit of detection (LOD) (based on 7 concentration points of the STD curve in the ICP-MS) was set at 0.079 µg/l for 58Ni, 0.082 µg/l 60Ni, 0.004 µg/l 59Co, and 0.19 µg/l 52Cr. All exposed samples analysed were above LOD. Nickel quantities found in samples was calculated as an average of 58Ni and 60Ni.

Any statistical relationship between the amount of metal retained in skin at exposures to nickel alone or in combination with cobalt and chromium in three exposure solvents for treated and untreated skin at two different time-points were evaluated using the Mann-Whitney U-test (GraphPad Prism version 9.5.0).

To determine which variable (TEWL, skin thickness, +/- SLS treatment, single nickel or Ni + Co + Cr co-exposure in Milli-Q water, 0.5% SLS or ethanol, and exposure time) affect metal retention in skin, linear regression with log-transformed amount of retained metal was performed using R (Version 4.4.1 (2024-06-14 ucrt)).