Inflammation and cell damage caused by bifenthrin in brain immune cells and hippocampus tissue samples

The use of experimental animals was conducted with caution according to the regulations of the Ethics Committee of the University of Freiburg Medical Center and with the official approval from Regierungspräsidium Freiburg (Nr. X-13/06A). Animals were obtained from the Center of Experimental Models and Transgenic Services in Freiburg (CEMT-Freiburg). Generally, careful procedures and experimental setup were carried out in this work to minimize animal numbers and their suffering.

An analytical standard of BF [2-methylbiphenyl-3-ylmethyl-(Z)-(1RS)-cis-3-(2-chloro-3,3,3-trifluoroprop-1-enyl)-2,2-dimethylcyclopropane carboxylate, 99.5%] was obtained from Sigma-Aldrich (Deissenhofen, Germany). BF was prepared as a stock solution of 100 mM dissolved in dimethyl sulfoxide (DMSO) and further diluted to appropriate concentrations in culture medium immediately prior to use [42]. 2′,7′-dichlorofluorescein diacetate (DCFH-DA) was obtained from Sigma-Aldrich (Deissenhofen, Germany). Lipopolysaccharide (LPS) from Salmonella typhimurium (Sigma-Aldrich, Deissenhofen, Germany) was resuspended in sterile phosphate-buffered saline (PBS) (5 mg/mL) as stock and subsequently used at a final concentration of 100 ng/mL. Clodronate disodium salt (Merck Chemicals, Darmstadt, Germany) was prepared as a stock solution of 1 mg/mL in ultrapure water and stored at − 20 °C. Propidium iodide (PI) dye was obtained from Life technologies (Darmstadt, Germany), prepared as a stock solution of 1 mg/mL in ultrapure water and stored at 4 °C. Dulbecco’s modified Eagle’s medium (DMEM), non-essential amino acids (NEAA), 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), Hank’s balanced salt solution (HBSS), and RPMI-1640 medium were purchased from Life Technologies. Penicillin, streptomycin, and 0.05% (w/v) trypsin/EDTA were obtained from Sigma-Aldrich. Fetal calf serum (FCS) was purchased from Bio&Sell (Feucht, Germany). Ethanol 100% (EthO) was obtained from Sigma-Aldrich (Deissenhofen, Germany). The CellTiter 96® AQueous One Solution Cell Proliferation Assay kit was obtained from Promega (Mannheim, Germany). EIAs for the determination of prostaglandin E₂ (PGE₂) were purchased from Cayman (distributed by Bertin Pharma, Montigny-le-Bretonneux, France) and ELISA kits for rat tumor necrosis factor-alpha (TNF-alpha) from eBioscience, Frankfurt, Germany.

Primary mixed glial cell cultures were established from the cerebral cortices of 1- to 3-day-old neonatal Sprague Dawley rats as previously described [43, 44]. Cerebral cortices were collected and meninges removed. Forebrains were then minced and gently dissociated by repeated pipetting in Dulbecco’s modified Eagle’s medium (DMEM) and filtered by passing through a 70-μm nylon cell strainer (BD Biosciences, Heidelberg, Germany). Cells were collected by centrifugation (1000×g, 10 min) and resuspended in DMEM (Life Technologies) containing 10% fetal calf serum (FCS) (Bio&Sell) and 1% penicillin and streptomycin antibiotics (40 and 40 μg/mL, respectively). Cells were then cultured on 10-cm cell culture dishes (Falcon, Heidelberg, Germany) with the density of 5 × 10⁵ cells/mL in 5% CO₂ at 37 °C. After 12–14 days in vitro, floating microglia were harvested from mixed glia (astrocyte-microglia) cultures and re-seeded into cell culture plates at the density of 2 × 10⁵ cells/well. On the next day, the medium was removed to get rid of non-adherent cells and fresh medium was added. After 1 h, microglial cells were exposed to different concentrations of BF (1–20 μM) for 4 and 24 h; control cells were incubated with the solvent DMSO (0.1%).

Cell viability was measured by quantitative colorimetric assay with MTT-tetrazolio (MTT), as described previously [45]. Primary microglial cells (3 × 10³ cells/well) were seeded on 96-well cell culture plates and incubated for 24 h in 5% CO₂ at 37 °C without treatment or treated with different concentrations of BF (0.1–100 μM). EthO 100% and DMSO 10% were used as positive controls to induce the cell death, thus, serving as positive controls (black columns). After treatment, cells were incubated with 20 μL of the CellTiter 96® AQueous One Solution Cell Proliferation Assay kit (Promega, Mannheim, Germany) for 4 h. Afterwards, supernatants were removed and cells were solubilized with DMSO to detect the intracellular formazan crystals formed in the viable cells. Finally, the absorbance of each well was measured at 490 nm by using an ELISA reader as described by the manufacturer.

Membrane integrity was evaluated by measuring lactate dehydrogenase (LDH) in the supernatants of untreated and treated cells. Primary microglial cells (3 × 10³ cells per well in 96-well plates) were incubated with different concentrations of BF (1–20 μM) for 4 and 24 h in 5% CO₂ at 37 °C. Cell culture medium was collected and LDH activity was measured following manufacturer’s protocol (Biomagreb, ref. 20012 Ariana, Tunis, Tunisia). The percentage of LDH activity was determined as a percent of enzyme activity in the incubation medium to the total LDH activity. The reduction in absorbance due to NAD (H) oxidation was measured at 340 nm. The absence of the viable cells corresponds to 100% of LDH activity in the incubation medium.

ROS levels in primary microglial cells were determined using 2,7-dichlorodihydrofluorescein (DCFH-DA). DCFH-DA enters cells passively and is deacetylated by esterase to non-fluorescent DCFH. DCFH reacts with ROS to form DCF, the fluorescent product. Primary microglial cells were seeded into 96-well culture plates (3 × 10³ cells/well) and treated with different concentrations of BF for 4 and 24 h. In brief, 5 μM H₂DCF–DA >was added to each well. The fluorescence was measured every 1 h during a 4-h time period using a FL800-BioTek spectrofluorometer (Bio-Tek Instruments INC, Germany) with excitation/emission wavelengths of 488 nm/525 nm. ROS levels were expressed as the mean percentage of fluorescence absorbance in treated versus control cells.

Primary microglial cells were left untreated or incubated with different concentrations of BF (1–20 μM) for 4 and 24 h. Afterward, supernatants were collected and then centrifuged at 1000×g for 5 min at 4 °C. EIA for PGE₂ (Cayman, distributed by Bertin Pharma, Montigny-le-Bretonneux, France) and ELISA for TNF-alpha (eBioscience, Frankfurt, Germany) were performed according to the manufacturer’s instructions. For PGE₂, standard concentrations of 39–2500 pg/mL were used and sensitivity of the assay was 36 pg/mL. For TNF-alpha, the standard concentrations were used in the interval of 16–2000 pg/mL and the detection limit was 16 pg/mL.

Primary microglial cells (3 × 10³ cells/well) were pre-incubated with various concentrations of BF (1–20 μM) for 4 and 24 h. After the indicated time points, supernatants were collected and centrifuged at 10,000×g for 5 min. NO levels in the supernatants were determined using the Griess method [46]. One hundred microliters of supernatants were incubated with 100 μL of Griess reagent (Sigma-Aldrich, Deissenhofen, Germany) containing 1% sulfanilamide, 2% phosphoric acid, and 0.1% naphthyethylene diamide. After 20 min of incubation, absorbance was measured, at 540 nm. Nitrite concentrations in the cells were determined based on a sodium nitrite standard curve. Results were expressed as mean nitrite concentration (μM) ± SEM for each group.

Primary microglial cells (3 × 10³ cells/ well) were left untreated or treated with different concentrations of BF (1–20 μM) for 4 and 24 h in 5% CO₂ at 37 °C. Afterwards, hydrogen peroxide (H₂O₂) formed by the microglial cells was determined by the ferrous ion oxidation-xylenol orange (FOX1) method [47]. The FOX1 reagent consists of 25 mM sulfuric acid, 250 μM ferrous ammonium sulfate, 100 μM xylenol orange, and 0.1 M sorbitol. Briefly, 100 μL of cell supernatants were added to 900 μL of FOX1 reagent, vortexed, and incubated for 30 min at room temperature. Solutions were then centrifuged at 12,000×g for 10 min. The amount of H₂O₂ in the supernatants was spectrophotometrically determined at 560 nm. Results were expressed as nmol per mg protein.

Primary microglial cells (3 × 10³ cells/ well) were left untreated or treated with different concentrations of BF (1–20 μM) and incubated for 4 and 24 h in 5% CO₂ at 37 °C. The extent of lipid peroxidation in homogenates of primary microglial cells was then determined by measuring the release of a thiobarbituric acid reactive substance (TBARS) in terms of malondialdehyde (MDA) formation content and was measured using the thiobarbituric acid (TBA) colorimetric assay according to the Draper and Hadley (1990) method [48]. Briefly, cultures were washed with ice-cold PBS, pooled in 0.1 mol/L PBS/5% Triton X-100 buffered solution, and incubated for 1 h at 37 °C. Then, trichloroacetic acid (350 μl; 20% w/v) was added to 250 μL of cellular lysate and centrifuged (1000×g at 4 °C for 10 min). Aliquots (450 μL) of each supernatant were mixed with an equal volume of 0.5% (w/v) TBA. The mixture was boiled at 100 °C for 30 min. After cooling, MDA formation was measured at 520 nm and results were expressed as nmol of MDA/mg of protein.

Microglial cells (3 × 10³ cells/well) were plated and treated with different concentrations of BF (1–20 μM) for 4 and 24 h and then collected for the measurement of antioxidant enzyme activities. Catalase (CAT) activity was assayed by the decomposition of hydrogen peroxide according to the method of Aebi et al. [49]. A decrease in absorbance due to H₂O₂ degradation was monitored at 240 nm for 1 min, and the enzyme activity was expressed as nmol H₂O₂ consumed/min/mg protein. Superoxide dismutase (SOD) (MnSOD and Cu/ZnSOD) activities were evaluated by measuring the inhibition of pyrogallol activity as described by Marklund et al. [50]. This method is based on the competition between pyrogallol oxidation by superoxide radicals and superoxide dismutation by SOD. The specific Cu/Zn-SOD inhibition by potassium cyanide allows the Mn-SOD determination in the same conditions. Assays were monitored by spectrophotometry at 420 nm. One unit (U) corresponds to the enzyme activity required to inhibit half of pyrogallol oxidation. SOD activity was expressed as U/mg protein. Glutathione peroxidase (GPx) activity was measured according to Flohe and Gunzleret al. [51]. The enzyme activity was expressed as nmol of GSH oxidized/min/mg protein.

Primary microglial cells were treated with BF (1–20 μM) for 4 and 24 h. Microglia RNA was then isolated using the Qiagen RNeasy Mini kit according to the manufacturer’s instructions (Qiagen, Germany), and cDNA was synthesized from 500 ng of total RNA using MMLV reverse transcriptase and random hexamers (Promega, Mannheim, Germany). The synthesized cDNA was used as template for the real-time qPCR amplification carried out by the CFX96 real-time PCR detection system using iQ™ SYBR supermix (LifeTechnologies, Darmstadt, Germany). Specific primer sequences were designed by using the Universal ProbeLibrary (Roche) and were obtained from Biomers (Ulm, Germany). These include primers for Nrf-2, NF-kappaBp65, TNF-alpha, COX-2, mPGES-1, and IL-6 (Table 1). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) served as an internal control for sample normalization, and the comparative cycle threshold Ct method was used for data quantification as described previously [52].

Primary microglial cells were treated with BF and then incubated for 4 and 24 h with or without LPS (100 ng/mL), which was used as a positive control (black column). Cells were then washed with PBS and lysed in 1.3× sodium dodecyl sulfate (SDS) containing sample buffer without dithiothreitol (DTT). Protein concentrations were measured using the bicinchoninic acid (BCA) assay (Thermo Fischer Scientific, Waltham, MA). The optical density was read at 570 nm using a microplate reader, and concentrations were given based on the bovine serum albumin (BSA) protein standard curve. Immediately before electrophoresis, bromophenol blue and DTT (final concentration of 10 mM each) were added to the samples and 30 μg of protein from each sample were subjected to SDS-PAGE (polyacrylamide gel electrophoresis) on a 12% gel under reducing conditions. Proteins were then transferred onto a polyvinylidene fluoride (PVDF) membrane (Merck Chemicals, Schwalbach, Germany) by semi-dry blotting. The membrane was blocked for 1 or 2 h at room temperature using Rotiblock (Roth, Karlsruhe, Germany) for COX-2 or 5% blocking milk (BioRad, München, Germany) for the other proteins. Primary antibodies used were anti-COX-2 (1:500; Santa Cruz Biotechnology, Heidelberg, Germany), anti-mPGES-1 (1:6000; Agrisera, Vännas, Sweden), anti-Nrf-2 (1:1000; Santa Cruz Biotechnology), anti-beta-actin (1:5000, Sigma-Aldrich), and anti-NF-kappaB-p65 (1:1000; Santa Cruz Biotechnology). Primary antibodies were diluted in TBS-T and 1% BSA. Membranes were incubated with the primary antibody overnight at 4 °C, followed by extensive washing (three times for 15 min each in TBS containing 0.1% Tween 20). Proteins were detected by a 1-h incubation with horseradish peroxidase (HRP)-coupled rabbit anti-goat IgG (Santa Cruz, 1:100000), anti-Mouse IgG (GE Healthcare, 1:20000), or HRP-coupled donkey anti-rabbit (GE Healthcare, 1:25000) using chemiluminescence (ECL) reagents (GE Healthcare). All Western blot experiments were performed at least three times. The densitometric analysis of the Western blots was performed using Image J software 1.47v (National Institute of Health). The beta-actin signals served as normalization controls for the target proteins.

Organotypic hippocampal slice cultures (OHSCs) were prepared as described earlier [43, 44, 53] with minor modifications. In brief, slice cultures were prepared from 1- to 3-day-old C57BL/6 wild-type mice under sterile conditions. After decapitation, brains were removed and the hippocampi from both hemispheres were acutely isolated in ice-cold serum-free Hank’s balanced salt solution (HBSS), supplemented with 0.5% glucose (Sigma-Aldrich) and 15 mM HEPES. Isolated hippocampi were cut into 350–375-μM-thick slices using a tissue chopper (McIlwain) and were transferred to 0.4 μM culture plate inserts (Merck Chemicals, PICM03050). The inserts containing four to six slices were placed in six-well plates containing 1.2 mL of culture medium per well. Culture medium (pH 7.2) consisted of 0.5× minimum essential medium (MEM) containing 25% heat-inactivated horse serum, 25% BME basal medium without glutamate, 2 mM glutamax (all from LifeTechnologies), and 0.65% glucose (Sigma-Aldrich). The slice cultures were kept at 37 °C in a humidified atmosphere (5% CO₂), and the culture medium was refreshed the first day after preparation and every consecutive 2 days.

To deplete microglia from OHSCs, slice cultures were placed immediately after preparation either on culture medium containing clodronate encapsulated in liposomes (Chlodronate Liposomes, Haarlem, Netherlands) or culture medium containing 100 μg/mL clodronate disodium salt (Merck Chemicals, Darmstadt, Germany) [43, 44]. Twenty-four hours after preparation, the slice cultures were washed briefly with prewarmed PBS and reincubated in fresh culture medium for 7 days before use. The OHSC culture medium was changed every second day.

For immunohistochemical analysis, controls and BF-challenged slice cultures were shortly rinsed in PBS and fixed with 4% paraformaldehyde (PFA) overnight at 4 °C. After fixation, the slice cultures were rinsed in PBS and pre-incubated with 5% normal goat serum (NGS, Vector) in PBS containing 0.3% Triton X-100 (PBS+) overnight. Subsequently, the slice cultures were incubated with the appropriate primary antibodies overnight in 1% NGS/ PBS+ at 4 °C. The following primary antibodies were used: rabbit-anti-Iba-1 (1:1000, Wako 019-19741) for detection of microglia, mouse-anti-GFAP (1:600, Chemicon MAB3402) for detection of astrocytes, and mouse-anti-NeuN (1:1000, Chemicon MAB377) for detection of neuronal nuclei. The secondary antibodies used were a donkey-anti-mouse-Alexa488 (Molecular Probes) for NeuN, donkey-anti-rabbit-Alexa633 (Molecular Probes) for Iba-1, and goat-anti-mouse-Cy3 (Jackson IR Laboratories) for GFAP. Imaging was carried out using laser scanning ZEISS LSM 510 META microscope. To quantify neuronal cell death in response to BF-induced neurotoxicity, slice cultures were incubated with 2 μg/mL PI during and after the BF challenge [44, 54]. Immunofluorescently stained OHSCs were analyzed by confocal laser scanning microscopy using a ZEISS LSM 510 META upright microscope (objective, C-apochromat × 10), and the number of PI-positive cells were quantified using ImageJ software (as described in [54]). The percentage of cell death was determined by calculation of PI area fraction as a percentage of NeuN area fraction. Quantifications of Iba-1 and GFAP area fractions after BF treatment were calculated as a percentage of untreated control area fraction using ImageJ software [54, 55].

All statistical analyses were carried out using GraphPad Prism 6.0 for Windows (GraphPad Software, San Diego, CA). Results were expressed as mean ± standard errors (mean ± SEM). Multiple comparisons data were analyzed using one-way ANOVA followed by post hoc Student–Newman–Keuls test. Two-way ANOVA with Bonferroni post hoc test was used for grouped analyses. For each experiment, a p value < 0.05 was considered statistically significant.

The cytotoxicity of BF on primary microglia was assessed using MTT assay (Fig. 1). Microglial cells were left untreated or exposed to various concentrations (0.1–100 μM) of BF for 24 h. Furthermore, ethanol (100%) and DMSO (10%) were used as positive controls to induce cell death. Exposure to high concentrations (10–100 μM) of BF for 24 h induced a significant decrease in the cell viability with maximal effects at 100 μM (p < 0.05) (Fig. 1). In contrast, no significant cell death was observed at lower concentrations (0.1, 1, and 5 μM) of BF (Fig. 1). The positive control ethanol induced 100% cell death and DMSO approximately 55% (Fig. 1, black columns).

Primary microglial cells were chosen to investigate the effect of BF on oxidative stress and inflammatory mediators. In a first step, microglial cells were treated with different concentrations of BF (Fig. 2a–d). Treatment for 4 h with BF (1–20 μM) did not affect LDH activity and ROS production (Fig. 2a, c). In contrast, a significant increase of LDH and ROS production was observed 24 h after BF treatments at concentrations of 1–20 μM and 10–20 μM, respectively (Fig. 2b, d).

To further investigate the direct effects of BF on neuroinflammatory mediators, primary microglial cells were exposed to different concentrations of BF (1, 5, 10, and 20 μM) for 4 and 24 h and levels of PGE₂, NO, and TNF-alpha were determined in the cell supernatants (Fig. 3). Our data showed that incubation with BF for 24 h induced a marked increase in PGE₂, NO, and TNF-alpha production as compared to the vehicle-treated controls, especially at the high concentrations (10–20 μM) (Fig. 3b, d, and f). However, 4 h after BF treatment, only the highest concentration (20 μM) induced PGE₂ release and increased NO levels, while TNF-alpha production was not affected (Fig. 3a, c, and e).

We next investigated the effects of BF on two common oxidative stress markers MDA and H₂O₂. Our results demonstrate that 24 h of BF treatment significantly increased the levels of MDA (at 10–20 μM) and H₂O₂ (at 1–20 μM) (Fig. 4b, d). No significant effects were observed following 4 h of BF treatment on MDA and H₂O₂ levels (Fig. 4a, c).

Intracellular levels of antioxidant enzymes, including SOD, CAT, and GPx, were investigated in primary microglial cells incubated with various concentrations of BF (1, 5, 10, and 20 μM) at 4 and 24 h (Fig. 5a–f). Our data demonstrated a decrease in SOD activity only with the high concentrations of BF (10 and 20 μM) at 24 h (Fig. 5b). GPx activity was significantly decreased only with the highest concentration of BF (20 μM) at 24 h (Fig. 5f). Furthermore, a marked decrease in the activity of CAT by BF was observed at different concentrations (5, 10, and 20 μM) after 24 h of treatment (Fig. 5d). In contrast, no significant decrease in the activities of endogenous antioxidant enzymes (SOD, CAT, and GPx) was observed after 4 h of BF treatment (Fig. 5a, c, and e).

It was further examined whether BF exposure increased mRNA expression of pro-inflammatory cytokines in primary microglia. Treatment with BF (1–20 μM) for 4 and 24 h increased IL-6 and TNF-alpha mRNA expression in a concentration-dependent manner (Fig. 6a, b; Additional file 1 A1-A and B). In addition, BF mediated a significant concentration-dependent increase of COX-2 and mPGES-1 at the mRNA and protein levels at 24 h (Fig. 6c–h). After 4 h of incubation with BF, a marked increase in mRNA and protein levels of COX-2 was observed starting from 5 μM (see Additional file 1 A1-C, E, and F). Although a significant increase of mPGES-1 at the mRNA and protein levels was observed with 20 μM BF at 4 h, lower concentrations of BF (1–10 μM) did not exert any effects on this enzyme at the same time point (Additional file 1).

Inflammation and oxidative stress are controlled by several cellular aspects; among them, Nrf-2 and NF-kappaB are the two key transcription factors with a major function in these processes [56]. Therefore, we evaluated the effects of BF on Nrf-2 and NF-kappaB pathways by qPCR and Western blot analysis. We observed a significant increase in mRNA expression of Nrf-2 and NF-kappaBp65 after 24 h of incubation with BF in primary microglia (Fig. 7a, b), whereas at 4 h, only NF-kappaBp65 expression was increased (Additional file 2). Increased mRNA expression was detectable starting at 5 μM and reached its maximum at 20 μM of BF.

Four hours of incubation with BF had no significant effect on the protein level of Nrf-2 (Additional file 2), whereas NF-kappaBp65 protein level was significantly increased by BF starting at 5 μM and maximal levels were identified at 20 μM (Additional file 2). Similar to the results obtained from mRNA analysis, Western blot data of the 24-h experiment showed a dose-dependent increase of Nrf-2 and NF-kappaBp65 proteins starting at 5 μM of BF (Fig. 7c–f). The highest levels of Nrf-2 and NF-kappaBp65 protein levels were observed at 20 μM of BF.

To examine possible neurocytotoxic effects of BF ex vivo, we treated OHSCs for 24 h with BF (1–20 μM) and performed PI and NeuN immunohistochemistry. Confocal images showed that BF treatment at higher concentrations (10 and 20 μM) induced neuronal death (colocalization of PI with NeuN) in all hippocampus layers (CA1, CA3, and DG) with 12% (10 μM) and 31% (20 μM) of PI uptake. However, BF concentrations lower than 10 μM did not induce significant cell death in any of the neuronal regions (Fig. 8a, c, and d).

We next aimed to examine whether microglial depletion in OHSCs might affect BF-induced neuronal cell death. We eliminated microglia by clodronate liposomes and subsequently examined neuronal cell death. We recently showed that clodronate treatment successfully depletes microglia without affecting other OHSC cells [44].

OHSCs with and without microglia were treated with different concentrations (1–20 μM) of BF for 24 h. Following microglial depletion, severe neuronal death was observed with all BF concentrations (1–20 μM) as compared to untreated controls and microglia-containing OHSCs (Fig. 8b–d). In microglia-depleted OHSCs, BF induces neurotoxicity even at the low concentrations of 1 μM with 13% and of 5 μM with 20% differences in increased cell death between microglia-depleted and microglia-containing OHSCs, respectively. Cytotoxic effects of BF were severe and further increased at the high concentrations of 10 and 20 μM BF in the microglia-depleted OHSCs compared to microglia-containing OHSCs (Fig. 8a–d).

Confocal images of Iba-1 staining showed no visible differences between BF treatments and controls in microglia (Fig. 9a). In addition, images of the astrocytic marker GFAP demonstrated no changes between BF treatments and controls (Fig. 9a). Quantifications of GFAP and Iba-1-positive staining area confirmed that BF does not induce microglial or astrocytic cell death in OHSCs (Fig. 9b, c).

Brahim Gargouri received fellowships from BMBF/MESRS (TUNGER-36) and Alzheimer Forschungsinitiative (AFI). The project was in part supported by the AIF project GmbH (BMWi) (AGEsense) and BMBF/MESRS (TUNGER-36). Nizar M. Yousif was supported by a scholarship as per the State Law on Graduate Funding (LGFG-IGA, Freiburg). The article processing charge was funded by the University of Freiburg Library in the funding program “Open Access Publishing.”

All data are provided in the manuscript and in the additional files.

All procedures used in this study were approved and performed according to the guidelines of the Ethics Committee of the University of Freiburg Medical School with the official approval from Regierungspräsidium Freiburg (Nr. X-13/06A).

The authors declare that they have no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

All data are provided in the manuscript and in the additional files.