Female C57BL/6 mice 5-7 weeks old were obtained from Taconic (Ry, Denmark). The mice were allowed to acclimatize for 1-3 weeks before the experiment. All mice were given food (Altromin no. 1324, Christian Petersen, Denmark) and water ad libitum during the whole experiment. The mice were group housed in polypropylene cages with sawdust bedding and enrichment at controlled temperature 21 ± 1°C and humidity 50 ± 10% with a 12-h light:12-h dark cycle. Female mice were studied at 8 weeks of age. The experiments were approved by the Danish "Animal Experiments Inspectorate" and carried out following their guidelines for ethical conduct and care when using animals in research.
Particles and sanding dusts
The Danish paint and lacquer industry provided an indoor acrylic paint product with 10% content of NanoTiO2 (referred as Indoor-NanoTiO2) and a corresponding product without nanoparticles (referred to as Indoor R) (Table 1). The tested pure nanomaterials comprised of a nanosized TiO2 material (UV-Titan L181, code: NanoTiO2), and carbon black (Printex 90), which was included as an internal reference particle. Printex 90 was a gift from Degüssa (Germany). As reported previously, the specific surface areas of the particles were 295 m2/g for Printex 90  and 107.7 m2/g for NanoTiO2 . The nanoparticles and the paints are described in detail in  and .
Generation of dusts
Dusts were generated by sanding of boards painted with the two selected paints paints as described previously . Briefly, dust samples were collected by an electrostatic precipitor. The device was modified from a commercial air cleaner and is described in details in . Paint dust was deposited on the silver coated plates. After sampling the plate holder system was covered with Aluminum-foil and carefully transported to the laboratory where dust was removed from the plates using a silicon scraper. The samples were stored in the freezer until toxicological testing.
Preparation of exposure stock
Particles were suspended by sonication in 0.9% NaCl MilliQ water containing 10% v/v acellular BAL collected from C57BL/6 mice. The BAL fluid was prepared by flushing unexposed mice twice to 0.6 ml 0.9% NaCl yielding approximately 1 ml of BAL fluid. Acellular BAL was prepared by centrifugation of BAL fluid at 400 g (10 min, 4°C). The particles (4.05 mg/ml) and dust suspensions (12.15 mg/ml) were sonicated using a Branson Sonifier S-450D (Branson Ultrasonics Corp., Danbury, CT, USA) equipped with a disruptor horn (Model number: 101-147-037). Total sonication time was 16 min, with alternating 10 s pulses and 10 s pauses at amplitude of 10% (8 min sonication total). During the sonication procedure the samples were continuously cooled on ice. These suspensions were used for the high dose (486 μg (dust) and 162 μg (NanoTiO2/Printex 90)) and diluted 1:3 for the medium dose and diluted further 1:3 for the low dose. Between the dilutions the suspensions were pipetted. Vehicle control solutions were prepared containing 90% 0.9% NaCl MilliQ water and 10% acellular BAL fluid.
Characterization of exposure
Characterization of the materials in instillation vehicle have been published previously [12, 14]. Briefly, the average size of the materials in instillation vehicle were determined by Dynamic Light Scattering (DLS) and the shapes of the materials and the extent of agglomeration/aggregation in instillation vehicle were characterized by transmission electron microscopy (TEM) and scanning electron microscopy (SEM). TEM pictures of the NanoTiO2 suspension used for instillation showed that the individual nanoparticles mainly occurred in partially open to dense aggregates of ca. 100 nm size or larger and the size-distribution plot by DLS indicated a multimodal size-distribution with two modes in the μm-range (peaks at 5.5 and 2.3 μm) and one in the sub-μm range (825 nm) . SEM analysis of the Indoor-NanoTiO2 and Indoor-R suspensions used for instillation suggests that most of the nanoparticles were still fully or partially encapsulated in the paint matrix after sanding. However, a considerable amount of the pigment particles were generally liberated from the matrix. Due to the size of the coarse dust fraction and sedimentation, DLS analysis was not reliable for the un-filtered dispersions of dusts .
Measurement of endotoxin
An amount of 4.05 mg/ml of each type of nanoparticle and dust was suspended in pyrogen free water with 0.05% Tween 20 by orbital shaking (300 rpm) at room temperature. The particles were suspended by sonification as described above. The suspensions were centrifuged (20.000 rpm) for 6 min, and the supernatant was used for endotoxin assay. The endotoxin contents were analysed in duplicate using the kinetic Limulus Amebocyte Lysate test (Kinetic-QCL endotoxin kit, Lonza, Walkersville, MD, USA). A standard curve (ranging from 0.05 to 50 EU/ml) obtained from an Escherichia coli O55:B5 reference endotoxin was used to determine the concentrations in terms of endotoxin units (EU) (15.0 EU = 1 ng). In addition inhibition/enhancement controls were prepared by spiking supernatants from sample Indoor-NanoTiO2, Indoor-R, TiO2 and Printex 90 with 10 μl of a 0.5 EU/ml solution of endotoxin. As a control of the method NanoTiO2 and Printex 90 were tested twice (on two plates) and less than 16% variance was found.
Exposure of mice
The mice were treated with a single intratracheal instillation with 18, 54 and 162 μg for the nanoparticles and 54, 162 and 486 for the paint dusts. To eliminate day to day variation, 3-4 materials including vehicle controls were instilled on each exposure day and each material dose was instilled on three separate days. Before the intratracheal instillation, the mice were anesthetized using Hypnorm® (fentanyl citrate 0.315 mg/ml and fluanisone 10 mg/ml from Janssen Pharma) and Dormicum® (Midazolam 5 mg/mL from Roche). Both were mixed with equal volume sterile water. A volume of 0.2 ml was injected subcutaneously into the loose skin over the neck of each mouse. The sedated mice were kept on 37°C heating plates. During instillation the mice were placed on their backs on a 40 degree slope. A one diode cold light source was placed touching the larynx. The tongue was pressed towards the lower jaw by a small spatula. The trachea was intubated using a 24 gauge BD Insyte catheter (Ref: 381212, Becton Dickinson, Denmark) with a shortened needle. The correct location of each intubation was tested by a small but highly sensitive pressure transducer (pneutachymeter), developed by our laboratory in collaboration with John Frederiksen (FFE/P, Copenhagen, Denmark). A 40 μl suspension was instilled followed by 150 μl air with a 250 μl SGE glass syringe (250F-LT-GT, MicroLab, Aarhus, Denmark). Control animals were instilled with vehicle (0.9% NaClwith 10% BAL). After the intubation catheter was removed, the mousebreathing was observed in order to assure that the delivered material did not block the airways. The mice were placed on to the 37°C heating plate until they recovered from anaesthesia.
Preparation of tissue and cells from the mice
One, 3 or 28 days after intratracheal instillation, the mice were anaesthetised with Hypnorm/Dormicum as described above. Immediately after withdrawing the heart blood, a bronchoalveolar lavage (BAL) was performed four times with 0.8 mL of 0.9% sterile saline through the trachea. The BAL was immediately put on ice until BAL fluid and BAL cells were separated by centrifugation at 4°C and 400 g for 10 min. The BAL cells were resuspended in 100 μL medium (HAMF12 with 10% fetal bovine serum). Part of the suspension (40 μL) was mixed with 160 μL medium containing 10% DMSO and stored at -80°C for later analysis in the comet assay. For differential count, cells from 50 μL suspension were collected on microscope slides by centrifugation at 10,000 rpm for 4 min in a Cytofuge 2 (StatSpin, Bie and Berntsen, Rødovre, Denmark). The slides were fixed with 96% ethanol and stained with May-Grünwald-Giemsa stain. The cellular composition of BAL cells was determined on 200 cells. The total number of cells was determined by using the NucleoCounter (Chemometec, Allerød, Denmark) live/dead assay according to the manufacturer's instructions. The lungs and a piece of liver tissue were snap frozen in cryotubes (NUNC) in liquid N2 and stored at -80°C. From vehicle controls and high dose groups exposed to each of the test materials (3-6 per group) another piece of liver tissue from the left lobe was kept in formaldehyde (4%) until liver histology was performed.
Preparation of RNA and cDNA from lung tissue
RNA was prepared using the NucleoSpin 96 RNA kit (Macherey-Nagel). RNA from the entire left lung of each mouse was prepared by lysing the tissue in 2 ml RLT buffer, while vigorously disrupting the sample with a Tissuelyser (Qiagen, Denmark) with a 5 mm stainless steel bead for 2 × 60 seconds and run through a QIAshredder (Qiagen, USA). The rest of the purification was performed as described by the manufacturer. cDNA was prepared using TaqMan reverse transcription reagents (Applied Biosystems, USA) as described by manufacturer.
The Tgf-β1 gene expression was determined using real-time RT-PCR with 18S RNA as reference gene (Kit nr. Mm03024053_m1 from Applied Biosystems). RT-PCR was performed using Universal Mastermix (Applied Biosystems, Nærum, Denmark). PCR was performed on an ABI PRISM® 7500 sequence detector (PE Biosystems, Foster City, CA, USA) as described previously . 18S (4310893E, Applied Biosystems, Nærum, Denmark) was used as the reference gene. Expression for each gene was quantified in separate wells. Samples were quantified in triplicates; standard deviation was below 20%. Run to run variation was controlled by quantifying the mRNA levels for the same control sample. The standard deviation in separate runs was lower than 25%. No template and -RT controls were included in all runs.
The level of DNA strand breaks in frozen BAL and liver tissue was determined by the alkaline comet assay as described in [40, 41] based on a protocol by . The strand breaks measured by the assay represent a mixture of direct strand breaks, alkaline labile sites and transient breaks in the DNA due to repair processes . The samples were analyzed using a high throughput allowing 48 samples per GelBondfilm, as recently described  and (Gützkow et al: A high throughput comet assay using 96 A high throughput comet assay using 96 minigelsminigels, in preparation). BAL cell suspensions in freezing medium with 10% DMSO were thawed quickly. For liver, deep frozen samples (ca. 40 mg) were pressed through a metal stapler (diameter 0.5 cm, mesh size 0.4 mm) into Merchant's media (0.14 M NaCl, 1.47 mM KH2PO4, 2.7 mM KCl, 8.1 mM Na2HPO4, 10 mM NaEDTA, pH 7.4) for inhibiting endogenous DNA cleaving enzymes (first described by )). Samples were embedded in agarose, lysed, subjected to alkaline electrophoresis, fixed and later stained and scored as previously described . Due to preparation time, the lysing procedure varied between 1-3.5 hour for samples in the present study. In order to minimize the effects of this we normalized the results to the positive assay control. As a positive assay control and to estimate the electrophoresis-to-electrophoresis variation, 0 and 30 μM H2O2 exposed A549 cells were included on each Gelbond film in all electrophoresis runs.
Specimens were taken from the liver of three to six mice from the vehicle control from the high dose groups of all test materials and euthanized 1, 3 or 28 days after instillation. The specimens were fixed in 4% neutral buffered formaldehyde, paraffin-embedded, and sections 4-6 μm were made and stained with hematoxylin and eosin for histological examination.
The data were assessed by non-parametric three-way ANOVA with post-hoc Tukey-type multiple comparison test for effects showing statistical significance in the overall ANOVA test. Statistical significances were tested at P < 0.05 level. The statistical analyses were performed in SAS version 9.2 (SAS Institute Inc., Cary, NC, USA).