Impact of exposure to wood dust on genotoxicity and

Mutagenesis Advance Access published May 9, 2015 Mutagenesis, 2015, 1–9 doi:10.1093/mutage/gev034 Original Article

Original Manuscript

Impact of exposure to wood dust on genotoxicity and cytotoxicity in exfoliated buccal and nasal cells Downloaded from http://mutage.oxfordjournals.org/ at Library MedUni Vienna (10076821) on May 10, 2015

Georg Wultsch, Armen Nersesyan, Michael Kundi1, Karl-Heinz Wagner2, Franziska Ferk, Robert Jakse3 and Siegfried Knasmueller* Department of Medicine I, Institute of Cancer Research, Comprehensive Cancer Center Medical University of Vienna, Borschkegasse 8a, A-1090 Vienna, Austria, 1Institute for Environmental Health, Center for Public Health, Medical University of Vienna, Kinderspitalgasse 15, 1090 Vienna, Austria, 2Department of Nutritional Sciences, University of Vienna, Althanstrasse 14, 1090 Vienna, Austria, 3Private Clinics Leech, 8010 Graz, Austria *To whom correspondence should be addressed. Tel: +43 1 40160 57561; Fax: +43 1 40160 957500; Email: siegfried. [email protected] Received October 24 2014; Revised March 20 2015; Accepted March 26 2015.

Abstract Wood dust was classified by the IARC as a human carcinogen which causes sinonasal tumours. However, the exposure in different industries varies strongly and the risks of workers depend on the specific situation which can be assessed by the use of biomonitoring methods. The aim of this study was to investigate the workers who are exposed to low dust levels (below the permitted concentrations) with cytogenetic and biochemical methods. Micronuclei (MNi) which are indicative for genomic damage, nuclear buds which reflect gene amplification, binucleated cells which are caused by mitotic disturbances and acute cytotoxicity parameters (pyknosis, karyorrhexis, condensed chromatin, karyolysis) were monitored in buccal and nasal cells of workers of a veneer factory (n = 51) who are exposed to volatile wood-derived compounds, in carpenters of a furniture factory which use no synthetic chemicals (n=38) and in a control group (n = 65). Additionally, markers were measured in blood plasma which reflect inflammations (C-reactive protein, CRP) and the redox status, namely malondialdehyde (MDA) and oxidised low density proteins (oxLDL). No induction of micronucleated cells was observed in both epithelia in the two exposure groups while all other nuclear anomalies except pyknosis were increased; also one health-related biochemical marker (MDA) was significantly elevated in the workers. Taken together, the results of our study show that exposure to low levels of wood dust does not cause formation of MNi indicating that the cancer risks of the workers are not increased as a consequence of genetic damage while positive results were obtained in earlier studies with workers who are exposed to high dust levels. However, our findings indicate that wood dust causes cytotoxic effects which may lead to inflammations.

Introduction In 1968, the first investigation was published which indicated that workers in the furniture industry have increased rates of nasal cancer (1). These findings were confirmed in a number of subsequent studies (2–5) and finally led to the classification of wood dust as a human

(group I) carcinogen by the International Agency for Research on Cancer (IARC) (6,7). At present, about 3.6 million workers are exposed to wood dust in Europe (8). Apart from furniture makers also other groups such as cabinet makers, carpenters, factory joiners and workers in the plywood and veneer production are exposed. Their health risks depend strongly on

© The Author 2015. Published by Oxford University Press on behalf of the UK Environmental Mutagen Society. All rights reserved. For permissions, please e-mail: [email protected].

1

G. Wultsch et al.

2

exposed to air-borne particles which are suspected to cause inflammations (39–41) and also in a recent study with wood dust exposed carpenters (42). The aim of this study was to investigate individuals who are exposed to low levels of wood dust which are in the frame of the permitted limits. In Austria (where the study was realised) and also in some European countries (Switzerland, the Netherlands), the permitted level is ≤2.0 mg/m3 (43–45). In the USA, the limit is even lower, namely 1.0 mg/m3 (46). All earlier investigations have been conducted with workers from factories with higher exposure levels and positive results were obtained in all these studies. To find out if the type of exposure has an impact on genotoxic and cytotoxic effects caused by wood dust, two groups of workers were monitored in this study, namely (i) individuals in a furniture factory who are solely exposed to wood dust and not to chemicals except to organic glue and (ii) workers at a veneer factory who are exposed to dust and in addition to volatile organic compounds (VOC) which are released during the production process (47). The different nuclear anomalies mentioned above [micronucleated cells (MNC), total number of MNi, NB, BN, CC, KR, KL, P and BasC] were monitored in buccal and in parallel in nasal cells of both groups of workers and additionally also in a matched control group. Furthermore, biomarkers which reflect the redox status (oxLDL, MDA) and inflammation (CRP) were monitored in all participants.

Materials and methods Subjects The study was carried out in accordance with the Declaration of Helsinki for human studies (48). It was conducted with workers from two factories. Subjects in factory I (n = 38) are carpenters who produce furniture from different woods without synthetic chemicals (paints, solvents). These subjects are solely exposed to wood dust and organic glue. They work with 94% press boards made solely of spruce wood, 6% of the boards which consist of 60% spruce, 20% oak and 20% beech wood. All press boards are formaldehyde-free. Workers from factory II produce veneer from wood logs. They are exposed to wood dust and additionally to VOC which are released during the cooking process. The veneer is made from both, softwood (55%) and hard wood (45%) dust. The softwood species which are used are stone pines (20%), spruce (19.5%), larch (13%), acacia (0.94%) and yews (0.02%). The hardwood fraction consists of oak (30%), maple (6.5%), beech (6.5%), the rest are apple and pear, chestnut and cherry. Inhalative wood dust concentrations were measured in both factories according to the recommendations of the DIN EN 689 (49) and the BGIA 7284 (50) as well as the BGI 505–42 (51). For the sampling of the inhalable wood dust fraction, a Gillian HFS pump (Personal Air Sampler, Gillian, Heimsheim, Germany) was used for personal air sampling and a Gravikon PM4 Ströhlein pump (VEBEG GmbH, Berlin, Germany) for stationary measurements. A PGP (“Personengetragene GefahrstoffProbenahmesystem”) and a GSP (“Gesamtstaub-Probenahmesystem “, Gesellschaft fur Schadstoffmesstechnik, GmbH, Neuss-Norf, Germany) devices in connection with a low volume sampler (3.5 and 10 l/min flow rate for 8 h) which were equipped with glass fiber filters (diameter 37 and 70 mm, average pore diameter 8 μm) were used. The collected dust concentrations were determined by gravimetric analyses (52). Control subjects were matched with the participants of two exposed groups in regard to sex, age, body mass index (BMI) smoking and alcohol consumption. The male participants were recruited from jail wardens, female controls were employees from the Medical University of Vienna and the Occupational Medical Center (AMEZ), Graz. The demographic data of the participants are summarised in Table 1.

Downloaded from http://mutage.oxfordjournals.org/ at Library MedUni Vienna (10076821) on May 10, 2015

the extent of exposure to dust and specific chemicals. It was postulated that the odds ratios (OR) for sinonasal tumours caused by moderate levels of dust (≤1.0 mg/m3) is around 3, while concentrations ≥5.0 mg/ m3 cause more dramatic effects (i.e. OR ≥ 45) (9). Furthermore, it is still a matter of debate, if the adverse effects are caused by the dust particles themselves, by chemicals released by woods such as aldehydes (10), quercetin and tannins or by synthetic compounds such as dyes and paints (7,10). The currently available data indicate that the adverse health effects of wood dust depend strongly on the specific occupational exposure scenarios. In order to assess the risks of workers in different settings, human biomonitoring methods can be used, which provide information on changes of health-related parameters. One of these approaches, which was used to assess the risks of wood workers in the past are micronucleus (MN) assays with both exfoliated epithelial cells (11–13) and peripheral lymphocytes (13–16) and chromosomal aberration test with lymphocytes (17). Micronuclei (MNi) reflect structural and numerical chromosomal aberrations (18) and it was postulated that increased frequencies of MNi in lymphocytes are associated with cancer risks (19). Experiments with cells from the respiratory tract may be particularly useful for inhalation studies, since these epithelia come in direct contact with the dust and chemicals. The cytome assay with buccal and nasal cells used in this study was performed also in numerous earlier occupational studies (20–22) is a minimally invasive system which allows not only to detect MNi but also nuclear buds (NB) which are caused by gene amplifications and disturbances of the mitotic cycle leading to the formation of binucleated cells (BN) (23,24). Additionally, also anomalies which reflect cytotoxic effects (KR, karyorrhexis; KL, karyolysis; CC, condensed chromatin and P, pyknosis) and the regenerative potential of epithelial tissue [basal cells (BasC)] can be monitored (23,24). Another important mechanism which may lead to adverse longterm effects in wood dust exposed workers is oxidative stress which is causally related to inflammations (25–27). It was emphasised by an expert group of the IARC (7) that the cancer risk of wood dust exposed workers are not necessarily caused by the direct action of genotoxins but possibly by inflammatory reactions. These processes can be monitored by the use of different biomarkers [for review see Knasmueller et al. (28)] Oxidation of low density lipoproteins (oxLDL) is a marker of the redox status which is associated with the risk for cardiovascular diseases (25). This parameter has not been measured in investigations with wood dust exposed individuals according to our knowledge, but it was monitored in a recent study with welders because it was found that they have increased oxidative stress (29). It is well documented that wood dust exposure leads to release of proinflammatory cytokines and elevated COX-2 expression [for review see (7)]. These processes are associated with the formation of reactive oxygen species which cause oxidation of lipoproteins (28). Malondialdehyde (MDA) is widely used as a marker for of this process which is associated with cardiovascular diseases (25,26) and cancer (30). This parameter was measured in a number of inhalation studies with workers who are exposed to particles which cause inflammations, e.g. in personnel exposed to photocopiers (31), in brick kiln workers (32), in individuals who inhale bentonite (33) and diesel engine exhaust particles (34) and also in welders (29). It is of particular interest in studies which concern alterations of the DNA stability as a number of findings show that MDA is DNA-reactive and induces chromosomal aberrations (35,36). C-reactive protein (CRP) is a widely used clinical marker for inflammations (27) and regarded as a valid parameter for the prediction of cardiovascular diseases (37). Furthermore, its association with cancer is well documented in a number of studies [for review see (38)]. CRP was monitored in a number of investigations with workers who were

Impact of exposure to wood dust

3

Table 1. Demographic data of the study participants Parameter

Factory I Males (n = 35)

Age (years) BMI, kg/m2 Smoking, n (cig/day) Alcohol, n (drinks/week)a Exposure Duration of exposure (years)

Factory II Females (n = 3)

41.5 ± 12.9 42.3 ± 11.5 26.9 ± 3.2 25.9 ± 3.5 4 (10.0 ± 7.7) 2 (9.0 ± 8.5) 11 (2.2 ± 0.9) 2 (1.5 ± 0.7) Wood dust and organic glue 24.1 ± 12.5 22.3 ± 7.7

Males (n = 25)

Control Females (n = 26)

40.4 ± 10.8 38.3 ± 10.8 26.3 ± 4.8 23.5 ± 4.6 14 (18.9 ± 7.7) 13 (16.5 ± 6.5) 10 (2.0 ± 1.0) 9 (2.8 ± 1.3) Wood dust and VOC 22.3 ± 7.7 9.1 ± 9.1

Males (n = 43)

Females (n = 22)

38.8 ± 7.9 26.8 ± 2.0 16 (17.2 ± 9.1) 13 (3.1 ± 1.3) None None

35.1 ± 5.5 23.6 ± 2.7 6 (11.7 ± 6.8) 11 (1.9 ± 0.8)

n, number; VOC, volatile organic compounds. a One drink: 0.5 l of beer or one glass (0.25 l) of wine.

Both groups of workers were sampled within a period of 2 days; negative control were sampled 1 day later (sampling period 2 days). Buccal cells were collected as described in our previous articles (21,22,53). The participants were asked to rinse their mouth twice with tap water immediately before the sampling. The cells were collected with wooden spatulas (Paul Hartmann Ges.m.b.H., Neudorf, Austria) from both cheeks. The material from each donor was smeared on two slides together with 2–3 drops of distilled water. Nasal cells were collected as described recently (21,22) from the middle turbinates with a cytobrush (Heinz Herenz, Hamburg, Germany). Cells from each participant were transferred to two slides. Subsequently, 2–3 drops of distilled water were added and the cells were smeared. ml of blood were collected in heparinised In addition, 10 Vacutainer tubes (Becton-Dickinson, Plymouth, UK). Subsequently, rpm, 10 min, the plasma was isolated by centrifugation (3000 Sorvall, ThermoScientific, Vienna, Austria) and stored at −80°C before the analyses. All samples (exfoliated cells and blood) were coded before analyses.

Fixation, staining and evaluation of nuclear anomalies in exfoliated cells The cells were fixed on the day of the collection with chilled (−20°C) 80% methanol, Ten minutes later, they were placed in glass beakers with 5.0 M HCl at room temperature for 30 min, rinsed with distilled water for 3 min and subsequently stained with Schiff’s reagent (Sigma–Aldrich, Steinheim, Germany) for 90 min, washed 5 min with running tap water and then counterstained with 0.2% (w/v) Light Green (Sigma–Aldrich) for 20 s. From each subject, ≥2000 buccal and ca. 1500 nasal cells were evaluated. Both buccal and nasal cells were scored under 1000-fold magnification (Nikon Photophot-FXA, Tokyo, Japan). As described in earlier articles, nasal smears consist of different cell types (54–56). MNi and all other anomalies were evaluated in all cell types of epithelial origin (i.e. in ciliated and non-ciliated and also in squamous and basal cells) but not in leucocytes (57,58). The morphology of the exfoliated cells which were analysed is described in detail in two previous publications (23,24). The follo wing endpoints were monitored: MNC, the total number of micronuclei (MNi), the rates of NB, BN, CC, KR, KL and P. All results are expressed as the number of anomalies per 1000 cells. In buccal cells also, the number of basal cells (BasC) was scored which is a marker of the mitotic activity of basal layer of the epithelium (23,24). The cells were examined under bright light; when MNi were detected, they were confirmed under fluorescent light. The nuclear anomalies were evaluated in both cell types according to the criteria

defined by Thomas et al. (23). The slides were analysed by one experienced scorer and cross-checked by another experienced scorer without knowledge of subject identity.

Determination of biochemical parameters in blood Plasma oxLDL concentrations were measured with a commercially available ELISA kit (Mercodia AB, Uppsala, Sweden). Absorbance of samples and standards was determined with a fluorimeter (BMG Lab Technologies, Offenburg, Germany). MDA levels were determined in plasma according to the method of Ramel et al. (59). The samples were neutralised after hea ting (60 min, 100°C) with methanol/NaOH, centrifuged (3 min, 3000 rpm), subsequently MDA was measured with high-performance liquid chromatography (HPLC) (excitation: λ 532 nm, emission: λ 563 nm, LaChrom Merck Hitachi Chromatography System, Vienna, Austria). Each sample was analysed in duplicate. CRP was determined in the plasma with a turbidimetric immunoassay (60). Photometric measurements were conducted with an Olympus AU5400 automated analyzer (Melville, NY, USA).

Statistical analyses Nuclear anomalies (the major variable of interest in nasal and buccal cells) were analysed by Poisson regression with individuals who were exposed to wood dust in factory I or II or belonged to the control group. Analyses were controlled for age, gender, smoking and alcohol consumption. The total number of cells counted was used as an offset variable. Results are expressed as odds ratios (OR) and 95% confidence intervals (CI) for factories I and II with controls as reference. In addition, for the total number of MN, the impact of duration of employment in the exposed area was evaluated in addition to other covariates for comparison with an earlier study (11). Blood chemical parameters were log transformed to obtain homogeneity of variances and to normalise the distribution. Groups were compared by application of the general linear model with the same covariates as included in the analyses of nuclear anomalies. Normality of residuals was tested by Kolmogorov–Smirnov tests with Lilliefors’ correction of P values. Homogeneity of variances was tested by Bartlett’s test. For all statistical analyses, a P value below 0.05 was considered significant.

Results MN and other nuclear anomalies in nasal cells The results of the cytome assays with nasal exfoliated cells are summarised in Figure 1. It can be seen that the number of cells with MNi was not significantly altered in both study groups while all


Blood and cell sampling

G. Wultsch et al.

4

other parameters were increased in comparison with the controls. The rates of NB were significantly elevated by 107% and the frequencies of BN cells were higher by 54% in the employees of the veneer factory compared to the controls. The strongest increase in regard to markers which reflect cytotoxicity was found with KR cells which were elevated in both exposed groups by a factor of 2.5. KL cells were significantly increased in the workers of the furniture factory (Factory I) by 54 and by 38% in the workers of veneer factory. All NB in the nasal cells were projections of the main nuclei while their morphology in buccal cells was typical for so-called ‘broken eggs’ which are attached to the main nuclei [for details see (23,24)]. MNi were detected in the nasal smears only in ciliated and squamous cells. The same observation has been made earlier by Italian groups (57,58) and confirms also the findings of our previous studies (21,22).

Figure 2. The rates were significantly increased as a consequence of wood dust exposure; however, it is unclear if this observation has any biological consequence because the effect was only moderate (increase from 5‰ basal cells in controls to 7‰ in exposed subjects).

MN and other nuclear anomalies in buccal cells

Figures 3A–C summarise the results of the biochemical measurements. It can be seen that no significant differences of the oxLDL levels were found in the three groups while the MDA concentrations were significantly increased in workers from both factories. Also the CRP levels were in the wood dust exposed individuals higher than in non-exposed subjects but this effect was significant only in workers of the veneer factory.

The findings which were obtained with cells from the oral cavity are summarised in Figure 2. It can be seen that the results are similar to those obtained in the nasal cells. The rates of most anomalies in buccal cells were in the control group higher as in the nasal cells. This observation is in agreement with the results of our earlier investigations (21,22). As in the nasal cells, a non-significant increase of MN was observed in both study groups while BN and NB and all other mar kers which are indicative for cytotoxicity were elevated. Notable differences between the results obtained in the two factories are the higher levels of P and of KL in workers of factory I.

Impact of wood dust exposure on the frequencies of basal cells in buccal mucosa The numbers of basal (undifferentiated) cells which were found in the controls and in the two exposure groups are also shown in

Impact of demographic factors on the formation of MN and other anomalies To find out if the duration of the exposure and if factors such as gender, age, smoking, alcohol consumption and the body mass index have an impact on the frequencies of MNi in both cell types, multiple Poisson regression analyses were applied. The results are presented in Table 2. It can be seen that these factors were neither associated with the MNC rates, nor with the frequencies of other nuclear anomalies.

Impact of wood dust exposure on biochemical parameters

Discussion Taken together, the results of this study show that MNC which reflect chromosomal damage are not increased in cells from the respiratory tract of wood dust exposed workers. Different other endpoints which are caused by cytotoxic effects (CC, KR and KL), gene amplifications (NB) and also BN, which reflect mitotic disturbances, were significantly elevated in workers. Furthermore, we also found evidence for an increase of biochemical parameters which are indicative for inflammations.


Figure 1. Means and 95% confidence intervals for nuclear anomalies in nasal cells of controls (open bars), in furniture carpenters (factory I, light grey) and veneer production workers in (factory II, dark grey). *P