Introduction

Formaldehyde (FA) is an important chemical with a wide commercial use. FA is used in molding compounds, glass wool and rock wool insulation, decorative laminates, textiles treatments, and in the production of resins, plastics, plywood, and chemical intermediates. FA has applications in medicine as a tissue preservative and bactericide. Occupational exposure involves individuals in the manufactures of FA and FA containing products, and users (IPCS 1989; WHO 2000).

Human studies have shown that chronic exposure to FA by inhalation is associated with upper airway irritation (DFG 2000; Kim et al. 2011; Salonen et al. 2009). Epidemiological studies of industrial workers, embalmers, and pathologists have provided some evidence of elevated risks for cancers at various sites, including nasal cavities, lung, and hematopoietic system (McGregor et al. 2006; Thompson and Grafström 2011). The International Agency for Research on Cancer (IARC) has classified FA as being carcinogenic to humans (Group I) (IARC 2006).

FA is a DNA-reactive chemical that does not require biotransformation for this reactivity. Inhaled FA undergoes chemical reactions with organic compounds such as DNA, nucleotides, nucleosides, and proteins, by addition and condensation. FA is mutagenic in Drosophila larvae, in bacteria, and in yeast (Benyajati et al. 1983; Chanet and von Borstel 1979; Takahashi et al. 1985).

A genotoxic and carcinogenic potential of FA has been documented in mammalian cells and in rodents (Sul et al. 2007; Ren et al. 2012). Increased chromosomal aberration frequencies and sister chromatid exchanges (SCEs) have been observed in Chinese hamster ovary (CHO) cells and in cultured human lymphocytes (Kreiger and Garry 1983; Natarajan et al.1983). Previous studies have shown that DNA–protein cross-links (DPX) occur in the nasal mucosa and in the upper respiratory tracts of humans and animals exposed to FA, but increased amounts of DPX were not found in tissues other than the respiratory tract (Heck and Casanova 2004; Lu et al. 2010; Speit et al. 2009; Ye et al. 2005; Zeller et al. 2012).

Recently, it has been argued that the genotoxicity and carcinogenicity of FA has a practical threshold, and health-based exposure limits were derived for occupational settings (SCOEL 2008) and for indoor exposures (Nielsen and Wolkoff 2010). It was anticipated that an 8-h time-weighted average exposure to 0.2 ppm FA with a short-term exposure limit of 0.4 ppm would be not irritating and not genotoxic in humans. These conditions were therefore proposed as Occupational Exposure Limit (OEL; TWA and STEL, respectively) by SCOEL (2008). The purpose of the present field study was to verify this in persons exposed occupationally under exactly these conditions. The study also aimed at evaluating immune alterations and genetic damage in peripheral lymphocytes of workers with low-level FA exposure at proposed the OEL. In order to assess the immune competence of workers occupationally exposed to FA, several subsets of peripheral blood mononuclear lymphocytes, for example, CD3+ (T cells), CD4+ (helper T cells), CD8+ (suppressor T cells), and CD20+ (B cells), and natural killer cells (NK) cells have been analyzed, and immunoglobulins (IgG, IgA, IgM), complement proteins, and tumor necrosis factor-alpha (TNF-α) levels have been determined. DNA damage in the lymphocytes of the workers was evaluated by single-cell gel electrophoresis (Comet) assay.

Materials and methods

Subjects

The study population consisted of 46 male workers occupationally exposed to FA from two medium density fibreboard (MDF) producing plants in Gebze, a town located in the Marmara region of Turkey, and 46 non-exposed male controls of comparable age, sex, lifestyles, and smoking habits living in the same area and with no history of occupational exposure to FA and other chemicals, employed in administrative government offices and in maintenance services (Table 1).

Table 1 Characteristics of the study population
Full size table

Health conditions, medical history, drug and alcohol usage, and smoking habits were inquired for each worker and control person using questionnaires. Use of protective measures and years of employment, any specific symptoms related to FA exposure, skin reactions, and respiratory diseases were also recorded.

All subjects participated voluntarily in the study and were fully informed about the procedures and the aim of the study. Prior to the study, each subjects signed an informed consent form. Ethical approval for this study was obtained from the local ethical commission of Hacettepe University. The study has been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki.

Monitoring of occupational exposure of formaldehyde

Twenty-four samples were collected at stationary points at the workplaces. Also, air samples were collected from breathing zone of workers for the representative working periods (from 6 am to 2 pm, about 8 h) using a passive, personal air sampler. The 8-h time-weighted average (TWA8h) was determined according to NIOSH method 3500 (NIOSH 1994).

Momentary values of FA concentration in the air samples were measured using an air-monitoring Formaldemeter 400™ (PPM Technology limited Wales, UK), based on electrochemical sensing technology. The sampling time was about 3 min. The momentary values were measured just before the biological samplings.

Blood sampling

A 10 ml peripheral blood sample was taken from each individual. All blood samples were stored at +4 °C and processed within 6 h. Five ml of each sample was collected in preservative-free heparin tubes and used for the analysis of peripheral blood mononuclear cells (PBMC) and DNA damage; 2 ml was collected in EDTA containing tubes for the analysis of total and differential blood cell counts; and 3 ml was allowed to clot for the measurement of serum immunoglobulins, complement proteins, and TNF-α concentrations. The serum samples were kept at −80 °C until the day of analysis.

Lymphocytes were separated using the Ficoll-Hypaque density gradient technique by centrifugation of heparinized peripheral blood samples (Boyum, 1976). After washing with phosphate-buffered saline (PBS) buffer twice, the cell concentrations were adjusted to approximately 2×105cells/ml PBS. Cell viability was determined by trypan blue and was found to be higher than 85 % in all cases. Lymphocytes were divided into two portions, one for the assessment of DNA damage and the other for the evaluation of mononuclear cells.

Alkaline single-cell gel electrophoresis (Comet assay)

The basic alkaline technique of the Comet assay of Singh et al. (1988), as described by Anderson et al. (1998) and Collins et al. (1997), was applied. A total of 50 μl of the cells, mixed with 75 μl of 0.65 % low melting point agarose (LMA), was embedded onto slides pre-coated with a layer of 1 % normal melting point agarose (NMA). Slides were allowed to solidify on ice for 5 min. Coverslips were then removed. The slides were immersed in fresh cold lysing solution (2.5 M NaCl,100 mM EDTA, 100 mM Tris,1 % sodium sarcosinate, pH 10), with 1 % Triton X-100 and 10 % DMSO added just before use for a minimum of 1 h at 4 °C. Then, they were removed from the lysing solution, drained, and left in the electrophoresis solution (1 mM sodium EDTA and 300 mM NaOH, pH 13) for 20 min at 4 °C to allow unwinding of the DNA and expression of alkali-labile damage. Electrophoresis was conducted at a low temperature (4 °C) for 20 min, using 24 V and adjusting the current to 300 mA by rising or lowering the buffer level. The slides were neutralized by washing 3 times in 0.4 M Tris-HCl (pH 7.5) for 5 min at room temperature. After neutralization, the slides were incubated in 50, 75, and 98 % alcohol for 5 min, successively.

The dried microscopic slides were stained with ethidium bromide (EtBr, 20 μg/ml in distilled water, 60 μl/slide) and covered with a cover-glass prior to analysis with a Leica® fluorescence microscope under green light. The microscope was connected to a charge-coupled device camera and a personal computer-based analysis system (Comet Analysis Software, version 3.0, Kinetic Imaging Ltd., Liverpool, UK) to determine the extent of DNA damage after electrophoretic migration of the DNA fragments in the agarose gel. Slides were examined at 40× in order to visualize DNA damage.

One-hundred cells from each of 2 replicate slides were counted, and results were expressed as tail intensity, tail moment (% of DNA in tail × tail length), and tail migration.

Flow-cytometric analysis of peripheral blood mononuclear cells (PBMC)

Lymphocytes were separated from heparinized peripheral blood as described before, and the cells were analyzed for cell surface phenotypes by the direct immunofluorescence technique (Hudson and Hay, 1991; Winchester and Ross 1986). Fluorescein-isothiocyanate (FITC)-conjugated monoclonal antibodies directed against human CD3+ (T cells), CD4+ (helper T cells), CD8+ (suppressor T cells), and CD20+ (B cells) cells were obtained from a Beckman Coulter. Natural killer (NK) cells were determined by indirect immunofluorescence using CD16+/CD56+ and FITC-conjugated antihuman polyclonal immunoglobulin antiserum (Beckman Coulter), respectively. The numbers of peripheral blood mononuclear cells (PBMC) were analyzed by flow cytometry (EPICS XLMCL-Coulter Electronics). The same batch of antibodies was used during the entire study, and the flow cytometry was carried out always by the same person.

Determination of immunoglobulins and complement proteins

Serum concentrations of immunoglobulins (IgG, IgA, IgM), and the C3 and C4 components of complement proteins were measured by turbidimetry using a Dade-Behring Turbitimer with reagents Turbiquant IgG (anti-human IgG), Turbiquant IgA (anti-human IgA), Turbiquant IgM (anti-human IgM), Turbiquant complement C3 (anti-human complement C3), and Turbiquant complement C4 (anti-human complement C4) from Dade-Behring (Marburg GmbH--A Siemens Company, Germany) (Hudson and Hay, 1991).

Determination of human tumor necrosis factor-alpha (TNF-α)

The quantitative determination of human TNF-α in serum was performed by Quantikine HS Immunoassay kit purchased from Quantikine®-HS. All reagents and samples were kept at room temperature before use. All samples, standards, and controls were assayed in duplicate. All reagents, samples, and working standards were prepared according to the kit procedure. The optical density of each sample was determined within 30 min, using a micro-plate reader at 490 nm. The mean recovery of the kit was 93 % (85–98 %), and the detection limit was 0.106 pg/ml.

Statistical analysis

For statistical analysis, the “SPSS for Windows 10.0’’ computer program was used. The results were expressed as mean ± standard error mean (SEM). The distribution of data was checked for normality by the Kolmogorov–Smirnov test. Depending on the distribution, the statistical analyses were carried out by the one was variances (ANOVA) test or Kruskal-Wallis H test, setting the probability level to p < 0.05. An analysis of linear regression was used to estimate the effects of the duration of occupational FA exposure and the smoking habits on the extent of DNA damage and immune parameters.

Results

The mean age of the workers and controls was 33.4 ± 0.9 years (range 22–52) and 38.4 ± 1.2 years (range 24–53 years), respectively. The mean work duration of the workers was 7.3 ± 0.8 years (range 0.33–30 years). Eighteen workers and 23 controls were smokers, and the average consumption of cigarettes in the workers’ group and that in the controls were 11.2 ± 1.2 cigarettes/day and 11.9 ± 1.1 cigarettes/day, respectively (Table 1). The persons had not suffered from acute respiratory symptoms like bronchitis or coughing. Only 26 workers of the total 40 workers had used protective measures, such as masks.

The exposure levels of FA in the MDF producing plants were given in Table 2. The mean levels of FA exposure (TWA8 h) in the 1st and the 2nd MDF producing plant were 0.19 ± 0.07 ppm (0.11–0.33 ppm) and 0.20 ± 0.05 ppm (0.10–0.28 ppm), respectively. The mean level of FA exposure (TWA8 h) of the 46 individuals in the MDF producing plants was 0.20 ± 0.06 ppm (0.10–0.33 ppm), which was in compliance with permitted occupational exposure limit (OEL) of 0.2 ppm in Turkey and the OEL recommended by SCOEL (2008). The exposure levels of FA (TWA8h) in 37.5 % of studied areas in the plants (stationary measurements) were higher than 0.2 ppm. Momentary FA levels in 16.6 % of study areas were higher than the value of 0.3 ppm recommended by ACGIH (2003). The highest momentary value (0.35 ppm) was detected in the MDF pressing section of the 1st MDF producing plant. This was still in compliance with a STEL of 0.4 ppm.

Table 2 FA exposure in the MDF producing plants
Full size table

The DNA damage in the blood cells of workers exposed to FA was not higher than in the control group, as seen in Table 3 and Fig. 1. DNA damage in the lymphocytes of the workers was even significantly lower than the controls (p < 0.05). Smokers, both in the worker and control groups, had significantly higher DNA damage than the non-smoker workers and non-smoker controls. There was no statistically significant difference on DNA damage in the peripheral lymphocytes between the workers using protective masks and the workers taking no protective measures.

Table 3 Assessment of DNA damage in peripheral lymphocytes of formaldehyde-exposed workers and controls by the Comet assay
Full size table
Fig. 1
figure 1

DNA damage in peripheral lymphocytes of workers exposed to low levels of formaldehyde and controls, expressed as (a) tail migration, (b) tail moment, and (c) tail intensity. The results are given as mean ± standard error mean. a p < 0.05, workers (n = 46) compared to controls (n = 46), b p < 0.05, smoker controls (n = 23) compared to non-smoker controls (n = 23), c p < 0.001, smoker workers (n = 18) compared to non-smoker workers (n = 28)

Full size image

No significant differences in peripheral blood cells in terms of white blood cells (WBC), red blood cells (RBC), hemoglobins (Hb), neutrophils, and monocytes were observed between workers and controls (Table 4). However, the percentage of lymphocytes was increased significantly in the workers (p < 0.05). Also, the absolute numbers and percentages of T lymphocytes and NK cells were higher in the workers, compared to controls (p < 0.05). No significant differences were found in the levels of helper T, suppressor T, and B lymphocytes in the workers, compared to controls.

Table 4 Values of WBC, RBC, Hb, neutrophils, monocytes, and lymphocytes and the numbers of the lymphocytes and lymphocytes subpopulations in formaldehyde-exposed workers and controls
Full size table

The levels of IgG and IgM in exposed workers were statistically lower than in the controls (p < 0.05). The levels of IgA and complements C3 and C4 were not different between the groups. The levels of TNF-α were also statistically higher in the workers than in the controls (p < 0.05) (Table 5). No relationship between the duration of exposure and the studied immune parameters was observed.

Table 5 Levels of the immunoglobins (Ig G, Ig A, and Ig M) and complement C3 and C4 proteins and TNF alpha in formaldehyde-exposed workers and controls
Full size table

No significant differences were found in the levels of the parameters studied such as numbers of hematologic cells, lymphocyte subpopulations, IgG, IgA, IgM, complement C3 and C4 except TNF-α levels between the workers using protective measures and the workers without protective measures. It was only found that the levels of TNF-α in the workers using protective measures were statistically lower in than the workers using no protective measures (p < 0.05) (Table 6).

Table 6 Levels of the immunoglobins (Ig G, Ig A, and Ig M), complement C3 and C4 proteins, and TNF alpha in workers exposed to formaldehyde, related to the use of protective measures
Full size table

Discussion

Formaldehyde (FA) is a ubiquitous potentially toxic compound, and many countries (e.g., the U.S., Germany, the Netherlands, Scandinavian countries) have adopted or proposed permissible concentration levels (ACGIH 2003; DECOS 2003; DFG 2000; Nordic Expert Group 2003). The carcinogenicity of FA and the derivation of a safe occupational exposure limit have been matters of documentations by a number of official bodies and scientific expert panels (Bolt et al. 2010). The Scientific Committee on Occupational Exposure Limits of the EU (SCOEL 2008) has considered FA to be a “genotoxic carcinogen, for which a practical threshold is supported” and has recommended a health-based Occupational Exposure Limit of 0.2 ppm. Regarding indoor exposures, Nielsen and Wolkoff (2010) have provided a new documentation on FA concluding that the guideline value of WHO (World Health Organization) (2000) of 0.08 ppm FA was preventive of carcinogenic effects.

Available data on the immunotoxicity and the hematoxicity of FA in exposed populations are sparse and inconsistent. No detailed and comprehensive immunotoxicity study has been performed in workers occupationally exposed to FA. Our study, in which immune parameters in workers with long-term low levels exposure to FA, is a contribution to the assessment of immunotoxicity of FA in humans. In the present study, the percentages of lymphocytes, the absolute number and the percentages of T lymphocytes (CD3+) and of NK (CD56 +) cells, and the levels of TNF-α were found significantly higher, but the blood levels of IgG and IgM were significantly lower in the workers compared to their controls (p < 0.05). The findings of increased numbers and the percentages of NK cells and the levels of TNF-α might suggest an immune activation and potentially an increased susceptibility to inflammation of these subjects exposed to FA levels at current OELs. A more detailed account on the immunotoxicity of FA can be found in the Electronic Supplementary Material (Annex I) accompanying this article.

Although chromosomal damage by FA exposure in human peripheral blood cells was been claimed to occur (Ladeira et al. 2011; Santovito et al. 2011), in our study an increase in DNA damage, as determined by the Comet assay, was not found in long-term FA-exposed workers. This data are consistent with the results of others. It had been suggested that DNA–protein cross-links could arrest DNA replication, but such an arrest would not be detectable at low FA concentrations (Heck and Casanova 1999; Merk and Speit 1998). Several studies have shown that short-term (8 weeks) exposure to higher levels of FA (0.41–0.8 ppm) increased micronuclei (MN) frequency in nasal epithelial cells (Knasmueller et al. 2011; Ye et al. 2005). Significant increases in the frequency of MN in peripheral lymphocytes and in buccal cells of workers from pathology and anatomy laboratories who had been exposed to very high levels of FA were also reported (Costa et al. 2008; Ladeira et al. 2011; Orsiere et al. 2006). For instance, Shaham et al. (2002) found higher of SCE in peripheral lymphocytes of 90 workers from 14 hospital pathology departments who were exposed to FA for 15 or more years. Also chromosomal damage leading to micronucleated lymphocytes was found to be more frequent in highly exposed pathology and anatomy laboratory workers than in controls. The difference was suggested to be due to a higher frequency of chromosome loss, suggesting FA-induced defects in the mitotic apparatus (Orsiere et al. 2006). It is also being debated whether FA-induced genetic alterations, in conjunction with proliferation of immune cells, are the key components for the proposed mode of action for FA-induced lymphohematopoietic malignancies (DeVoney et al. 2006). In general, positive genotoxicity results were reported in studies at exposure levels much higher than currently proposed OELs. In our study, there were no other confounding work-places chemicals. The study, therefore, confirms that the currently proposed OEL (SCOEL 2008) is safe with regard to genotoxicity. On the other hand, our study also suggests that additional studies on possible immunotoxic effects of FA in exposed populations are warranted. It has been claimed that the likelihood for the development of allergic asthma increases proportionately with indoor FA concentration, when levels exceed 0.08 ppm (Kim et al. 2011). In general, it appears that the aspect of immunotoxicity of formaldehyde at low-level exposures should be further investigated.