Detection of Hormone-Like and Genotoxic Activities in Indoor Dust from Taiwan Using a Battery of in Vitro Bioassays

Indoor dust serves as a potential sink for various synthetic chemicals used in our daily lives, while exposure to these anthropogenic contaminants via dust contact, ingestion, or inhalation may pose potential threats to human health. In this study, in vitro biological assays were used to investigate the endocrine disrupting activity and genotoxicity in dust samples collected from a university located in southern Taiwan.


INTRODUCTION
A variety of synthetic chemicals are used in our daily lives to promote the health and wealth of human beings, whilst chronic exposure to these substances raises concern owing to their potential adverse effects.Indoor dust ingestion, inhalation, and dermal contact are important routes of exposure to household contaminants (Butte and Heinzow, 2002).In particular, infants that stay at home for a longer time and ingest a sufficient amount of house dust may be more vulnerable to indoor contaminants (Roberts et al., 2009).In the past decade, indoor dust has been reported to contain numerous legacy pollutants and emerging contaminants, including polycyclic aromatic hydrocarbons (PAHs), polychlorinated dibenzo-p-dioxins, phthalates, brominated flame retardants, benzotriazole ultraviolet stabilizers, antibacterial agents, polyfluoroalkyl phosphate esters, and so on (Fan et al., 2010;Deziel et al., 2012;Peng et al., 2012;Kubwabo et al., 2013;Wang et al., 2013;Chao et al., 2014;Eriksson and Kärrman, 2015;Król et al., 2014).Rudel et al. (2003) have also detected 52 and 66 endocrine disrupting compounds in residential air and dust collected from 120 homes, whereas banned chlorinated pesticides were frequently found in dust as well, indicating poor indoor degradation.
Among different household contaminants, PAHs are frequently found in the toxic fine particulate matter smaller than 2.5 µm (PM 2.5 ).Their outdoor sources include steel plants, coal fired power plants, vehicle emissions, municipal solid waste incinerators, biomass burning, whereas indoor sources can be from building and furnishing materials, heating and cooking facilities, cigarette smoke, incense burning, and etc. (Madany et al., 1994, Li et al., 2005;Fang et al., 2004;Kume et al., 2007;Amodio et al., 2013;Da et al., 2013;Park et al., 2013;Yang et al., 2013;Li et al., 2014;Kim et al., 2015).The presence of PAHs in indoor dust, in particular the carcinogenic and mutagenic PAHs with 4-7 rings, may pose serious threats to human health.
Although levels of target contaminants in indoor dust can be analysed by instrumental analysis, it is not appropriate to assess the toxic effects caused by a mixture of pollutants if only the contents are available.On the other hand, bioassays that are designed to evaluate specific modes of toxic actions may provide valuable information on mixture toxicity (Eggen et al., 2003).Bioassays, especially in vitro cell-based assays are suitable for fast screening of the presence of toxicity in environmental matrices.In addition, relationship between toxicity and chemical contamination could be more elucidated by combining bioassay and chemical analyses.Maertens et al. (2004) summarized that mutagenic activities could be found in settled house dust, in which PAHs ranged from 0.5 and to 500 µg/g dust equivalent.Tue et al. (2013) reported that a significant portion of the dioxin-like activities in indoor dust could be related to unknown pollutants by using a combination of Dioxin-Related Chemical-Activated LUciferase gene eXpression (DR-CALUX) assay and instrumental chemical analysis.Vandermarken et al. (2015) have also used an estrogen responsive elements-CALUX assay to evaluate the estrogenic activities in dust samples from kindergartens in Belgium, and the results showed that phthalate concentrations were well correlated with the estrogenic activities.
In this study, indoor dust from a university located in southern Taiwan and outdoor dust from roads surrounding the university campus were collected and subjected to bioassay analysis.In vitro biological assays were used to investigate whether aryl hydrocarbon receptor (AhR) agonist, androgen receptor (AR) agonist/antagonist, thyroid hormone receptor (TR) agonist/antagonist, and genotoxic activities were present in the dust samples.PAHs in indoor dust samples were also analysed by gas chromatography-mass spectrometry (GC-MS) to assess their contributions to the bioassay-derived AhR agonist activities.Samples showing high AhR agonist activities were further fractionated by high performance liquid chromatography (HPLC) to isolate potential AhR agonist contaminants in active fractions.Numerous environmental pollutants have been shown to disrupt AhR, AR, and TR via receptor binding.For example, dioxins and PAHs are typical AhR agonist contaminants, whereas phthalates elicited AR agonist, AR antagonist, and TR antagonist activities in reporter gene assays (Shen et al., 2009;Murray et al., 2014).Binding to receptors by xenobiotic agonist and antagonist contaminants may lead to altered activation and inhibition of receptor activity, respectively.

Dust Sample Collection
Indoor dust samples were collected from different sites in a university located in southern Taiwan during 2013, including photocopy room (Ph), office (Of), classroom (Cl), computer room (Co), and chemical laboratory (Lab).Outdoor dust samples were also obtained from roads surrounding the university campus (Fig. 1).Sampling was conducted by sweeping using a broom, and dust samples were placed in pre-cleaned amber glass bottles and were stored at -20°C until extraction.

Dust Sample Extraction
One gram of each dust sample was homogenized and freeze-dried using a freeze dryer (FDU-1200, EYELA, Japan), and then was Soxhlet-extracted using 200 mL of hexane:acetone (1:1, v:v) mixed solution for 24 h.After extraction, the mixed solution was concentrated by a rotary evaporator (N-1000, EYELA, Japan) to less than 3-5 mL, and was added with activated copper for desulfurization.The extract solution was concentrated using a purified nitrogen stream (MG-2200, EYELA, Japan) after adding DMSO to obtain a final concentration of 1 g dust equivalent (dust EQ)/mL DMSO.

Yeast-Based Assay for AhR Agonist Activity Detection
A yeast-based reporter gene assay using the recombinant YCM3 strain was applied to measure AhR agonist activity in each dust sample (Miller III, 1999).The assay was undertaken as described in Chou et al. (2014).In brief, 5 µL of an overnight culture of yeast cells and 193 µL of a synthetic medium (2% galactose, lacking tryptophan) were mixed with 2 µL of β-NF (a typical AhR agonist), DMSO (blank), or a dust sample extract in a 96-well microplate.The mixture was further incubated at 30°C for 12 h, and the optical density at 595 nm (OD 595 ) was determined using a microplate reader (xMark, Bio-Rad, USA).Then, 10 µL of the cell suspension was mixed with 140 µL of Z-buffer and 50 µL of ONPG solution (4 mg/mL in Z-buffer) in a new microplate, and the microplate was incubated at 37°C for 1 h.The optical density at 405 nm (OD 405 ) was measured after 1-h incubation for calculating AhR agonist activity, which was shown as the percentage of maximal response induced by 10 µM β-NF (AhR agonist activity (%) = (OD 405 / OD 595 ) SAMPLE /(OD 405 /OD 595 ) 10 µM β-NF × 100%).AhR agonist activity of each sample was also converted to BaP equivalent concentration (BaP EQ) using the concentration-activity curves of β-NF and BaP (Supplementary material, Fig. S1).

Yeast-Based Assays for AR/TR Agonist and Antagonist Activity Detection
AR/TR agonist and antagonist activities were evaluated by yeast assays, in which yeast strains co-transfected with human nuclear receptor expression plasmids (AR or TRβ) and a lacZ reporter plasmid carrying the response element for the nuclear receptor-ligand complex were utilized (Shiizaki et al., 2010).Measurement of AR/TR disrupting activity was carried out similar to that of AhR agonist activity test with several modifications (Chen et al., 2014), and DHT, T3, FLU were used as AR agonist, TR agonist, and AR antagonist standards, respectively (Fig. S1).AR/TR agonist activity was calculated as the percentage of maximal response induced by 10 µM DHT/T3 (AR/TR agonist activity (%) = (OD 405 /OD 595 ) SAMPLE /(OD 405 /OD 595 ) 10 µM DHT/T3 × 100%).AR/TR antagonist activity was shown as fold of induction of 25 nM DHT/T3 (AR/TR antagonist activity (%) = (OD 405 /OD 595 ) SAMPLE /(OD 405 /OD 595 ) 25 nM DHT/T3 × 100%), and AR antagonist activity was transformed to FLU equivalent concentrations (FLU EQ) using the concentration-activity curves of FLU.

Rec Assay for Genotoxicity Evaluation
Genotoxicity in each indoor dust sample was analysed using the Rec assay (Takigami et al., 2002), which was carried out as described in Chou et al. (2014).Genotoxicity was calculated as the quotient of median inhibitory concentrations (IC 50 , survival rate (%) = 50%) of Rec(+) and Rec(-) strains (wild-type and DNA repair-deficient strains of Bacillus subtilis) (R 50 = (IC 50 ) Rec(+) /(IC 50 ) Rec(-) ).4-NQO and DMSO was used as the positive control (Fig. S1) and the negative control.Genotoxicity was confirmed when R 50 of a sample was greater than 1.5 (Oksuzoglu et al., 2008).

PAH Analysis by Using GC-MS
Contents of 36 PAHs in each indoor dust sample were determined as described in Ko et al. (2014) PAH contents were analysed using a Varian 3800 GC/ Saturn 4000 ion trap mass spectrometer equipped with a 30 m DB-5 column (i.d.: 0.25 mm, film thickness: 0.25 µm, Agilent, USA).Five perdeuterated PAHs (d10-acenaphthene, d10-phenanthrene, d12-benz[a]anthracene, d12benzo[a]pyrene, d12-benzo[g,h,i]perylene) was added to each dust sample extract as internal standards prior to analysis.PAHs were measured under the selected ion monitoring mode and were identified by retention times identical to those of PAH standards.Quantification of PAHs was carried out by comparing the integrated area of the molecular ion-extracted chromatogram to that of the analogous internal standards.Four perdeuterated PAH surrogates (d8-napthalene, d10-fluorene, d10-fluoranthene, d12-perylene) were added to the procedural blanks, which were analyzed with indoor dust sample extracts to assess quality control.The method detection limits (MDLs) of PAHs were calculated as: MDLs = the average mass of each individual PAH in the blanks + 3 × standard deviation (Table S1).

HPLC Fractionation
Indoor dust samples showing greater AhR agonist activities were subjected to HPLC fractionation by a HPLC (Hitachi L-2130, Japan) equipped with a diode array detector (Hitachi L-2455, Japan).A Supelco-Ascentis C18 column (5 µm, 150 × 4.6 mm, Sigma, USA) was eluted with a gradient condition of 10-100% methanol in water from 0-20 min, and maintained at 100% methanol for another 20 minutes.The flow rate was set at 0.8 mL/min, and the injection volume was 50 µL.Fractions were collected every 10 minutes from 0-10 and 30-40 minutes, and every 5 minutes from 11-15, 16-20, 21-25, and 26-30 minutes.Fractions were then evaporated to dryness and redissolved in DMSO for bioassay analysis.

AhR Agonist Activity in Indoor and Outdoor Dust Samples
AhR agonist activity was detected in indoor and outdoor dust from a university in southern Taiwan using a yeastbased reporter gene assay.As shown in Fig. 2, AhR agonist activities in the concentrated indoor and outdoor dust samples (concentration: 10 mg dust EQ/mL DMSO) ranged from 36.3-71.2% and 32.1-68.9% of the maximal AhR agonist response induced by 10 µM β-NF, respectively.Higher AhR agonist activities were detected in the indoor dust samples collected from computer room (16112 ng BaP EQ/g dust dry weight (dw)) and laboratory (9686 ng BaP EQ/g dust dw), which were further fractionated by HPLC to isolate active HPLC fractions.The AhR agonist activities in Taiwanese indoor dust (3448-16112 ng BaP EQ/g dust dw) were similar to or greater than those detected in indoor dust from New York State (30-8000 pg TCDD EQ/g, equivalent to 67-17778 ng BaP EQ/g) and Japan (38-1400 pg TCDD EQ/g, equivalent to 84-3111 ng BaP EQ/g), but lower than those in street dust collected from Vietnam and India (11-68 ng TCDD EQ/g, equivalent to 24-151 µg BaP EQ/g) using DR-CALUX assay (Suzuki et al., 2007;Tue et al., 2013;Tuyen et al., 2014).The variations in AhR agonist activities suggested area-specific sources of AhR agonist contaminants in different regions.

AR/TR Agonist and Antagonist Activity in Indoor and Outdoor Dust Samples
AR agonist and TR agonist activities were not found in dust samples collected in this study (Fig. S2).In contrast, indoor dust samples exhibited significant AR antagonist activities (Fig. 3(a), 728-1125 µg FLU EQ/g dust dw), and outdoor dust samples showed lower AR antagonist activities (Fig. 3(b), N.D-438 µg FLU EQ/g dust dw).TR antagonist activities were also detected in indoor and outdoor dust samples (Figs.3(c) and 3(d)).Suzuki et al. (2013) reported that AR antagonist activities were frequently found in indoor dust from Japan, the United States, Vietnam, the Philippines, and Indonesia, and several flame retardants exhibited antiandrogenic activity in the AR-CALUX assay.In addition, several synthetic compounds widely used in our daily lives, such as triclosan, parabens, and phthalates have been shown to elicit antiandrogenic or antithyroid hormonal activities (Chen et al., 2007;Shen et al., 2009;Christen, 2012), and these compounds have been detected in indoor dust as well (Bornehag et al., 2005;Fan et al., 2010).Quantification of these potential AR and TR antagonists may be useful in characterizing major AR/TR antagonist contaminants in indoor dust from Taiwan.

Genotoxic Activity in Indoor and Outdoor Dust Samples
Genotoxicity in indoor and outdoor dust samples were assessed using Rec assay.As demonstrated in Fig. 4, significant genotoxic activities (R 50 > 1.5) were detected in indoor dust from classroom, computer room, and laboratory, and the R 50 values were > 3.3, 2.9, and > 3.2, respectively.In contrast, only one outdoor dust sample collected from Sheng-Li Road elicited genotoxic potential (R 50 > 1.7).Previous studies have reported that indoor dust from houses using coal for heating and cooking in China showed genotoxic potential in in vitro and in vivo assays by forming DNA adducts (Naufal et al., 2007).Kang et al. (2011) also detected mutagenicity and genotoxicity in workplace dust and settled house dust from Hong Kong, Shenzhen and Concentration (mg dust EQ/mL DMSO) Guangzhou (Kang et al., 2011).Settled house dust from homes in Canada exhibited mutagenic potencies in Salmonella test as well, and PAHs were potential contributors to the mutagenic activity (Maertens et al., 2008).

Contents of PAHs in Indoor Dust Samples
In this study, 36 PAHs in indoor dust samples were analysed using GC-MS, and the total PAH contents were 277 (photocopy room), 108 (classroom), 88 (office), 207 (computer room), and 229 (laboratory) ng/g dust dw.The PAH contents in indoor dust in Taiwan were lower than those found in house dust from Shanghai's universities (9.84-21.44µg/g), major cities around the Pearl River Delta (1.63 to 29.2 µg/g) and Italy (0.036-34.5 µg/g) (Mannino et al., 2008;Kang et al., 2011;Peng et al., 2012).Significant correlation could not be found among total PAH content, AhR agonist activity, AR antagonist activity, and genotoxic activity, whereas samples with higher PAH contents, such as indoor dust from photocopy room, computer room, and laboratory also elicited stronger AhR agonist activities (Table 1).In addition, although several PAHs have been shown to exhibit AhR agonist, AR antagonist, and genotoxic potencies, their contributions to AhR agonist and AR antagonist activities in diesel exhaust particles or sediment have been reported to be low (Kizu et al., 2003;Louiz et al., 2008).Also, the contribution of genotoxic PAH metabolites formed via metabolic activation may be insignificant since the genotoxicity tests were performed without S9 activation in this study (Chou et al., 2014).
Some studies have shown that outdoor sources can be great contributors to indoor PAHs, in particular 5-to 7ring PAHs (Naumova et al., 2002;Ohura et al., 2004;Li et al., 2005).For example, the road dust containing PAHs from vehicular gasoline, diesel combustion, and coal and oil combustion is suggested to be a major outdoor source contributed to high indoor PAHs (Dong and Lee, 2009).Few studies have reported that 4-to 7-ring PAHs showed greater indoor-outdoor correlations compared to 2-to 3ring PAHs (Li et al., 2005;Kume et al., 2007).In this study, we only analysed the PAHs in indoor dust, and further GC-MS analysis will be carried out to elucidate the correlation between PAHs in indoor and outdoor dust and to identify predominant sources of PAHs.

Contributions of PAHs to the AhR Agonist Activity in Indoor Dust Samples
Among the 36 PAHs analysed in this study, fluoranthene, benzo  4. Concentration-survival rates of Rec(+) and Rec(-) strains exposed to different dust samples.R 50 > 1.5 indicates significant genotoxic activity (N.A: not available, IC 50,Rec(+) and IC 50,Rec(-) were greater than 10 mg dust EQ/mL DMSO).c,d]pyrene exhibited significant induction of AhR agonist activity in the yeast-based reporter gene assay (Murahashi et al., 2007).Table 2 shows the BaP EQ concentrations of individual AhR agonistic PAHs and the sum of BaP EQ concentrations of 9 AhR agonistic PAHs (BaP EQ 9 PAHs ) detected in indoor dust samples.The toxic equivalents were estimated using the toxic equivalent factors (TEFs) obtained from Murahashi et al. (2007).The 4-and 5-ring benz[a]anthracene and benzo[k]fluoranthene were principal AhR agonist activity contributors owing to their higher TEFs (7 and 11, respectively).Nonetheless, 9 AhR agonistic PAHs only accounted for a small portion of the AhR agonist activities in indoor dust (2.1-6.2%), which were similar to other studies (Louiz et al., 2008;Muusse et al., 2012;Tue et al., 2013).Our results indicated that further investigation is necessary to identify unknown xenobiotic AhR agonist contaminants in indoor dust from Taiwan.

HPLC Fractionation of AhR-Active Indoor Dust Samples
Indoor dust samples from computer room and laboratory were further fractionated by HPLC to investigate potential AhR agonists in these sample extracts.Fig. 5 shows the HPLC chromatograms (absorption wavelength: 254 nm) and the AhR agonist activities in corresponding HPLC fractions.According to Fig. 5, higher AhR agonist activities were detected in fractions collected from 15-20 and 20-25 minutes after fractionation.Several AhR agonist PAHs, including benz[a]anthracene, benzo[b]fluoranthene, benzo[k]fluoranthene, and benzo[a]pyrene were eluted in fractions collected from 25-30 minutes using similar HPLC conditions in our previous study (Chou et al., 2014).In this work, potential AhR agonist contaminants were detected in more polar fractions, which may contain AhR-active oxygenated and nitrated PAHs generated via atmospheric reactions or other unknown AhR active substances (Misaki et al., 2007;Bekki et al., 2009).Several peaks observed in the HPLC chromatograms (Fig. 5(a)) may be further isolated to investigate whether these substances are potential xenobiotic AhR agonists.

CONCLUSIONS
In this study, a battery of in vitro recombinant cell assays were used to detect AhR agonist, AR antagonist, TR antagonist, and genotoxic activities in dust samples from a university located in southern Taiwan.Contents of PAHs in indoor dust samples were also analysed, whereas PAH contents were not correlated to the AhR agonist activities or genotoxic activities found in dust sample extracts.Indoor dust samples collected from computer room and laboratory elicited higher AhR agonist activity, and potential AhR agonist candidates were present in HPLC fractions collected from 15-20 and 20-25 min after fractionation.By contrast, higher AR antagonist and genotoxic activities were found in indoor dust samples from office and classroom, suggesting that different types of household contaminants were present in different indoor dust samples.In conclusion, AhR agonistic PAHs were minor contributors to the AhR agonist activities detected in indoor dust samples, thus further identification of novel AhR agonists as well as other endocrine disrupting compounds is essential to better protect human health and the environment.

Fig. 1 .
Fig. 1.Sampling locations of indoor and outdoor dust from a university located in southern Taiwan.

Fig. 2 .Fig. 3 .
Fig. 2. Concentration-activity curves of (a) AhR agonist activity in indoor dust (b) AhR agonist activity in outdoor dust from a university located in southern Taiwan.
Photocopy Room Classroom Office Computer Room Laboratory PhotocopyRoom Computer Room

Fig. 5 .
Fig. 5. (a) HPLC chromatograms (UV absorption at 254 nm) (b) AhR agonist activity in HPLC fractions of indoor dust collected from computer room and laboratory in a university located in southern Taiwan (1 g dust EQ/mL DMSO).
This work was supported by the Ministry of Science and Technology, Taiwan (NSC 100-2221-E-006-037-MY2 and

Table 1 .
PAH content, AhR agonist activity, AR antagonist activity, and genotoxicity in indoor dust collected from a university located in southern Taiwan.

Table 2 .
BaP EQ concentrations of PAHs in indoor dust collected from a university located in southern Taiwan.TEF: toxic equivalent factor = BaP EC 50 /PAH EC 50 ; b N.D: not detected. a

Table S1
Contents of PAHs (ng/g dust dw) in indoor dust sample extracts a N.D: lower than MDL S3