Ac-DEVD-CHO

OPFRs and BFRs induced A549 cell apoptosis by caspase-dependent mitochondrial pathway

a b s t r a c t
Organophosphate flame retardants (OPFRs) and brominated flame retardants (BFRs) are frequently detected in indoor environment at high levels, posing health risks to humans. However, the potential cytotoxicity mediated by OPFRs and BFRs in relevant human cell models is limited. In current study, non- small cell lung cancer A549 cell was employed to investigate toxicity mechanisms of typical OPFRs (i.e., tris (2-chloroethyl) phosphate (TCEP), tris-(2-chloropropyl) phosphate (TCPP), tricresy phosphate (TCP), triphenyl phosphate (TPHP) and BFRs (i.e., 2,2’,4,4’-tetrabromodiphenyl ether (BDE-47), 3,3’, 5,5’-tetra- bromobisphenol A (TBBPA)). It was found that BDE-47 exhibited the strongest cytotoxicity, followed by TBBPA, TPHP, TCP, TCPP and TCEP. OPFRs and BFRs could cause the reduction of cell viability of A549 cell in both dose- and time-dependent manner after exposure for 24 and 48 h. Simultaneously, excessive generation of reactive oxygen species (ROS), mitochondrial membrane potential (MMP) dysfunction, cell apoptosis and overload of intracellular free Ca2+ demonstrated that cytotoxicity induced by OPFRs and BFRs were mediated by oxidative stress. Of note, the survival rate of cell significantly increased when pretreated with Ac-DEVD-CHO, suggesting that caspase-3 dependent mitochondrial pathway may have played a primary role in the process of A549 cell apoptosis.

1.Introduction
As a class of commercial chemical additives, organophosphate flame retardants (OPFRs) are extensively applied in household and industrial products, including upholstered furniture, hydraulic fluids, lacquer, anti-foam agents, building materials, baby products, electronics and other manufactured chemicals (Van der Veen and de Boer, 2012). In recent years, the phase-out and restrictions of polybrominated diphenyl ethers (PBDEs) contributed to excessive utilization of OPFRs (Stapleton et al., 2012; Stockholm Convention on POPs, 2013). The global consumption of OPFRs was approxi- mately achieved to 500,000 tons in 2011 and the usage increased to 680,000 tons in 2015 (Ou, 2011; van der Veen and de Boer, 2012). Specifically, the usage in China was estimated at approximately 300,000 tons in 2013 and increased rapidly (Zhang, 2014). Most of OPFRs basically exist in end-products by physical addition that are merely mixed with polymer substrates and not covalently bounded to the host matrix, thus these compounds can leach easily out into ambient environment through volatilization or abrasion, and exist persistently in the various matrix. (Marklund et al., 2003; Zhang et al., 2016). Many researches have reported that OPFRs were detected in various biotic and abiotic matrices including indoor air and dust (Ali et al., 2012; Cristale et al., 2018; He et al., 2018), sur- face water (Gustavsson et al., 2018), drinking water (Lee et al., 2016), sediments (Cao et al., 2017; Tan et al., 2016) and even found in human breast milk (Kim et al., 2014; Sundkvist et al., 2010).

Numerous studies identified that the higher levels of OPFRs were detected in indoor environments compared with BFRs, implying their potential human exposure risk (He et al., 2018; Larsson et al., 2018; Vojta et al., 2017). Chlorinated-OPFRs, such as tris (2-chloroethyl) phosphate (TCEP), tris-(2-chloropropyl) phos- phate (TCPP) have been concerned mostly due to their character- istics of potential suspected carcinogens (WHO, 1998), and TCEP was also reported as endocrine disruptors (Chen et al., 2015; Krivoshiev et al., 2018). Besides, the adverse effects of triphenyl phosphate (TPHP), tricresy phosphate (TCP), 2,2’,4,4’-tetra- bromodiphenyl ether (BDE-47), 3,3’, 5,5’-tetrabromobisphenol A (TBBPA) have also been revealed, comprising of reproductive toxicity, neurotoxicity, and immunotoxicity (Costa et al., 2015; Su et al., 2014; Xu et al., 2017; Yang and Chan, 2015). Chen et al. (2015) revealed that the growth of TM3 cells was significantly inhibited with treatment of TPP and TCEP at high concentration, and main genes expression related to testosterone (T) synthesis was down-regulated. Besides, TCEP could induce senescence like growth arrest via the p21Waf1/Cip1-Rb pathway in a p53- independent manner, and the regulators responsible for p38MAPK-NF-kB pathways were down-regulated in hepatocytes (Zhang et al., 2017).

TCEP and TCPP were capable of decreasing peripheral blood mononuclear cells (PBMCs) viability and altering PBMCs size and granulation (Mokra et al., 2018). Li et al. (2017b) clarified that excessive production of ROS induced by TCPP was detected in L02 cell, and TCPP might interfere with cell growth/ division and gene expression, energy and material metabolism, and signal transduction. Specifically, the rising levels of TPHP in indoor dust or air might disrupt endocrine levels and decrease semen quality (Meeker and Stapleton, 2010). As a consequence, the risk of OPFRs and BFRs to human health cannot be neglected. The routes of human exposure to OPFRs were similar to that of BFRs, including inhalation pathways (Cequier et al., 2014; Yadav et al., 2017), ingestion of indoor dust (Abdallah and Covaci, 2014) or dermal absorption (Abdallah et al., 2016). OPFRs exposure in indoor envi- ronments have also demonstrated that the intake rates of OPFRs for children and toddlers via inhalation were several orders of magnitude higher than that reported for adults (Xu et al., 2017). However, there have been a limited number of reports regarding the toxicological assessment of OPFRs and BFRs to human respi- ratory system up to now. As the common model in vitro assay, non-small cell lung cancer A549 cells were employed for systematical toxicology research of the respiratory system (An et al., 2016). In the present work, the cytotoxicity of TCEP, TCPP, TPHP, TCP, BDE-47 and TBBPA along with the further related toxic mechanism were investigated. The aim of present study was to (i) evaluate the cytotoxicity of OPFRs and BFRs; (ii) elucidate the potential cellular apoptosis mechanism of these compounds; (iii) and provide new evidence of the health risk of OPFRs and BFRs on human.

2.Materials and methods
Tris-(2-chloroethyl) phosphate (TCEP, 99% purity), tris-(2- chloropropyl) phosphate (TCPP, 99% purity), and tricresyl phos- phate (TCP, >98% purity), were purchased from Dr. EhrenstorferGmbH (Augsburg, Germany). Triphenyl phosphate (TPHP), 2,20,4,40- tetrabromodiphenyl ether (BDE-47, 100% purity) and tetra- bromobisphenol A (TBBPA, 100% purity) were purchased from Sigma-Aldrich (Gillingham, Dorset, UK). Dulbecco’s modified Ea- gle’s medium (DMEM), fetal bovine serum (FBS) and 0.25% trypsin- EDTA were purchased from Gibco-BRL (Carlsbad, CA, USA). All other reagents were presented in Text S1.Human non-small cell lung cancer A549 cell lines, obtained from ATCC, were grown in DMEM supplemented with 10% (v/v) FBS, 1% penicillin (100 U/mL) and streptomycin (100 mg/mL), and maintained in a 37 ◦C incubator (Thermo Electron Corporation,USA) with a humidified atmosphere containing 5% (v/v) CO2. The cells were trypsinized and subcultured every 3 days.Four OPFRs (ie., TCEP, TCPP, TPHP, TCP) and two typical BFRs (ie., BDE-47,TBBPA) were dissolved in dimethyl sulfoxide (DMSO) as a stock solution and then further diluted with DMSO to the con- centration of 50, 100, 200, 300 mM OPFRs and 10, 30, 50, 100 mM BFRs, respectively. Hereafter, A549 cells were seeded in 96-well or 6-well microplate at an initial density of 1.0 × 104 or 1.0 × 105 cells/ well respectively, containing a defined volume of growth medium (DMEM +10% FBS). After adherent incubation for 24 h, the cells separately exposed to 50, 100, 200, and 300 mM of OPFRs and 10, 30, 50, and 100 mM of BFRs for 24 and 48 h. Control groups received an equivalent amount of vehicle. The final concentration of DMSO in each treatment was lower than 0.3% (v/v); and DMSO at this con- centration did not affect A549 cell viability.Cell viability and cytotoxicity of OPFRs and BFRs were quanti- tatively evaluated with Cell Counting Kit-8 (CCK-8) and Neutral Red uptake assay kit, respectively (Jin et al., 2010). Briefly, A549 cells were seeded in 96-well microplate with a total volume of 200 mL growth medium and exposed to 50, 100, 200 and 300 mM of OPFRs and 10, 30, 50 and 100 mM of BFRs for 24 and 48 h, individually. For CCK-8 assay, after treatment with chemicals of different concen- trations, 20 mL of CCK-8 solution was added to each well, and then the wells were incubated at 37 ◦C for 2 h.

The absorbance was measured on a microplate reader (Bio-TEK instrument, USA) at awavelength of 450 nm. The cells viability were calculated as the percentage of OD treatment compared with OD control. Neutral Red uptake assay was conducted according to the manufacturer’s protocol. Briefly, after exposure terminated, 20 mL of neutral red solution was added to each well, and then the wells were incubated at 37 ◦C for 2 h. Afterward, growth medium containing neutral redsolution was removed, and then each well was rinsed with appropriate volume of PBS twice. Finally, the wells were placed on the shaker for 15 min after 200 mL of neutral red lysis solution was added to each of them. The absorbance was measured at a wave- length of 540 nm. The percentage of Neutral red uptake was expressed by the cell survival rate compared with the control. All experiments were run at least triplicate.A549 cells apoptosis was quantified with an Annexin V-FITC apoptosis detection kit (Beyotime Institute of Biotechnology, China), which was used to detect phosphatidyl serine that appeared on the surface of the cell membrane by FITC-labeledAnnexin V during apoptosis. Briefly, after exposure to various chemicals, cells were harvested and re-suspended with 50 mL Annexin-V binding buffer, and then stained simultaneously with2.5 mL Annexin V-FITC and 2.5 mL PI and incubated in the dark for 15 min. Subsequently, each sample was added with 150 mL Annexin-V binding buffer for the final detection. The apoptotic cells were ascertained by measuring the fluorescence intensity of fluo- rescein FITC and PI using a flow cytometer (Beckman Coulter Gal- lios, USA). More than 10000 cells for per dose was counted and analyzed using Kaluza Analysis 1.3 software (Beckman Coulter, USA). The cell apoptotic rates were calculated by the total sum of early apoptosis and late apoptosis.Mitochondrial membrane potential (MMP) was assessed with an ideal fluorescent probe JC-1 (Sigma-Aldrich, USA), which was widely used as an indicator of early apoptosis and depolarization of MMP.

Briefly, the manner of cell culture and exposure was consis- tent with that in Section 2.4. Afterward, cells were collected and rinsed twice with ice-PBS, and then loaded with a fresh JC-1 fluo-rescent probe containing medium (Sigma-Aldrich, USA) at 25 ◦C inthe dark for 30 min. Finally, the changes of MMP were ascertained by measuring the fluorescence intensity of JC-1 using a flow cy- tometer (Beckman Coulter Gallios, USA) (Yang et al., 2017). The percentage of B2 as the indicator of alteration on MMP was adopted for the final data.The 2’-7-’dichlorodihydrofluoresceindiacetate (DCFH-DA) fluo- rescent probe was applied to evaluate and quantify the generation of intracellular reactive oxygen species (ROS). DCFH-DA is cell- permeable and hydrolyzed by intracellular non-specific esterase to form DCFH, and then further oxidized by intracellular ROS to generate fluorescent product 2’,7’dichlorofluorescin (DCF). After co-culture with different concentration of six chemicals for defined exposure time, the cells were trypsinized, harvested and rinsed twice in PBS. Next, the cells were re-suspended in DCFH-DA (2 mM) for 15 min in the dark and washed again with PBS. Subsequently, each sample received 200 mL binding buffer and then the fluores- cence intensity of DCF of samples was immediately detected by flow cytometer (Beckman Coulter Gallios, USA) with excitation wavelength at 488 nm and emission wavelength at 525 nm. Kaluza Analysis 1.3 software (Beckman Coulter, USA) was employed to analyze data from at least 10000 cells.Caspase-3 inhibitory assay was determined using Ac-DEVD- CHO (Beyotime Institute of Biotechnology, China) that is a specific inhibitor for cell apoptosis via caspase-3 pathway.

To evaluate the suppression effect of Ac-DEVD-CHO, cells were seeded in 96-well plate with a total volume of 100 mL growth medium. Cells were pretreated with defined dosage of Ac-DEVD-CHO for 4 h, and then exposed to defined concentrations of chemicals for 24 and 48 h, respectively. Afterward, cells survival rates were performed as the indirect characterization to assess whether the A549 cells apoptosis was correlated to activating caspase-3 pathway.The intracellular Ca2+ concentration was quantified using aFluo-3AM fluorescent probe (Beyotime Institute of Biotechnology, China), which is cell-permeable and cleaved by intracellular esterase to form Fluo-3 that binds with Ca2+ to excite strong fluo- rescence. After exposure to OPFRs and BFRs, the cells were collected and loaded with 5 mM Fluo-3 AM in PBS for 40 min in the dark at25 ◦C. The cells were immediately determined using flow cytome-ter (Beckman Coulter Gallios, USA) with excitation wavelength at 488 nm and emission wavelength at 525 nm. The data of intracel- lular free Ca2+ level was expressed by the mean fluorescent in- tensity of Fluo-3.Quantitative results from three independent duplicate experi- ments are expressed as the mean ± standard deviation. Statistical analysis was conducted using GraphPad Prism 7.0 software. Sig- nificance level between various exposure groups and control groups was performed by ANOVA tests. P-values of less than 0.05 were considered statistically significant.

3.Results and discussion
The quantified analyses of cell viability was performed by the CCK-8 and NR uptake assays. The detailed results in Fig. 1 depicted that the cell viability and NR uptake rates dropped. The concentration-and time-dependent effects in both assays were observed as the concentration of target pollutants increased and the exposure time extended, which is similar to the results of previous study (An et al., 2016). As shown in Fig. 1A a and b, Fig. 1B a and b, after 24 h exposure, all dosages could cause certain decrease of cell viability and NR uptake, whereas the significant inhibitory effects on cell viability and NR uptake occurred as extending exposure for 48 h (p < 0.05) (Fig. 1A c and d, Fig. 1B c and d), sug- gesting that cell death was induced in dose-and time-dependent manner. In particular, when the cells treated separately with 300 mM OPFRs and 100 mM BFRs for 48 h, the cell viability and NR uptake dropped to 83% and 75.7% for TCEP, 61.8% and 52.6% forTCPP, 53.4% and 49.8% for TCP, 40.8% and 47.8% for TPHP, 50.9% and51.6% for TBBPA, and 39.1% and 42.2% for BDE-47, respectively. It needed to be noted that TPHP had the strongest inhibitory effect on cell viability and NR uptake compared with other OPFRs, while TCEP seemed to have no obvious suppression on cell viability. It was also found that similar cytotoxicity was obtained in 100 mM of BDE- 47 and TBBPA, and 300 mM of TPHP, inferring that the stronger cytotoxicity could be induced by BFRs rather than OPFRs. The discrepancy of toxicity, particularly with respect to the loss of cell viability was also observed between the OPFRs and BFRs. TCEP and TCPP with similar alky-pattern exhibited potential lower cytotox- icity than other four flame retardants (FRs) with benzene ring in their molecular structure. Furthermore, we found that TCP and TPHP posed relatively lower cytotoxicity in comparison with BFRs. Dishaw et al. (2011) elucidated that the toxic discrepancy of different FRs may be caused by different halogenated-substitution patterns. Moreover, similar halogenation substitution patterns be- tween target contaminants can exhibit similar toxicity due to the presence of structure-activity relationship. Based on this point of view, we deduced that the similar halogenated-substitution type, like TCEP and TCPP might cause similar toxic effects, which is inaccordance with the findings of Krivoshiev et al. (2018). TCP or TPHP with sole benzene ring structure might display lower cyto- toxicity than that of BFRs containing benzene ring structure with Br-substituted.In addition, the significant reduction of cell viability observed inthis study was associated with A549 cell apoptosis and death induced by OPFRs and BFRs (An et al., 2016; Jin et al., 2010; Tang et al., 2018). Yet, caspase-dependent mitochondrial pathway was considered as a crucial pathway responsible for cell apoptosis (Tang et al., 2013; Wang et al., 2014). To assess whether OPFRs and BFRs activate mitochondrial pathways to induce A549 cells apoptosis and further reveal the underlying toxicological mechanism, intra- cellular reactive oxygen species (ROS) levels, mitochondrial mem- brane potential (MMP), cell apoptosis, free Ca2+ release and Ac- DEVD-CHO assay were conducted.Reactive oxygen species (ROS) is confirmed to have been involved in multiple biological processes, the generation of which can significantly reduce cell viability and induce cell apoptosis, even death (Wang et al., 2016). The intracellular ROS formation was presented in Fig. 2. It was seen that a limited elevation of intra- cellular ROS levels occurred with the decrease of cell viability and NR uptake when cells were exposed to six FRs with indicated dosage for 24 h (Fig. 2a and b). Specifically, ROS levels in cells treated with individual TCEP, TCPP, TCP, TPHP of 300 mM and TBBPA, BDE-47 of 100 mM were 269.8%, 303.2%, 331.1%, 375.2%, 359.5% and397.5% higher than DMSO groups. In contrast with the 24 h results, except for TCEP exposure groups, a significant concentration- dependent increase of ROS levels was observed after 48 h treat- ment (Fig. 2c and d), with each intracellular ROS going up to 462.2%, 478.7%, 566.7%, 909.2%, 499.2% and 719.6%, much higher than that of control groups, which suggested that OPFRs and BFRs stimulatedredundant ROS generation in A549 cells (p < 0.05). These results were in agreement with the previous researches that OPFRs and BFRs could stimulate the excessive generation of ROS in cells, resulting in oxidative stress that played a crucial role in suppressing cell viability (An et al., 2016; Jin et al., 2010; Li et al., 2017a). ROS is produced by stimulation from exogenous sources. How- ever, excessive production of ROS can cause oxidative stress, which will trigger adverse effects including mitochondrial dysfunction and cell death (Tsai et al., 2012). OPFRs and BFRs significantly induced the excessive production of ROS, resulting in the loss of cell viability when exposure for 48 h (p < 0.05), which inferred that the oxidative stress might be an important factor for its cell viability inhibition (Fig. 1). In detail, compared to OPFRs, more ascent of ROS levels was received in cells treated with BFRs, demonstrating that BFRs possessed stronger capacity to induce ROS production in A549 cells, all of which implied that more powerful oxidative stress happened by BFRs stimulation. Moreover, due to their congenial molecular pattern, TCEP and TCPP had similar toxic effects, which was well in line with a previous study by Krivoshiev et al. (2018).Apoptosis is commonly considered to be one of the most crucial factors responsible for cell death (Son et al., 2010; Jiang et al., 2017). As such, to further study the cell apoptosis induced by FRs, the experiment was conducted by Annexin V-FITC/PI staining combining with flow cytometer (Beckman Coulter Gallios, USA). After exposure to OPFRs and BFRs for 24 and 48 h, cell apoptosis induced in dosage- and time-dependent manner by target chem- icals as well as the declining proportion of intact cells weredescribed in Fig. 3 and Fig. S1. The proportion of cells death after exposed for 48 h to individual TCEP, TCPP, TCP, TPHP of 300 mM and TBBPA, BDE-47 of 100 mM conspicuously increased by 17.6%, 28.7%, 29.5%, 36.3%, 29.8% and 33.1%, respectively (Fig. 3A c and d), much higher than that of DMSO groups (p < 0.05), while no significant apoptosis was observed for all tested groups with 24 h exposure (Fig. 3A a and b) (p < 0.05). It is worth noting that cellular apoptosis induced by TCEP was still the lowest in comparison with other FRs, whilst TBBPA and BDE-47 still exhibited stronger capacity of inducing cell death compared with OPFRs of the same dosage, which was consistent with the findings by Tang et al. (2018). Combining with the current results of ROS, excessive generation of ROS could induce lipid peroxidation and cellular damage including DNA and protein. Thus, unsaturated fatty acids and cholesterol as vital cellular components were attacked by ROS, resulting in the elevation of Malondialdehyde (MDA) level, which was deemed as a critical indicator for cellular damage (Gu et al., 2015). On the other hand, ROS have been confirmed to mediate apoptosis by regulating phosphorylation and activation of the MAPK pathways. The MAPK pathway is ubiquitous in mammalian cells, where it plays a vital role in the transmission of extracellular signals to the intracellular targets, with subsequent effects on proliferation, apoptosis, adhe- sion, and differentiation (Park et al., 2012; Zhan et al., 2012). Therefore, oxidative stress induced by OPFRs and BFRs evoked the process of cell apoptosis and even death.It is documented that collapse of mitochondrial membrane potential (MMP) irreversibly induced cell apoptosis (Tang et al., 2013; Wang et al., 2011) and also deemed as another indicator to reveal the underlying mechanisms of cellular apoptosis. Therefore, the MMP alteration was investigated by employing with JC-1 fluorescent dye with flow cytometer (Beckman Coulter Gallios, USA). OPFRs and BFRs caused the increase of B2% in a dose- and time-dependent manner (Fig. S2). The increase of B2% value prac- tically stood for MMP reduction. After exposure to individual OPFRs of 50 and 100 mM, and BFRs of 10 and 30 mM for 24 h, a slight decline of MMP was found (Fig. S2 a and b). Yet, MMP of cells significantly dropped when exposing to six chemicals at higher doses after 48 h and the value of B2% increased by 1.7%, 13.5%, 37.8%, 41.7% for 300 mM TCEP, TCPP, TCP, TPHP and 19.1%, 19.7% for 100 mM BDE-47,TBBPA, respectively (Fig. S2 c and d) (p < 0.05). Once again, thelowest value of B2% from TCEP proved that it might pose weaker cytotoxicity to A549 cells, which was consistent with previous study (Cheng et al., 2016; Krivoshiev et al., 2018). BDE-47 and TBBPA still exhibited the powerful cytotoxicity which was in agreement with the results of ROS formation, cell apoptosis rates and cell viability.The mitochondria-mediated pathway is considered as one of the apoptosis types, especially in ROS-induced apoptosis. Intracellular ROS overload can give rise to the dysfunction of mitochondria and thereby promote further release of ROS into the cytosol (Maria et al., 2013; Lin et al., 2011). Guo et al. (2016) and Tsai et al. (2012) have reported that excessive production of ROS could facilitate free radical damage to the components of the respiratory chain, thereby destroying mitochondrial function. Furthermore, the mitochondrial dysfunction would trigger the opening of mito- chondrial permeability transition pore (MPTP), some molecules would non-selectively diffuse into mitochondria that resulted in mitochondria depolarization, disturbing oxidative phosphorylation process and accelerating adenosine triphosphate (ATP) depletion, all of which could cause the decrease of MMP, increase of intra- cellular Ca2+ load and activation of the apoptosis signal (Gao et al.,2006; Halestrap et al., 2004; Li et al., 2012a). The relationship be- tween MMP reduction and ROS formation was also exhibited in the following aspects: mitochondrial membrane lipids and membrane proteins were attacked by excessive ROS accumulated in the cell, resulting in inactivation of membrane receptors, which in turn led to the increase of mitochondrial membrane permeability and the reduction of membrane potential (Ma et al., 2011). Wang et al. (2017) reported that pro-apoptotic protein can homodimerize and activate caspases by altering the mitochondrial function, which will result in the release of cytochrome c into the cytosol. On the con- trary, the anti-apoptotic protein protects mitochondria from mitochondrial dysfunction through inhibiting the opening of MPTP. Besides, the thiol structure of adenine nucleotide translocase (ANT) on the mitochondrial inner membrane and the disulfide bond are confirmed to be main site susceptible to ROS attack (Schriewer et al., 2013), and overloaded ROS could cause conformational changes of mitochondrial permeability transition pore by oxidizing the mitochondrial inner membrane and some sensitive sites on the mitochondrial permeability transition pore, thereby inducing pore opening and leading to a decrease in MMP. In addition, ROS can also directly destroy the membrane potential established by mito- chondrial electron transport chain complex I and complex II, resulting in the decrease in MMP (Chen et al., 2003). Collectively, these results suggested that OPFRs and BFRs were able to induce mitochondria dysfunction and affect a series of biological process.In order to further assess whether tested flame retardants induced apoptosis was correlated with the activation of caspase-3 pathway in A549 cells, Ac-DEVD-CHO, a caspase-3 inhibitor was applied to pretreat the cells before exposure to indicated doses of OPFRs and BFRs. The inhibitory effect was described indirectly by measuring the cell survival rates. As clearly seen in Fig. S3, cell survival improved by the pretreatment of caspase-3 inhibitor over 24 and 48 h exposure. A trifling elevation of cell survival after exposure to 24 h (Fig. S3 a and b), and a remarkable increase of cell viability and cell number after 48 h (Fig. S3 c and d) occurred in A549 cells. In particular, when cells underwent pretreatment of caspase-3 inhibitor, cell survival after exposure to 300 mM OPFRs and 100 mM BFRs for 48 h significantly increased by 20.4%, 21.8%, 18.9%, 23.0%, 30.4% for TCPP, TCP, TPHP, TBBPA and BDE-47 testedgroups, respectively, and the highest cell survival (88%) for TCEP was also observed (p < 0.05). Caspase-3 was confirmed to be a down-stream effector and a critical induction factor of the cell apoptosis (Wang et al., 2016). The depolarization of MMP can lead to cytochrome c release, resulting in the caspase-3 activation and further PARP cleavage, thereby triggering a series of apoptotic re- actions such as nuclear DNA fragmentation (Wang et al., 2014). In current study, the cell apoptosis was alleviated by the inhibitory effects of Ac-DEVD-CHO on the caspase-3 biosynthesis, but the adverse impacts driven via OPFRs and BFRs were not completely offset, suggesting that caspase-3 dependent pathway might play an primary role in the process of A549 apoptosis (Wang et al., 2014). In other words, the present results indicated that the apoptosis pro- cess induced by target chemicals might mainly depend on the caspase-dependent mitochondrial pathways (Tang et al., 2013).A remarkable sustaining elevation of intracellular Ca2+ level is considered as one of the most critical stimulating factor responsible for cell death and cell apoptosis (Wang et al., 2011). A specific cell- permeable fluorescent dye, Fluo-3AM was employed to assess theeffects of OPFRs and BFRs on intracellular Ca2+ level in A549 cells. As clearly depicted in Fig. 4, OPFRs and BFRs significantly gave rise to the increase of intracellular Ca2+ content that showed positive correlation with chemicals at dose and exposure time dependent manner (p < 0.05). In detail, after exposing individually to 300 mM of TCEP, TCPP, TCP, TPHP and 100 mM of TBBPA, BDE-47 for 24 h, thelevel of intracellular free Ca2+ was 118.5%, 135.7%, 163.1%, 179.6%, 153.4% and 161.9% higher than that of control cells separately (Fig. 4a and b). Similarly, when cells underwent 48 h treatment with FRs, the intracellular free Ca2+ level was 171.3%, 207.1%, 230%, 271.3%, 301.3% and 341.6% significantly higher in contrast with control groups, respectively(p < 0.05) (Fig. 4c and d). These results demonstrated that the significant elevation of intracellular free Ca2+ levels was induced by the chemicals when exposed to 24 and 48 h with the lowest Fluo-3 fluorescence intensity by TCEP still being observed, all of which was in accordance with a series of response to oxidative stress results such as the disruption of MMP, the increase of ROS, decrease of cell viability, and elevation of cell apoptosis rates, indicating that the imbalance of intracellular Ca2+ level was associated with the loss of integrity of mitochondrial membrane and eventual apoptosis (Lin et al., 2017; Yang et al., 2013). Zhang et al. (2012) clarified that calcium was considered as an indispensable secondary messenger for its key role in tuning apoptosis. In general, an increase of Ca2+ concentration could occur through release from intracellular stores via Ca2+-selective chan- nels at the plasma membrane, or be mediated by the calcium dependent cysteine protease either in cytosol or in mitochondria (Araújo et al., 2010; Guo et al., 2016; Li et al., 2012b). Additionally, excessive accumulation of intracellular Ca2+ could activate some Ca2+-dependent proteolytic enzymes, endonucleases, which caused the decomposition of a large amount of cytoskeleton, DNAand nucleic acid, and inhibited cell division and growth, yet, the degradation of DNA by Ca2+-dependent endonuclease was considered to be the key factor for DNA fragmentation in cell apoptosis (Robertson et al., 2000). Furthermore, the overload of intracellular free Ca2+ caused cell membrane and mitochondrial inner membrane damage, the activity of mitochondrial electron transport chain was inhibited, and a large amount of ROS was generated in a short time, thereby leading to further aggravated mitochondrial damage, and finally resulting in cell apoptosis (Wang et al., 2011). Ca2+ release is one of the common characteristics for A549 cell apoptosis and death, the stronger signal of Ca2+ overload induced by BFRs compared with OPFRs under same dose suggests that BFRs exhibited more powerful cytotoxicity than OPFRs. 4.Conclusion In current study, the cytotoxicity variance and potential toxic mechanism induced by TCEP, TCPP, TCP, TPHP, TBBPA and BDE-47 were investigated. The results proved that all tested OPFRs and BFRs could give rise to toxic effect on A549 cells via inhibiting cell viability, forming excessive ROS, causing MMP dysfunction, inducing intracellular Ca2+ overload and cell apoptosis. Addition- ally, we also found six flame retardants induced cell apoptosis was likely correlative with oxidative stress mediated by intracellular ROS generation. Further studies on cell apoptosis pathway revealed that apoptosis induced by OPFRs and BFRs was caspase-dependent mitochondrial pathway. Moreover, it was found that BDE-47 possessed the most powerful cytotoxicity on A549 cells followed by TBBPA, TPHP, TCP, TCPP, and TCEP. Of note, the lower cytotoxicity induced by TCEP should not be neglected. Given the Ac-DEVD-CHO relatively short exposure time in our current study, chronic exposure investigation is necessary for revealing the potential persistent toxicity effect of TCEP and TCPP from the perspective of cell and molecular biology.