- Email: [email protected]
Phototoxic risk assessment on benzophenone UV filters: In vitro assessment and a theoretical model
Toxicology in Vitro 60 (2019) 180–186
Contents lists available at ScienceDirect
Toxicology in Vitro journal homepage: www.elsevier.com/locate/toxinvit
Phototoxic risk assessment on benzophenone UV filters: In vitro assessment and a theoretical model Lidan Xionga,b, Jie Tanga,b, Yiming Lic, Li Lia,b,c,
T
⁎
a
Cosmetics Safety and Efficacy Evaluation Center, West China Hospital, Sichuan University, No. 5, Gong Xing Road, Chengdu, Sichuan, PR China Sichuan Engineering Technology Research Center of Cosmetic, Chengdu, Sichuan, PR China c Department of Dermatology, West China Hospital, Sichuan University, No. 37, Guo Xue Xiang, Chengdu, Sichuan, PR China b
A R T I C LE I N FO
A B S T R A C T
Keywords: Benzophenone UV filters Phototoxic In vitro assessment A theoretical model
Benzophenones (BPs), filtering out both UVA and UVB rays, are widely used in a great variety of sunscreens and personal care products. However, they have not been extensively studied for the mechanisms of UV-absorbing toxicity. In this study, we used CPZ (chlorpromazine) as a positive control and SDS (sodium dodecyl sulfate) as a negative control, and the phototoxic of BP-1, BP-3 and BP-4 were investigated in vitro assays using three cell types under different UV exposure conditions. This was followed by setting up a theoretical model, which was adopted to predict and compare the phototoxicity. It was found that Balb/c 3T3 (Balb/c 3T3 fibroblast cell lines) showed sensitivity to UVA+ and UVB+ exposure, while the HS68 (human HS68 fibroblast cell lines) to UVA+ and the HaCaT (human HaCaT keratinocyte cell lines) to UVB+. The test compound, BP-1, was detected to be phototoxic at UVA+ conditions, but BP-3 and BP-4 were discovered to be non-phototoxic at UVA+ conditions. This demonstrated that BP-1, BP-3 and BP-4 remained low-risk chemicals under UVB+ condition. The theoretical calculation of the energy gap (EGAP) showed BP-1(EGAP) > BP-3(EGAP) > BP-4(EGAP).
1. Introduction UV radiation, including UVB and UVA spectra, can apparently penetrate human skin tissues to a considerable depth, which promotes the development of melanoma and other skin diseases (Bowden, 2004). Phototoxicity is of increased concern in dermatology, since people in modern society pay more attention to sun radiation. The definition of phototoxicology is given, as well as an illustration of the primary mechanisms for different phototoxic endpoints including energy transfer, electron transfer and formation of toxic photoproducts (Wondrak et al., 2006; Carbonare and Pathak, 1992). Generally, phototoxicity is increasingly assessed by the 3T3 NRU (neutral red uptake) in vitro phototoxicity test as described in the OECD Guideline No. 432 (OECD, 2004). Besides, different cell lines are needed for studying discrepancies in vitro phototoxicity experiments (Gülden et al., 2005), including HEKn, HDF fibroblasts (Rai et al.,
2011), Caco-2 (Broeders et al., 2013), HepaRG (Broeders et al., 2013) and artificial skin constructs (morphological changes, immunochemistry) (Lelièvre et al., 2007). In addition, the use of different set-up of the cell assay should be considered, including differences in exposure conditions such as cell concentration, solvents and end-points (Tanneberger et al., 2010). At present, there are several assays to evaluate of the toxic end-points in vitro. Previous studies have indicated that there is a high correlation between CCK-8(Cell Counting Kit-8), MTT(3-(4,5)-dimethylthiahiazo(-z-yl)-3,5-di-phenytetrazoliuromide) and NRU (Neutral Red Uptake) assays (Ishiyama et al., 1997; Lou et al., 2010). Compared with other two, CCK-8 assay is more rapid and sensitive, yielding accurate method for cytotoxicity detection. Therefore, the procedure is more suitable for this study. This study employs a minor deviation from the OECD No. 432 guideline using CCK-8 assay as a substitute for NRU. The discrepancies of UV spectral within the different cell types are
Abbreviations: CPZ, chlorpromazine hydrochloride; SDS, sodium dodecyl sulfate; BP-1, 2,4-dihydroxybenzophenone; BP-3, oxybenzone,2-hydroxy-4-methoxybenzophenone; BP-4, 5-benzoyl-4-hydroxy-2-methoxybenzenesulfonic acid; Balb/c 3T3, mouse embryonic fibroblast cell line; HS68, neonatal human foreskin fibroblast lines; HaCaT, neonatal human keratinocytes; PIF, photo irritancy factor; MPE, mean phototoxic effect; CCK-8, Cell Counting Kit-8; ROS, reactive oxygen species; QSAR, quantitative structure-activity relationships; ICH, International council for harmonization of technical requirements for pharmaceuticals for human use; OECD, organization for economic cooperation and development; HOMO, highest occupied molecular orbital; LUMO, lowest unoccupied molecular orbital ⁎ Corresponding author at: Cosmetics Safety and Efficacy Evaluation Center, West China Hospital, Sichuan University, No. 5, Gong Xing Road, Chengdu, Sichuan, PR China. E-mail address: [email protected] (L. Li). https://doi.org/10.1016/j.tiv.2019.05.021 Received 9 December 2018; Received in revised form 27 May 2019; Accepted 29 May 2019 Available online 30 May 2019 0887-2333/ © 2019 Published by Elsevier Ltd.
Toxicology in Vitro 60 (2019) 180–186
L. Xiong, et al.
Fig. 1. Chemical structures of Benzophenone UV filters, CPZ (positive) and SDS (negative), with the range of absorption.
pathogenetic mechanisms of drug-induced phototoxicity. (Ringeissen et al., 2011).Available tools include in silico prediction systems such as DEREK allows toxicity prediction of chemicals based on structures known to be associated with the incidence of toxicity(Onoue et al., 2017).The OECD (Q)SAR is another initiative model which promotes the use of non-testing information for many diverse toxicological endpoints through the use of existing knowledge, categories based on mechanistic rationale and/or structural similarity, read-across and mathematical models based on structure activity relationships (OECD (Q)SARs projects, 2010; Ringeissen et al., 2011).Besides, HOMO-LUMO gap evaluation provides a measure of the photoreactive potential of chemicals (Betowski et al., 2002; Onoue et al., 2017). The internal effects conditioning the light absorbance and stability of chemicals are found to be best described by the energy gap (EGAP) between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) (Parr and Zhou, 1993), and EGAP evaluation provides a measure of the photoreactive potential of chemicals (Ringeissen et al., 2011). Although these are not completely specific for chemical phototoxicity in silico and photochemical prediction tools, the prediction methods, perhaps deployed in a high throughput system, may be appropriate for initial screening of chemicals to predict risk of phototoxicity. Given the above facts, this study was to examine the phototoxic potential of the BP-1, BP-3, and BP-4 using the Balb/c 3T3 cells and CPZ as positive control, SDS as negative control following the modified guidelines from OECD No. 432 (OECD, 2004). The studies were further extended from Balb/c 3T3 to human skin cells HS68 and HaCaT to compare the inter-cellular differences and sensitivity of normal human dermal and epidermal cells for phototoxicity analysis. This may allow us to understand the correlation of toxicity levels of compounds in mouse cell lines and human cell lines. Next, the study was further extended from UVA to UVB in order to observe the response of the tested cells against different UV spectra. Subsequently, this study was to examine the toxic/phototoxic potential of a set of common BP-type UVfilters. It is essential that such sunscreen products are safe, effective, and useful. The present results would provide useful information for evaluating the potential skin health risk of BPs. Finally, this study used
usually ignored. However, UVA and UVB affect different layers of the skin, and induce damage to cells by different mechanisms. Epidermis and dermis cells constantly exposed to high energetic UVB and less energetic but more penetrating UVA radiation react in different ways (Herrling et al., 2006). Compared with UVA filters, UVB filters are readily available, while most phototoxicities were regarded as caused and tested by UVA previously (OECD, 2004). One aim of our study is to verify which kinds of skin cells (fibroblasts or keratinocytes) were more susceptible to UVA/UVB irradiation. Sunscreens contain organic and inorganic UV filters as active ingredients. The organic UV filters absorb UV radiation within a specific range of wavelengths depending on their chemical structure, and usually possess aromatic structures that can absorb the solar UV radiation (Crovara Pescia et al., 2012; de Oliveira et al., 2015). Benzophenones (BPs) are among well-known organic UV filters due to theirs broad absorption range in the UVA and UVB spectra (Benevenuto et al., 2015). BPs contain an aromatic ketone with the ability to absorb UVB and UVA, whose absorption is influenced by hydrogen bonding (Saroj Kumar et al., 2015) showed in Fig. 1. However, BPs may penetrate through stratum corneum into deeper skin layers for their small size, which may induce contact dermatitis and skin sensitization by enhancing the ROS generation and electron delocalization in presence of UV radiation (Saroj Kumar et al., 2015). However, there is no study reported in the literature which compares phototoxicity of BP-1, BP-3 and BP-4 in skin cells. In addition, CPZ (Chlorpromazine hydrochloride) has been shown to cause photosensitivity induced damage to membranes, protein and DNA by both direct and indirect mechanisms (Kochevar, 1981; Du et al., 2017). This phototoxic potential is exhibited both in vitro and in vivo conditions. Therefore, this parameter of phototoxicity is selected as the positive in our study. SDS (Sodium dodecyl sulfate), proven in experimental and clinical trials, is adopted as the negative control for this study (OECD, 2004). Today, there is an increasing need for filtering compounds to be tested in the cells, however, this requires since testing requires resources both in terms of cost and time. Various methodologies, including in silico prediction models, have been developed to predict the phototoxic potential of chemicals over the past few years, based on the 181
Toxicology in Vitro 60 (2019) 180–186
L. Xiong, et al.
Subsequently, the plates were washed with PBS (150 μl/well) and the cells were exposed to eight different concentrations of test solutions/ controls prepared in PBS of each test material (100 μl/well) for 1 h. One plate was then exposed to UVA (UVA+) at 5 J/ cm2, or UVB(UVB+) at 30 mJ/cm2, while the other plates were placed in the dark environment and incubated for observing cytotoxicity caused solely by BPs. The UV irradiation conditions were determined by the instrumental programmed software (Sigma, Shanghai, China). Next the chemical solutions were decanted and washed with PBS followed by addition of cell media (100 μl/well). After 18–22 h of incubation, the cells were washed with PBS, and DMEM (100 μl/well) containing CCK-8 (10% v/v) was added to each well following incubation time of 3 h. Absorbance was measured at 450 nm using a microplate spectrophotometer (Bio-Red, Model 680, USA).
theoretical calculation, which was conducted using structure activity relationship and molecular docking methods. Our study allowed for a better understanding of the mechanisms occurring in the human skin, and may provide a better choice to cosmetic formulators for selecting UV filter in the effective products. 2. Material and methods 2.1. Chemicals and instrumentation BP-1 (Benzophenone 1, 2,4-Dihydroxybenzophenone, with a purity 99%, CAS NO. 131–56-6), BP-3 (Benzophenone 3, Oxybenzone,2Hydroxy-4-methoxybenzophenone, with a purity 98%, CAS NO. 13157-7), BP-4 (Benzophenone 4, 5-Benzoyl-4-hydroxy-2-methoxybenzenesulfonic acid, with a purity ≥97.0%, CAS NO.4065-45-6), SDS (Sodium dodecyl sulfate, with a purity ≥99.0%, CAS NO.151-21-3) and CPZ (Chlorpromazine hydrochloride, CAS NO.69-09-0) were obtained from Sigma Chemical Co. (SIGMA-ALDRICH, USA). Fetal bovine serum (FBS) was procured from Gibco (Thermo Fisher, USA). CCK-8 (Cell Counting Kit-8) was obtained from Dojindo (Dojindo Laboratories, Japan). PBS (phosphate buffers 10X) was purchased from Zsbio Commerce CO. (Zsbio, China). DMEM (Dulbecco's Minimum Essential Medium), penicillin and streptomycin solution, trypsin (0.25%), and DMSO (dimethylsulfoxide) were acquired from HyClone (GE Health care Life Science, USA). UVA Bio-Sun instrument (320-340 nm, SS-04A, Sigma, Shanghai, China), UVB irradiation instrument (290–315 nm, SS04B, Sigma, Shanghai, China), BIO RED Model 680 Microplate Reader (BIO RED Inc., USA).
2.4.2. HS68 phototoxicity assays For the cytotoxicity and phototoxicity assays, HS68 cells were cultured in 96-well plates (Corning, USA) at a seeding density of 8 × 103 cells/well. The cells were cultured in the plate for 24 h, and the following steps were the same as 2.4.1. 2.4.3. HaCaT phototoxicity assays HaCaT cells were transferred to 96-well plates at a seeding density of 104 cells/well, and were cultured for 24 h at 37 °C in 5% CO2.The following steps were the same as 2.4.1. 2.4.4. Statistical analysis The Phototox Version 2.0 software was employed for concentration–-response analysis by PIF and MPE. The Phototox Version 2.0 software, available by OECD, was used to calculate Photo-IrritationFactor (PIF) and Mean Photo Effect (MPE). Each detected material was tested for three times in dependent assays.
2.2. Cell culture The murine Balb/c 3T3 fibroblast cell lines were purchased from Kunming Cell Bank of Chinese Academy of Sciences. These cells were cultured in supplemented DMEM (with 10% FBS, 100 U/ml penicillin and 100 μg/ml streptomycin). The human HS68 fibroblast cell lines were obtained from American Type Culture Collection (ATCC, CRL1635, USA) and were cultured in DMEM (with 10% FBS, 100 U/ml penicillin and 100 μg/ml streptomycin). The human HaCaT keratinocyte cell lines were ordered from Kunming Cell Bank of Chinese Academy of Sciences and cultured in DMEM (with 10% FBS, 10 ng/ ml HKGS (human keratinocyte growth factor), 100 U/ml penicillin and 100 μg/ml streptomycin).
PIF =
IC 50(UV −) IC 50(UV +)
PIF < 2 (NO phototoxic); 2 < PIF < 5 (PROBABLE Phototoxic); PIF > 5 (Phototoxic). When IC50 cannot be obtained, MPE is calculated as following equation n
MPE =
∑i = 1 wi PEci n
∑i = 1 wi
PIF < 2 or an MPE < 0.1 (NO phototoxic); 2 < PIF < 5 or 0.1 < MPE < 0.15 (PROBABLE Phototoxic); PIF > 5 or MPE > 0.15 (Phototoxic) (OECD, 2004). Results were obtained after calculating the mean of at least two independent experiments.
2.3. Preparation of test materials Benzophenones (BPs) have limited solubility in water, hence Dimethylsulphoxide (DMSO) was selected as the solvent based on OECD, 2004 recommendations. Stock solutions of BP-1, BP-3, BP-4, SDS and CPZ were prepared at 1 M using DMSO. For the assay, the chemicals were diluted in PBS. Concentrations of 0.1–500 μM in PBS were achieved which is the maximum concentration of DMSO was safe for the cells. The final solutions with concentrations of 500, 100, 50, 10, 5,1, 0.5 and 0.1 μM were prepared in PBS. All test materials were freshly prepared prior to use.
2.5. Prediction of phototoxicity All DFT calculation were performed with DMol3 program package of Materials Studio (Delley, 1990, 2000; Perdew et al., 1996). PerdewBurke-Ernzerh (PBE) of correlation function of generalized gradient approximation (GGA) level was conducted. A DFT semi-core pseudopotentials (DSPP) core treatment was chosen, which has been used in many DFT studies to optimize structures. The localized double numerical polarization (DNP) basis set was applied to expand the KohnSham orbitals, which included a polarization p-function on all hydrogen atoms. Self-consistent field (SCF) calculations were carried out with a tolerance of 1*10−5 a.u. on the total energy. The real-space global orbital cutoff was chosen to be 4.4 Å in the computation. The value of smearing was 0.005 Hartree. COSMO solvent calculation with the DCPBE functional was also performed using the dielectric constant of DMSO (46.7). The energy gap (EGAP) is defined as followed: EGAP = E (LUMO)-E (HOMO). Therefore, the more positive the value is, the higher the EGAP. The electronic parameters in a molecule are strongly affected by their conformational flexibility. Three molecules were used
2.4. Test procedure 2.4.1. 3T3 phototoxicity assays 3 T3 cells culture was carried out as described by the OECD 432 (OECD, 2004) adjusted for the actual conditions. In brief, 3T3 cells were plated on 96-well plates (Corning, USA) at a seeding density of 104 cells/well. Initially, a cell suspension of 105 cells/ml was prepared in the culture medium and 100 μl cell suspension was seeded exclusively into the central 60 wells of the plate. In the peripheral wells of the 96well plate, 100 μl of the culture medium alone (without cells) was added (blank). The cells were cultured in the plate for 24 h. 182
Toxicology in Vitro 60 (2019) 180–186
L. Xiong, et al.
Fig. 2. Effects of the three cell systems under UVA and UVB irradiation. (A) human skin cells lines (HS68 and HaCaT) and mouse cell line Balb/c 3T3 under UVA (5 J/ cm2) and UVB (30 mJ/cm2) irradiation, (B) Cell viability was evaluated in 3T3, HS68 and HaCaT cells with UVA and UVB. Mean and SD are shown for three independent experiments. *p < 0.05; **p < 0.01.
under exposure and non-exposure (dark conditions) of UVA/UVB light (Fig. 2B). When 3T3 cells were exposed to UV at 5 J/cm2 (UVA) and 30 mJ/cm2 (UVB), 3T3 cell viability significantly declined to 56.25% (UVA+) and 61.39% (UVB+) respectively compared with non- exposure control. While the relative survival rate of HS68 cells decreased to 68.22% (UVA+) and 81.20% (UVB+), HaCaT cell viability decreased to 82.50% (UVA+) and 67.00% (UVB+).
to explore the bell-shape relationship between phototoxicity and EGAP: the most reactive (i.e., those with maximum EGAP). 3. Results 3.1. Measurement of the three cell lines exposing to UVA and UVB irradiation We first screened whether UVA and UVB irradiation affected the cell proliferation on human skin cells lines (HS68 and HaCaT) and mouse cell line Balb/c 3T3 (Fig. 2). Cell images were taken in presence and absence of UV radiation and these images were provided qualitative visual information on the overall effect of UV radiation on the cells (Fig. 2A). CCK-8 assay was performed to investigate cell viability, both
3.2. Evaluating phototoxicity of CPZ and SDS in the three cell systems under UVA and UVB exposure CPZ and SDS, as the reference chemicals for positive and negative controls, were tested in vitro for phototoxicity in human skin cell lines (HS68 and HaCaT) and mouse cell line Balb/c 3 T3 (Table 1). Briefly, 183
Toxicology in Vitro 60 (2019) 180–186
L. Xiong, et al.
Table 1 Photo irritation factor (PIF) and mean photo effect (MPE) values for CPZ and SDS generated by phototox software in Balb/c 3T3, HS68 and HaCaT cells. Cell type
3T3 3T3 3T3 3T3 HS68 HS68 HS68 HS68 HaCaT HaCaT HaCaT HaCaT
Substance
CPZ SDS CPZ SDS CPZ SDS
Run
1 2 1 2 1 2 1 2 1 2 1 2
UVA
UVB
PIF mean
MPE mean
Phototoxic potential
PIF mean
MPE mean
Phototoxic potential
76.592 79.943 1.031 0.982 52.602 43.838 0.896 0.850 24.267 46.209 1.996 1.456
0.584 0.601 0.002 0.035 0.435 0.419 0.003 0.053 0.205 0.387 0.086 0.023
Yes Yes No No Yes Yes No No Yes Yes No No
16.524 18.707 0.217 0.188 9.657 13.508 1.547 0.734 18.621 22.702 1.000 0.611
0.227 0.235 0.026 0.022 0.198 0.377 0.008 0.022 0.248 0.284 0.024 0.014
Yes Yes No No Yes Yes No No Yes Yes No No
under UVB were calculated to be < 2, MPE < 0.15, respectively. BP-3 and BP-4 were showed to be non-phototoxic in UVA+ conditions, because PIF and MPE values were calculated to be < 2, MPE < 0.15, respectively. Thus, BP-1 was identified as phototoxic (UVA) molecules, while BP-3, and BP-4 were determined as non-phototoxic (UVA) ones. Besides, BP-1, BP-3 and BP-4 remained low-risk chemicals in UVB+ condition.
photo Irritation Factor (PIF) and Mean Photo Effect (MPE) values for CPZ generated by Phototox Software in 3T3 cells are shown in Fig. S1 which present the sources of data (Fig. S1a and b). The results of CPZ and SDS coincided with the range recommended by OECD No.432 (OECD, 2004). In Table 1, the results obtained were summarized. The result of PIF Mean and MPE for SDS in 3T3 cells both under UVA and UVB were all negative, which served as a negative control in the following tests. SDS was also found to be completely non-phototoxic in HS68 and HaCaT cells. The result of PIF and MPE for CPZ in 3T3, HS68 and HaCaT cells were found to exhibit the phototoxicity potential of CPZ in presence of UVA and UVB.
3.4. Prediction of phototoxicity In order to further illustrate the potential mechanism of the phototoxicity associated with these three substances, we mainly used the theoretical calculations. According to the report (Ringeissen et al., 2011), we know that the higher EGAP, the more reactive the substance is. The HOMO and LUMO orbitals (Table 3) were used to calculate the EGAP value. From the results, the EGAP of BP-1 was higher than BP-3 and much higher than BP-4, and we could speculate that the phototoxicity of BP-1 was stronger than that of BP-3 and BP-4.Note that we could speculate that the phototoxicity of BP-1 provide a close representation to BP-3, while the in vitro results suggested that BP-1 was tested for their phototoxicity in comparison to BP-3. That may be another limitation of the cell test.
3.3. Measurement of exposuring BP-1, BP-3 and BP-4 in the three cell systems under UVA and UVB irradiation The cell viability curves of BPs with or without UVA/UVB irradiation were obtained, and the PIF and MPE values were calculated (Table 2). Compared with the non-irradiated groups, UVA-irradiated BP-1 exhibited significant cytotoxicity at lower doses. The PIF values of BP-1 were estimated to be > 2.0 and the MPE values of BP-1 were higher than 0.1 in 3 T3, HS68 and HaCaT cells, which identified as phototoxic or probable phototoxic molecules. However, compared but compared with positive control CPZ, PIF and MPE values of BP-1 were lower and more uncertain. In contrast, the PIF and MPE values of BP-1
Table 2 Photo irritation factor (PIF) and mean photo effect (MPE) values for BP-1, BP-3 and BP-4 generated by phototox software in Balb/c 3T3, HS68 and HaCaT cells. Cell type
3T3
Substance
BP-1
HS68 HaCaT 3T3
BP-3
HS68 HaCaT 3T3 HS68 HaCaT
a b
BP-4
Run
1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2
UVA
UVB
PIF mean
MPE mean
Phototoxic potential
PIF mean
MPE mean
Phototoxic potential
5.977 5.899 3.501a 5.687 3.192a 4.910a 1.000 1.000 1.000 1.000 1.051 1.000 1.129 1.207 1.000 1.000 1.000 1.000
0.313 0.309 0.177b 0.198 0.165b 0.263b 0.130 0.067 0.122 0.123 0.198b 0.136 0.045 0.069 0.011 0.024 0.103 0.085
Yes Yes Yes Yes Yes Yes No No No No
1.396 1.000 0.818 0.847 1.045 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000
0.087 0.112 −0.100 −0.119 −1.250 −0.550 0.033 0.088 0.049 0.091 0.082 0.103 −0.097 −0.296 −0.007 −0.087 −1.642 −1.427
No No No No No No No No NO No No No No No No No No No
No No No No No No No
Probable phototoxic. Phototoxic (according to OECD, 2004). 184
Toxicology in Vitro 60 (2019) 180–186
L. Xiong, et al.
Table 3 The HOMO and LUMO orbital and EGAP of BP-1, BP-3 and BP-4. Substance name
Structural formula
HOMO
LUMO
EGAP (eV)
EGAP (J/mol)
BP-1
2.891
278.937
BP-3
2.832
273.244
BP-4
0.014
1.351
et al.) after electron transfer (type I reaction) or singlet oxygen (1O2) following energy transfer (type II reaction). The organic UV filters are also called ultraviolet absorbent substances. With their potential of UV absorbance, BPs offer sun protection, but it may lead to phototoxicity. BPs contain chromophore which on absorption triggers photophysical or photochemical phenomena leading to a direct or indirect damage of biological macromolecules (Bignon et al., 2017). The protective mechanism of BPs against harmful effects of ultraviolet light functions by chemically absorbing light energy (photons). As this occurs, the benzophenone molecule becomes activated to higher energy levels. As the excited molecule returns to its ground state, the energy is released in the form of thermal energy. Photosensitization cascade may involve energy or electron-transfer, or activation of molecular oxygen via production of the reactive singlet oxygen (1O2) that ultimately produces oxidative reactions (Bignon et al., 2017). Then, the historical parabolic relationships between phototoxicity and the energy gap (EGAP) between energies of the highest occupied molecular orbital and the lowest unoccupied molecular orbital was confirmed (Parr and Zhou, 1993; Ringeissen et al., 2011). We set up the mechanistic model using DMol3 program package of Materials Studio (Delley, 2000; Perdew et al., 1996), and showed that the calculating results of the model were similar with experiment results. The EGAP of BP-1 was higher than BP-3 and much higher than BP-4. Note that we could speculate that the phototoxicity of BP-1 provide a close representation to BP-3, but in vitro results suggested that BP-1 were tested for their phototoxicity in comparison to BP-3. That may be another limitation with the cell test. In vitro test is only employed for hazard identification while its utility for the assessment of phototoxic potency is not warranted. In particular, this assay system lacks metabolic activity which is critical in the manifestation of systemically exposed chemicals. It has been reported that BP-3 is metabolized principally to BP-1, via metabolic activities of cytochrome P450 (CYP) enzymes (Kunisue et al., 2012; Jeon et al., 2008; Kasichayanula et al., 2005; Nakagawa and Suzuki, 2002; Okereke et al., 1994). There were also reports that intramolecular hydrogen bond formation (e.g. BP-3), reversible photoisomerization (e.g. ecamsule, octyl methoxycinnamate) and conjugated bonds were present in almost all UV filters (Kockler et al., 2012; Shaath, 2010). However, the phototoxicity of BPs is far less
4. Discussion In the present study, we summarized the phototoxicity results obtained from the modified OECD No. 432 assay (OECD, 2004), which showed some a few divergent scenarios that could be encountered. The original SOP developed for the 3T3 NRU assay had to be adapted to the UVA irradiation with a dose of 5 J/ cm2.Our assay systems were found to be applicable for the prediction of UVB phototoxicity, and CPZ could be selected as a positive control because the PIF and MPE were calculated to test and verify in the three cell lines, consistent with the previously reported data (Ray et al., 2008; Murli et al., 2010). Organic sunscreens often protect us by absorbing environmental UVA and UVB radiation, so we need an integrated testing strategy to predict phototoxicity involving both UVA and UVB. Moreover, 3T3 NRU test has many limitations, for instance, the test exhibits highly sensitive but low specific predictive capacity (Kim et al., 2015). In this study, we found that irradiated HS68 and HaCaT were less susceptible than 3T3 when exposed to UV radiation. As provisional prediction model, we could find it similar to light injury, UVB could cause superficial injury such as in stratum epidermis and epidermal cells, but UVA can penetrate deep into dermis of the skin eventually causing premature photo-aging by stimulating the over-expression of MMPs (matrix Metalloproteinases) by fibroblasts (Kang et al., 2003). Therefore, we suggested to use UVA radiation in fibroblasts to and UVB radiation in keratinocytes for future studies, making the model more close to reality. Recently, concerns about chemical phototoxicity are on the rise, and most UV filters are widely used in sunscreens generate ROS when skin is exposed to UV radiation. (Biniek et al., 2012). We tried to develop a new mechanistic model to predict photo-toxicity. Photo-induced toxicity by PAH was first modeled in 1987 (Newsted and Giesy, 1987). They investigated many of the structural factors of phototoxicity of the PAH, among which were the lowest singlet excitation energy of the excited molecule, the lowest triplet excitation energy of the excited molecule, and the phosphorescence lifetime (Newsted and Giesy, 1987). As mentioned above, the most commonly reported phototoxic process is oxidative reactions. The chromophore in its excited state transfers energy or electrons to oxygen and generates ROS (O2%-, NO%, OH%, ROO%
185
Toxicology in Vitro 60 (2019) 180–186
L. Xiong, et al.
established and should be properly characterized and understood through one or multiple methods. In conclusion, we rationalized the UV filtering efficiency of BPs by performing cell viability assays over different cell lines exposed to UVB and UVA light. Indeed, the highly phototoxic substances (such as CPZ, Angelicin, 8-methoxypsoralen) are found in many cell types selectively affect. It's hard to determine a very weak phototoxic potential in human skin as of BPs in vitro assays. In this study, BP-1 appeared to be phototoxic under UVA+ condition. However, BP-3 and BP-4 were found to be non-phototoxic under UVA+ condition. It verified that BP-1, BP-3 and BP-4 remained low-risk chemicals in UVB+ condition. The theoretical calculation showed BP-1(EGAP) > BP-3(EGAP) > BP-4(EGAP). The methods established can be used for testing phototoxicity of chemical components. For systemically exposed chemicals that require metabolic activation more integrated testing strategies are recommended, and this study could provide a better choice to cosmetic formulators for selecting UV filter in the potentially effective products. Supported by these conclusions, we recommend BP-4 as an effective and safer UV protective compound in sunblock formulations.
Phys. 113, 7756–7764. Du, Y., Wang, W.Q., Pei, Z.T., Ahmad, F., Xu, R.R., Zhang, Y.M., Zhang, Y.M., Sun, L.W., 2017. Acute toxicity and ecological risk assessment of Benzophenone-3 (BP-3) and Benzophenone-4 (BP-4) in ultraviolet (UV)-filters. Inter. J. Env. Res. Pub. Heal. 14, 1414. Gülden, M., Mörchel, S., Seibert, H., 2005. Comparison of mammalian and fish cell line cytotoxicity: impact of endpoint and exposure duration. Aquat. Toxicol. 71, 229–236. Herrling, T., Jung, K., Fuchs, J., 2006. Measurements of UV-generated free radicals/reactive oxygen species (ROS) in skin. Spectrochim. Acta A 63, 840–845. Ishiyama, M., Miyazono, Y., Sasamoto, K., Ohkura, Y., Ueno, K., 1997. A highly watersoluble disulfonated tetrazolium salt as a chromogenic indicator for NADH as well as cell viability. Talanta 44, 1299–1305. Jeon, H., Sarma, S.N., Kim, Y., Ryu, J., 2008. Toxicokinetics and metabolisms of benzophenone-type UV filters in rats. Toxicol 248, 89–95. Kang, S., Chung, J.H., Lee, J.H., Fisher, G.J., Wan, Y.S., Duell, E.A., Voorhees, J.J., 2003. Topical N-acetyl cysteine and genistein prevent ultraviolet-light-induced signaling that leads to photoaging in human skin in vivo. J. Invest. Dermatol. 120, 835. Kasichayanula, S., House, J.D., Wang, T., Gu, X., 2005. Simultaneous analysis of insect repellent DEET, sunscreen oxybenzone and five relevant metabolites by reversedphase HPLC with UV detection: application to an in vivo study in a piglet model. J. Chromatogr. B 822, 271–277. Kim, K., Park, H., Lim, K.M., 2015. Phototoxicity: its mechanism and animal alternative test methods. Toxicol. Res. 31, 97–104. Kochevar, I.E., 1981. Phototoxicity mechanisms: chlorpromazine photosensitized damage to DNA and cell membranes. J. Invest. Dermatol. 77, 59–64. Kockler, J., Oelgemöller, M., Robertson, S., Glass, B.D., 2012. Photostability of sunscreens. J Photochem Photobiol C: Photochem Rev 13, 91–110. Kunisue, T., Chen, Z., Louis, G.M.B., et al., 2012. Urinary concentrations of benzophenone-type UV filters in U.S. women and their association with endometriosis. Environ. Sci. Technol. 46, 4624–4632. Lelièvre, D., Justine, P., Christiaens, F., Bonaventure, N., Coutet, J., Marrot, L., TinoisTessonneaud, E., Cotovio, J., 2007. The episkin phototoxicity assay (epa): development of an in vitro tiered strategy using 17 reference chemicals to predict phototoxic potency. Toxicol. in Vitro 21, 977–995. Lou, J., Zhou, G., Chu, G., Jiang, J., Huang, F., Zheng, S., Lu, Y., Li, X., He, J., 2010. Studying the cyto-genotoxic effects of 12 cigarette smoke condensates on human lymphoblastoid cell line in vitro. Mutat. Res. 696, 48–54. Murli, H., Aardema, M., Lawlor, T., Spicer, C., 2010. Photoclastogenicity?An improved protocol, its validation, and investigation of the photogenotoxicity of DMBA. Environ. Mol. Mutagen. 40, 41–49. Nakagawa, Y., Suzuki, T., 2002. Metabolism of 2-hydroxy-4methox- ybenzophenone in isolated rat hepatocytes and xenoestrogenic effects of its metabolites on MCF-7 human breast cancer cells. Chem. Biol. Interact. 139, 115–128. Newsted, J.L., Giesy, J.P., 1987. Predictive models for photoinduced acute toxicity of polycyclic aromatic hydrocarbons to daphnia magna, Strauss, Cladocera, Crustacea. Environ. Toxicol. Chem. 6, 445–461. OECD, 2004. Test No. 432: In Vitro 3T3 NRU Phototoxicity Test. Publishing, OECD. OECD (Q)SARs projects, 2010. Available at: http://www.oecd.org/document/23/0, 3343,en_2649_34379_33957015_1_1_1_1,00.html, Accessed date: 14 June 2010. Okereke, C.S., Abdel-Rhaman, M.S., Friedman, M.A., 1994. Disposition of benzophenone3 after dermal administration in male rats. Toxicol. Lett. 73, 113–122. Onoue, S., Seto, Y., Sato, H., Nishida, H., Hirota, M., Ashikaga, T., Api, A.M., Basketter, D.A., Tokura, Y., 2017. Chemical photoallergy: photobiochemical mechanisms, classification, and risk assessments. J. Dermatol. Sci. 85, 4–11. Parr, R., Zhou, Z., 1993. Absolute hardness: unifying concept for identifying shells and subshells in nuclei, atoms, molecules, and metallic clusters. Acc. Chem. Res. 26, 256–258. Perdew, J.P., Burke, K., Ernzerhof, M., 1996. Generalized gradient approximation made simple. Phys. Rev. Lett. 77, 3865–3868. Rai, V., Dayan, N., Michniakkohn, B., 2011. A comparative evaluation of photo-toxic effect of fractionated melanin and chlorpromazine hydrochloride on human (dermal fibroblasts and epidermal keratinocytes) and mouse cell line/s (fibroblast balb/c 3t3). Toxicol. in Vitro 25, 538–544. Ray, R.S., Agrawal, N., Sharma, A., et al., 2008. Use of L-929 cell line for phototoxicity assessment[J]. Toxicol. in Vitro 2008, 1775–1781. Ringeissen, S., Marrot, L., Labarussiat, A., Imbert, S., Todorov, M., Mekenyan, O., Meunier, J., 2011. Development of a mechanistic SAR model for the detection of phototoxic chemicals and use in an integrated testing strategy. Toxicol. in Vitro 25, 324–334. Saroj Kumar, A., Goyal, S., Dubey, D., Srivastav, A.K., Chopra, D., Singh, J., Shankar, J., Chaturvedi, R.K., Ray, R.S., 2015. Benzophenone 1 induced photogenotoxicity and apoptosis via release of cytochrome c and Smac/DIABLO at environmental UV radiation. Toxicol. Lett. 239, 182–193. Shaath, N.A., 2010. Ultraviolet filters. Photochem. Photobiol. Sci. 9, 464–469. Tanneberger, K., Rico-Rico, A., Kramer, N.I., Busser, F.J.M., Hermens, J.L.M., Schirmer, K., 2010. Effects of solvents and dosing procedure on chemical toxicity in cell- based in vitro assays. Environ. Sci. Technol. 44, 4775–4781. Wondrak, G.T., Jacobson, M.K., Jacobson, E.L., 2006. Endogenous UVA-photosensitizers: mediators of skin photodamage and novel targets for skin photoprotection. Photochem. Photobiol. Sci. 5, 215–237.
Declaration of competing interests There are no conflicts of interest. Acknowledgements We gratefully acknowledge the financial support provided by Sichuan Science and Technology Program (No. 2017SZ0070 and No. 2018SZ0104), China National Natural Science Foundation(No. 81673084) and 1.3.5 project for disciplines of excellence, West China Hospital, Sichuan University. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.tiv.2019.05.021. References Benevenuto, C.G., Guerra, L.O., Gaspar, L.R., 2015. Combination of Retinyl palmitate and UV-filters: phototoxic risk assessment based on Photostability and in vitro and in vivo Phototoxicity assays. Eur. J. Pharm. S.C. 68, 127–136. Betowski, L.D., Enlow, M., Riddick, L., 2002. The phototoxicity of polycyclic aromatic hydrocarbons: a theoretical study of excited states and correlation to experiment. Comput. Chem. 26, 371–377. Bignon, E., Marazzi, M., Besancenot, V., Gattuso, H., Drouot, G., Morell, C., Eriksson, L.A., Grandemange, S., Dumont, E., Monari, A., 2017. Ibuprofen and ketoprofen potentiate UVA-induced cell death by a photosensitization process. Sci. Rep. 7 (8885). Biniek, K., Levi, K., Dauskardt, R.H., 2012. Solar UV radiation reduces the barrier function of human skin. P. Nat. Acad. Sci. 109, 17111–17116. Bowden, G.T., 2004. Prevention of non-melanoma skin cancer by targeting ultraviolet-Blight signalling. Nat. Rev. Cancer 4, 23–35. Broeders, J.J., Blaauboer, B.J., Hermens, J.L., 2013. In vitro biokinetics of chlorpromazine and the influence of different dose metrics on effect concentrations for cytotoxicity in balb/c 3t3, caco-2 and heparg cell cultures. Toxicol. in Vitro 27, 1057–1064. Carbonare, M.D., Pathak, M.A., 1992. Skin photosensitizing agents and the role of reactive oxygen species in photoaging. J. Photochem. Photobiol. B 14, 105–124. Crovara Pescia, A., Astolfi, P., Puglia, C., Bonina, F., Perrotta, R., Herzog, B., Damiani, E., 2012. On the assessment of photostability of sunscreens exposed to UVA irradiation: from glass plates to pig/human skin, which is best? Int. J. Pharm. 427, 217–223. de Oliveira, C.A., Peres, D.D., Rugno, C.M., Kojima, M., de Pinto, C.A.S.O., Consiglieri, V.O., Kaneko, T.M., Rosado, C., Mota, J., Velasco, M.V.R., Baby, A.R., 2015. Functional photostability and cutaneous compatibility of bioactive UVA sun care products. J. Photochem. Photobiol. B Biol. 148, 154–159. Delley, B., 1990. An all-electron numerical method for solving the local density functional for polyatomic molecules. J. Chem. Phys. 92, 508–517. Delley, B., 2000. From molecules to solids with the DMol [sup 3] approach. J. Chem.
186
Contents lists available at ScienceDirect
Toxicology in Vitro journal homepage: www.elsevier.com/locate/toxinvit
Phototoxic risk assessment on benzophenone UV filters: In vitro assessment and a theoretical model Lidan Xionga,b, Jie Tanga,b, Yiming Lic, Li Lia,b,c,
T
⁎
a
Cosmetics Safety and Efficacy Evaluation Center, West China Hospital, Sichuan University, No. 5, Gong Xing Road, Chengdu, Sichuan, PR China Sichuan Engineering Technology Research Center of Cosmetic, Chengdu, Sichuan, PR China c Department of Dermatology, West China Hospital, Sichuan University, No. 37, Guo Xue Xiang, Chengdu, Sichuan, PR China b
A R T I C LE I N FO
A B S T R A C T
Keywords: Benzophenone UV filters Phototoxic In vitro assessment A theoretical model
Benzophenones (BPs), filtering out both UVA and UVB rays, are widely used in a great variety of sunscreens and personal care products. However, they have not been extensively studied for the mechanisms of UV-absorbing toxicity. In this study, we used CPZ (chlorpromazine) as a positive control and SDS (sodium dodecyl sulfate) as a negative control, and the phototoxic of BP-1, BP-3 and BP-4 were investigated in vitro assays using three cell types under different UV exposure conditions. This was followed by setting up a theoretical model, which was adopted to predict and compare the phototoxicity. It was found that Balb/c 3T3 (Balb/c 3T3 fibroblast cell lines) showed sensitivity to UVA+ and UVB+ exposure, while the HS68 (human HS68 fibroblast cell lines) to UVA+ and the HaCaT (human HaCaT keratinocyte cell lines) to UVB+. The test compound, BP-1, was detected to be phototoxic at UVA+ conditions, but BP-3 and BP-4 were discovered to be non-phototoxic at UVA+ conditions. This demonstrated that BP-1, BP-3 and BP-4 remained low-risk chemicals under UVB+ condition. The theoretical calculation of the energy gap (EGAP) showed BP-1(EGAP) > BP-3(EGAP) > BP-4(EGAP).
1. Introduction UV radiation, including UVB and UVA spectra, can apparently penetrate human skin tissues to a considerable depth, which promotes the development of melanoma and other skin diseases (Bowden, 2004). Phototoxicity is of increased concern in dermatology, since people in modern society pay more attention to sun radiation. The definition of phototoxicology is given, as well as an illustration of the primary mechanisms for different phototoxic endpoints including energy transfer, electron transfer and formation of toxic photoproducts (Wondrak et al., 2006; Carbonare and Pathak, 1992). Generally, phototoxicity is increasingly assessed by the 3T3 NRU (neutral red uptake) in vitro phototoxicity test as described in the OECD Guideline No. 432 (OECD, 2004). Besides, different cell lines are needed for studying discrepancies in vitro phototoxicity experiments (Gülden et al., 2005), including HEKn, HDF fibroblasts (Rai et al.,
2011), Caco-2 (Broeders et al., 2013), HepaRG (Broeders et al., 2013) and artificial skin constructs (morphological changes, immunochemistry) (Lelièvre et al., 2007). In addition, the use of different set-up of the cell assay should be considered, including differences in exposure conditions such as cell concentration, solvents and end-points (Tanneberger et al., 2010). At present, there are several assays to evaluate of the toxic end-points in vitro. Previous studies have indicated that there is a high correlation between CCK-8(Cell Counting Kit-8), MTT(3-(4,5)-dimethylthiahiazo(-z-yl)-3,5-di-phenytetrazoliuromide) and NRU (Neutral Red Uptake) assays (Ishiyama et al., 1997; Lou et al., 2010). Compared with other two, CCK-8 assay is more rapid and sensitive, yielding accurate method for cytotoxicity detection. Therefore, the procedure is more suitable for this study. This study employs a minor deviation from the OECD No. 432 guideline using CCK-8 assay as a substitute for NRU. The discrepancies of UV spectral within the different cell types are
Abbreviations: CPZ, chlorpromazine hydrochloride; SDS, sodium dodecyl sulfate; BP-1, 2,4-dihydroxybenzophenone; BP-3, oxybenzone,2-hydroxy-4-methoxybenzophenone; BP-4, 5-benzoyl-4-hydroxy-2-methoxybenzenesulfonic acid; Balb/c 3T3, mouse embryonic fibroblast cell line; HS68, neonatal human foreskin fibroblast lines; HaCaT, neonatal human keratinocytes; PIF, photo irritancy factor; MPE, mean phototoxic effect; CCK-8, Cell Counting Kit-8; ROS, reactive oxygen species; QSAR, quantitative structure-activity relationships; ICH, International council for harmonization of technical requirements for pharmaceuticals for human use; OECD, organization for economic cooperation and development; HOMO, highest occupied molecular orbital; LUMO, lowest unoccupied molecular orbital ⁎ Corresponding author at: Cosmetics Safety and Efficacy Evaluation Center, West China Hospital, Sichuan University, No. 5, Gong Xing Road, Chengdu, Sichuan, PR China. E-mail address: [email protected] (L. Li). https://doi.org/10.1016/j.tiv.2019.05.021 Received 9 December 2018; Received in revised form 27 May 2019; Accepted 29 May 2019 Available online 30 May 2019 0887-2333/ © 2019 Published by Elsevier Ltd.
Toxicology in Vitro 60 (2019) 180–186
L. Xiong, et al.
Fig. 1. Chemical structures of Benzophenone UV filters, CPZ (positive) and SDS (negative), with the range of absorption.
pathogenetic mechanisms of drug-induced phototoxicity. (Ringeissen et al., 2011).Available tools include in silico prediction systems such as DEREK allows toxicity prediction of chemicals based on structures known to be associated with the incidence of toxicity(Onoue et al., 2017).The OECD (Q)SAR is another initiative model which promotes the use of non-testing information for many diverse toxicological endpoints through the use of existing knowledge, categories based on mechanistic rationale and/or structural similarity, read-across and mathematical models based on structure activity relationships (OECD (Q)SARs projects, 2010; Ringeissen et al., 2011).Besides, HOMO-LUMO gap evaluation provides a measure of the photoreactive potential of chemicals (Betowski et al., 2002; Onoue et al., 2017). The internal effects conditioning the light absorbance and stability of chemicals are found to be best described by the energy gap (EGAP) between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) (Parr and Zhou, 1993), and EGAP evaluation provides a measure of the photoreactive potential of chemicals (Ringeissen et al., 2011). Although these are not completely specific for chemical phototoxicity in silico and photochemical prediction tools, the prediction methods, perhaps deployed in a high throughput system, may be appropriate for initial screening of chemicals to predict risk of phototoxicity. Given the above facts, this study was to examine the phototoxic potential of the BP-1, BP-3, and BP-4 using the Balb/c 3T3 cells and CPZ as positive control, SDS as negative control following the modified guidelines from OECD No. 432 (OECD, 2004). The studies were further extended from Balb/c 3T3 to human skin cells HS68 and HaCaT to compare the inter-cellular differences and sensitivity of normal human dermal and epidermal cells for phototoxicity analysis. This may allow us to understand the correlation of toxicity levels of compounds in mouse cell lines and human cell lines. Next, the study was further extended from UVA to UVB in order to observe the response of the tested cells against different UV spectra. Subsequently, this study was to examine the toxic/phototoxic potential of a set of common BP-type UVfilters. It is essential that such sunscreen products are safe, effective, and useful. The present results would provide useful information for evaluating the potential skin health risk of BPs. Finally, this study used
usually ignored. However, UVA and UVB affect different layers of the skin, and induce damage to cells by different mechanisms. Epidermis and dermis cells constantly exposed to high energetic UVB and less energetic but more penetrating UVA radiation react in different ways (Herrling et al., 2006). Compared with UVA filters, UVB filters are readily available, while most phototoxicities were regarded as caused and tested by UVA previously (OECD, 2004). One aim of our study is to verify which kinds of skin cells (fibroblasts or keratinocytes) were more susceptible to UVA/UVB irradiation. Sunscreens contain organic and inorganic UV filters as active ingredients. The organic UV filters absorb UV radiation within a specific range of wavelengths depending on their chemical structure, and usually possess aromatic structures that can absorb the solar UV radiation (Crovara Pescia et al., 2012; de Oliveira et al., 2015). Benzophenones (BPs) are among well-known organic UV filters due to theirs broad absorption range in the UVA and UVB spectra (Benevenuto et al., 2015). BPs contain an aromatic ketone with the ability to absorb UVB and UVA, whose absorption is influenced by hydrogen bonding (Saroj Kumar et al., 2015) showed in Fig. 1. However, BPs may penetrate through stratum corneum into deeper skin layers for their small size, which may induce contact dermatitis and skin sensitization by enhancing the ROS generation and electron delocalization in presence of UV radiation (Saroj Kumar et al., 2015). However, there is no study reported in the literature which compares phototoxicity of BP-1, BP-3 and BP-4 in skin cells. In addition, CPZ (Chlorpromazine hydrochloride) has been shown to cause photosensitivity induced damage to membranes, protein and DNA by both direct and indirect mechanisms (Kochevar, 1981; Du et al., 2017). This phototoxic potential is exhibited both in vitro and in vivo conditions. Therefore, this parameter of phototoxicity is selected as the positive in our study. SDS (Sodium dodecyl sulfate), proven in experimental and clinical trials, is adopted as the negative control for this study (OECD, 2004). Today, there is an increasing need for filtering compounds to be tested in the cells, however, this requires since testing requires resources both in terms of cost and time. Various methodologies, including in silico prediction models, have been developed to predict the phototoxic potential of chemicals over the past few years, based on the 181
Toxicology in Vitro 60 (2019) 180–186
L. Xiong, et al.
Subsequently, the plates were washed with PBS (150 μl/well) and the cells were exposed to eight different concentrations of test solutions/ controls prepared in PBS of each test material (100 μl/well) for 1 h. One plate was then exposed to UVA (UVA+) at 5 J/ cm2, or UVB(UVB+) at 30 mJ/cm2, while the other plates were placed in the dark environment and incubated for observing cytotoxicity caused solely by BPs. The UV irradiation conditions were determined by the instrumental programmed software (Sigma, Shanghai, China). Next the chemical solutions were decanted and washed with PBS followed by addition of cell media (100 μl/well). After 18–22 h of incubation, the cells were washed with PBS, and DMEM (100 μl/well) containing CCK-8 (10% v/v) was added to each well following incubation time of 3 h. Absorbance was measured at 450 nm using a microplate spectrophotometer (Bio-Red, Model 680, USA).
theoretical calculation, which was conducted using structure activity relationship and molecular docking methods. Our study allowed for a better understanding of the mechanisms occurring in the human skin, and may provide a better choice to cosmetic formulators for selecting UV filter in the effective products. 2. Material and methods 2.1. Chemicals and instrumentation BP-1 (Benzophenone 1, 2,4-Dihydroxybenzophenone, with a purity 99%, CAS NO. 131–56-6), BP-3 (Benzophenone 3, Oxybenzone,2Hydroxy-4-methoxybenzophenone, with a purity 98%, CAS NO. 13157-7), BP-4 (Benzophenone 4, 5-Benzoyl-4-hydroxy-2-methoxybenzenesulfonic acid, with a purity ≥97.0%, CAS NO.4065-45-6), SDS (Sodium dodecyl sulfate, with a purity ≥99.0%, CAS NO.151-21-3) and CPZ (Chlorpromazine hydrochloride, CAS NO.69-09-0) were obtained from Sigma Chemical Co. (SIGMA-ALDRICH, USA). Fetal bovine serum (FBS) was procured from Gibco (Thermo Fisher, USA). CCK-8 (Cell Counting Kit-8) was obtained from Dojindo (Dojindo Laboratories, Japan). PBS (phosphate buffers 10X) was purchased from Zsbio Commerce CO. (Zsbio, China). DMEM (Dulbecco's Minimum Essential Medium), penicillin and streptomycin solution, trypsin (0.25%), and DMSO (dimethylsulfoxide) were acquired from HyClone (GE Health care Life Science, USA). UVA Bio-Sun instrument (320-340 nm, SS-04A, Sigma, Shanghai, China), UVB irradiation instrument (290–315 nm, SS04B, Sigma, Shanghai, China), BIO RED Model 680 Microplate Reader (BIO RED Inc., USA).
2.4.2. HS68 phototoxicity assays For the cytotoxicity and phototoxicity assays, HS68 cells were cultured in 96-well plates (Corning, USA) at a seeding density of 8 × 103 cells/well. The cells were cultured in the plate for 24 h, and the following steps were the same as 2.4.1. 2.4.3. HaCaT phototoxicity assays HaCaT cells were transferred to 96-well plates at a seeding density of 104 cells/well, and were cultured for 24 h at 37 °C in 5% CO2.The following steps were the same as 2.4.1. 2.4.4. Statistical analysis The Phototox Version 2.0 software was employed for concentration–-response analysis by PIF and MPE. The Phototox Version 2.0 software, available by OECD, was used to calculate Photo-IrritationFactor (PIF) and Mean Photo Effect (MPE). Each detected material was tested for three times in dependent assays.
2.2. Cell culture The murine Balb/c 3T3 fibroblast cell lines were purchased from Kunming Cell Bank of Chinese Academy of Sciences. These cells were cultured in supplemented DMEM (with 10% FBS, 100 U/ml penicillin and 100 μg/ml streptomycin). The human HS68 fibroblast cell lines were obtained from American Type Culture Collection (ATCC, CRL1635, USA) and were cultured in DMEM (with 10% FBS, 100 U/ml penicillin and 100 μg/ml streptomycin). The human HaCaT keratinocyte cell lines were ordered from Kunming Cell Bank of Chinese Academy of Sciences and cultured in DMEM (with 10% FBS, 10 ng/ ml HKGS (human keratinocyte growth factor), 100 U/ml penicillin and 100 μg/ml streptomycin).
PIF =
IC 50(UV −) IC 50(UV +)
PIF < 2 (NO phototoxic); 2 < PIF < 5 (PROBABLE Phototoxic); PIF > 5 (Phototoxic). When IC50 cannot be obtained, MPE is calculated as following equation n
MPE =
∑i = 1 wi PEci n
∑i = 1 wi
PIF < 2 or an MPE < 0.1 (NO phototoxic); 2 < PIF < 5 or 0.1 < MPE < 0.15 (PROBABLE Phototoxic); PIF > 5 or MPE > 0.15 (Phototoxic) (OECD, 2004). Results were obtained after calculating the mean of at least two independent experiments.
2.3. Preparation of test materials Benzophenones (BPs) have limited solubility in water, hence Dimethylsulphoxide (DMSO) was selected as the solvent based on OECD, 2004 recommendations. Stock solutions of BP-1, BP-3, BP-4, SDS and CPZ were prepared at 1 M using DMSO. For the assay, the chemicals were diluted in PBS. Concentrations of 0.1–500 μM in PBS were achieved which is the maximum concentration of DMSO was safe for the cells. The final solutions with concentrations of 500, 100, 50, 10, 5,1, 0.5 and 0.1 μM were prepared in PBS. All test materials were freshly prepared prior to use.
2.5. Prediction of phototoxicity All DFT calculation were performed with DMol3 program package of Materials Studio (Delley, 1990, 2000; Perdew et al., 1996). PerdewBurke-Ernzerh (PBE) of correlation function of generalized gradient approximation (GGA) level was conducted. A DFT semi-core pseudopotentials (DSPP) core treatment was chosen, which has been used in many DFT studies to optimize structures. The localized double numerical polarization (DNP) basis set was applied to expand the KohnSham orbitals, which included a polarization p-function on all hydrogen atoms. Self-consistent field (SCF) calculations were carried out with a tolerance of 1*10−5 a.u. on the total energy. The real-space global orbital cutoff was chosen to be 4.4 Å in the computation. The value of smearing was 0.005 Hartree. COSMO solvent calculation with the DCPBE functional was also performed using the dielectric constant of DMSO (46.7). The energy gap (EGAP) is defined as followed: EGAP = E (LUMO)-E (HOMO). Therefore, the more positive the value is, the higher the EGAP. The electronic parameters in a molecule are strongly affected by their conformational flexibility. Three molecules were used
2.4. Test procedure 2.4.1. 3T3 phototoxicity assays 3 T3 cells culture was carried out as described by the OECD 432 (OECD, 2004) adjusted for the actual conditions. In brief, 3T3 cells were plated on 96-well plates (Corning, USA) at a seeding density of 104 cells/well. Initially, a cell suspension of 105 cells/ml was prepared in the culture medium and 100 μl cell suspension was seeded exclusively into the central 60 wells of the plate. In the peripheral wells of the 96well plate, 100 μl of the culture medium alone (without cells) was added (blank). The cells were cultured in the plate for 24 h. 182
Toxicology in Vitro 60 (2019) 180–186
L. Xiong, et al.
Fig. 2. Effects of the three cell systems under UVA and UVB irradiation. (A) human skin cells lines (HS68 and HaCaT) and mouse cell line Balb/c 3T3 under UVA (5 J/ cm2) and UVB (30 mJ/cm2) irradiation, (B) Cell viability was evaluated in 3T3, HS68 and HaCaT cells with UVA and UVB. Mean and SD are shown for three independent experiments. *p < 0.05; **p < 0.01.
under exposure and non-exposure (dark conditions) of UVA/UVB light (Fig. 2B). When 3T3 cells were exposed to UV at 5 J/cm2 (UVA) and 30 mJ/cm2 (UVB), 3T3 cell viability significantly declined to 56.25% (UVA+) and 61.39% (UVB+) respectively compared with non- exposure control. While the relative survival rate of HS68 cells decreased to 68.22% (UVA+) and 81.20% (UVB+), HaCaT cell viability decreased to 82.50% (UVA+) and 67.00% (UVB+).
to explore the bell-shape relationship between phototoxicity and EGAP: the most reactive (i.e., those with maximum EGAP). 3. Results 3.1. Measurement of the three cell lines exposing to UVA and UVB irradiation We first screened whether UVA and UVB irradiation affected the cell proliferation on human skin cells lines (HS68 and HaCaT) and mouse cell line Balb/c 3T3 (Fig. 2). Cell images were taken in presence and absence of UV radiation and these images were provided qualitative visual information on the overall effect of UV radiation on the cells (Fig. 2A). CCK-8 assay was performed to investigate cell viability, both
3.2. Evaluating phototoxicity of CPZ and SDS in the three cell systems under UVA and UVB exposure CPZ and SDS, as the reference chemicals for positive and negative controls, were tested in vitro for phototoxicity in human skin cell lines (HS68 and HaCaT) and mouse cell line Balb/c 3 T3 (Table 1). Briefly, 183
Toxicology in Vitro 60 (2019) 180–186
L. Xiong, et al.
Table 1 Photo irritation factor (PIF) and mean photo effect (MPE) values for CPZ and SDS generated by phototox software in Balb/c 3T3, HS68 and HaCaT cells. Cell type
3T3 3T3 3T3 3T3 HS68 HS68 HS68 HS68 HaCaT HaCaT HaCaT HaCaT
Substance
CPZ SDS CPZ SDS CPZ SDS
Run
1 2 1 2 1 2 1 2 1 2 1 2
UVA
UVB
PIF mean
MPE mean
Phototoxic potential
PIF mean
MPE mean
Phototoxic potential
76.592 79.943 1.031 0.982 52.602 43.838 0.896 0.850 24.267 46.209 1.996 1.456
0.584 0.601 0.002 0.035 0.435 0.419 0.003 0.053 0.205 0.387 0.086 0.023
Yes Yes No No Yes Yes No No Yes Yes No No
16.524 18.707 0.217 0.188 9.657 13.508 1.547 0.734 18.621 22.702 1.000 0.611
0.227 0.235 0.026 0.022 0.198 0.377 0.008 0.022 0.248 0.284 0.024 0.014
Yes Yes No No Yes Yes No No Yes Yes No No
under UVB were calculated to be < 2, MPE < 0.15, respectively. BP-3 and BP-4 were showed to be non-phototoxic in UVA+ conditions, because PIF and MPE values were calculated to be < 2, MPE < 0.15, respectively. Thus, BP-1 was identified as phototoxic (UVA) molecules, while BP-3, and BP-4 were determined as non-phototoxic (UVA) ones. Besides, BP-1, BP-3 and BP-4 remained low-risk chemicals in UVB+ condition.
photo Irritation Factor (PIF) and Mean Photo Effect (MPE) values for CPZ generated by Phototox Software in 3T3 cells are shown in Fig. S1 which present the sources of data (Fig. S1a and b). The results of CPZ and SDS coincided with the range recommended by OECD No.432 (OECD, 2004). In Table 1, the results obtained were summarized. The result of PIF Mean and MPE for SDS in 3T3 cells both under UVA and UVB were all negative, which served as a negative control in the following tests. SDS was also found to be completely non-phototoxic in HS68 and HaCaT cells. The result of PIF and MPE for CPZ in 3T3, HS68 and HaCaT cells were found to exhibit the phototoxicity potential of CPZ in presence of UVA and UVB.
3.4. Prediction of phototoxicity In order to further illustrate the potential mechanism of the phototoxicity associated with these three substances, we mainly used the theoretical calculations. According to the report (Ringeissen et al., 2011), we know that the higher EGAP, the more reactive the substance is. The HOMO and LUMO orbitals (Table 3) were used to calculate the EGAP value. From the results, the EGAP of BP-1 was higher than BP-3 and much higher than BP-4, and we could speculate that the phototoxicity of BP-1 was stronger than that of BP-3 and BP-4.Note that we could speculate that the phototoxicity of BP-1 provide a close representation to BP-3, while the in vitro results suggested that BP-1 was tested for their phototoxicity in comparison to BP-3. That may be another limitation of the cell test.
3.3. Measurement of exposuring BP-1, BP-3 and BP-4 in the three cell systems under UVA and UVB irradiation The cell viability curves of BPs with or without UVA/UVB irradiation were obtained, and the PIF and MPE values were calculated (Table 2). Compared with the non-irradiated groups, UVA-irradiated BP-1 exhibited significant cytotoxicity at lower doses. The PIF values of BP-1 were estimated to be > 2.0 and the MPE values of BP-1 were higher than 0.1 in 3 T3, HS68 and HaCaT cells, which identified as phototoxic or probable phototoxic molecules. However, compared but compared with positive control CPZ, PIF and MPE values of BP-1 were lower and more uncertain. In contrast, the PIF and MPE values of BP-1
Table 2 Photo irritation factor (PIF) and mean photo effect (MPE) values for BP-1, BP-3 and BP-4 generated by phototox software in Balb/c 3T3, HS68 and HaCaT cells. Cell type
3T3
Substance
BP-1
HS68 HaCaT 3T3
BP-3
HS68 HaCaT 3T3 HS68 HaCaT
a b
BP-4
Run
1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2
UVA
UVB
PIF mean
MPE mean
Phototoxic potential
PIF mean
MPE mean
Phototoxic potential
5.977 5.899 3.501a 5.687 3.192a 4.910a 1.000 1.000 1.000 1.000 1.051 1.000 1.129 1.207 1.000 1.000 1.000 1.000
0.313 0.309 0.177b 0.198 0.165b 0.263b 0.130 0.067 0.122 0.123 0.198b 0.136 0.045 0.069 0.011 0.024 0.103 0.085
Yes Yes Yes Yes Yes Yes No No No No
1.396 1.000 0.818 0.847 1.045 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000
0.087 0.112 −0.100 −0.119 −1.250 −0.550 0.033 0.088 0.049 0.091 0.082 0.103 −0.097 −0.296 −0.007 −0.087 −1.642 −1.427
No No No No No No No No NO No No No No No No No No No
No No No No No No No
Probable phototoxic. Phototoxic (according to OECD, 2004). 184
Toxicology in Vitro 60 (2019) 180–186
L. Xiong, et al.
Table 3 The HOMO and LUMO orbital and EGAP of BP-1, BP-3 and BP-4. Substance name
Structural formula
HOMO
LUMO
EGAP (eV)
EGAP (J/mol)
BP-1
2.891
278.937
BP-3
2.832
273.244
BP-4
0.014
1.351
et al.) after electron transfer (type I reaction) or singlet oxygen (1O2) following energy transfer (type II reaction). The organic UV filters are also called ultraviolet absorbent substances. With their potential of UV absorbance, BPs offer sun protection, but it may lead to phototoxicity. BPs contain chromophore which on absorption triggers photophysical or photochemical phenomena leading to a direct or indirect damage of biological macromolecules (Bignon et al., 2017). The protective mechanism of BPs against harmful effects of ultraviolet light functions by chemically absorbing light energy (photons). As this occurs, the benzophenone molecule becomes activated to higher energy levels. As the excited molecule returns to its ground state, the energy is released in the form of thermal energy. Photosensitization cascade may involve energy or electron-transfer, or activation of molecular oxygen via production of the reactive singlet oxygen (1O2) that ultimately produces oxidative reactions (Bignon et al., 2017). Then, the historical parabolic relationships between phototoxicity and the energy gap (EGAP) between energies of the highest occupied molecular orbital and the lowest unoccupied molecular orbital was confirmed (Parr and Zhou, 1993; Ringeissen et al., 2011). We set up the mechanistic model using DMol3 program package of Materials Studio (Delley, 2000; Perdew et al., 1996), and showed that the calculating results of the model were similar with experiment results. The EGAP of BP-1 was higher than BP-3 and much higher than BP-4. Note that we could speculate that the phototoxicity of BP-1 provide a close representation to BP-3, but in vitro results suggested that BP-1 were tested for their phototoxicity in comparison to BP-3. That may be another limitation with the cell test. In vitro test is only employed for hazard identification while its utility for the assessment of phototoxic potency is not warranted. In particular, this assay system lacks metabolic activity which is critical in the manifestation of systemically exposed chemicals. It has been reported that BP-3 is metabolized principally to BP-1, via metabolic activities of cytochrome P450 (CYP) enzymes (Kunisue et al., 2012; Jeon et al., 2008; Kasichayanula et al., 2005; Nakagawa and Suzuki, 2002; Okereke et al., 1994). There were also reports that intramolecular hydrogen bond formation (e.g. BP-3), reversible photoisomerization (e.g. ecamsule, octyl methoxycinnamate) and conjugated bonds were present in almost all UV filters (Kockler et al., 2012; Shaath, 2010). However, the phototoxicity of BPs is far less
4. Discussion In the present study, we summarized the phototoxicity results obtained from the modified OECD No. 432 assay (OECD, 2004), which showed some a few divergent scenarios that could be encountered. The original SOP developed for the 3T3 NRU assay had to be adapted to the UVA irradiation with a dose of 5 J/ cm2.Our assay systems were found to be applicable for the prediction of UVB phototoxicity, and CPZ could be selected as a positive control because the PIF and MPE were calculated to test and verify in the three cell lines, consistent with the previously reported data (Ray et al., 2008; Murli et al., 2010). Organic sunscreens often protect us by absorbing environmental UVA and UVB radiation, so we need an integrated testing strategy to predict phototoxicity involving both UVA and UVB. Moreover, 3T3 NRU test has many limitations, for instance, the test exhibits highly sensitive but low specific predictive capacity (Kim et al., 2015). In this study, we found that irradiated HS68 and HaCaT were less susceptible than 3T3 when exposed to UV radiation. As provisional prediction model, we could find it similar to light injury, UVB could cause superficial injury such as in stratum epidermis and epidermal cells, but UVA can penetrate deep into dermis of the skin eventually causing premature photo-aging by stimulating the over-expression of MMPs (matrix Metalloproteinases) by fibroblasts (Kang et al., 2003). Therefore, we suggested to use UVA radiation in fibroblasts to and UVB radiation in keratinocytes for future studies, making the model more close to reality. Recently, concerns about chemical phototoxicity are on the rise, and most UV filters are widely used in sunscreens generate ROS when skin is exposed to UV radiation. (Biniek et al., 2012). We tried to develop a new mechanistic model to predict photo-toxicity. Photo-induced toxicity by PAH was first modeled in 1987 (Newsted and Giesy, 1987). They investigated many of the structural factors of phototoxicity of the PAH, among which were the lowest singlet excitation energy of the excited molecule, the lowest triplet excitation energy of the excited molecule, and the phosphorescence lifetime (Newsted and Giesy, 1987). As mentioned above, the most commonly reported phototoxic process is oxidative reactions. The chromophore in its excited state transfers energy or electrons to oxygen and generates ROS (O2%-, NO%, OH%, ROO%
185
Toxicology in Vitro 60 (2019) 180–186
L. Xiong, et al.
established and should be properly characterized and understood through one or multiple methods. In conclusion, we rationalized the UV filtering efficiency of BPs by performing cell viability assays over different cell lines exposed to UVB and UVA light. Indeed, the highly phototoxic substances (such as CPZ, Angelicin, 8-methoxypsoralen) are found in many cell types selectively affect. It's hard to determine a very weak phototoxic potential in human skin as of BPs in vitro assays. In this study, BP-1 appeared to be phototoxic under UVA+ condition. However, BP-3 and BP-4 were found to be non-phototoxic under UVA+ condition. It verified that BP-1, BP-3 and BP-4 remained low-risk chemicals in UVB+ condition. The theoretical calculation showed BP-1(EGAP) > BP-3(EGAP) > BP-4(EGAP). The methods established can be used for testing phototoxicity of chemical components. For systemically exposed chemicals that require metabolic activation more integrated testing strategies are recommended, and this study could provide a better choice to cosmetic formulators for selecting UV filter in the potentially effective products. Supported by these conclusions, we recommend BP-4 as an effective and safer UV protective compound in sunblock formulations.
Phys. 113, 7756–7764. Du, Y., Wang, W.Q., Pei, Z.T., Ahmad, F., Xu, R.R., Zhang, Y.M., Zhang, Y.M., Sun, L.W., 2017. Acute toxicity and ecological risk assessment of Benzophenone-3 (BP-3) and Benzophenone-4 (BP-4) in ultraviolet (UV)-filters. Inter. J. Env. Res. Pub. Heal. 14, 1414. Gülden, M., Mörchel, S., Seibert, H., 2005. Comparison of mammalian and fish cell line cytotoxicity: impact of endpoint and exposure duration. Aquat. Toxicol. 71, 229–236. Herrling, T., Jung, K., Fuchs, J., 2006. Measurements of UV-generated free radicals/reactive oxygen species (ROS) in skin. Spectrochim. Acta A 63, 840–845. Ishiyama, M., Miyazono, Y., Sasamoto, K., Ohkura, Y., Ueno, K., 1997. A highly watersoluble disulfonated tetrazolium salt as a chromogenic indicator for NADH as well as cell viability. Talanta 44, 1299–1305. Jeon, H., Sarma, S.N., Kim, Y., Ryu, J., 2008. Toxicokinetics and metabolisms of benzophenone-type UV filters in rats. Toxicol 248, 89–95. Kang, S., Chung, J.H., Lee, J.H., Fisher, G.J., Wan, Y.S., Duell, E.A., Voorhees, J.J., 2003. Topical N-acetyl cysteine and genistein prevent ultraviolet-light-induced signaling that leads to photoaging in human skin in vivo. J. Invest. Dermatol. 120, 835. Kasichayanula, S., House, J.D., Wang, T., Gu, X., 2005. Simultaneous analysis of insect repellent DEET, sunscreen oxybenzone and five relevant metabolites by reversedphase HPLC with UV detection: application to an in vivo study in a piglet model. J. Chromatogr. B 822, 271–277. Kim, K., Park, H., Lim, K.M., 2015. Phototoxicity: its mechanism and animal alternative test methods. Toxicol. Res. 31, 97–104. Kochevar, I.E., 1981. Phototoxicity mechanisms: chlorpromazine photosensitized damage to DNA and cell membranes. J. Invest. Dermatol. 77, 59–64. Kockler, J., Oelgemöller, M., Robertson, S., Glass, B.D., 2012. Photostability of sunscreens. J Photochem Photobiol C: Photochem Rev 13, 91–110. Kunisue, T., Chen, Z., Louis, G.M.B., et al., 2012. Urinary concentrations of benzophenone-type UV filters in U.S. women and their association with endometriosis. Environ. Sci. Technol. 46, 4624–4632. Lelièvre, D., Justine, P., Christiaens, F., Bonaventure, N., Coutet, J., Marrot, L., TinoisTessonneaud, E., Cotovio, J., 2007. The episkin phototoxicity assay (epa): development of an in vitro tiered strategy using 17 reference chemicals to predict phototoxic potency. Toxicol. in Vitro 21, 977–995. Lou, J., Zhou, G., Chu, G., Jiang, J., Huang, F., Zheng, S., Lu, Y., Li, X., He, J., 2010. Studying the cyto-genotoxic effects of 12 cigarette smoke condensates on human lymphoblastoid cell line in vitro. Mutat. Res. 696, 48–54. Murli, H., Aardema, M., Lawlor, T., Spicer, C., 2010. Photoclastogenicity?An improved protocol, its validation, and investigation of the photogenotoxicity of DMBA. Environ. Mol. Mutagen. 40, 41–49. Nakagawa, Y., Suzuki, T., 2002. Metabolism of 2-hydroxy-4methox- ybenzophenone in isolated rat hepatocytes and xenoestrogenic effects of its metabolites on MCF-7 human breast cancer cells. Chem. Biol. Interact. 139, 115–128. Newsted, J.L., Giesy, J.P., 1987. Predictive models for photoinduced acute toxicity of polycyclic aromatic hydrocarbons to daphnia magna, Strauss, Cladocera, Crustacea. Environ. Toxicol. Chem. 6, 445–461. OECD, 2004. Test No. 432: In Vitro 3T3 NRU Phototoxicity Test. Publishing, OECD. OECD (Q)SARs projects, 2010. Available at: http://www.oecd.org/document/23/0, 3343,en_2649_34379_33957015_1_1_1_1,00.html, Accessed date: 14 June 2010. Okereke, C.S., Abdel-Rhaman, M.S., Friedman, M.A., 1994. Disposition of benzophenone3 after dermal administration in male rats. Toxicol. Lett. 73, 113–122. Onoue, S., Seto, Y., Sato, H., Nishida, H., Hirota, M., Ashikaga, T., Api, A.M., Basketter, D.A., Tokura, Y., 2017. Chemical photoallergy: photobiochemical mechanisms, classification, and risk assessments. J. Dermatol. Sci. 85, 4–11. Parr, R., Zhou, Z., 1993. Absolute hardness: unifying concept for identifying shells and subshells in nuclei, atoms, molecules, and metallic clusters. Acc. Chem. Res. 26, 256–258. Perdew, J.P., Burke, K., Ernzerhof, M., 1996. Generalized gradient approximation made simple. Phys. Rev. Lett. 77, 3865–3868. Rai, V., Dayan, N., Michniakkohn, B., 2011. A comparative evaluation of photo-toxic effect of fractionated melanin and chlorpromazine hydrochloride on human (dermal fibroblasts and epidermal keratinocytes) and mouse cell line/s (fibroblast balb/c 3t3). Toxicol. in Vitro 25, 538–544. Ray, R.S., Agrawal, N., Sharma, A., et al., 2008. Use of L-929 cell line for phototoxicity assessment[J]. Toxicol. in Vitro 2008, 1775–1781. Ringeissen, S., Marrot, L., Labarussiat, A., Imbert, S., Todorov, M., Mekenyan, O., Meunier, J., 2011. Development of a mechanistic SAR model for the detection of phototoxic chemicals and use in an integrated testing strategy. Toxicol. in Vitro 25, 324–334. Saroj Kumar, A., Goyal, S., Dubey, D., Srivastav, A.K., Chopra, D., Singh, J., Shankar, J., Chaturvedi, R.K., Ray, R.S., 2015. Benzophenone 1 induced photogenotoxicity and apoptosis via release of cytochrome c and Smac/DIABLO at environmental UV radiation. Toxicol. Lett. 239, 182–193. Shaath, N.A., 2010. Ultraviolet filters. Photochem. Photobiol. Sci. 9, 464–469. Tanneberger, K., Rico-Rico, A., Kramer, N.I., Busser, F.J.M., Hermens, J.L.M., Schirmer, K., 2010. Effects of solvents and dosing procedure on chemical toxicity in cell- based in vitro assays. Environ. Sci. Technol. 44, 4775–4781. Wondrak, G.T., Jacobson, M.K., Jacobson, E.L., 2006. Endogenous UVA-photosensitizers: mediators of skin photodamage and novel targets for skin photoprotection. Photochem. Photobiol. Sci. 5, 215–237.
Declaration of competing interests There are no conflicts of interest. Acknowledgements We gratefully acknowledge the financial support provided by Sichuan Science and Technology Program (No. 2017SZ0070 and No. 2018SZ0104), China National Natural Science Foundation(No. 81673084) and 1.3.5 project for disciplines of excellence, West China Hospital, Sichuan University. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.tiv.2019.05.021. References Benevenuto, C.G., Guerra, L.O., Gaspar, L.R., 2015. Combination of Retinyl palmitate and UV-filters: phototoxic risk assessment based on Photostability and in vitro and in vivo Phototoxicity assays. Eur. J. Pharm. S.C. 68, 127–136. Betowski, L.D., Enlow, M., Riddick, L., 2002. The phototoxicity of polycyclic aromatic hydrocarbons: a theoretical study of excited states and correlation to experiment. Comput. Chem. 26, 371–377. Bignon, E., Marazzi, M., Besancenot, V., Gattuso, H., Drouot, G., Morell, C., Eriksson, L.A., Grandemange, S., Dumont, E., Monari, A., 2017. Ibuprofen and ketoprofen potentiate UVA-induced cell death by a photosensitization process. Sci. Rep. 7 (8885). Biniek, K., Levi, K., Dauskardt, R.H., 2012. Solar UV radiation reduces the barrier function of human skin. P. Nat. Acad. Sci. 109, 17111–17116. Bowden, G.T., 2004. Prevention of non-melanoma skin cancer by targeting ultraviolet-Blight signalling. Nat. Rev. Cancer 4, 23–35. Broeders, J.J., Blaauboer, B.J., Hermens, J.L., 2013. In vitro biokinetics of chlorpromazine and the influence of different dose metrics on effect concentrations for cytotoxicity in balb/c 3t3, caco-2 and heparg cell cultures. Toxicol. in Vitro 27, 1057–1064. Carbonare, M.D., Pathak, M.A., 1992. Skin photosensitizing agents and the role of reactive oxygen species in photoaging. J. Photochem. Photobiol. B 14, 105–124. Crovara Pescia, A., Astolfi, P., Puglia, C., Bonina, F., Perrotta, R., Herzog, B., Damiani, E., 2012. On the assessment of photostability of sunscreens exposed to UVA irradiation: from glass plates to pig/human skin, which is best? Int. J. Pharm. 427, 217–223. de Oliveira, C.A., Peres, D.D., Rugno, C.M., Kojima, M., de Pinto, C.A.S.O., Consiglieri, V.O., Kaneko, T.M., Rosado, C., Mota, J., Velasco, M.V.R., Baby, A.R., 2015. Functional photostability and cutaneous compatibility of bioactive UVA sun care products. J. Photochem. Photobiol. B Biol. 148, 154–159. Delley, B., 1990. An all-electron numerical method for solving the local density functional for polyatomic molecules. J. Chem. Phys. 92, 508–517. Delley, B., 2000. From molecules to solids with the DMol [sup 3] approach. J. Chem.
186