Advertisement
Research ArticleToxicology

Phototoxic Risk Assessments on Benzophenone Derivatives: Photobiochemical Assessments and Dermal Cassette-Dosing Pharmacokinetic Study

Yoshiki Seto, Hiroto Ohtake, Masashi Kato and Satomi Onoue
Journal of Pharmacology and Experimental Therapeutics August 2015, 354 (2) 195-202; DOI: https://doi.org/10.1124/jpet.115.223644

Abstract

This study aimed to qualify photosafety screening on the basis of photochemical and pharmacokinetic (PK) data on dermally applied chemicals. Six benzophenone derivatives (BZPs) were selected as model compounds, and in vitro photochemical/phototoxic characterization and dermal cassette-dosing PK study were carried out. For comparison, an in vivo phototoxicity test was also conducted. All of the BZPs exhibited strong UVA/UVB absorption with molar extinction coefficients of over 2000 M−1 × cm−1, and benzophenone and ketoprofen exhibited significant reactive oxygen species (ROS) generation upon exposure to simulated sunlight (about 2.0 mW/cm2); however, ROS generation from sulisobenzone and dioxybenzone was negligible. To verify in vitro phototoxicity, a 3T3 neutral red uptake phototoxicity test was carried out, and benzophenone and ketoprofen were categorized to be phototoxic chemicals. The dermal PK parameters of ketoprofen were indicative of the highest dermal distribution of all BZPs tested. On the basis of its in vitro photochemical/phototoxic and PK data, ketoprofen was deduced to be highly phototoxic. The rank of predicted phototoxic risk of BZPs on the basis of the proposed screening strategy was almost in agreement with the results from the in vivo phototoxicity test. The combined use of photochemical and cassette-dosing PK data would provide reliable predictions of phototoxic risk for candidates with high productivity.

Introduction

Drug-induced phototoxicity is caused after exposure of light-exposed tissues to topically and/or systemically administered chemicals, including pharmaceuticals, cosmetics, and food ingredients, followed by exposure to sunlight, consisting of UVB (290–320 nm), UVA (320–400 nm), and UV-visible light (400–700 nm) (Moore, 1998, 2002; Onoue et al., 2009). Recently, interest in phototoxic events due to photoreactive substances has markedly increased owing to high-intensity UV rays from the sun reaching the Earth’s surface upon the destruction of the ozone layer (Onoue et al., 2009). Perceptible adverse events would leave patients and consumers with negative impressions of products, and the pharmaceutical and cosmetic industries have struggled to predict and/or avoid phototoxic events. In the early 2000s, regulatory agencies, including the Food and Drug Administration, the European Medicines Agency, and the Organization for Economic Cooperation and Development (OECD), established guidance on photosafety testing of medicinal products (EMEA/CPMP, 2002, 2008; FDA/CDER, 2002; OECD, 2004). In 2014, the International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH) established the ICH S10 guideline (ICH, 2014), and this guidance describes photosafety assessment strategies on the basis of photochemical/photobiological properties and in vivo pharmacokinetic (PK) behaviors in UV-exposed tissues, such as skin and eyes.

Over the past few years, a number of efforts have been made to develop effective phototoxic assessments, and analytical and biochemical methodologies for evaluating phototoxic risk have been proposed, including in silico prediction models, photochemical screening tools, and in vitro phototoxicity assessments (Seto et al., 2012). In particular, UV-visible measurement (Henry et al., 2009), 3T3 neutral red uptake phototoxicity test (3T3 NRU PT) (Spielmann et al., 1994), and/or reactive oxygen species (ROS) assay (Onoue and Tsuda, 2006; Onoue et al., 2008) are currently described as recommended methodologies for evaluating the photosafety of test substances, and these photosafety assessment tools are also useful in the early stages of product development. Although effective in vitro photosafety assessments have been developed on the basis of photochemical/photobiological mechanisms, the data obtained from these methodologies cannot fully reveal the occurrence of phototoxic events in a clinical context due to a lack of data on distribution and retention in UV-exposed tissues. Previously, combined use of in vitro photochemical/phototoxic data and in vivo PK behaviors was proposed for evaluating the in vivo phototoxic risk of chemicals (Seto et al., 2009), and a cassette-dosing PK study was employed with the aim of reducing the use of research resources and enhancing throughput (Seto et al., 2011; Onoue et al., 2014a). In these previous studies, the predicted in vivo phototoxic risk of systemically exposed chemicals on the basis of the in vitro photochemical/photobiological and in vivo PK data exhibited a strong association with clinical phototoxicity findings (Seto et al., 2011; Onoue et al., 2014a). In contrast, the proposed screening strategy has not been applied to phototoxic risk assessment on dermally applied chemicals, the phototoxicity of which may lead to a no-go decision in product development.

The present study aimed to clarify the applicability of the photosafety screening using photochemical and PK data for dermally administered chemicals. Six benzophenone (BZ) derivatives (BZPs), namely, BZ, dioxybenzone (DO), ketoprofen (KT), mexenone (MX), oxybenzone (OX), and sulisobenzone (SB), were selected as model chemicals (Fig. 1). The phototoxic skin reactions by KT, MX, and OX have been reported on the basis of the results from human photopatch tests (Chuah et al., 2013; Infante Hernando et al., 2013), although these are directly applied to the skin. OX, SB, DO, and MX were developed as sunscreens (Darvay et al., 2001; Baughman et al., 2009), and KT is one of the most well known nonsteroidal anti-inflammatory drugs given via the dermal route. In this investigation, UV spectral analysis and ROS assay were carried out to evaluate the photochemical properties of BZPs, and 3T3 NRU PT was also undertaken to verify the in vitro phototoxicity of BZPs. Cassette-dosing PK study in rats after the dermal coadministration of BZPs was conducted to determine the PK behaviors of BZPs in the skin, and the skin concentration of BZPs was assessed by ultra-performance liquid chromatography equipped with electrospray ionization mass spectrometry (UPLC/ESI-MS). On the basis of the obtained in vitro photochemical/photobiochemical and in vivo dermal PK data, the in vivo phototoxic risk of the tested BZPs was rated. To compare the predicted in vivo phototoxic risk with the in vivo phototoxic observations, in vivo phototoxicity tests were also performed on each BZP.

Fig. 1.
Fig. 1.

Chemical structure of tested BZPs. aClog P calculated using ChemBioDraw Ultra 13.0 software.

Materials and Methods

Chemicals

BZ was purchased from Junsei Chemical (Tokyo, Japan). DO, KT, MX, OX, erythromycin (EM), dimethylsulfoxide (DMSO), imidazole, nitroblue tetrazolium, p-nitrosodimethylaniline, propylene glycol, disodium hydrogen phosphate 12-water, and sodium dihydrogen phosphate dihydrate were obtained from Wako Pure Chemical Industries (Osaka, Japan). Quinine HCl (QN) and SB were purchased from Sigma-Aldrich Japan (Tokyo, Japan). Acetonitrile was obtained from Honeywell International (Morristown, NJ).

UV-Visible Spectral Analysis

Each chemical was dissolved in 20 mM sodium phosphate buffer (NaPB; pH 7.4) at a final concentration of 20 μM. UV-visible absorption spectra were recorded with a Hitachi U-2010 spectrophotometer (Hitachi, Tokyo, Japan) interfaced to a PC for data processing (Spectra Manager; JASCO, Easton, MD). A spectrofluorimeter quartz cell with 10 mm path length was employed.

ROS Assay

Irradiation Conditions.

Chemicals were stored in an Atlas Suntest CPS+ solar simulator (Atlas Material Technology, Chicago, IL) equipped with a xenon arc lamp (1500 W) and cooling unit SR-P20FLE (Hitachi). A UV special filter (56052371; Atlas Material Technology) was installed to adapt the spectrum of the artificial light source to natural daylight, and the Atlas Suntest CPS series had a high irradiance capability that met CIE85/1989 daylight simulation requirements. The irradiation test was carried out at 25°C with irradiance of approximately 2.0 mW/cm2, as determined using the calibrated UVA detector Dr. Hönle 0037 (Dr. Hönle, Munich, Germany).

Assay Procedure.

For the detection of both singlet oxygen and superoxide generation from irradiated compounds, the ROS assay was carried out according to the validated protocol with minor modification (Onoue et al., 2013, 2014b). Briefly, each tested compound was dissolved in DMSO at 10 mM for stock solution. For the determination of singlet oxygen generation, compounds (200 µM), p-nitrosodimethylaniline (50 µM), and imidazole (50 µM) were dissolved in 20 mM NaPB (pH 7.4). To monitor the generation of superoxide, compounds (200 µM) and nitroblue tetrazolium (50 µM) were dissolved in 20 mM NaPB (pH 7.4). Each sample was mixed in a tube, and then a 200 μl sample was transferred into a well of a plastic 96-well microplate (clear, untreated, flat-bottomed; Asahi Glass, Tokyo, Japan). The samples were checked for precipitation before irradiation, and the plate fixed in the reaction container with a quartz cover was irradiated with simulated sunlight for 60 minutes. Before and after irradiation, absorbance levels at 440 nm and 560 nm were measured using SAFIRE (TECAN, Männedorf, Switzerland) for determination of singlet oxygen and superoxide generation, respectively.

3T3 NRU PT

The in vitro 3T3 NRU PT was carried out as described in the OECD test guidelines 432 with minor modification (OECD, 2004). Briefly, Balb/c 3T3 cells were maintained in culture for 24 hours for the formation of monolayers. Two 96-well plates per test chemical were then preincubated with six different concentrations of the chemical dissolved in Earle’s balanced salt solution for 1 hour in duplicate. One plate was then exposed to a dose of 5 J/cm2 UVA (+Irr experiment), whereas the other plate was kept in the dark by covering it with aluminum foil (−Irr experiment). UVA irradiation was performed using a sol 500 Sun simulator (Dr. Hönle) equipped with a 500 W metal halide lamp and a H-1 filter to remove potentially cytotoxic UVB wavelengths. The treatment medium was then replaced with culture medium and, after 24 hours, cell viability was determined by neutral red uptake for 3 hours. After that, cells were lysed in eluate (ethanol:water:acetic acid, 50:49:1), and the neutral red uptake was measured at the absorbance of 540 nm using the Benchmark Plus microplate spectrophotometer (Bio-Rad, Hercules, CA). Cell viability obtained with each of the six concentrations of the test chemical was compared with that of untreated controls, and the percent inhibition was calculated. For evaluating in vitro phototoxicity, the concentration responses obtained in the presence and in the absence of UVA irradiation were compared, usually at the IC50 level, that is, the concentration inhibiting cell viability by 50% compared with that of untreated controls. The photoirritancy factor (PIF) was determined using the following equation:

Embedded Image

In Vivo Preparation

Male Sprague-Dawley rats at 10–12 weeks of age (approximately 309–440 g body weight) were purchased from SLC (Hamamatsu, Japan), housed in the laboratory with free access to food and water, and maintained on a 12-hour dark/light cycle in a room with controlled temperature (24 ± 1°C) and humidity (55 ± 5%). On the day before experiments, rats were anesthetized with pentobarbital Na (50 mg/kg), and then the hair on the abdomen was shaved (5 cm × 5 cm). All of the procedures used in the present study were conducted according to the guidelines approved by the Institutional Animal Care and Ethical Committee of the University of Shizuoka.

Skin Deposition of BZPs after Dermal Coadministration

A cocktail solution containing all six BZPs (final concentration of each: 1 mg/ml) was prepared with the use of propylene glycol and applied to the rat skin over the abdomen (0.1 mg/rat each, n = 5) using an adhesive plaster for a patch test (Libatape Pharmaceutical, Kumamoto, Japan) under anesthesia with pentobarbital Na (50 mg/kg). At the indicated times (2, 4, 6, 8, 12, and 24 hours) after the dermal coadministration of BZPs, rats were humanely killed by taking blood from the descending aorta under anesthesia with pentobarbital Na (50 mg/kg), and the tissues were then perfused with cold saline from the aorta. The skin was dissected, minced with scissors, and homogenized using Physcotron (Microtec, Chiba, Japan) in 4 ml acetonitrile. After sonication for 10 minutes and shaking for 10 minutes, the samples were centrifuged (3000 rpm, 10 minutes). Extraction was repeated twice with acetonitrile, and the supernatants were pooled. The collected eluents were pooled with acetonitrile extracts, and the samples were evaporated to dryness under a gentle stream of nitrogen at 45°C. The residues were dissolved in 50% acetonitrile, including fenofibric acid (500 ng/ml) as an internal standard for UPLC analysis.

UPLC Analysis

The concentration of BZPs in rat skin was determined by UPLC/ESI-MS analysis. The UPLC/ESI-MS system consisted of a Waters Acquity UPLC system (Waters, Milford, MA), which included a binary solvent manager, a sample manager, a column compartment, and a micromass SQ detector connected with Waters Masslynx, version 4.1. A Waters Acuity UPLC BEH C18 (particle size: 1.7 μm and column size: Φ2.1 × 50 mm; Waters) was used, and the column temperature was maintained at 40°C. The standards and samples were separated using a gradient mobile phase consisting of Milli-Q containing 0.1% formic acid (A) and acetonitrile (B). The gradient conditions of the mobile phase were 0–1 minute, 50% B; 1–3 minutes, 50–75% B (linear gradient curve); 3–4 minutes, 75–95% B (linear gradient curve); 4–5.5 minutes, 95% B; and 5.5–6 minutes, 50% B, and the flow rate was set at 0.25 ml/min. Analysis was carried out using selected ion recording for specific m/z 183.2 for BZ [M + H]+, 255.2 for KT [M + H]+, 229.2 for OX [M + H]+, 307.1 for SB [M − H]−, 245.2 for DO [M + H]+, 243.3 for MX [M + H]+, and 319.2 for fenofibric acid [M + H]+ as an internal standard. Peaks for BZ, KT, OX, SB, DO, MX, and fenofibric acid were detected at retention times of 1.96, 1.14, 2.54, 0.57, 2.54, 3.18, and 2.00, respectively. The newly developed UPLC/ESI-MS method for the determination of BZPs was validated in terms of linearity, accuracy, and precision according to the ICH guidelines “Q2B Validation of Analytical Procedures: Methodology.” On the basis of the skin concentration obtained, PK analysis was performed, and the area under concentration versus time curve from 0 to 24 hours and mean residence time (MRT) were estimated using a trapezoid formula.

In Vivo Phototoxicity Testing

Each BZP or control (QN and EM) was dissolved in DMSO at 100 mg/ml and was applied to two application sites on rat skin at the abdomen (10 mg/site, n = 4–5) using filter paper (2 cm × 2 cm) under anesthesia with pentobarbital Na (50 mg/kg). The exposure time before irradiation was set at 3 hours after dermal administration to avoid a long time of restraint and to reduce the risk of death by supplemental anesthesia. At 3 hours after dermal administration, the filter papers containing chemicals on the application sites were removed and wiped using cotton soaked with distilled water. Then rats were irradiated individually using black light (FL15BL-B; National, Tokyo, Japan) as a UVA light source with an irradiance of approximately 2.7 mW/cm2 for about 3 hours until the UV irradiance level reached 30 J/cm2. Because UVB light is highly cytotoxic, a UVA light source was employed for the in vivo phototoxicity testing. During the UVA irradiation, rats were restrained on a sunbed under anesthesia with pentobarbital Na (50 mg/kg) to ensure uniform irradiation of their abdomen, and nonirradiated sites were wrapped in aluminum foil for protection from UV rays. UV intensity was monitored using the calibrated UVA detector Dr. Hönle 0037. A colorimeter equipped with a data processor (NF333; Nippon Denshoku, Tokyo, Japan) was used as a measure of skin color. This instrument records three-dimensional color reflectance, so-called L*a*b* system, as recommended by the Commission Internationale de l’Eclairage. The luminance (L*) gives the relative brightness ranging from total black (L* = 0) to total white (L* = 100). The hue (a*) axis represents the balance between red (positive values up to 100) and green (negative values up to −100), and the chroma (b*) axis represents the balance between yellow (positive values up to 100) and blue (negative values up to −100). The differences in skin color (ΔE) between before and after irradiation were described as follows (Westerhof et al., 1986; Pierard and Pierard-Franchimont, 1993):

Embedded Image

Data Analysis

For statistical comparisons, one-way analysis of variance with pairwise comparison by Fisher’s least significant difference procedure was used. A P value of less than 0.05 was considered significant for all analyses. Theoretical calculations of the lipophilicity as calculated log P (Clog P) were performed using ChemBioDraw Ultra 13.0 (PerkinElmer, Waltham, MA) with chemical structure inputs.

Results

Photochemical Characterization of BZPs.

As a first step of phototoxic events, phototoxic compounds are transferred from their ground state to an excited state by absorption of photon energy from sunlight, and UV-visible spectral analysis in the range of 290–700 nm is recommended by guidelines as a rapid and simple method for predicting the photoreactive/phototoxic potential of chemicals (Seto et al., 2012). Thus, the UV-visible spectral patterns of six BZPs (20 μM) were recorded in 20 mM NaPB (pH 7.4). All BZPs exhibited intense absorption in the UVA/B range (Fig. 2A); in contrast, absorption spectral patterns of BZPs could not be observed in the visible light region (data not shown). On the basis of the spectral patterns obtained, molar extinction coefficient (MEC) values of BZPs at 290 nm were calculated as follows: 2317 (BZ), 10,833 (DO), 4350 (KT), 2850 (MX), 6817 (OX), and 11,700 (SB) M−1⋅cm−1. According to the previous report (Henry et al., 2009) and ICH S10 photosafety guidance (ICH, 2014), compounds with MEC values greater than 1000 M−1⋅cm−1 at any wavelength (290–700 nm) are considered to be of greater photosafety risk. In this study, all tested BZPs were thus suggested to have photoexcitability and photoreactive potential, possibly leading to phototoxic reactions.

Fig. 2.
Fig. 2.

In vitro photochemical properties of BZPs. (A) UV absorption spectra of BZPs (20 μM) in 20 mM NaPB (pH 7.4). Thick solid line, BZ; thin solid line, KT; thin dashed line, OX; thick dashed line, SB; thin dotted line, DO; and thick dotted line, MX. (B) Generation of singlet oxygen (filled bars) and superoxide (open bars) from BZPs (200 μM) exposed to simulated sunlight (about 2.0 mW/cm2). QN and EM as positive and negative controls, respectively. Data represent mean ± SD for three experiments. N.A., not applicable; N.D., not detected.

In the indirect process of phototoxic events, ROS, including singlet oxygen and superoxide, are well known as principal intermediate species, and oxidative damage against biomolecules by chemicals after sunlight exposure mainly arises via ROS, resulting in various phototoxic events (Epstein, 1983). Thus, monitoring ROS generation from photoirradiated chemicals would be useful for evaluating the phototoxic risk of chemicals. To clarify the photoreactivity of BZPs, the ROS assay was undertaken with the use of six BZPs at 200 μM, and QN and EM (200 μM) were employed as positive and negative controls for ROS generation, respectively (Fig. 2B). KT and QN exhibited significant generation of both singlet oxygen and superoxide under simulated sunlight exposure, and BZ and OX yielded singlet oxygen via a type II photochemical reaction. In contrast, ROS generation from DO, SB, and EM was negligible, and the ROS data on MX could not be obtained owing to the appearance of precipitation of MX. As for poorly water-soluble chemicals, a lower concentration of chemicals can be applied to the ROS assay for evaluating photoreactivity; therefore, for further investigation, ROS assay on MX at 100 μM was also carried out. The generation of singlet oxygen from irradiated MX could be detected without the precipitation of MX (ΔA440 nm⋅103: 67), whereas the generation of superoxide anion from irradiated MX (100 μM) was negligible. The chemicals can be determined to be photoreactive if the obtained ROS data surpass the previously defined criteria in the ROS assay [singlet oxygen (ΔA440 nm⋅103): 25; and superoxide (ΔA560 nm⋅103): 20] (Onoue et al., 2008, 2014b). On the basis of these criteria, BZ, KT, MX, and OX were found to be photoreactive, and these BZPs would cause phototoxic reactions in the skin. DO and SB were judged to be less photoreactive by ROS assay, whereas, interestingly, their MEC values were higher than those of the other BZPs generating ROS. These findings might suggest that UV-visible absorption properties would not always reflect the rank-order photoreactivity/phototoxicity of chemicals.

In Vitro Phototoxicity of BZPs.

For evaluating in vitro phototoxicity, 3T3 NRU PT, which is widely used as an alternative in vitro methodology to various in vivo photosafety assessments, was conducted. The 3T3 NRU PT was developed and validated under the auspices of European Centre for the Validation of Alternative Methods from 1992 to 1997 (Liebsch and Spielmann, 2002), and the OECD recommends this in vitro methodology for evaluating the photosafety of chemicals (OECD, 2004). The 3T3 NRU PT assesses the concentration-dependent cytotoxicity on the Balb/c 3T3 mouse fibroblast cell line of UVA-irradiated chemicals using the uptake of neutral red by living cells. The cell viability curves of BZPs with or without UVA irradiation were obtained, and the PIF values were calculated (Table 1). Compared with the nonirradiated groups, UVA-irradiated BZ and KT exhibited significant cytotoxicity at lower doses, and the viability curves of UVA-irradiated groups were shifted to the left. The PIF values of BZ and KT were estimated to be 49.5 or more and 68.9, respectively. In contrast, the PIF values of DO, OX, and SB were calculated to be about 1.0 because these chemicals produced similar cell viability curves between UVA-irradiated and nonirradiated groups. The cytotoxicity of MX could not be fully monitored owing to its poor solubility to the assay mixture, and the PIF values of MX could not be calculated; therefore, 3T3 NRU PT would not be suitable for evaluating the phototoxicity of MX. According to the OECD test guidelines 432, classification criteria of 3T3 NRU PT on the basis of PIF values are defined as follows: 1) phototoxicity (PIF ≥ 5); 2) probable phototoxicity (2 ≤ PIF < 5); and 3) nonphototoxicity (PIF < 2) (OECD, 2004). Thus, BZ and KT were identified as phototoxic molecules, and DO, OX, and SB were determined as nonphototoxic ones.

View this table:
TABLE 1

Decision matrix

Pharmacokinetic Behaviors of BZPs.

The deposition of chemicals to sunlight-exposed tissues, including skin and eyes, was shown to be closely associated with the incidence of drug-induced phototoxic reactions (Seto et al., 2012), and the ICH S10 photosafety guidance describes in vivo PK behaviors as important factors of drug-induced phototoxicity (ICH, 2014); therefore, monitoring the skin concentration of dermally administered chemicals should be a key consideration for phototoxic risk prediction. In the present investigation, dermal cassette-dosing PK study was employed for improving the throughput of experiments and reducing the use of various research resources. In particular, considering implementation of the 3Rs principles (refinement, reduction, and replacement), cassette-dosing approaches can make a strong contribution to reducing the number of animals humanely killed (Allen et al., 1998). The concentration versus time curves in the skin after dermal coadministration of the six BZPs (0.1 mg/rat each) were obtained by UPLC/ESI-MS analysis (Fig. 3), and the PK parameters of the BZPs were also calculated on the basis of the concentration-time profiles obtained (Table 1). With regard to BZ, the skin concentration reached the maximum concentration (Cmax) at 4 hours after dermal coadministration (about 6.1 μg/g tissue) and then decreased steadily. As for DO, MX, and OX, these three BZPs gradually increased their skin concentration, and each of their skin concentrations almost reached Cmax by approximately 4 hours after dermal coadministration. The exposure levels of the skin to these three BZPs were maintained around each of their Cmax until about 8 hours after dermal coadministration, and then they were eliminated from the skin. From these findings, the phototoxic risk of the three BZPs would persist for 4–8 hours. Unlike the PK behaviors of the other four BZPs, the skin depositions of KT and SB reached their Cmax at 24 hours after dermal coadministration, and their elimination phase could not be observed under the present experimental conditions. The skin concentration of KT, in particular, remained around its Cmax value (8.7 μg/g tissue) from 8 to 24 hours after dermal coadministration, and the area under the concentration versus time curve from 0 to 24 h and MRT of KT were estimated to be 160.0 hours⋅μg/g tissue and over 14.2 hours, respectively; therefore, KT would be associated with higher and longer exposure of the skin after dermal administration than the other BZPs. Previously, long skin retention of KT after dermal administration was observed in guinea pigs (Nakazawa et al., 2006), and the present outcomes are in agreement with this previous report. SB also exhibited the highest MRT value (>14.9 hours) among the tested BZPs, affecting the duration of exposure risk of the skin to SB.

Fig. 3.
Fig. 3.

Skin concentration-time profiles after dermal coadministration of BZPs (0.1 mg/rat each). ▪, BZ; △, KT; ▽, OX; □, SB; ○, DO; and ⋄, MX. Data represent mean for five rats.

In Vivo Phototoxicity on BZPs.

To confirm the assessment capability of the photosafety screening, the in vivo phototoxicity of BZPs was also evaluated. In this study, in vivo phototoxicity testing in rat skin after dermal administration of each BZP, QN (positive control), and EM (negative control) at a dose of 10 mg/site was also conducted for screening purposes. On the basis of the dermal cassette-dosing PK study, most of tested BZPs reached Cmax (BZ) or near Cmax (DO, MX, and OX) at about 4 hours; therefore, exposure of the skin to tested BZPs for 3 hours was undertaken, and then UVA irradiation was conducted for about 3 hours in the in vivo phototoxicity test. In contrast, in the present experiment sequence, the skin deposition of KT and SB might not be sufficient for induction of phototoxicity on the basis of the obtained PK data. To avoid insufficient skin exposure to tested BZPs, the 100-fold higher doses for dermal application were employed in the in vivo phototoxicity test compared with those in the cassette-dosing PK study. UVA-induced cutaneous inflammation was monitored by changes in skin color, and skin color was measured before and after UVA irradiation using a colorimeter with the L*a*b* system (Fig. 4). In the L*a*b* system, L* represents brightness of shade, and a* and b* represent the amount of red-green and yellow-blue color, respectively. In colorimetrical evaluation of the skin surface, QN was associated with a significant difference in ΔE values between irradiated and nonirradiated sites. The difference of ΔE values was calculated to be approximately 7.5 and was almost the same value as for orally administered QN in the previous in vivo phototoxicity test (Onoue et al., 2014a). In contrast, the difference in ΔE values of EM between irradiated and nonirradiated sites was not significant, and EM might not induce phototoxic events in the present conditions; therefore, the in vivo phototoxicity test would have been adequately carried out, and the obtained outcomes may be reliable to some extent. The ΔE values of BZ and KT in the irradiated groups were significantly different from those in the corresponding nonirradiated groups, and irradiated MX tended to exhibit a high ΔE value compared with nonirradiated MX (P = 0.0607). Thus, BZ and KT would have potent in vivo phototoxicity, and MX was determined to be a moderately phototoxic compound. In contrast, for DO, OX, and SB, a significant difference between the irradiated and nonirradiated groups could not be observed, and the in vivo phototoxicity of these three BZPs would be extremely low.

Fig. 4.
Fig. 4.

Colorimetrical evaluation of skin inflammation induced by irradiated BZPs. Differences in skin color (ΔE) between before and after irradiation were estimated on the basis of L*, a*, and b* values. QN and EM as positive and negative controls, respectively. Filled bars, UVA-irradiated group; open bars, nonirradiated group. Data represent mean ± S.E. of four to five rats. #P < 0.05, significantly different from corresponding nonirradiated group.

Discussion

In the present investigation, photosafety screening with the combined use of in vitro photochemical assessments and a dermal cassette-dosing approach was applied to in vivo phototoxicity evaluation on six dermally administered BZPs. From the present in vitro photochemical characterization, all BZPs indicated strong UV absorption on the basis of their MEC values (Henry et al., 2009); however, ROS generation from irradiated DO and SB was negligible. Referring to ICH S10 guideline, a negative result in the ROS assay supersedes a positive indication in UV-visible spectral analysis (ICH, 2014). BZ, KT, MX, and OX were found to have potent photoreactivity, possibly leading to phototoxic events in skin. By contrast, DO and SB were judged to be less phototoxic chemicals because ROS generation from irradiated DO and SB was negligible. Hence, the photoreactivity of BZPs was ranked on the basis of the present photochemical characterization as follows: KT>BZ>OX>MX≫SB≒DO. The 3T3 NRU PT was also conducted, and, on the basis of the results obtained, the rank of in vitro phototoxicity of BZPs was as follows: KT>BZ≫OX≒SB≒DO. According to the dermal cassette-dosing PK study using UPLC/ESI-MS, skin exposure levels of KT and SB were higher and lasted longer than those of the other tested BZPs, and the order for skin deposition properties on the basis of the Cmax and MRT was determined as follows: KT>SB≫OX>MX>BZ>DO.

The ICH S10 guideline (ICH, 2014) has described the critical characteristics of a phototoxicant as follows: 1) absorption of sunlight ranging from 290 to 700 nm; 2) generation of reactive species including ROS; and 3) sufficient distribution to light-exposed tissues. The in vitro photoreactivity/phototoxicity and in vivo PK profiles in the skin were obtained as crucial factors for evaluating the in vivo cutaneous phototoxic risk of tested BZPs. To integrate these results for phototoxic risk prediction on the six BZPs tested, a decision matrix was built using the experimental outcomes, including MEC, ROS data, PIF, Cmax, and MRT (Table 1). A decision matrix is a summarized schematic model of qualitative or quantitative data, which enables systemic identification, analysis, and evaluation of complicated sets of experimental outcomes (Seto et al., 2011). In the present decision matrix, if both in vitro photoreactivity/phototoxicity and PK profiles are high, the tested chemical is determined to have high phototoxic potential. In contrast, a low level of either of these two factors would lead to a judgment of a mild phototoxic risk of the tested chemical. As weighting for the phototoxic prediction, in vitro photoreactivity/phototoxicity has precedence over in vivo PK profiles because photoreactivity is reported as a key trigger for phototoxic reactions (Onoue et al., 2009); therefore, low-photoreactive chemicals were determined to be less phototoxic even if their PK parameters were high. ICH S10 guideline (ICH, 2014) has described that a photoreactive chemical with longer residence time in light-exposed tissues is more likely to produce a phototoxic reaction than a photoreactive compound having a shorter residence time. On the basis of the decision matrix, KT was deduced to have the most potent phototoxic risk because of its high in vitro photoreactivity/phototoxicity and high levels of skin deposition. Additionally, long-term dermal exposure risk of KT was suspected from the MRT value, and the cutaneous phototoxic risk might persist longer. BZ would also have potent phototoxic risk owing to significant singlet oxygen generation and potent in vitro phototoxicity even though its skin deposition was low. Although the in vitro phototoxicity of MX could not be evaluated by 3T3 NRU PT because of its poor solubility, MX determined to have phototoxic risk on the basis of its high photoreactivity and moderate skin exposure. In contrast, OX was deduced to be a less phototoxic agent due to the negative outcome from the in vitro phototoxicity test even if the positive results were obtained from in vitro photochemical assessments. SB would have less phototoxic potential due to its low in vitro photoreactivity/phototoxicity, although the MRT value of SB was almost the same as that of KT. DO was also determined to be less phototoxic because its low levels of both in vitro photoreactivity/phototoxicity and in vivo PK behaviors in the skin were observed. Overall, on the basis of the decision matrix, the order for in vivo phototoxic risk of tested BZPs was deduced as follows: KT≫BZ>MX>OX≫SB≒DO (Table 2).

View this table:
TABLE 2

The outcomes from photosafety assessments on BZPs

To clarify the in vivo phototoxicity of tested BZPs, in vivo phototoxicity test was also conducted on the six BZPs. According to the phototoxic outcomes obtained from the experiment, the ranks of phototoxicity of BZPs were as follows: BZ>KT≫MX>SB≒OX≒DO for in vivo phototoxicity (Table 2). From these findings, the order of in vivo phototoxic risk on six BZPs determined by photosafety screening was likely to show agreement with the in vivo phototoxicity of BZPs. As for MX, the 3T3 NRU PT was not applicable owing to its poorly aqueous solubility; however, the photoreactivity of MX could be obtained using the ROS assay, and in vivo phototoxic risk of MX could be deduced on the basis of the photoreactivity and dermal PK behaviors. Thus, the photoreactivity data from the ROS assay would be of help for in vivo photosafety prediction if the in vitro phototoxicity information on a chemical is not available. Phototoxic skin reactions have been divided into at least three types, namely, photoirritancy, photogenotoxicity, and photoallergy, on the basis of the mechanisms of drug-induced phototoxicity, and ROS generation from an irradiated chemical can be a key trigger for all types of phototoxic events; therefore, the ROS assay is theoretically available for evaluating the risk of all types of phototoxicity (Onoue and Tsuda, 2006). Hence, the proposed photosafety screening might be able to demonstrate all types of in vivo phototoxic risk of compounds with high accuracy, and, in the case of negative outcomes on chemicals, it would lead to the avoidance of in vivo phototoxicity test, achieving a reduction of the number of animals humanely killed.

For further investigation, Clog P values of BZPs were calculated using ChemBioDraw Ultra 13.0 software as an indicator of their lipophilicity (Fig. 1) because lipophilicity has been thought to be a dominant factor determining the skin permeability of dermally administered chemicals (Abraham and Ibrahim, 2007). On the basis of the Clog P values and the obtained concentration-time profiles in the skin, there appeared to be some relationship between lipophilicity and skin absorption behaviors of BZPs. From these findings, lipophilicity and related parameters might be useful for estimating the skin absorption properties of chemicals after dermal administration. By contrast, the prediction of the skin distribution of orally administered phenothiazines was challenging on the basis of Clog P values (Onoue et al., 2014a). In this study, for screening purposes, the lipophilicity and related parameters might be of help to establish an in silico approach as an alternative methodology for in vivo dermal PK study, and development of a nonanimal photosafety assessment with combined use of in silico skin permeability and in vitro photoreactivity/phototoxicity might make a major contribution to animal welfare. However, in vivo PK profiling would still be important for estimating temporal changes in the phototoxic risk of compounds, including appearance time and duration of retention.

From the present findings, the proposed photosafety screening on the basis of the in vitro photoreactive/phototoxicity and in vivo PK data could evaluate the phototoxic risk of BZPs and would be applicable to the prediction of in vivo phototoxic risk, including all types of phototoxicity, on dermally administered chemicals with high accuracy. Recently, interest in the photosafety of chemicals has increased in both regulatory agencies and industry, and regulatory agencies have recommended the implementation of the 3Rs principle (refinement, reduction, and replacement). Considering these trends surrounding product development, the proposed screening would be useful for evaluating the in vivo phototoxic risk of chemicals in product development on a large scale.

Authorship Contributions

Participated in research design: Seto, Ohtake, Kato, Onoue.

Conducted experiments: Ohtake, Kato.

Contributed new reagents or analytic tools: Seto, Ohtake, Kato, Onoue.

Performed data analysis: Seto, Ohtake, Kato, Onoue.

Wrote or contributed to the writing of the manuscript: Seto, Ohtake, Onoue.

Footnotes

    • Received February 13, 2015.
    • Accepted May 21, 2015.

  • This work was supported in part by a Health Labour Sciences Research Grant from the Ministry of Health, Labour, and Welfare, Japan [H25-iyaku-wakate-024]; a grant from the Cosmetology Research Foundation [527]; and a grant from the Hoyu Science Foundation [31].

  • dx.doi.org/10.1124/jpet.115.223644.

Abbreviations

3T3 NRU PT
3T3 neutral red uptake phototoxicity test
BZ
benzophenone
BZP
benzophenone derivative
DMSO
dimethylsulfoxide
DO
dioxybenzone
EM
erythromycin
ESI-MS
electrospray ionization mass spectrometry
ICH
International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use
KT
ketoprofen
MEC
molar extinction coefficient
MRT
mean residence time
MX
mexenone
NaPB
sodium phosphate buffer
OECD
Organization for Economic Cooperation and Development
OX
oxybenzone
PIF
photoirritancy factor
PK
pharmacokinetic
QN
quinine HCl
ROS
reactive oxygen species
SB
sulisobenzone
UPLC
ultra-performance liquid chromatography

References

  1. ↵
    1. Abraham MH and
    2. Ibrahim A
    (2007) Blood or plasma to skin distribution of drugs: a linear free energy analysis. Int J Pharm 329:129–134.
    OpenUrlCrossRefPubMed
  2. ↵
    1. Allen MC,
    2. Shah TS, and
    3. Day WW
    (1998) Rapid determination of oral pharmacokinetics and plasma free fraction using cocktail approaches: methods and application. Pharm Res 15:93–97.
    OpenUrlCrossRefPubMed
  3. ↵
    1. Baughman BM,
    2. Stennett E,
    3. Lipner RE,
    4. Rudawsky AC, and
    5. Schmidtke SJ
    (2009) Structural and spectroscopic studies of the photophysical properties of benzophenone derivatives. J Phys Chem A 113:8011–8019.
    OpenUrlPubMed
  4. ↵
    1. Chuah SY,
    2. Leow YH,
    3. Goon AT,
    4. Theng CT, and
    5. Chong WS
    (2013) Photopatch testing in Asians: a 5-year experience in Singapore. Photodermatol Photoimmunol Photomed 29:116–120.
    OpenUrlCrossRefPubMed
  5. ↵
    1. Darvay A,
    2. White IR,
    3. Rycroft RJ,
    4. Jones AB,
    5. Hawk JL, and
    6. McFadden JP
    (2001) Photoallergic contact dermatitis is uncommon. Br J Dermatol 145:597–601.
    OpenUrlCrossRefPubMed
  6. ↵
    1. Epstein JH
    (1983) Phototoxicity and photoallergy in man. J Am Acad Dermatol 8:141–147.
    OpenUrlCrossRefPubMed
  7. ↵
    European Agency for the Evaluation of Medicinal Products, Evaluation of Medicines for Human Use, Committee for Proprietary Medicinal Products (EMEA/CPMP) (2002) Note for Guidance on Photosafety Testing, CPMP/SWP/398/01, EMA, London.
  8. ↵
    European Agency for the Evaluation of Medicinal Products, Evaluation of Medicines for Human Use, Committee for Proprietary Medicinal Products (EMEA/CPMP) (2008) Concept Paper on the Need for Revision of the Note for Guidance on Photosafety Testing, CPMP/SWP/398/01, EMA, London.
  9. ↵
    1. Food and Drug Administration, Center for Drug Evaluation and Research (FDA/CDER)
    (2002) Guidance for Industry, in Photosafety Testing, FDA, Silver Springs, MD.
  10. ↵
    1. Henry B,
    2. Foti C, and
    3. Alsante K
    (2009) Can light absorption and photostability data be used to assess the photosafety risks in patients for a new drug molecule? J Photochem Photobiol B 96:57–62.
    OpenUrlCrossRefPubMed
  11. ↵
    1. Infante Hernando L,
    2. Serra-Baldrich E,
    3. Dordal T, and
    4. Puig Sanz L
    (2013) Photoallergic contact dermatitis caused by benzophenones in magazine inks. Contact Dermat 69:124–126.
    OpenUrlCrossRefPubMed
  12. ↵
    International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH) (2014) ICH Guideline S10 Guidance on Phototoxicity Evaluation of Pharmaceuticals, ICH, Geneva, Switzerland.
  13. ↵
    1. Liebsch M and
    2. Spielmann H
    (2002) Currently available in vitro methods used in the regulatory toxicology. Toxicol Lett 127:127–134.
    OpenUrlCrossRefPubMed
  14. ↵
    1. Moore DE
    (1998) Mechanisms of photosensitization by phototoxic drugs. Mutat Res 422:165–173.
    OpenUrlCrossRefPubMed
  15. ↵
    1. Moore DE
    (2002) Drug-induced cutaneous photosensitivity: incidence, mechanism, prevention and management. Drug Saf 25:345–372.
    OpenUrlCrossRefPubMed
  16. ↵
    1. Nakazawa T,
    2. Shimo T,
    3. Chikamatsu N,
    4. Igarashi T,
    5. Nagata O, and
    6. Yamamoto M
    (2006) Study on the mechanism of photosensitive dermatitis caused by ketoprofen in the guinea pig. Arch Toxicol 80:442–448.
    OpenUrlCrossRefPubMed
  17. ↵
    1. Onoue S and
    2. Tsuda Y
    (2006) Analytical studies on the prediction of photosensitive/phototoxic potential of pharmaceutical substances. Pharm Res 23:156–164.
    OpenUrlCrossRefPubMed
  18. ↵
    1. Onoue S,
    2. Kawamura K,
    3. Igarashi N,
    4. Zhou Y,
    5. Fujikawa M,
    6. Yamada H,
    7. Tsuda Y,
    8. Seto Y, and
    9. Yamada S
    (2008) Reactive oxygen species assay-based risk assessment of drug-induced phototoxicity: classification criteria and application to drug candidates. J Pharm Biomed Anal 47:967–972.
    OpenUrlCrossRefPubMed
  19. ↵
    1. Onoue S,
    2. Seto Y,
    3. Gandy G, and
    4. Yamada S
    (2009) Drug-induced phototoxicity; an early in vitro identification of phototoxic potential of new drug entities in drug discovery and development. Curr Drug Saf 4:123–136.
    OpenUrlCrossRefPubMed
  20. ↵
    1. Onoue S,
    2. Hosoi K,
    3. Wakuri S,
    4. Iwase Y,
    5. Yamamoto T,
    6. Matsuoka N,
    7. Nakamura K,
    8. Toda T,
    9. Takagi H,
    10. Osaki N,
    11. et al.
    (2013) Establishment and intra-/inter-laboratory validation of a standard protocol of reactive oxygen species assay for chemical photosafety evaluation. J Appl Toxicol 33:1241–1250.
    OpenUrlPubMed
  21. ↵
    1. Onoue S,
    2. Kato M,
    3. Inoue R,
    4. Seto Y, and
    5. Yamada S
    (2014a) Photosafety screening of phenothiazine derivatives with combined use of photochemical and cassette-dosing pharmacokinetic data. Toxicol Sci 137:469–477.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    1. Onoue S,
    2. Hosoi K,
    3. Toda T,
    4. Takagi H,
    5. Osaki N,
    6. Matsumoto Y,
    7. Kawakami S,
    8. Wakuri S,
    9. Iwase Y,
    10. Yamamoto T,
    11. et al.
    (2014b) Intra-/inter-laboratory validation study on reactive oxygen species assay for chemical photosafety evaluation using two different solar simulators. Toxicol In Vitro 28:515–523.
    OpenUrlCrossRefPubMed
  23. ↵
    Organization for Economic Cooperation and Development (OECD) (2004) OECD Guideline for Testing of Chemicals, 432, In Vitro 3T3 NRU Phototoxicity Test, OECD, Paris.
  24. ↵
    1. Piérard GE and
    2. Piérard-Franchimont C
    (1993) Dihydroxyacetone test as a substitute for the dansyl chloride test. Dermatology 186:133–137.
    OpenUrlCrossRefPubMed
  25. ↵
    1. Seto Y,
    2. Onoue S, and
    3. Yamada S
    (2009) In vitro/in vivo phototoxic risk assessments of griseofulvin based on photobiochemical and pharmacokinetic behaviors. Eur J Pharm Sci 38:104–111.
    OpenUrlCrossRefPubMed
  26. ↵
    1. Seto Y,
    2. Inoue R,
    3. Ochi M,
    4. Gandy G,
    5. Yamada S, and
    6. Onoue S
    (2011) Combined use of in vitro phototoxic assessments and cassette dosing pharmacokinetic study for phototoxicity characterization of fluoroquinolones. AAPS J 13:482–492.
    OpenUrlCrossRefPubMed
  27. ↵
    1. Seto Y,
    2. Hosoi K,
    3. Takagi H,
    4. Nakamura K,
    5. Kojima H,
    6. Yamada S, and
    7. Onoue S
    (2012) Exploratory and regulatory assessments on photosafety of new drug entities. Curr Drug Saf 7:140–148.
    OpenUrlCrossRefPubMed
  28. ↵
    1. Spielmann H,
    2. Liebsch M,
    3. Döring B, and
    4. Moldenhauer F
    (1994) First results of an EC/COLIPA validation project of in vitro phototoxicity testing methods. ALTEX 11:22–31.
    OpenUrlPubMed
  29. ↵
    1. Westerhof W,
    2. van Hasselt BA, and
    3. Kammeijer A
    (1986) Quantification of UV-induced erythema with a portable computer controlled chromameter. Photodermatol 3:310–314.
    OpenUrlPubMed
View Abstract
Back to top

In this issue

Download PDF
Article Alerts
Email Article
Citation Tools
Research ArticleToxicology

Phototoxic Risk Assessments on Benzophenone Derivatives

Yoshiki Seto, Hiroto Ohtake, Masashi Kato and Satomi Onoue
Journal of Pharmacology and Experimental Therapeutics August 1, 2015, 354 (2) 195-202; DOI: https://doi.org/10.1124/jpet.115.223644
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Jump to section