False negatives in the modified ESR-based photosafety test (ESR-PT) caused by ultraviolet-C light

Abstract

The modified electron spin resonance (ESR)-based photosafety test (ESR-PT) is a non-animal prediction test using ESR spectroscopy that is applicable to hydrophobic and colored chemicals and the results show high concordance with existing photosafety reference information. The modified ESR-PT is based on the detection of singlet oxygen and free radical photoproducts generated from chemicals in the presence of 4-hydroxy-2,2,6,6-tetramethyl-piperidine (4-hydroxy-TEMP). We obtained false-negative results caused by signal increment of the control solution, which is used as the denominator of the classifier in the modified ESR-PT, when we used a different type of lighting unit (type I) to the previously used type of lighting unit (type II). Spectral measurement of the irradiated light from the light sources revealed that the type I lighting unit emitted stronger UV-C light than the type II lighting unit. Consequently, as UV-C absorption of 4-hydroxy-TEMP (λ_max<210 nm) was confirmed, we repeated the modified ESR-PT using a type I lighting unit equipped with a UV-C cut filter, which led to an apparent decrease of the signal increment in the control solution, and all the false-negative-judged chemicals correctly tested positive but a false positive result was also noted. Therefore, installation of a UV-C cut filter in the lighting unit in modified ESR-PT appears to be a reliable solution for avoiding UV-C light-mediated false-negative results. However, it may also be necessary to reconsider the classifier used in ESR-PT to avoid obtaining false-positive results when using a UV-C cut filter.

INTRODUCTION

Exposure to ultraviolet (UV) or visible light irradiation after oral intake or dermal application of phototoxins or photoallergens is known to cause phototoxicity or photoallergic reactions. These reactions are sometimes caused by cosmetic and pharmaceutical products via formation of singlet oxygen or free radicals (Moore, 1998), so that photosafety evaluation of these consumer products is important to prevent phototoxic or photoallergic reactions to these products. The electron spin resonance (ESR)-based photosafety test (ESR-PT) was developed as a method using ESR spectroscopy with 4-hydroxy-2,2,6,6-tetramethyl-piperidine (4-hydroxy-TEMP) as the spin trap reagent (Hinoshita et al., 2021). When phototoxic or photoallergic chemicals are exposed to UV or visible light, they generate singlet oxygen, which reacts with 4-hydroxy-TEMP to form a stable radical, 4-hydroxy-2,2,6,6-tetramethyl-piperidine-1-oxyl (TEMPOL), which is detected by the ESR spectrometer as a triplet signal. The method was further improved to detect free radical photoproducts generated from chemicals simultaneously by introducing a new classifier, the photoreactivity index (PRI). The modified ESR-PT showed a prediction accuracy of 91.1% as compared with the existing photosafety reference information, and applicability of 100% to a variety of chemicals, including hydrophobic and colored chemicals, and was confirmed as being better than other non-animal photosafety tests, such as the reactive oxygen species (ROS) assay (Hinoshita et al., 2024).

Light source is an indispensable element in ESR-PTs. The prototype ESR-PT and modified ESR-PT were developed using the xenon lamp with a collimator lens as the light source. However, we found that the modified ESR-PT yielded false-negative results for the well-known phototoxic chemicals, acridine and 8-methoxypsoralen, because of signal increment in the control solution when a xenon lamp with a quartz light guide was used. This report describes the causes of the false-negative results for these chemicals yielded by the modified ESR-PT performed using the xenon lamp with a quartz light guide, and possible solutions that could be applied to general light sources used in ESR-PT for avoiding UV-C light-mediated false-negative reactions.

MATERIALS AND METHODS

Anthracene, chlorpromazine HCl, quinine HCl, acridine and 8-methoxypsoralen were used as positive chemicals, and sulisobenzone was used as a negative chemical based on the photosafety reference data obtained from humans and the in vitro 3T3 neutral red uptake phototoxicity test (ROS Assay Validation Management Team, 2013), and the data reported from the modified ESR-PT (Hinoshita et al., 2024). Anthracene and quinine HCl were obtained from FUJIFILM Wako Pure Chemical (Osaka, Japan), acridine, 8-methoxypsoralen and 4-hydroxy-2,2,6,6-tetramethyl-piperidine (4-hydroxy-TEMP) were obtained from Sigma-Aldrich (St. Louis, MO, USA), and chlorpromazine HCl and sulisobenzone were obtained from Tokyo Chemical Industry (Tokyo, Japan). Ethanol (analytical grade) was purchased from FUJIFILM Wako Pure Chemical (Osaka, Japan).

Two types of lighting units, a 150 W xenon lamp (LC8, L9566-04; Hamamatsu Photonics, Shizuoka, Japan) with a quartz light guide (Type I) and 500 W xenon lamp with a collimator lens (UXL-500SX; Ushio, Tokyo, Japan) (Type II), that have been previously used in the modified ESR-PT were tested in this study. In addition, an optical filter (LU0275; Asahi Spectra, Tokyo, Japan) was used to evaluate the effect of UV-C light (<275 nm) on the results of the modified ESR-PT.

The ESR spectrum measurement was performed under the same experimental conditions as those described previously (Hinoshita et al., 2024). Test chemicals were prepared at concentrations of 10 mmol/L with 100 mmol/L 4-hydroxy-TEMP in ethanol; the test mixtures were transferred into a flat quartz ESR cell (FMC-LC07; Flashpoint, Tokyo, Japan) and irradiated at a total light dose of approximately 20 J/cm², as previously reported (corresponding to a UV-A dose of 1 J/cm²) from the xenon lamp. Then, the ESR spectra were recorded with an ESR spectrometer at room temperature before and after irradiation (Magnettech ESR5000; Bruker, Massachusetts, USA). The ESR measurements were conducted under the following condition: microwave frequency, 9.45 GHz; microwave power, 4.0 mW; modulation width, 0.1 mT; magnetic field, 336 ± 5 mT; sweep width, 30 sec; and accumulation, 4. The control solution (ethanol containing 100 mmol/L 4-hydroxy-TEMP) was also measured in the same manner. Then, the photoreactivity index (PRI) was calculated using the equation shown in Fig. 1. The PRI is a parameter based on the intensity of all signals observed in the magnetic field between the third and fourth signals of the Mn²⁺ marker, and it aids the detection of various free radical photoproducts in addition to the singlet oxygen. The test mixtures were measured in triplicate on separate days concurrently with the control solution used to calculate the PRI. The mean PRI and standard deviation were then determined from the three values. The cutoff value (=2.0) determined in a previous study (Hinoshita et al., 2024) was used to discriminate between a positive and negative judgement in the modified ESR-PT.

Fig. 1

Schematic diagram of calculation of the PRI in the modified ESR-PT. Relative intensity (RI) can be obtained by dividing the intensity obtained from the third signal to the fourth signal of the Mn²⁺ marker (I₁) by the intensity of the third signal of the Mn²⁺ marker (I₀). The control solution is ethanol containing 100 mmol/L 4-hydroxy-TEMP without test chemicals.

Spectral patterns of two types of lighting units were measured with a CCD spectrometer (C10082CA; Hamamatsu Photonics, Shizuoka, Japan) in the range of 200 nm to 800 nm.

UV/visible light absorption spectrum was recorded with a quartz cell (10 mm pathlength) for 4-hydroxy-TEMP ethanol solution (2 mmol/L) from 210 nm to 800 nm by using a spectrophotometer (UV-2600; Shimadzu, Kyoto, Japan).

RESULTS AND DISCUSSION

The modified ESR-PT using the type I lighting unit was performed for five well-known positive chemicals and a negative chemical (Table 1). When the PRI cutoff value (=2.0) was used, three out of the five positive chemicals tested positive, whereas the remaining two chemicals, acridine and 8-methoxypsoralen, tested negative, even though all the five positive chemicals tested positive with the test performed using the type II lighting unit in a previous study (Hinoshita et al., 2024). The negative chemical, sulisobenzone, tested negative with both types of lighting units.

Table 1. Outcomes of the modified electron spin reosnance-based photosafety test (ESR-PT) obtained with the type I and the type II lighting unit.

All the PRIs obtained with the type I lighting unit were relatively low as compared with those obtained with the type II lighting unit, ranging from one-tenth to one-half, likely due to the relatively high ΔRI in the control solution, which is used as the denominator in the calculation of the PRI.

The spectral patterns of the two types of lighting units were measured to confirm the difference between the type I and type II lighting units (Fig. 2). The type I unit had a similar spectrum to the type II unit, but relatively high emission was observed in the UV-C region (from 220 nm to 250 nm). To consider the effect of UV-C light on formation of TEMPOL in the control solution, the UV/visible absorption spectrum of 4-hydroxy-TEMP (2 mmol/L) in ethanol was recorded, and 4-hydroxy-TEMP showed an absorption band in the UV-C region (λ_max<210 nm) (Fig. 3). Upon excitation, secondary saturated amines undergo photolysis including dissociation of the N—H bond (Halpern, 1982), which may lead to the formation of various products. For example, tinuvin 770, which is a light stabilizer for polymer materials having a hindered piperidine structure similar to 4-hydroxy-TEMP, was reported to generate a nitroxyl radical in aerated benzene solution after UV irradiation (Lucki et al., 1984). Therefore, UV-C light contained in the irradiated light may trigger the formation of TEMPOL from 4-hydroxy-TEMP. In addition, the UV-C intensity from the lighting unit may vary depending on the type of apparatus or an individual product, even if the UV-A dose is the same. Consequently, the stronger UV-C light emitted from lighting unit I as compared with that from lighting unit II used in the previous study could lead to a high ΔRI of the control solution, resulting in low PRIs.

Fig. 2

Spectral patterns of the xenon lamp adopted in this study (LC8) provided by Hamamatsu Photonics (type I, solid line), its filtered light to remove the UV-C light (dashed line) and previously reported (UXL-5005SX) provided by Ushio (type II, dotted line). The inset is the spectra in an enlarged ordinate scale.

Fig. 3

Absorption spectrum of 4-hydroxy-TEMP (2 mmol/L) in ethanol from 210 nm to 800 nm. The inset is the chemical structure of 4-hydroxy-TEMP.

Since UV-C light may cause false negatives by the production of TEMPOL in the control solution, we considered that installing a UV-C cut filter might be effective for avoiding false negatives. Therefore, we adopted an optical filter to cut UV-C light from the xenon lamp (Fig. 2) and repeated the ESR-PT measurements. This results in a marked decrease of the ΔRI of the control solution from 0.81±0.15 to 0.01±0.01 (Table 2). This result suggests that UV-C light contributes to the generation of TEMPOL from 4-hydroxy-TEMP even in the absence of any test chemical in ethanol. The mean PRIs under the filtered light became approximately 80 times higher than the values under unfiltered light from the type I lighting unit, since the ΔRIs of the test mixtures are divided by the value of the control solution in the calculation of the PRI. Basically, the classifier, such as PRI, used in the ESR-PT is based on the detection of singlet oxygen or free radicals in the presence of 4-hydroxy-TEMP. Therefore, installation of a UV-C cut filter would be necessary to avoid the effect of UV-C interference with 4-hydroxy-TEMP. When we used the UV-C cut filter, all the chemicals, including sulisobenzone, tested positive. Sulisobenzone is a negative control chemical for the ESR-PT and ROS assay, so that this result should be considered as a false positive caused by the extremely low ΔRI value of the control solution.

Table 2. Outcomes of the modified electron spin reosnance-based photosafety test (ESR-PT) obtained with and without the optical filter to remove UV-C light from the type I lighting unit.

In this present study, we examined typical positive and negative chemicals by the modified ESR-PT using a xenon lamp with a quartz light guide (type I) and obtained false negative results for the chemicals, acridine and 8-methoxypsoralen, because of the high ΔRI of the control solution. Analysis of the spectral patterns revealed relatively high emission in the UV-C region (from 220 nm to 250 nm) from the type I lighting unit as compared with the type II lighting unit reported from a previous study. Since we confirmed UV-C absorption by 4-hydroxy-TEMP (λ_max<210 nm), we considered that exposure to UV-C might result in the production of TEMPOL from 4-hydroxy-TEMP via photolysis. Consequently, we repeated the modified ESR-PT using the type I lighting unit equipped with a UV-C cut filter, which resulted in a significant decrease of the ΔRI of the control solution and positive results for all the chemicals tested. Considering the above findings, installation of a UV-C cut filter in the lighting unit in the modified ESR-PT is a feasible solution for avoiding UV-C light-mediated false negative results. These modifications could elevate the modified ESR-PT as a light-source-independent method. However, the classifier in the ESR-PT would also require to be reconsidered, such as the ratio of the ΔRI of the test mixture to that in not the control solution but in the positive control, to avoid false positive results obtained with the use of the UV-C cut filter.

Funding

No funding was provided for the work.

Conflict of interest

The authors declare that there is no conflict of interest.

Data availability

The data in this study are included in the article/supplementary materials. Contact the corresponding author(s) directly to request the underlying data.

Author Contribution Statement

Conceptualization: Masumi Hinoshita, Yosuke Maeda

Investigation: Masumi Hinoshita

Supervision: Yosuke Maeda

Visualization: Masumi Hinoshita

Writing – original draft: Masumi Hinoshita

Writing – review & editing: Akio Kawai, Masahiro Takeyoshi

Ethics approval and consent to participate

Not applicable.

Patient consent for publication

Not applicable.

REFERENCES

• Halpern, A.M. (1982). The spectroscopy, photophysics and photochemistry of saturated amines. In: The Chemistry of Amino, Nitroso, and Nitro Compounds and Their Derivatives. (Patai, S. ed.), pp.155–180, Wiley, New York.
•  Hinoshita,  M.,  Abe,  T.,  Sato,  A.,  Maeda,  Y. and  Takeyoshi,  M. (2021): Development of a new photosafety test method based on singlet oxygen generation detected using electron spin resonance. J. Appl. Toxicol., 41, 247-255.
•  Hinoshita,  M.,  Abe,  T.,  Sato,  A.,  Maeda,  Y. and  Takeyoshi,  M. (2024): Modified ESR-based photosafety test (ESR-PT) detecting singlet oxygen and free radical formation. J. Appl. Toxicol., 44, 651-662.
•  Lucki,  J.,  Rabek,  J.F.,  Rånby,  B. and  Dai,  G.S. (1984): ESR study of interaction between hindered piperidine stabilizers and low molecular weight ketones, hydroperoxides and peroxides. Polym. Photochem., 5, 385-409.
•  Moore,  D.E. (1998): Mechanisms of photosensitization by phototoxic drugs. Mutat. Res., 422, 165-173.
• ROS Assay Validation Management Team. (2013). Validation report for the international validation study on ROS (Reactive Oxygen Species) assay as a test evaluating phototoxic potential of chemicals (Atlas Suntest version), http://www.jacvam.jp/files/news/ROS_Assay_full%2020130920%20atlas_fourth%20data.pdf