The present study was a randomized, placebo-controlled, double-blinded study, with a parallel group test conducted by Derma Labo, Inc. This trial conformed to the Declaration of Helsinki (31) and was approved by the Ethics Committee of Tactics (Hokkaido Activation Center) on December 14, 2016 (approval no. 2016-100). Written informed consent was obtained from all potential participants prior to participant selection. Fig. 1 shows the test schedule. The test was conducted between January 16 and May 1, 2017.

The individual in charge of test materials allocation performed random allocation, such that the primary background factors were not biased amongst each group. The allocation of the test materials was performed based on an allocation table (key code) prepared from a random number table in advance by the individual in charge of the test materials allocation. The allocation table was sealed by the person in charge of test materials allocation and kept secret until opening. After confirming that all data was fixed, the individual in charge of test materials' allocation opened the key after confirming the sealing status of the allocation table and emergency key code. The individual in charge of test materials allocation submitted the allocation table after opening the key and created a record.

The purpose of the study was to clarify the effects of LFE ingestion on UVB-induced skin erythema. After screening, participants were recruited from among 118 individuals in the Exam's volunteer bank, which was managed by Derma Labo. Inc, who agreed to participate. Participants were Japanese nationals aged between 20 and 60 years of age.

Individuals were excluded from the study if any of the following circumstances applied: i) Pigmentation, inflammation, or other significant reactions were present at the skin site; ii) the participant had a history of photo sensitivity or presented symptoms of such; iii) the participant was suspected to be allergic to the test materials (for example, the participant had experienced allergic reactions to the applied food in the past); iv) sunburn had occurred in the skin site during the 3 months preceding the study; v) they had chronic skin conditions (such as atopic dermatitis) present at the skin site; vi) the participant was pregnant (or planned to get pregnant during the study) or lactating; vii) the participant had asthma or a similar chronic condition and regularly took medication for the condition; viii) the participant was taking medication that may affect the test results, including medication for freckles, UV-induced pigmentation, or liver spots (e.g., drugs containing L-cysteine, vitamin C, or tranexamic acid); ix) during the 3 months preceding the study, the participant was regularly taking (>5 days a week) skin whitening supplements or other nutritional supplement with claims of antioxidant properties (e.g., supplements rich in catechins, flavonoids, or polyphenols); x) the participant was currently undergoing, or had undergone during the past three months, a trial that involved the ingestion of food or drugs, the use of cosmetics, or something similar; xi) the participant was a regular smoker; xii) the participant had applied topical medication to the skin site during the past week; xiii) the participant intended to travel abroad or to swim in the sea during the study period (between screening and end of the study); xiv) the participant was undergoing hormone replacement therapy following the start of menopause or a menopausal disorder; and/or xv) the participant was otherwise deemed ineligible for the study by the physician-in-charge or principal investigator.

During the study, participants were required to adhere to the following rules and prohibitions: i) Participants were not allowed to apply topical medication to the skin site, take medication for freckles, UV-induced pigmentation, or liver spots (e.g., drugs containing L-cysteine, vitamin C, or tranexamic acid), or use non-medical drugs (officially known in Japan as ʻquasi-drugsʼ) or cosmetics. Participants were permitted to use personal hygiene products such as soap and shower gel, but they were not allowed to switch these products during the study period. If participants unavoidably needed to use the above medication, quasi-drugs, or cosmetics, they were instructed to report the product name, volume used, period used, and reason for use in a journal; ii) participants were not allowed to take medication (or apply topical medication) or take or apply newly designated ʻquasi-drugsʼ or traditional Chinese medicines; iii) participants were not allowed to ingest newly designated health foods or foods carrying health or nutritional claims (officially known in Japan as ʻfood for specified health usesʼ or ʻfood with function claimsʼ). If participants unavoidably needed to eat such foods, they were instructed to either report the product name, ingredients, and volume consumed in their journal or to affix the label (or a copy thereof) to their records. Participants were also instructed to report the name and amount consumed of any health food, ʻfood for specified health uses,ʼ or ʻfood with function claimsʼ that they were regularly consuming; iv) For a 2-week period preceding the study, participants had to refrain from any activity that could cause sunburn, such as participating in outdoor sports or using a tanning bed. Participants were asked to take steps to avoid UV exposure altogether (from both indoor and outdoor sources). They were instructed, for example, to wear sun-protective hats and clothing and to use sunscreen. If participants nonetheless suffered sunburn, they were instructed to report the time, duration of exposure, and severity (e.g., ʻskin red, feels soreʼ) in their journal; v) on the evening prior to a test day, participants were required to bathe or shower before going to bed. They were not allowed to bathe or shower before testing on the test day; vi) on a test day, participants were not allowed to engage in intense exercise until the test was over; vii) on a test day, participants were not allowed to consume any spicy food or drink such as curry, chilies, or hot sauce (e.g., Tabasco); vii) participants were not allowed to scrub the skin site with abrasive personal hygiene products; ix) for a 1-month period preceding the study, participants were not allowed to undergo beauty treatments such as a chemical peel or spa treatment; and x) participants were not allowed to start any regular supplement regimens.

The supplement used in this study was a tablet containing LFE, which was provided by Matsuura Yakugyo Co., Ltd., and consisted of microcrystalline cellulose, carmellose calcium, and calcium stearate. This supplement was comparable to a placebo containing the same components, except that the LFE was substituted with potato starch (Table I). A total of six LFE tablets (the daily dose) contained 900 mg LFE; the amount of rutin and chlorogenic acid contained in the extract of LFE in this human study was 926 and 876 µg/day, respectively, with 966 mg microcrystalline cellulose, 36 mg carmellose calcium, and 18 mg calcium stearate. A total of 6 placebo tablets contained 933 mg microcrystalline cellulose, 36 mg carmellose calcium, 18 mg calcium stearate, and 933 mg potato starch. The LFE and placebo tablets were supplied by Shiseido Pharmaceutical. The tablets were administered orally at a rate of six per day for 8 weeks.

The candidates reported their skin type with reference to the Fitzpatrick scale (32). Those with Type I (always burns, never tans) and Type II (usually burns, tans minimally) were selected for screening. Solar Light's Model 601 Multiport solar simulator was used to produce six stages of erythema on the selected candidates' backs at increments of ~20% (11.4, 15.0, 19.2, 21.9, 27.6, and 30.6 mJ/cm²). Each irradiated surface area was 0.5 cm² (Φ=8 mm). UV irradiance was measured using a Solar Light's PMA2100 radiometer and PMA2108 biologically weighted UV-B detector. The following day, the irradiated areas were examined to determine each participant's MED. Minimal tanning dose (MTD) was determined on the 7th day following irradiation. Based on the MED and MTD results, 44 candidates were selected for participation in the study. These candidates were randomly assigned to be divided into two groups to avoid bias in major background factors among the groups. Allocation of test materials was carried out based on the allocation table prepared in advance from the random number table. The assignment table was sealed by the person in charge of the trial food assignment and was strictly stored until opening.

Other screening processes included a lifestyle survey [with items on medical history, present symptoms, lifestyle, medication, use of cosmetics and supplements, and physical characteristics such as height (cm), weight (kg), BMI, systolic blood pressure (mmHg), diastolic blood pressure (mmHg), and pulse (beats/min)], a general biochemical blood test, a blood test for in vitro antioxidant potential test, and a questionnaire about diet (brief-type self-administered diet history questionnaire) (33).

Of the candidates judged eligible to participate by the physician-in-charge, 44 individuals were finally selected. All the candidates were highly UV-sensitive (8 men, 36 women; median age was 50 years old; age range 22 to 59 years).

During the study, MED was determined as follows. Before and after an 8-week course of the active agent or placebo (intervention), the skin sites were exposed to six stages of irradiation as in the screening. A total of 24 h after irradiation, the pigmentation in the testing sites was measured, and the pigmentation after intervention was compared with that before intervention.

Three skin sites on the back were irradiated. The irradiated UV intensity was 1.5x higher than that of the participant's MED. The participants were divided into two groups: An experimental (LFE) group and control (placebo) group. To control for MED differences, the participants were assigned to groups such that the groups had a similar average MED. Before intervention, pigmentation was measured four times: Before irradiation, and on days 1, 3, and 7 after irradiation. The participants then took their tablets for 8 weeks; those in the LFE group took the supplement containing LFE and those in the placebo group took the placebo. After the 8-week intervention, the participants underwent irradiation with the UV intensity at 1.5x their MED. As before, pigmentation was measured before irradiation, and on days 1, 3, and 7 after irradiation (Fig. 1).

The devices used to measure skin pigmentation were a Minolta CR-200 Chromameter (Konica Minolta), a Mexameter MX 16 (Courage + Khazaka Electric GmbH), and a C-Cube (Pixience SA) (34). The measurements were conducted in a room with a constant temperature (23±2˚C) and humidity (45±5%), and the participants were given 20 mins to acclimatize themselves to the room before measurement. Two skin hydration values were measured: Stratum corneum hydration (SCH) and transepidermal water loss (TEWL). The former was measured using a Corneometer CM825 (Courage + Khazaka Electric GmbH). Each irradiated site was measured five times and the average SCH was determined as arbitrary units (AU). Non-irradiated sites were measured as well to provide a control value.

The latter value, TEWL, was measured using a Tewameter TM300 (Courage + Khazaka Electric GmbH). The TM300 was placed on the target skin site for 15 secs and then activated for 20 secs. Five readings were taken after 10 secs following activation, and the average (where the deviation was least) was defined as the TEWL (g/hm²).

Blood samples were taken just before trial and at 8 weeks. In each case, the samples were taken before irradiation by a nurse under a physician's supervision. The blood sample was used to determine the in vivo antioxidative capacity. Each sample was collected using a vacuum blood collection tube with a separating agent. The blood was centrifugally separated for 15 min at 2,000 x g at room temperature. The separated blood was then stored at -30˚C before measurement. Antioxidant volume was used as a measure of antioxidative capacity. The blood samples were sent to Hoken Kagaku's Sapporo laboratory for biochemical and hematological analysis. Four markers of in-serum antioxidative capacity were used. The first was total glutathione disulfide (GSH+GSSG), measured using an OxiSelect total glutathione assay kit (cat. no. STA-312). The second was carbonylated protein, measured using an OxiSelect protein carbonyl ELSIA kit (cat. no. STA-310-T). The third was lipid peroxides, measured using an OxiSelect 8-iso-prostaglandin F2 α Elisa kit (cat. no. STA-337). These three OxiSelect devices were provided by Cell Biolabs. The fourth marker was 8-hydroxy-2-deoxyguanosine (8-OHdG), measured using a highly sensitive 8-OHdG check ELISA kit (cat. no. KOG-HS10E) available on the market by the Japan Institute for the Control of Aging (Nikken Seil Co. Ltd.).

The participants' daily diet was ascertained using the brief-type self-administered diet history questionnaire (BDHQ). The results were analyzed by the survey provider's support center (the ʻDHQ Support Centerʼ).

An immortalized human keratinocyte cell line (HaCaT) was used to analyze the antioxidative effect. HaCaT cells were obtained from Deutsches Krebsforschungszentrum with a material transfer agreement. The HaCaT cells were cultured using DMEM (Thermo Fisher Scientific Inc.) supplemented with 10% FBS (Thermo Fisher Scientific Inc.) and antibiotics (50 units/ml penicillin and 50 µg/ml streptomycin, both provided on the market by Fujifilm Wako Pure Chemical Industries). Alongside LFE, two reagents were used on the HaCaT cells: Chlorogenic acid and quercetin, both of which were provided by Sigma-Aldrich; Merck KGaA. To obtain the LFE, hydrous ethanol was dehydrated, and the remnants were dissolved in DMSO (final concentration 0.1% (v/v). The protective effect of LFE was measured in terms of cell viability after exposure to UVB. The HaCaT cells were treated with the reagent for 24 h 1 day after pre-culture. Then, the culture medium was replaced with Hank's Balanced Salt Solution (Thermo Fisher Scientific Inc.) and exposed to 50 mJ/cm² UVB. The cells were incubated for an additional 24 h after which the cell viability was determined using an MTT (FUJIFILM Wako) assay.

Glutathione (GSH), a marker of intracellular antioxidative capacity, was measured using the DTNB method (Total Glutathione Quantification Kit, Dojindo Molecular Technologies, Inc.). A total of 24 h after inoculation, the cells were treated with LFE for 24 h. Cells were then collected, and the intracellular GSH was measured.

Statistical analysis was performed using Microsoft Excel 2013 (Microsoft Corporation). A paired Student's t-test was used for intragroup comparisons. Multivariate regression analysis was used to compare changes in MED between the two groups. Covariance analysis was used for erythema time series results, including for time point comparisons. An unpaired Student's t-test was used for the corresponding time points. All values are presented as the mean ± SD. P<0.05 was considered to indicate a statistically significant difference.

This study was registered at the University Hospital Medical Information Network (registration no. UMIN000025593).

All 44 healthy volunteers selected in the screening process participated in this study. However, six participants were subsequently discontinued; four were voluntarily withdrawn, one was excluded after taking medication, and another was excluded after the MED results indicated a concerning reaction to UV exposure. The final trial consisted of 38 remaining participants, 20 of whom were in the placebo group and 18 of whom were in the experimental group. There were no significant inter-group differences in these items (the placebo group's average MED was 17.5±4.1 mJ/cm², and the LFE group's was 18.2±3.8 mJ/cm²; the placebo group's average BMI was 22.5±3.1 kg/m², and the LFE group's was 22.6±3.5 kg/m²). There were no harmful effects of test material based on general biochemical blood tests.

Table II shows the placebo and LFE groups' MED values at week 0 (pre-intervention) and week 8 (post-intervention), as well as the change between the two-time points. The LFE group's average MED was significantly higher post-intervention (P=0.002). The week 8 figure was 20.6±3.9 mJ/cm², which marks a 13% increase from the week 0 figure of 18.2±3.8 mJ/cm². There was no significant change in the placebo group's MED (P=0.38). The week 0 figure was 17.5±4.1 mJ/cm² and the week 8 figure was 18.0±5.1 mJ/cm². The degree of change significantly differed between the groups (P=0.02). In the LFE group, MED changed by 2.4±2.6 mJ/cm², while in the placebo group, it only changed by 0.5±3.5 mJ/cm².

This section outlines the results for the erythemas that formed at the three sites on the participants after exposure to UV at 1.5x their MED. The data included 20 placebo participants and the 18 remaining participants in the LFE group. Table III shows the changes in pigmentation over the four time points before intervention (Day 0, Day 1, Day 3, and Day 7) and after intervention (Day 63, Day 64, Day 66, and Day 70). The a* values indicate CR-200 readings, while ʻmelanin indexʼ indicates MX 16 readings. Table IV shows the changes in erythema index as measured by the C-Cube. Fig. 2 shows the changes before and after intervention. In Tables IV and V, values in parentheses indicate change relative to the pre-irradiation figure (Day 0/Day 63), which was scaled at 100.

As shown in Table III, the a* value peaked at the first day (Day 1/Day 64) and declined over the next two time points. Conversely, the melanin index rose across the four time points. The LFE group did not differ significantly from the placebo group in terms of the score differences for either of the two indices before and after intervention.

As shown in Table IV, the C-Cube-measured erythema index rose between Day 0/Day 63 and Day 1/Day 64 before declining. Before intervention, the LFE group's results for the C-Cube-measured erythema index did not differ significantly from that of the placebo group (P=0.72). After intervention however, the LFE group's results for the C-Cube-measured erythema index differed significantly from that of the placebo group (P=0.001). On Days 66 and 70 (the 4^th and 7^th days after irradiation), the placebo group's values were significantly higher than that of the LFE group. In addition, UV exposure also affected SCH and TEWL, but the LFE group's post-intervention results did not significantly differ from that of the placebo group after 8 weeks of LFE ingestion (data not shown).

Table V shows the results for the four markers of in-serum antioxidative capacity. Total GSSG increased in both groups after intervention, but not significantly. In the placebo group, total GSH was 3.09±1.67 nM at pre-intervention and 3.27±1.44 nM at post intervention; in the LFE group, it was 3.04±1.68 nM at pre-intervention and 4.17±3.38 nM at post intervention. While 8-OHdH in serum were significantly suppressed after intervention in both the Placebo and LFE groups, there were no differences between the two groups. Carbonylated protein measurements showed no differences before or after intervention, or between groups. Carbonylated protein levels declined following intervention in both groups, but the decline was not significant. The remaining two markers, lipid peroxides and 8-iso-pro F2 alpha, were significantly lower after intervention, but there was no significant intergroup difference.

The HaCaT, a spontaneously transformed human keratinocyte cell line derived from the epidermis was used to determine the protective effects of LFE as well as chlorogenic acid (a major component of LFE) and quercetin (an intestinal metabolite of rutin, another major component of LFE). Pretreatment with LFE for 24 h exhibited a significant protective effect on the UVB-induced cytotoxicity at a concentration of 50 µg/ml, but this effect decreased at 1,000 µg/ml (Fig. 3A). Similarly, chlorogenic acid showed a significant effect at 5 µM, but this effect was absent at 100 µM (Fig. 3B). Quercetin also exhibited an inhibitory effect from 0.5-5 µM (Fig. 3C). The pretreatment of LFE, chlorogenic acid or quercetin significantly inhibited the hydrogen peroxide-induced cytotoxicity (data not shown). Thus, chlorogenic acid and quercetin as well as LFE are effective for inhibiting cell damage resulting from UV exposure, possibly through their antioxidative effects.

Fig. 4 shows the effect of LFE upon intracellular total GSH, which was determined using the DTNB method. Quercetin increased total GSH at concentrations of 2.5 and 10 µM, LFE and chlorogenic acid showed no significant effect for increasing total GSH in HaCaT cells, even though LFE has been reported to enhance the cellular GSH level in hepatocytes (30).

In the BDHQ survey, we asked the participants how much tea and coffee they consumed. The responses indicated that participants, who consumed plenty of tea or coffee prior to the start of the test, tended to frequently have these beverages in the second half of the study (at 8 weeks). Based on this, we assumed that the increasing response in serum total GSH levels comes from the participants who consumed low amounts of tea or coffee per day. There were 10 participants in the LFE group who consumed <220 g of tea or coffee a day and 12 such participants in the placebo group (total GSH: from 2.7±1.6 to 4.7±3.2 nM). The increase in serum GSH in the 10 LFE participants (total GSH: from 2.7±1.7 to 3.2±1.4 nM) was not significantly greater than in the placebo group (P=0.11). It would be necessary to control tea/coffee consumption to confirm the cause-effect relationship.