A total of 49 male Sprague-Dawley rats (age: 8 weeks, mean weight: 180 ± 20 g) purchased from Firat University Experimental Research Center (FUDAM) were used in the study. The rats were maintained in a controlled environment with a 12:12 light-dark cycle, room temperature of 22 ± 2 C and humidity of 55% ± 10, and had ad libitum access to feed and water. All experiments were carried out in accordance with the standard ethical guidelines for laboratory animal use and care as defined in the European Economic Community rules (Anonymous, 2010) and with the approval of the Animal Ethics Committee of Firat University (2018/11-126).

The dorsal hair of rats was clipped under anesthesia (2 cm in width x 6 cm in length) prior to the experiment. Forty-nine rats were randomly divided into seven groups with seven rats each, as shown in Supplementary Table S1) NC (normal control; no UVB radiation), 2) SC (shaved control; shaved and no UVB radiation), 3) UVB group (shaved and received UVB radiation), 4) ASI + MgB-L [shaved, received UVB radiation, ASI (140 mg human equivalent dose (HED) = 2.2 mg ASI/rat/d) and low dose MgB (1.5 mg HED of biotin = 24°μgMgB/rat/d], 5) ASI + MgB-H [shaved, received UVB radiation, ASI and high dose MgB (10 mg HED of biotin = 160 μg MgB/rat/d)], 6) ASI + MgB-L + MgB-C [shaved received UVB radiation, ASI, low dose MgB and MgB cream application (2% Mg)], 7) ASI + MgB-H + MgB-C (shaved, received UVB radiation, ASI and high dose MgB and MgB cream application).

The diet used in the study was adapted to comprise spray-dried egg whites as the single protein source. Avidin protein in egg white binds approximately 1.44 mg biotin/kg purified diet, inhibiting biotin absorption (Lewis et al., 2001). All treatments (ASI and MgB) were supplemented daily as an oral supplement, and UVB irradiation was performed five times per week for ten weeks. The normal control shaved control, and UVB rats were orally administered with saline. The oral and topical dosages of both agents were determined as proposed in the literature (Rathman et al., 2003; Kalman et al., 2015; Nourbakhsh et al., 2016).

ASI contains 49.47% arginine, 8.2% silicon, and 25% inositol. Arginine (3.8 g, 21.8 mmol) was added to a vigorously stirred solution of inositol (1.25°g, 6.9 mmol) in potassium silicate [5°ml, 29.8°Be, 8.3% K₂O (0.52°g, 5.5 mmol), 20.8% SiO₂ (1.3°g, 21.8 mmol)]. The ASI complex and MgB were provided by JDS Therapeutics, LLC (Harrison, NY, United States ). Both ASI and MgB were dissolved in drinking water and were administered via oral gavage. In the control group, saline was administered by oral gavage after being dissolved in drinking water. Oral gavage and 2% MgB cream were administered 2°h and 30 min prior to the experiment, respectively (Supplementary Table S1).

For the photoaged rat model, rats were exposed to UVB lamps with UVB with emission spectrums between 290 and 320 nm (peak at 310–315 nm) (Waldman UV800, Germany) and UVA emissions in the range of 320–400 nm (peak at 365) four times per week as described in an earlier study (Fan et al., 2015).To determine the minimal erythema dose (MED) of rats in each group, rats were exposed to UVB at doses of 20, 30, 40, 60, 80, and 100 mj/cm² (approximate dosage increase = 40%) on a total area of 24 cm² on the lower back, which comprised a total of six sites measuring 2 cm× 2 cm. MED was accepted as the amount of UV radiation that produced a minimal erythematous reaction with well-demarcated borders 24 h after UVB exposure. Prior to the experiment’s initiation, the average MED of the seven rats in each group was accepted as the baseline MED for each group. Throughout UVB exposure, the rats were kept in the incubator at a 20–30 cm distance from the UVB lamp, and UVB exposure was performed four times a week. The initial radiation dose was 1 MED in the first and second weeks and was increased by 0.2 MED/week over the remaining 10 weeks (range between weeks 3 and 1=1.2–2.6 MED). UV energy was detected with a UV radiometer. The same procedure was performed in the non-irradiated control group; however, the UVB was switched off. ASI and MgB were administered by oral gavage 2 h prior to the study, and topical 2% MgB cream was administered 30 min before the study.

After successful modeling, the rats were anesthetized; their dorsal skin was photographed. At the end of the experiment, all the rats were sacrificed by cervical dislocation, and blood and dorsal skin samples removed from the irradiated area were collected for further analyses. Blood samples were centrifuged at 3000 ×°g for 10°min, and the serum was carefully removed for further analysis. The sera and tissue samples were stored in a deep freeze (Hettich, Germany) at 80°C until analysis. The skin of the back of the animals was observed daily during the irradiation period.

A blinded observer assessed noticeable variation of rats’ dorsal skin according to assessment criteria shown in Supplementary Table S2 (Kong et al., 2015). Skin elasticity was measured by performing the pinch test according to the adapted protocol described by Tsukahara et al. (Tsukahara et al., 2005). The dorsal skin at the midline of the rat was picked up with fingers as much as possible until the lower limbs were almost off the ground, and then the rat was released from the fingers. Skin elasticity was defined as skin recovery time between the onset and disappearance of the pinch, whereby longer recovery time was accepted to indicate lower skin elasticity (Supplementary Figure S3).

Serum glucose, urea-N (BUN), and creatinine concentrations and activities of alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were analyzed with an automated chemistry analyzer (Samsung LABGEO PT10V, Samsung Electronics Co., Suwon, Korea). For hydroxyproline analyses, the dorsal skin tissues were homogenized using 2 ml of 1 N acetic acid, and the homogenates were centrifuged (3.000 × g, 10 min). The hydroxyproline levels were measured in the supernatants using a commercially available assay kit according to the manufacturer’s instructions (Diagnostic Systems Laboratories, TX, United States ). The intra- and inter-assay coefficients of variations (CV) were 4.3 and 6.5%. The tissue was also homogenized and solubilized in 0.1% Triton-X 100, 0.2 M Tris–HCl (pH 8.0) buffer, followed by ultrasonication and by centrifugation (2000 g × 20 min) to obtain supernatants for the elastase enzyme assay. The activity was detected using a commercially available assay kit (Diagnostic Systems Laboratories, TX, United States ) according to the manufacturer’s instructions. The inter-and intra- assay constants were 3.6 and 6.4%. Skin activity of biotinidase was detected using the colorimetric method by measuring p-aminobenzoate liberation from N-biotinyl-p-aminobenzoate as described in detail by Wolf et al. (Wolf et al., 1983). Biotinidase activity is expressed as mU/100 mg protein.

To determine the concentrations of serum levels of Mg, silicon (Si), iron (Fe), copper (Cu), and zinc (Zn) in the sera and tissue samples, 0.3 g of skin and 0.5 ml of serum samples were digested with 5 ml concentrated nitric acid in a Microwave Digestion System (Berghof, Eningen, Germany) for 30 min and diluted 1∶10 with distilled deionized water before analysis. Mineral levels in each sample were measured by flame AAS (AAS, Perkin-Elmer, Analyst 800, Norwalk, CT, United States ) via documented and fully established processes at the 285.2, 248.3, 213.9, 324,8 nm wavelength for Mg, Fe, Zn, and Cu, respectively. The method was confirmed with specialized reference materials, and the precision was 98%. Mineral concentrations in tissues were expressed as wet weights (Sahin et al., 2018).

The serum biotin was analyzed by HPLC (Shimadzu, Kyoto, Japan) as earlier defined, with minor modifications (Lewis et al., 2001; Grotzkyj Giorgi et al., 2012). The reversed-phase column used was a C18-ODS-3 column (250 × 4.6 mm, 5 m), and the biotin-containing chromatography fractions were dried under a stream of nitrogen before the assessment. Serum arginine concentrations were examined by high-performance liquid chromatography (HPLC, Shimadzu, Japan) as defined by Pieper and Dondlinger (Pieper and Dondlinger, 1997). Sera samples were extracted 1:1 in 35% (wt/vol) sulfosalicylic acid dihydrate. After mixing and centrifugation, the supernatant was mixed 1:1 with lithium-D buffer before analysis. The amino acid standard was obtained from Sigma-Aldrich Chemicals (St. Louis, United States ).

For malondialdehyde (MDA) analyses, skin tissues were homogenized in ice-cold phosphate buffer solution for 5 min using both an ultrasonic (Vibra-Cell 130 VCX; Sonics & Materials, Inc., Newtown, CT) and a mechanic (IKA Works, Inc; Wilmington, NC) homogenizers and then centrifuged at 7,000 g for 15 min and the protein content in the supernatant was determined using nanodrop spectrophotometry (MaestroGen, Las Vegas, NV, United States ). Then, an HPLC apparatus of Shimadzu UV–vis SPD-10 AVP detector, a CTO-10 AS VP column, and 30 mM KH2PO4 and methanol (82.5: 17.5, v/v, pH 3.6) at a flow rate of 1.2 ml/min were used (Shimadzu, Japan). Column waste was monitored at 250 nm.

Dorsal skin protein levels of biotin-dependent carboxylase including ACC1, ACC2, PC, PCC, MCC, and mammalian target of rapamycin 2 (mTOR2; ser2481), p-p70S6K, p4E-BP1, vascular endothelial growth factor (VEGF), sirtuin 1 (SIRT1), matrix metalloproteinase-1 and -3 (MMP-1 and MMP-3), TNFα, NFκB, IL-6, IL-8, COX-2, Bax, Bcl-2 and caspase-3 were measured using the Western blotting technique as defined earlier (Tong et al., 2015). Skin tissues were pooled and homogenized at 4 °C in an extraction buffer and centrifuged at 13,000 × g for 20 min at 4 °C. Protein samples were separated using 10% SDS-Page and transferred onto nitrocellulose membranes for 1 h prior to applying primary antibodies. The following primary antibodies were used: ACC1, ACC2, PC, PCC, MCC, p-mTORC2, p-p70S6K, p4E-BP1, VEGF, SIRT1, MMP-1, MMP-3, TNFα, NFκB, IL-6, IL-8, COX-2, Bax, Bcl-2, caspase-3, AP-1 (p-c-Jun and p-c-Fos), p-p38 MAPK and β-actin (Santa Cruz Biotechnology, CA, United States ). After washing, the membranes were incubated with secondary goat anti-mouse antibodies (Santa Cruz Biotechnology) in Tris-buffered saline containing 0.05% Tween 20 for 1 h, and protein levels were measured densitometrically.

Dorsal skin samples (approximately 1 × 0.4 cm in size) were quickly collected at the end of the study and were fixed in a 10% formaldehyde solution, fixed in paraffin, and then cut into sections of 5 µm thickness, as instructed in the literature (Kong et al., 2013). The slides were stained with hematoxylin and eosin (H&E) for routine histological examination and histological assessment of epidermal hyperplasia. The slides were also stained with Masson’s trichrome and Verhoeffelastic stain to evaluate skin elasticity. For epidermal hyperplasia assessment, the epidermal thickness was photographed using an Olympus DP73 camera, and mean epidermal thickness was calculated based on 10 random site measurements on each slide using an optic microscope (Olympus BX53). Fine lines and wrinkles, two of the most important signs of photoaging, were evaluated both macro- and microscopically by examining microfold formation in the tissues stained with hematoxylin and eosin. Solar elastosis is another principal histological sign of photoaging, characterized as dermal collagen breakdown and accumulation of elastotic material, a bluish stain with the appearance of elastin fibers dermal infiltration of inflammatory cells. These changes in collagen were evaluated using Masson’s trichrome stain and Verhoeff elastic stain. The intensity of inflammatory cell infiltration was expressed as the number of cells per mm².

For immunohistochemistry, collected skin tissues were fixed in 10% neutral formalin for 48 h at room temperature. Subsequently, tissue samples were embedded in paraffin and cut into sections (5 μm thick) using a microtome. The sections were then fixed onto slides. The sections were stained using rat-specific MMP-1 (Thermo Fisher Scientific, WA, United States ), COX-2, IL-6, and mTOR (Abcam, Cambridge, MA, United States ) antibodies with the help of an automatic immunohistochemical staining device (Ventana, Tucson, AZ, United States ). We performed quantitative analyses of the immunoreactivities against MMP-1, COX-2, IL-6, and mTOR antibodies using a light microscope (Olympus BX53) equipped with a digital camera (Olympus DP73) and connected to a PC monitor. For MMP-1, COX-2, IL-6, and mTOR, the staining (0, 0%; 1, <25%; 2, 25–50%; 3, 51–75%; and 4, >75%) were scored.

Statistical analyses were performed using SPSS for Windows version 21.0 (IBM Corp., Armonk, NY, United States ), with a one-way analysis of variance or Kruskal-Wallis test. Then Tukey post hoc test or Mann Whitney U test were used for multiple comparisons. A p value of <0.05 was considered significant.

As shown in Table 1, there were no significant changes in serum glucose, liver (ALT, AST), and kidney (Urea-N, creatinine) function among groups, indicating no hepato-and nephrotoxic effects of the combination of ASI and MgB in rats exposed to UVB (p > 0.05). UVB exposure caused a marked decrease in skin hydroxyproline levels and biotinidase activity by 46.1 and 42.8% compared to shaved rats (Table 1; p < 0.001). Compared to the UVB group, hydroxyproline and biotinidase levels increased in all ASI + MgB treatment groups (p < 0.05). The ASI + MgB-H + MgB-C group showed the greatest increase in hydroxyproline levels (65.8%) and biotinidase activity (68.2%) (p < 0.05) (Table 1). Elastase activity in the dorsal skin increased from UVB irradiation by 53.3% compared to shaved rats (p < 0.05), whereas ASI + MgB treatments exerted protection against the increase in dorsal skin elastase activity induced by UVB irradiation (p < 0.05). The effect was more pronounced in the ASI + MgB-H group compared to the ASI + MgB-L group. However, a combination of ASI + MgB oral administration and MgB cream provided even greater preventive effects, resulting in a 30.6% decrease in elastase levels compared to the UVB group. We also examined the effects of the combination of ASI and MgB on the MDA levels, a lipid peroxidation marker, induced by UVB radiation in the skin tissue of rats. The concentration of MDA was higher in the UVB-treated group than in the no radiation groups, whereas the concentration decreased significantly in the skin of ASI and MgB-treated groups (p < 0.001). The effect of the high dose of MgB combined with ASI and MgB cream was much better compared to the low doses (Table 1).

Serum Mg, Fe, Zn, Cu, and Si levels were significantly reduced by 24.9, 15.4, 19.0, 29.3, and 30.2% in the UVB group compared with the shaved group, respectively (p < 0.001; Table 2). ASI + MgB groups presented a significant increase in these parameters compared to the UVB group, particularly in the ASI + MgB-H + MgB-C group (p < 0.05). Mg, Fe, Zn, Cu, and Si levels increased 165.7, 15.8, 9.0, 23.9, and 237.0%, respectively, compared to the UVB group. In addition, serum biotin and arginine concentrations were decreased by 36.8 and 23.5% in the dorsal skin following UVB irradiation (p < 0.01). However, the combination of ASI and MgB treatment improved this reduction caused by UVB (p < 0.05; Table 2). This effect was greater in the ASI + MgB-H + MgB-C group (2.2 mg ASI + 160 μg MgB), with biotin and arginine levels increased by 179.5 and 89.8%, respectively.

Illustrative pictures of dorsal skin lesions caused by UVB exposure are shown in Supplementary Figures S1, S2. The intensity of fine lines and wrinkles, which is one of the most critical signs of photoaging, was increased by UVB irradiation, with the highest intensity observed in the UVB group (Supplementary Figures S1, S2). In addition, the dorsal skin of the rats in the UVB groups had an irregular and corrugated appearance compared to the control group. In contrast, non-irradiated rats in the normal and shaved groups showed neither wrinkles nor lesions. This not only further demonstrated UVB’s skin-damaging effects but also showed that shaving had no macroscopic damage to the skin. Compared to the UVB group, the skin of rats in the ASI + MgB-H + MgB-C, ASI + MgB-L + C, and ASI + MgB-H groups appeared much smoother without visible lesions. In contrast, the dorsal skin of rats in the ASI + MgB-L group showed numerous wrinkles and mild erythema. At the end of the experiment, rats exposed to UVB with no other treatment had greater visual scores, demonstrating greater impairment when compared to control rats (p < 0.001; Supplementary Figure S1). However, the visual scores of the skin of the rats were significantly reduced by 37.8–76.0% in the ASI + MgB groups, an effect that was greatest in the ASI + MgB-H + MgB-C group (−76.0%) (p < 0.05; Supplementary Figure S1B).

The pinch test evaluated rats’ dorsal skin elasticity, and the representative pictures of skin after being pushed are presented in Supplementary Figure S3A. The pinch test indicated no difference between the rats’ recovery times in the normal and shaved control groups (p > 0.05). However, skin recovery time, associated with photoaging and a loss of skin elasticity, increased in all UVB groups. Skin recovery time was longest in the UVB group, with a time 6.8 times longer than that of the control group (p < 0.05) (Supplementary Figure S3B). Compared with the UVB group, treatment with ASI + MgB significantly reduced recovery time by 34.8–62.8%, with the lowest duration found in the ASI + MgB-H + MgB-C group (−62.8%).

As established in H&E staining pictures (Figure 1A), Masson’s trichrome (Figure 1B), and elastic fiber staining images (Figure 1C), normal and shaved rats presented almost similar appearances and showed normal skin layers, including normal epidermises covered by thin layer stratum corneum, wavy dermal-epidermal junction lines, regularly disseminated hair follicles, closely refined sequential collagen bundles, as well as well-distributed elastic fibers. Besides, no inflammatory infiltration was detected in these groups. After UVB-irradiation, rats’ epidermises exhibited irregular hyperplasia with thickened stratum corneum. Thin and irregular collagen fibers were also observed in the dermis, showing prominent fractures (Figure 1A). In some instances, the elastic fibers were twisted, broken, or even irregularly accumulated, and the density was significantly reduced (Figure 1C).

Moreover, inflammatory infiltrations were visibly demonstrated throughout the dermis. However, ASI + MgB treatment orally or MgB topically (especially at a high dose with cream application) significantly alleviated these UV-induced skin damage features. The rats' skin in ASI + MgB-H + MgB-C revealed fairly complete epidermises and dermis, containing thin layer stratum corneum and well-regulated collagen and elastic fibers uniform thickness and distribution.

In Masson’s trichrome staining, collagen is indicated as the blue-stained parts of the dermis that are not exposed to UVB. The collagen dispersal in the control group was regularly concentrated in the dermis, but the collagen dispersal was changed and destructed with UVB exposure. However, the administration of ASI + MgB prevented collagen fibers from UVB damage. Particularly, the high dose of ASI + MgB with the application of MgB cream presented stronger blue staining as compared to the control group (Figure 1B).

Epidermal thicknesses and microfolds of rats in the UVB group dramatically increased after chronic UVB exposure, with levels 3.0 and 4.2 times that of shaved rats, respectively (Figure 1). However, the epidermal thickening was markedly decreased to 27.2, 40.3, 47.7, and 64.3% in the ASI + MgB-L, ASI + MgB-H, ASI + MgB-L + C, and ASI + MgB-H + C groups, respectively, compared to the UVB group (Figure 1D; p < 0.05). Microfolds decreased by 29.3, 42.9, 60.9, and 67.9% in the ASI + MgB-L, ASI + MgB-H, ASI + MgB-L + C and ASI + MgB-H + C groups, respectively, compared to the UVB group (Figure 1E; p < 0.05). The intensity of inflammatory cell infiltration in the skin was 7.7 times greater in the UVB irradiation group compared to shaved controls (Figure 1F; p < 0.001). The administration of a combination of ASI and MgB treatment partially prevented inflammatory cell infiltration (p > 0.05), particularly in the ASI + MgB-H + C group, where infiltration decreased by 79.0%.

The effect of MgB and ASI complex treatment on the expressions of mTOR, MMP-1, IL-6, and COX-2 proteins in the skin of rats was evaluated using immunohistochemistry after UVB exposure (Figure 2). The results showed that mTOR, MMP-1, IL-6, and COX-2 immunoreactivity was increased in the UVB treatment rats skin tissues compared to the control and shaved groups (Figures 2E–H; p < 0.001), whereas ASI + MgB-H treatments decreased the expression of mTOR, MMP-1, IL-6, and COX-2 in UVB treatment rats skin tissues. The highest level of decrease was detected in a high dose of MgB combined with ASI and MgB cream group.

Compared to shaved rats, dorsal skin ACC1, ACC2, PC, PCC, and MCC protein levels decreased by 63.0, 64.0, 76.8, 54.6, and 64.2, respectively, due to UVB irradiation (Figure 3: p < 0.001 for all). However, a combination of ASI and MgB treatment, especially the ASI + MgB-H with cream treatment, increased ACC1, ACC2, PC, PCC, and MCC relative to those in the UVB group (p < 0.05 for all). To investigate the mechanism underlying the protective effect of ASI + MgB on UVB-induced photoaging, dorsal skin levels of VEGF, SIRT1, mTORC2 (ser2481), p-p70S6K, p4E-BP1, MMP-1, and MMP-3 were also analyzed by Western blot (Figure 4). The results showed that UV irradiation-induced a decrease in protein levels of VEGF and SIRT1 by 81.4 and 63.4% and an increase in p-mTORC2 (ser2481), p-p70S6K, p4E-BP1, MMP-1, and MMP-3 protein levels by 160.4, 106.6, 141.6, 73.7 and 80.9% as compared to shaved rats, respectively. However, ASI + MgB treatment caused a significant increase in protein levels of VEGF and SIRT1 and a decrease in mTORC2 (ser2481), p-p70S6K, p4E-BP1, MMP-1, and MMP-3 protein levels in the dorsal skin of rats exposed to UVB. In addition, the UVB irradiation caused higher levels of inflammatory factors, including TNF-α, NFκB, IL-6, IL-8, and COX-2 protein levels in rats’ dorsal skin (p < 0.01; Figure 5). However, ASI + MgB treatment, especially in the high dose with MgB cream application, markedly inhibited their levels (p < 0.05).

In order to determine the mechanism of action of the combination of ASI and MgB, we examined the mitogen-activated protein kinases (MAPKs) and AP-1 signaling pathways known to be associated with MMP expression. MAPK signaling factors stimulate activator protein-1 (AP-1) expression, which is a transcription factor and consists of c-Jun and c-Fos subunits. For this purpose, we measured two major components of phosphorylated c-jun N-terminal kinase (p-JNK), p-p38, and AP-1 levels by western blot analysis. As seen in Figure 6, levels of p-JNK and p-p38 were increased by UVB exposure compared to control and shaved groups (p < 0.0001). ASI + MgB treatments downregulated p-JNK and p-p38 levels in rats exposed to UVB. Their levels are particularly reduced by ASI + MgB combined with MgB cream than other groups (p < 0.0001). Similarly, UVB irradiation significantly activated the skin tissue AP-1 (p-c-Jun and p-c-Fos) level in rats (p < 0.0001). However, ASI + MgB treatments, particularly with MgB cream application, blocked the activation of the AP-1 in UVB-irradiated rats (Figure 6, p < 0.0001).

As shown in Figure 7, UVB exposure significantly increased the expressions of Bax and caspase-3 and decreased the Bcl-2 levels compared to the control and shaved groups (p < 0.001). However, ASI + MgB treatments, especially in the MgB high dose with MgB cream application, markedly reversed their levels (p < 0.01).