Comparative Study on Epidermal Moisturizing Effects and Hydration Mechanisms of Rice-Derived Glucosylceramides and Ceramides

Ceramide (Cer) plays an important role in skin barrier functions in the stratum corneum (SC). The ingestion of food-derived glucosylceramides (GlcCer) attenuates transepidermal water loss (TEWL). However, the moisturizing effects of single molecules of GlcCer and Cer remain unclear. Therefore, we herein purified 13 GlcCer and 6 Cer, including elasticamide, which has the same structure as human Cer[AP], from rice and compared their epidermal moisturizing effects in a reconstructed human epidermal keratinization model. The results obtained showed that 10 µM of 5 GlcCer[d18:2] with a 4E,8Z sphingadienine and C18 to C26 fatty acids and 10 µg/mL of 3 Cer with C23 or C24 fatty acids significantly reduced TEWL. The moisturizing effects of these GlcCer were dependent on the length of fatty acids. Furthermore, 10 µg/mL of elasticamide increased the SC Cer contents by promoting the expression of GlcCer synthase. Electron microscopic observations revealed that 1 µM of GlcCer[d18:2(4E,8Z)/26:0] increased the number of keratohyalin granules and desmosomes. Immunostaining and Western blotting indicated that 1 µM of GlcCer[d18:2(4E,8Z)/26:0] up-regulated the expression of filaggrin and corneodesmosin, which contribute to epidermal hydration. This comparative study on epidermal moisturization by GlcCer and Cer isolated from rice revealed differences in their hydration mechanisms.


Introduction
The epidermis is the outermost layer of the skin and contributes to skin barrier functions and moisturization. It consists of the following four sublayers in order from the inner-to outermost layers: the stratum basal, stratum spinosum, stratum granulosum, and stratum corneum (SC). Approximately 95% of cells in the epidermis are keratinocytes. SC is formed by keratinization and performs barrier functions, such as inhibiting water evaporation and protecting against foreign substances and ultraviolet light [1][2][3]. The barrier functions of SC are maintained by the lamellar structure, which is formed by intercellular lipids and water molecules and fills the spaces between SC cells. Intercellular lipids are mainly composed of ceramide (Cer), which accounts for approximately 50%, and cholesterol, fatty acids, and cholesterol sulfates [4]. However, defective epidermal barrier functions cause skin dehydration due to poor water retention and the skin surface becomes more sensitive to non-specific stimuli, such as protein antigens and ultraviolet light. Defects also result in the release of various cytokines from inflammatory cells, leading to the development and worsening of dermatitis and itching. Therefore, it is important to retain the factors that contribute to barrier functions in order to maintain a healthy epidermis.   (14)(15)(16)(17)(18)(19) isolated from GlcCer-rich rice oil by-product. (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13) and Cer (14)(15)(16)(17)(18)(19)

Effects of GlcCer (1-13) and Cer (14-19) on SC Cer contents in the RHEK Model
To elucidate hydration mechanisms, we examined the effects of GlcCer (1-13) and Cer (14)(15)(16)(17)(18)(19) on SC Cer contents. Regarding GlcCer, SC Cer contents were not affected by any of the tested molecules at 10 µM ( Figure 4A). On the other hand, 14 and 18, which attenuated TEWL in the RHEK model, significantly increased SC Cer contents ( Figure 4B). Therefore, we confirmed the concentration dependency of 14, which showed the strongest activity on TEWL improvement among the Cer examined. Additionally, 14 increased SC Cer contents in a concentration-dependent manner, with 10 µg/mL of 14 significantly increasing total Cer and Cer[NS,NDS] ( Figure 5).

Effects of Elasticamide (14) on the Expression of Enzymes Related to SC Cer Synthesis in the RHEK Model
Elasticamide (14) significantly increased SC Cer contents in the RHEK model, as described above. Therefore, we examined the effects of 14 on the expression of enzymes involved in SC Cer synthesis. Figure 6A shows the effects of 14 on the mRNA expression of enzymes related to SC ceramide synthesis in the RHEK model: SPT2, CerS3, GCS, GBA, SMS2, and ASM. Furthermore, 14 significantly up-regulated the mRNA expression of GCS and SMS2 at 1 and 10 µg/mL, respectively. Figure 6B shows the effects of 14 on the protein expression of these enzymes; it significantly up-regulated the protein expression of GCS at 1 and 3 µg/mL, but not that of SMS2.  (14)(15)(16)(17)(18)(19). The RHEK model was treated for 7 days with GlcCer (1-13: 10 µ M) or Cer (14-19: 10 µ g/mL). The extraction of lipids from the SC of the RHEK model and a high performance thin-layer chromatography (HPTLC) analysis were performed as described in the Materials and Methods section. The effects of each GlcCer are corrected for the control value. Data are expressed as means ± S.E. (n = 3-4). Asterisks denote a significant difference from the control group, ** p < 0.01.     volved in SC Cer synthesis. Figure 6A shows the effects of 14 on the mRNA expression of enzymes related to SC ceramide synthesis in the RHEK model: SPT2, CerS3, GCS, GBA, SMS2, and ASM. Furthermore, 14 significantly up-regulated the mRNA expression of GCS and SMS2 at 1 and 10 µ g/mL, respectively. Figure 6B shows the effects of 14 on the protein expression of these enzymes; it significantly up-regulated the protein expression of GCS at 1 and 3 µ g/mL, but not that of SMS2.  As described above, GlcCer did not affect SC Cer contents, while elasticamide (14) increased SC Cer via the up-regulation of GCS expression. Therefore, we conducted further experiments to elucidate the epidermal hydration mechanism of 13, which exerted the strongest moisturizing effects. The concentration-dependent effects of 13 (1, 3, and 10 µM) on TEWL were confirmed ( Figure 7A). However, similar to the results shown in Figure 4, none of the concentrations of 13 increased SC Cer contents ( Figure 7B). Accordingly, we performed electron microscopic observations of cross-sections of the RHEK model. The results obtained indicated that 1 µM of 13 increased the intercellular density of the SC layer. Furthermore, 13 appeared to increase the number of keratohyalin granules and desmosomes ( Figure 7C).

Model
As described above, GlcCer did not affect SC Cer contents, while elasticamide (14) increased SC Cer via the up-regulation of GCS expression. Therefore, we conducted further experiments to elucidate the epidermal hydration mechanism of 13, which exerted the strongest moisturizing effects. The concentration-dependent effects of 13 (1, 3, and 10 µ M) on TEWL were confirmed ( Figure 7A). However, similar to the results shown in Figure 4, none of the concentrations of 13 increased SC Cer contents ( Figure 7B). Accordingly, we performed electron microscopic observations of cross-sections of the RHEK model. The results obtained indicated that 1 µ M of 13 increased the intercellular density of the SC layer. Furthermore, 13 appeared to increase the number of keratohyalin granules and desmosomes ( Figure 7C). Since electron microscopic observations indicated that 13 increased keratohyalin granules and desmosomes, we examined the effects of 13 on the protein expression of filaggrin and corneodesmosin. Figure 8A shows immunostaining and Western blotting images of filaggrin expression. Based on immunostaining images, 13 (1 µM) appeared to up-regulate the expression of filaggrin in SC and Western blotting showed significant increases in the protein expression of filaggrin by 13 (1 and 3 µM). As shown in Figure 8B, immunostaining images suggested the up-regulated expression of corneodesmosin in SC by 13 (1 µM) and Western blotting revealed that 13 (1 µM) significantly increased the protein expression of corneodesmosin.
images of filaggrin expression. Based on immunostaining images, 13 (1 µ M) appeared to up-regulate the expression of filaggrin in SC and Western blotting showed significant increases in the protein expression of filaggrin by 13 (1 and 3 µ M). As shown in Figure 8B, immunostaining images suggested the up-regulated expression of corneodesmosin in SC by 13 (1 µ M) and Western blotting revealed that 13 (1 µ M) significantly increased the protein expression of corneodesmosin.

Discussion
In the present study, we isolated 13 GlcCer and 6 Cer from by-products of rice bran oil and compared their effects on epidermal moisturization in a RHEK model. We initially investigated the effects of GlcCer (1-13) and Cer (14)(15)(16)(17)(18)(19) on TEWL. TEWL is the most widely used measurement to evaluate epidermal barrier functions [29]. Significant increases in TEWL are commonly observed in dry skin diseases, such as AD [30], xerosis, and psoriasis [31]. We previously reported the moisturizing effects of isolated steroidal saponins from tomato seeds and β-sitosterol 3-O-glucoside from rice on TEWL in RHEK models [32]. In the present study, all GlcCer

Discussion
In the present study, we isolated 13 GlcCer and 6 Cer from by-products of rice bran oil and compared their effects on epidermal moisturization in a RHEK model. We initially investigated the effects of GlcCer (1-13) and Cer (14-19) on TEWL. TEWL is the most widely used measurement to evaluate epidermal barrier functions [29]. Significant increases in TEWL are commonly observed in dry skin diseases, such as AD [30], xerosis, and psoriasis [31]. We previously reported the moisturizing effects of isolated steroidal saponins from tomato seeds and β-sitosterol 3-O-glucoside from rice on TEWL in RHEK models [32]. In the present study, all GlcCer[d18:2(4E,8Z)] (1, 3, 7, 10, and 13) significantly attenuated TEWL, whereas GlcCer[t18:1(8Z)] (2, 5, 9, and 12), GlcCer[d18:2(4E,8E)] (4, 8 and 11), and GlcCer[t18:1(4E)] (6) did not. These results suggest that the d18:2(4E,8Z) structure is required for the moisturizing effects of GlcCer. Furthermore, reductions in TEWL were dependent on the length of fatty acids, and the moisturizing effects of 10 and 13, which have C24 or more fatty acids, were detected in the earlier stages of the culture (Days 1 and  3). Therefore, a longer fatty acid length appeared to accelerate the epidermal hydration activity of GlcCer. Previous studies demonstrated the effects of plant-derived GlcCer on TEWL. Tsuji et al. reported that the oral administration of GlcCer from rice or maize attenuated TEWL in hairless mice [18], while Kuwata et al. showed that dietary GlcCer from pineapples reduced TEWL in hairless mice [20]. Our previous findings revealed that the oral administration of rice-derived GlcCer reduced TEWL in the skin of SDStreated hairless mice [17]. However, these experiments used a mixture of GlcCer as the test substance and a comparison of the effects of single molecules of GlcCer has not yet been conducted. Therefore, the present study is the first report to compare and evaluate the effects of single molecules of GlcCer on epidermal hydration. Regarding Cer species, the significant attenuation of TEWL was observed with Cer  (19), which have more than C24 fatty acids, did not affect TEWL. Moreover, although none of the tested GlcCer[t18:1(8Z)] (2, 5, 9, and 12) affected TEWL, Cer (18) with the t18:1(8Z) structure significantly attenuated TEWL. On the other hand, Cer (14) and GlcCer (13) both reduced TEWL in the later stages of the culture (Days 5 and 7), while only GlcCer (13) attenuated TEWL from the earlier stage (Day 1). Collectively, these results revealed differences in the hydration mechanisms of GlcCer and Cer.
To elucidate the hydration mechanisms of GlcCer and Cer, we quantified SC Cer contents in the RHEK model treated with GlcCer (1-13) and Cer (14)(15)(16)(17)(18)(19). In a profiling study on Cer species, 20 major classes of Cer were identified in the human SC [6]. SC Cer contents significantly decreased in patients with AD [13] and xerosis [14], and reductions in SC Cer contents resulted in epidermal water evaporation [15]. In the present study, 10 µg/mL of 14 and 18 significantly increased total SC Cer contents in the RHEK model. Moreover, 14 increased SC Cer contents in a concentration-dependent manner in concert with a significant increase in Cer[NS,NDS], which is the most abundant Cer species in the SC. Although the structure of 14 is the same as that of human Cer[AP], 14 did not increase SC Cer [AP] contents. This result suggests that the epidermal hydration mechanism of 14 does not involve a direct supply of Cer to the SC from the medium under the RHEK membrane. Therefore, we examined the effects of 14 on the mRNA and protein expression of SPT2, CerS3, GCS, GBA, SMS2, and ASM, which are involved in SC ceramide synthesis [7][8][9][10][11][12].
The results obtained showed that 14 significantly up-regulated the expression of GCS, which acts as a glucosyltransferase enzyme to synthesize GlcCer, the precursors of SC Cer [9]. Therefore, 14 increased SC Cer contents by up-regulating the expression of GCS, leading to reductions in TEWL. We demonstrated that tiliroside, the main component of strawberry seed extract, increased SC Cer[NS,NDS] by up-regulating GCS and GBA [33]. Although 14 did not significantly up-regulate the expression of GBA, its protein expression level was slightly increased. Consequently, the mechanisms by which tiliroside and 14 increase SC Cer contents are considered to be similar.
Regarding the hydration mechanism of rice GlcCer, 1, 3, 7, 10, and 13 did not affect SC Cer contents even though they significantly attenuated TEWL. Furthermore, between 1 and 10 µM of 13 did not increase SC Cer contents but dose-dependently reduced TEWL. Accordingly, the epidermal hydration mechanism of 13 differed from that of 14. Electron microscopic observations to investigate the hydration mechanism of 13 revealed that the intercellular density of SC layers was increased by 13, indicating better moisture retention. In addition, the number of keratohyalin granules appeared to be increased by 13. Keratohyalin granules are present in the stratum granulosum and consist of profilaggrin, which is a precursor of filaggrin. Although profilaggrin does not bind to keratin, it is dephosphorylated and proteolyzed into filaggrin monomers during the later stages of the keratinization process [34]. Filaggrin then provides natural moisturizing amino acids through its decomposition in the SC [35] and acts as a filament-aggregating protein [36]. Previous studies reported that the number of keratohyalin granules and filaggrin expression were reduced or absent in the skin of patients with ichthyosis vulgaris; therefore, the number of keratohyalin granules in the stratum granulosum reflect filaggrin production levels [37,38]. Herein, we examined the effects of 13 on filaggrin protein expression in the RHEK model. Immunostaining observations and Western blotting results revealed that the protein expression of filaggrin was significantly increased by 13. These results suggest that the increase in filaggrin, which functions as a natural moisturizing factor in the SC, may be the epidermal hydration mechanism of 13.
In addition, electron microscopic observations suggested an increase in the number of desmosomes by 13. Desmosomes are intercellular junctions that provide strong adhesion between epidermal cells [39,40] and their shape markedly changes from corneodesmosomes in the SC [41]. Corneodesmosin, which is specifically found in corneodesmosomes, plays a pivotal role in SC cell adhesion [42]. Furthermore, an analysis of corneodesmosin-knockout mouse skin revealed the detachment of the SC from the living layers of the epidermis, which resulted in prominent defects in barrier functions [43]. Therefore, increases in the expression of corneodesmosin are considered to be a reasonable mechanism for epidermal hydration. Immunostaining observations and Western blotting results revealed that the protein expression of corneodesmosin was significantly up-regulated by 13. Therefore, changes in corneodesmosin may be another hydration mechanism of 13 that reduces TEWL.

Conclusions
Herein, we demonstrated that among the 13 GlcCer isolated from rice, d18:2(4E,8Z)type GlcCer reduced TEWL in a fatty acid length-dependent manner. The hydration mechanism of GlcCer appears to involve increases in SC density and the expression of filaggrin and corneodesmosin. Regarding the 6 isolated Cer, Cer with C23 or C24 fatty acids also reduced TEWL. Moreover, elasticamide (14) has been suggested to increase SC Cer contents (particularly Cer[NS,NDS]) via the up-regulated expression of GCS. The present study is the first to compare the epidermal moisturizing effects of GlcCer and Cer species as single compounds and reveal differences in the hydration mechanisms of GlcCer and Cer.

Preparation of GlcCer and Cer
To isolate GlcCer and Cer, they were purified according to our previously described methods [27,28]. The gummy by-products obtained during the manufacturing process of rice bran oil were subjected to flash column chromatography equipped with a silica gel column and evaporative liquid scattered detector (ELSD) to obtain crude GlcCer and Cer fractions. The crude GlcCer fraction was separated and repeatedly purified by HPLC to obtain GlcCer  (13). The isolated yields from 1-13 were described in our previous study [28].
To obtain each Cer, the crude Cer fraction was purified by HPLC equipped with an ODS column (Capcell pak C18 SG-120, φ 20 × 250 mm, Osaka Soda, Osaka, Japan) and the RI detector, MeOH: tetrahydrofuran (9:1) as the mobile phase to obtain Cer

Measurement of TEWL in the RHEK Model
TEWL measurements were performed according to previously described methods [32,50,51]. Measurements were conducted before and 1, 3, 5, and 7 days after the treatment of samples with Tewitro TW24 (Courage + Khazaka, Cologne, Germany), which is the device specifically developed for the TEWL measurement in vitro testing and can directly and simultaneously measure the TEWL of the RHEK model in a 24-well plate. The RHEK model was placed on a thermal insulation mat (HIENAI Mat; Cosmo Bio Co., Ltd., Tokyo, Japan), which consistently kept the entire bottom surface at 32 • C, without a lid for 5 min before measurements. TEWL was measured for 30 min by inserting the sensor of Tewitro TW24 into each well of the RHEK model while maintaining the bottom surface at 32 • C, and the mean value for the last 10 min was used in the analysis. Measurements were performed under sterile conditions.

Lipid Extraction and Cer Determination
Lipid extraction and Cer determination were conducted according to our previously described methods [32,33,52]. SC lipids from the RHEK model were extracted using a mixture of CHCl 3 , MeOH, and PBS (1:2:0.8). Dried SC lipid samples were subjected to HPTLC to measure Cer contents.

Real Time RT-PCR
The mRNA expression of Cer synthesis-related enzymes in the RHEK model was measured by quantitative real-time RT-PCR according to our previously described method [32]. Extracted RNA from the whole RHEK model was reverse-transcribed to obtain cDNA. Information on specific primers was previously described [32].

Western Blotting Analysis
The Western blotting analysis described in our previous studies [33,52] was used to assess protein expression. Proteins extracted from the RHEK model were electrophoresed on 10% SDS gels and separated proteins were transferred to polyvinylidene difluoride membranes. After blocking with 5% skim milk, membranes were treated with a primary antibody followed by a secondary antibody

Microscopic Analysis and Immunostaining
In immunohistochemistry of the RHEK model, formalin/paraformaldehyde-fixed tissues were embedded in paraffin to obtain specimens at Aichi Pathology Laboratory LLC (Mihama, Japan). Deparaffinized specimens were treated with 0.3% H 2 O 2 in MeOH and soaked in L.A.B solution for 15 min to activate antigens. After blocking with 5% skim milk for 1 h, specimens were treated with the primary antibody followed by the secondary antibody. Anti-filaggrin (1:200) and anti-corneodesmosin (1:200) antibodies were used as the primary antibodies. HRP-conjugated goat anti-rabbit IgG (1:5000) was used as the secondary antibody. Specimens were then treated with DAB solution followed by staining with hematoxylin solution. Microscopic images were taken by a Leica DMI6000 (Leica Microsystems GmbH, Wetzlar, Germany).
Regarding electron microscopic observations of tissues, the preparation and imaging of pathological sections were consigned to the Hanaichi UltraStructure Research Institute, Co., Ltd. (Okazaki, Japan). The samples of tissues for TEM were fixed in phosphate buffered 2% glutaraldehyde, and subsequently post-fixed in 2% osmium tetra-oxide for 2 h in an ice bath. Then, the specimens were dehydrated in a graded ethanol and embedded in the epoxy resin. Ultrathin sections were obtained by ultramicrotome technique. Ultrathin sections stained with uranyl acetate for 15 min and lead staining solution for 2 min were submitted to TEM observation at 100 kV (H-7600, HITACHI Ltd., Tokyo, Japan).

Statistical Analysis
All statistical analyses were performed using Statcel (OMS Ltd., Tokyo, Japan) as an add-in to excel. All results are expressed as means ± standard errors (S.E.). The significance of differences was examined by a one-way analysis of variance followed by Dunnett's test, with p < 0.05 or p < 0.01 indicating a significant difference.

Data Availability Statement:
The data that support the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

Abbreviations
The following abbreviations are used in this manuscript. SC stratum corneum Cer ceramide GlcCer glucosylceramide SPT serine palmitoyltransferase CerS ceramide synthase GCS glucosylceramide synthase GBA β-glucocerebrosidase SMS sphingomyelin synthase