Epidermal hydration (water content in the epidermis) is largely maintained by natural moisturising factors (NMF) and the epidermal barrier in the epidermis(Reference Rawlings and Harding1, Reference Harding and Scott2). Compared with the epidermal barrier, in which the multilayered lamellar structure of ceramides, cholesterol and fatty acids provides a hydrophobic barrier against water loss in the epidermis(Reference Rawlings and Harding1), NMF consist primarily of amino acid (AA) and their derivatives together with lactate, sugars, urea and inorganic ions, which confer the hygroscopic ability of maintaining epidermal hydration as a complex mixture of water-soluble components(Reference Rawlings and Harding1, Reference Harding and Scott2). AA, the major components of NMF, are formed by the degradation of filaggrin during epidermal differentiation. Filaggrin is initially synthesised as profilaggrin, which consists of multiple repeats of filaggrin(Reference Harding and Scott2, Reference Harding and Scott3). Through the generation of two-repeat intermediate (2RI) or three-repeat intermediate (3RI) of filaggrin, profilaggrin is degraded to form monomeric filaggrin, and then undergoes subsequent degradation into free AA(Reference Harding and Scott3, Reference Scott, Harding and Barrett4), which consists of multistep processes, in which various regulators and proteases are involved(Reference Sandilands, Sutherland and Irvine5). The similarities in AA profiles between filaggrin and NMF indicate that filaggrin is the sole source of AA in NMF, as glutamate, serine and glycine accounted for about >50 % of the total free AA in either filaggrin or NMF(Reference Harding and Scott2, Reference Scott, Harding and Barrett4). The reduced expression of filaggrin protein is paralleled by decreased amounts of free AA in dry skin diseases such as atopic dermatitis (AD) and ichthyosis(Reference Howell, Kim and Gao6–Reference Seguchi, Cui and Kusuda8).
The recognition of dry skin condition in various skin diseases(Reference Howell, Kim and Gao6–Reference Seguchi, Cui and Kusuda8), coupled with the inconvenience of applying moisturiser and emollients frequently, and undesirable side effects of some pharmaceutical agents, has led to the development of alternatives for dry skin protection from dietary and natural sources. Silk protein purified from the cocoons of silkworms (Bombyx mori) has been reported to have beneficial effects for dry skin protection(Reference Padamwar, Pawar and Daithankar9, Reference Voegeli, Meier and Blust10). Silk consists of two types of proteins, fibroin and sericin. In silk textile processing, sericin, which envelops fibroin with successive sticky layers, is mostly removed and fibroin, a fibrous protein, is purified. Although fibroin has been reported to be a useful biomaterial in the skin(Reference Inouye, Kurokawa and Nishikawa11, Reference Mason12), sericin has been particularly reported to prevent dry skin condition(Reference Padamwar, Pawar and Daithankar9, Reference Voegeli, Meier and Blust10). Possibly due to its high serine content (30–33 % of total AA) and the AA profile similarity to NMF(Reference Kato, Sato and Yamanaka13), human studies have indicated that topically applied sericin increases epidermal hydration(Reference Padamwar, Pawar and Daithankar9, Reference Voegeli, Meier and Blust10). Sericin may therefore be a valuable source of natural dry skin protection for cosmetic and food industries(Reference Kato, Sato and Yamanaka13). However, the dry skin protection activities of sericin have been suggested mostly by topical application (in vitro studies)(Reference Padamwar, Pawar and Daithankar9, Reference Voegeli, Meier and Blust10), and little information is available on the dietary effects of sericin. In this regard, we previously reported that dietary sericin increases the total amount of ceramides(Reference Park, Choi and Kim14), the major lipid maintaining the lamellar structure of the epidermal barrier(Reference Rawlings and Harding1), in NC/Nga mice, an animal model of AD(Reference Aioi, Tonogaito and Suto15, Reference Suto, Matsuda and Mitsuishi16). To further investigate the dietary effect of silk protein on improving skin dryness, we compared the dietary effects of sericin and fibroin on epidermal hydration in NC/Nga mice, and delineated a possible mechanism of these effects in profilaggrin expression and its degradation into free AA in the present study. Specifically, altered levels of filaggrin species and free AA, as well as regulators and proteases involved in profilaggrin expression and degradation processing, were evaluated.
Experimental methods
Preparation of sericin and fibroin protein powders
Sericin and fibroin powders were provided by the Rural Development Administration, Republic of Korea(Reference Park, Choi and Kim14, Reference Madyarov, Lee and Yeo17). In brief, sericin was extracted twice by boiling small pieces of silk cocoons of B. mori (100 g) with 3 litres of distilled water at 110°C for 5 h, and filtering through Whatman no. 2 paper(Reference Park, Choi and Kim14, Reference Madyarov, Lee and Yeo17). Fibroin extracts were prepared by boiling small pieces of silk cocoons of B. mori (100 g) with 5 litres of 0·03 % sodium carbonate and 0·05 % marceillous soap for 30 min and washed twice with boiling water(Reference Park, Choi and Kim14, Reference Madyarov, Lee and Yeo17). An electrodialysis system (Acilyzer-02; Astom) was used to remove salt residues from fibroin extracts(Reference Park, Choi and Kim14). All sericin and fibroin extracts were lyophilised to yield sericin and fibroin powders. The molecular weights of sericin and fibroin protein powders obtained were confirmed to be about 65 300 (sericin) and 101 700 (fibroin) by gel electrophoresis(Reference Park, Choi and Kim14).
Animals and diets
Male BALB/c mice (5 weeks old, n 10) and male NC/Nga mice (5 weeks old, n 30) were purchased from SLC Japan (Shizuoka, Japan). NC/Nga mice, an animal model of AD, in which dry skin is raised by conventional environments(Reference Aioi, Tonogaito and Suto15, Reference Suto, Matsuda and Mitsuishi16) (not by genetic defects in FLG, profilaggrin gene)(Reference Heish, Imai and Katayama18, Reference Presland, Boggess and Lewis19), were assigned to three groups of ten mice each: an atopic control group (group CA) fed a control diet, and groups S and F fed diets supplemented with 1·0 % powdered extracts of sericin and fibroin, respectively. Mice were fed the experimental diets for 10 weeks to compare the dietary effects of sericin and fibroin on epidermal hydration. Simultaneously, group C, a normal control group which consisted of BALB/c mice, was fed the control diet for 10 weeks. The compositions of the experimental diets are shown in Table 1, and the AA compositions of sericin and fibroin are shown in Table 2.
Compositions of the experimental diets
C, BALB/c mice fed a control diet; CA, NC/Nga mice fed a control diet; S, NC/Nga mice fed a diet supplemented with 1 % sericin; F, NC/Nga mice fed a diet supplemented with 1 % fibroin.
* Casein (nitrogen × 6·25), 870 g/kg.
† Vitamin mix composition, AIN-93 vitamin mix #310025 (Dytes, Inc.): niacin, 3 g/kg; calcium pantothenate, 1·6 g/kg; pyridoxine HCl, 0·06 g/kg; thiamin HCl, 0·6 g/kg; riboflavin, 0·6 g/kg; folic acid, 0·2 g/kg; biotin, 0·2 g/kg; vitamin E acetate (500 IU/g), 15 g/kg; vitamin B12 (0·1 %), 2·5 g/kg; vitamin A palmitate (500 000 U/g), 0·8 g/kg; vitamin D3 (400 000 IU/g), 0·25 g/kg; vitamin K1/dextrose mix (10 mg/g), 7·5 g/kg; sucrose, 967·23 g/kg.
‡ Salt mix composition, AIN-93G salt mix #210025 (Dytes, Inc.): calcium carbonate, 357 g/kg; potassium phosphate (monobasic), 196 g/kg; potassium citrate H2O, 70·78 g/kg; NaCl, 74 g/kg; potassium sulphate, 46·6 g/kg; magnesium oxide, 24 g/kg; ferric citrate USP, 6·06 g/kg; zinc carbonate, 1·65 g/kg; manganous carbonate, 0·63 g/kg; cupric carbonate, 0·3 g/kg; potassium iodate, 0·01 g/kg; sodium selenate, 0·01025 g/kg; ammonium paramolybdate.4H2O, 0·00795 g/kg; sodium metasilicate·9H2O, 1·45 g/kg; chromium potassium sulphate·12H2O, 0·275 g/kg; lithium chloride, 0·0714 g/kg; boric acid, 0·0815 g/kg; sodium fluoride, 0·0635 g/kg; nickel carbonate, 0·0318 g/kg; ammonium vanadate, 0·066 g/kg; sucrose finely powdered sucrose, 221·026 g/kg.
Amino acid composition of casein and silk proteins
EAA, essential amino acids; NEAA, non-essential amino acids.
During the 10-week feeding period, all mice were maintained under conventional laboratory conditions without air filtration to induce AD as described previously(Reference Aioi, Tonogaito and Suto15, Reference Suto, Matsuda and Mitsuishi16). Mice were housed under conditions of controlled temperature (22–24°C), humidity (55–60 %) and light (lights on from 07.00 to 19.00 hours). Food intakes and body weights of all groups were monitored weekly over the 10-week feeding period, and there were no significant differences between the groups. Animal care and handling conformed to the guidelines provided by the Animal Care and Use Review Committee of Kyung Hee University. At the end of week 10, all mice were killed by cervical dislocation.
Assessment of epidermal hydration
Under standardised conditions of 22–24°C and 55–60 % humidity, epidermal hydration was measured on the dorsal skin surface at the end of week 10 using a corneometer (model MPA-5; Courage+Khazaka Electronic GmbH). Data are expressed as means with their standard errors (capacitance in arbitrary units (AU)).
Immunofluorescence analysis
Dissected dorsal skin was fixed in 4 % paraformaldehyde solution at 4°C for 24 h. Subsequently, skin was dehydrated and embedded in paraffin. Serial sections were obtained at a thickness of 5 μm. For antigen retrieval, sections were boiled in 0·05 m-citrate buffer for 10 min, and then completely cooled. After washing in 0·01 m-TBS, non-specific binding was blocked with 5 % goat serum and 0·1 % Triton-X in TBS for 1 h at room temperature. Skin sections were incubated at 4°C for 24 h with anti-rabbit antibody against filaggrin (ab81468; Abcam) diluted at 1:1000 in blocking buffer(Reference Kang, Park and Kim20). After washing again, the sections were incubated for 1 h in the dark at room temperature with fluorescein isothiocyanate (FITC)-conjugated anti-rabbit IgG (Sigma) at a dilution of 1:200(Reference Kang, Park and Kim20). Slides were rinsed and mounted with Permafluor™ aqueous mounting medium (Thermo Scientific). The sections were photographed under light microscope (200 × magnifications) and fluorescence images were captured using a Zeiss microscope (Zeiss LSM 510; Carl Zeiss). The intensity of fluorescence was graded visually on a scale from 1 to 5 (1, low; 3, moderate; 5, high, 0·5 scale) by a pathologist who was unaware of the experimental diets and groups.
Western blot analysis
For the detection of profilaggrin (>220 kDa), degradative intermediates of profilaggrin including 3RI (90 kDa), 2RI (60kDa) and filaggrin (28 kDa)(Reference Kam, Resing and Lim21), the epidermis was isolated after 18 h incubation of the whole skin in PBS (pH 7·4) containing 10 mm-EDTA at 4°C, and immediately homogenised in 50 mm-Tris buffer containing 9 m-urea, 2 % SDS, 2 mm-EDTA and a protease inhibitor cocktail with polytron (P-8340; Sigma)(Reference Ginger, Blachford and Rowland22). For the detection of PPARγ, peptidylarginine deiminase-3 (PAD3) and caspase-14, a regulator and proteases involved in profilaggrin expression and its degradation processing into free AA(Reference Sandilands, Sutherland and Irvine5), or detection of cytokeratin 5 (loading control), the epidermis was isolated after overnight incubation of the whole skin in an ice-cold 1:1 mixture of dispase II (2·4 units/ml; Roche) and Roswell Park Memorial Institute (RPMI) medium supplemented with 10 % fetal bovine serum at 4°C(Reference Maciejewski-Lenoir, Richman and Hakak23), and homogenised in 20 mm-Tris–HCl buffer (pH 7·5) containing 2 mm-EDTA, 0·5 mm-EGTA and 2 mm-phenylmethylsulfonyl fluoride with polytron(Reference Kim, Oh and Park24). Epidermal homogenates were centrifuged at 300 g for 5 min, and after the removal of debris, a portion of the homogenates was used to measure protein concentration by a modified Lowry method(Reference Lowry, Rosebrough and Farr25).
Protein extracts were electrophoretically separated on either 4–12 % Bis-Tris gel for various filaggrin species(Reference Kam, Resing and Lim21, Reference Ginger, Blachford and Rowland22) or 10 % SDS-PAGE gel for other proteins (NuPage Invitrogen)(Reference Kim, Oh and Park24). Gels were blotted onto nitrocellulose membranes and incubated with primary antibodies against either filaggrin (ab81468), PPARγ (sc-6285), caspase-14 (sc-56 041), PAD3 (ab50246) or keratin-5 (ab53121) (filaggrin, PAD3 and keratin-5 antibodies: Abcam; PPARγ and caspase-14 antibodies: Santa Cruz Biotechnology, Inc.) followed by incubation with IgG-horseradish peroxidase (HRP) secondary antibodies (Cell Signaling Technology, Inc.)(Reference Kim, Oh and Park24). Immunoreactive proteins were visualised with an enhanced chemiluminescence (ECL) detection system (GE Healthcare), and band intensity was quantified by densitometry.
Free amino acid analysis
Epidermis isolated with an ice-cold 1:1 mixture of dispase II (2·4 units/ml; Roche) and RPMI supplemented with 10 % fetal bovine serum at 4°C(Reference Maciejewski-Lenoir, Richman and Hakak23) was homogenised in 15 % TCA with polytron and placed at 4°C for 16 h followed by centrifugation at 12 000 g for 10 min to remove the acid-precipitated protein. After filtration (0·2 μm; Sartorius), the supernatant was subjected to free AA analysis by ion-exchange chromatography using an automatic AA analyser (Model LC-5A; Shimadzu). Free AA were eluted from a lithium high-performance column by sequential usage of lithium citrate buffers with increasing concentrations and pH (0·20 m, pH 2·80; 0·30 m, pH 3·00; 0·50 m, pH 3·15; 0·9 m, pH 3·50; 1·65 m, pH 3·55) in the mobile phase(Reference Park, Chang and Sung26, Reference Moore and Stein27). The column temperature was maintained at 47°C and the flow rate was 0·45 ml/min. Post-column derivatisation with ninhydrin was followed by spectrophotometric detection at 440 and 570 nm. The CV of multiple analyses was within 5 %. The internal standard method eliminated the error caused by the loss associated with the analytical technique. With measuring protein concentration of epidermal homogenates(Reference Lowry, Rosebrough and Farr25), free AA contents were reported as nmol/mg epidermal protein.
Statistical analysis
Data are expressed as means with their standard errors. All data were analysed by one-way ANOVA using SPSS software (SPSS version 13.0; SPSS, Inc.)(Reference Norusis28), and differences among the groups (C, CA, S and F) were determined by Duncan's multiple comparison test. P < 0·05 was considered significant. Data for epidermal hydration (Fig. 1) were further analysed by the unpaired Student's t test at P < 0·05 or P < 0·01.
Dot plot illustrating the epidermal hydration in BALB/c mice fed a control diet (group C) and NC/Nga mice fed a control diet (group CA) or diets supplemented with 1 % sericin (group S) or fibroin (group F) for 10 weeks. Each point in the dot plot represents one mouse, but occasionally overlapping data points obscure the sample size of ten in each group. The central line within the column of points represents the mean values (n 10), with standard errors represented by vertical bars. a,b Groups not sharing a common letter are significantly different (P < 0·05) using one-way ANOVA and Duncan's multiple comparison test. P < 0·05 or P < 0·01, between groups as indicated by the unpaired Student's t test. AU, arbitrary unit.
Results
Assessment of epidermal hydration
Decreased epidermal hydration has been reported in NC/Nga mice with the induction of AD-like skin lesions(Reference Aioi, Tonogaito and Suto15, Reference Suto, Matsuda and Mitsuishi16). When the distribution of the assessed level of epidermal hydration was illustrated as the dot plot, the dots in each group indicated normally distributed variances (Fig. 1). The epidermal hydration in group CA (27·2 (sem 1·67) AU) was significantly lower than in group C (34·1 (sem 2·32) AU), the normal control group. In group S, epidermal hydration (36·0 (sem 2·41) AU) was higher than that in group CA and was similar to that of group C. However, the epidermal hydration of group F (25·7 (sem 1·52) AU) was similar to that of group CA, indicating that dietary supplementation with sericin only improves epidermal hydration to similar levels of the normal control group and that dietary fibroin has no beneficial effects on epidermal hydration.
Expression of profilaggrin, degradative intermediates and filaggrin proteins
In immunofluorescence analyses (Fig. 2(A) and (B), the filaggrin proteins were highly expressed in the upper regions (i.e. stratum granulosum and stratum corneum) of the epidermis, in which terminal differentiation is processed, as reported previously(Reference Harding and Scott3), in the skin sections of all groups. The protein expressions of filaggrins in group CA became faint, when compared with those of group C. After dietary supplementation with sericin or fibroin, the protein expressions of filaggrins in groups S and F were higher than those of group CA. On a quantitative basis, the fluorescence intensity score of group CA was lower than those of other groups in relation to the level of epidermal hydration (Fig. 1). Although the fluorescence intensity scores of filaggrins in both groups S and F were higher than those in group CA, group F had lower scores than group S. The fluorescence intensity scores in group S were higher than those of group C.
Indirect immunofluorescence of filaggrins in the epidermis of BALB/c mice fed a control diet (group C) and NC/Nga mice fed a control diet (group CA) or diets supplemented with 1 % sericin (group S) or fibroin (group F) for 10 weeks. Skin specimens of the dorsal trunk were harvested for immunofluorescence staining with a primary antibody of filaggrin, which recognises an epitope within the filaggrin repeat domain and therefore detects profilaggrin, three-repeat or two-repeat intermediates and filaggrin, and using a fluorescein isothiocyanate (FITC)-conjugated secondary antibody. (A) Images in fluorescence microscopy (top) and differential interference contrast light microscopy (middle); co-captured images in fluorescence microscopy and differential interference contrast light microscopy (bottom). Arrows and dotted lines indicate the bottom layer of the epidermis. A representative section is shown from each group. Scale bar: 10 μm. (B) Score of fluorescence intensity. Values are means (n 10), with standard errors represented by vertical bars. a,b,c Mean values with unlike letters were significantly different (P < 0·05; one-way ANOVA and Duncan's multiple comparison test).
Further studies to characterise the altered filaggrin species demonstrated that the protein expressions of total filaggrins including profilaggrin, 3RI, 2RI and monomeric filaggrin in group CA were significantly lower than those in group C (Fig. 3), reflecting the faint fluorescence intensity of filaggrins shown in Fig. 2. The protein expressions of total filaggrins in both groups S and F were higher than those in group CA, and were similar to group C. However, the protein expressions of total filaggrins in group S were significantly higher than those in group F, and profilaggrin and 3RI were specifically expressed at higher levels than those in group C. In contrast, the protein expressions of 2RI and filaggrin in group S were lower than those in group C. In group F, the protein expressions of filaggrin species were similar or modestly lower than those in group C.
Altered protein expression of profilaggrin, three-repeat (3RI) or two-repeat intermediates (2RI), and filaggrin in the epidermis of BALB/c mice fed a control diet (group C, ■) and NC/Nga mice fed a control diet (group CA, 
) or diets supplemented with 1 % sericin (group S, 
) or fibroin (group F, 
) for 10 weeks. Protein extracts (5 μg each) from epidermal homogenates in 50 mm-Tris buffer containing 9 m-urea, 2 % SDS, 2 mm-EDTA and a protease inhibitor cocktail were electrophoretically separated on 4–12 % Bis-Tris gel and immunoblotted with anti-rabbit antibodies against filaggrins, which recognises an epitope within the filaggrin repeat domain and therefore detects profilaggrin, 3RI, 2RI and filaggrin. (A) Representative expressions of profilaggrin, 3RI, 2RI and filaggrin in the epidermis of mice. (B) The signal intensities from multiple experiments of (A) were quantified and the integrated areas were normalised to the corresponding value of keratin-5 (loading control). Values are means (n 10), with standard errors represented by vertical bars. a,b,c Mean values with unlike letters in each protein expression were significantly different (P < 0·05; one-way ANOVA and Duncan's multiple comparison test).
Expression of PPARγ, peptidylarginine deiminase-3 and caspase-14 proteins
Although multistep processes of profilaggrin expression and degradation into free AA are poorly understood, several regulators including activator protein-1 (AP-1) (Jun and/or Fos)(Reference Jang, Steinert and Markova29)and PPARγ(Reference Mao-Qiang, Fowler and Schmuth30) have been reported to enhance the profilaggrin expression during terminal differentiation of the epidermis(Reference Sandilands, Sutherland and Irvine5). Once synthesised, >220 kDa profilaggrin is initially dephosphorylated by phosphatase type 2A and rapidly cleaved to filaggrin (28 kDa) by a variety of proteases including matriptase, protasin and probably kallikrein-5, which then binds to and condenses the keratin cytoskeleton with transglutaminases and PAD, specifically PAD3(Reference Sandilands, Sutherland and Irvine5, Reference Prealand, Joseph and Lawrence31). Ultimately, filaggrin monomers undergo degradation by a variety of proteases including caspase-14, into free AA, the major constituents of NMF(Reference Sandilands, Sutherland and Irvine5, Reference Prealand, Joseph and Lawrence31).
Of the various regulators and proteases involved in profilaggrin expression and degradation processes(Reference Sandilands, Sutherland and Irvine5, Reference Jang, Steinert and Markova29–Reference Prealand, Joseph and Lawrence31), the protein expressions of PPARγ, PAD3 and caspase-14 are shown in Fig. 4. The protein expressions of PPARγ, PAD3 and caspase-14 in group CA were significantly less than those in group C. In group S, the expressions of these proteins were higher than those in group CA, and were similar (PAD3 and caspase-14) to or even higher (PPARγ) than those in group C. Although the protein expressions of PPARγ and PAD3 were similar to those in group C, the protein expression of caspase-14 in group F was less than that in group C. Although further studies are required to investigate the dietary modulation of profilaggrin expression and degradation processes with sericin supplements in depth, these results indicate that the increase in PPARγ, PAD3 and caspase-14 may, at least in part, contribute to an initial increase in profilaggrin expressions, and further to the enhanced degradation of the intermediates of filaggrins and filaggrin itself into free AA, ultimately recovering dry skin conditions that were onset with AD induction.
Altered protein expression of PPARγ, peptidylarginine deiminase 3 (PAD3) and caspase-14 in the epidermis of BALB/c mice fed a control diet (group C) and NC/Nga mice fed a control diet (group CA) or diets supplemented with 1 % sericin (group S) or fibroin (group F) for 10 weeks. Protein extracts were electrophoretically separated on 10 % SDS-PAGE gel and immunoblotted with primary antibodies against PPARγ (57/54 kDa), PAD3 (75 kDa) or caspase-14 (30 kDa). (A) Representative expressions of PPARγ, PAD3 and caspase-14 in the epidermis of mice. (B) The signal intensities from multiple experiments of (A) were quantified and the integrated areas were normalised to the corresponding value of keratin-5 (loading control). Values are means (n 8), with standard errors represented by vertical bars. a,b,c Mean values with unlike letters in each protein expression were significantly different (P < 0·05; one-way ANOVA and Duncan's multiple comparison test).
Free amino acid analysis
Comparison of free AA contents in the epidermis of each group is shown in Table 3 and Fig. 5. The total amounts of AA in groups S and F were higher than those in group CA, and were similar to those in group C (Table 3). Further analysis of the individual AA content in group C of BALB/c mice demonstrated that glutamate and serine were most abundant in the epidermis, followed by lysine, glycine, histidine, aspartate and alanine, revealing that glutamate and serine are the most prominent free AA, comprising almost 62·0 and 28·8 % of total free AA, respectively, in the epidermis (Table 3 and Fig. 5). Arginine, phenylalanine, valine, cysteine and tyrosine were barely detected, and the combined contents of these free AA were only 0·7 % of total free AA in the epidermis. Isoleucine, leucine, methionine, threonine, tryptophan and proline were not detected in the epidermis of group C, similar to the AA profile of filaggrin in BALB/c mice as well as the AA profile of the human epidermis, in which glutamate, serine and glycine are the major AA(Reference Scott, Harding and Barrett4).
Free amino acid (AA) contents in the epidermis of mice (Mean values with their standard errors, n 10)
C, BALB/c mice fed a control diet; CA, NC/Nga mice fed a control diet; S, NC/Nga mice fed a diet supplemented with 1 % of sericin; F, NC/Nga mice fed a diet supplemented with 1 % fibroin; EAA, essential AA; NEAA, non-essential AA; ND, not detected.
a,b,c Mean values within a row with unlike superscript letters were significantly different (P < 0·05; one-way ANOVA and Duncan's multiple comparison test).
Comparison of free amino acid contents in the epidermis of BALB/c mice fed a control diet (group C, ■) and NC/Nga mice fed a control diet (group CA, ) or diets supplemented with 1 % sericin (group S, 
) or fibroin (group F, 
) for 10 weeks. a,b,c Mean values with unlike letters for each amino acid were significantly different (P < 0·05; one-way ANOVA and Duncan's multiple range comparison test).
In group CA of NC/Nga mice, glutamate was not detected and the serine content was significantly lower than that in group C, which was only at about 25·8 % of that in group C (Table 3 and Fig. 5). In contrast, methionine, which was not detected in group C, was abundant and alanine was modestly, but significantly increased in group CA. After dietary supplementation of sericin in NC/Nga mice, serine as well as glycine, arginine and valine in group S were significantly higher than those in group CA. In addition, glutamate, which was not detected in group CA, was modestly increased at about 26·7 % of that in group C. Furthermore, methionine, which was highly detected in group CA, was decreased, and alanine was not detected in group S. In group F, serine and glutamate were detected at very low levels, and glycine, arginine and valine were either barely or not detected. Although methionine was not detected, tyrosine, cysteine, phenylalanine and isoleucine, minor free AA in NMF(Reference Scott, Harding and Barrett4), which were either barely or not detected in other groups, were highly detected in group F. These data demonstrated that recovery of serine and glutamate, the major AA in NMF(Reference Scott, Harding and Barrett4), coupled with decreases in methionine, normalised the free AA contents of group S, in parallel with enhancing the epidermal hydration (Fig. 1). Despite no detection of methionine, dietary supplementation of fibroin did not normalise the free AA contents in the epidermis without the recovery of glutamate and serine, or with abnormal increases of minor free AA in NMF.
Discussion
Although filaggrin is initially synthesised as profilaggrin and degraded into free AA, the major components of NMF(Reference Harding and Scott2, Reference Harding and Scott3), little is known about the dietary modulation of profilaggrin and filaggrin levels or of free AA levels related to improving the dry skin condition such as AD. The present study indicates that decreased levels of total filaggrins including profilaggrin, 3RI or 2RI of filaggrins, and filaggrin are paralleled by the reduction of epidermal hydration in group CA. The low protein expressions of filaggrins in group CA could be due to either the decreased synthesis of filaggrins or their enhanced degradation into free AA in dry skin conditions. However, although filaggrin degradation is further stimulated in the dry condition (stable in the hydrated condition)(Reference Scott and Harding32), the total free AA in group CA were significantly lower than those in group C, which is consistent with the results of previous studies, in which reduced amounts of total free AA are paralleled with the decreased immunostaining intensity or protein expression of filaggrins in the dry epidermis of either AD or ichthyosis patients(Reference Howell, Kim and Gao6–Reference Seguchi, Cui and Kusuda8). Therefore, the reduction of epidermal hydration in group CA is more probably due to the decreased expressions of filaggrins.
In the same manner, the recovery of epidermal hydration in group S to a level similar to that of group C could be explained by the increased expressions of filaggrins. In fact, the protein expressions of total filaggrins in group S were significantly higher than those in group CA and even higher (in immunofluorescence analysis) than those in group C. However, further analysis of filaggrin species in group S reveals that despite higher expressions of profilaggrin and 3RI proteins than those in group C, the protein expression of 2RI was lower than that in group C, and the protein expression of filaggrin was further decreased similar to that in group CA. In addition, total amounts of free AA in group S were significantly higher than those in group CA, and were increased to a similar level to that in group C, indicating that dietary sericin initially increases the protein expression of profilaggrins and enhances their degradation into free AA.
Sericin has low digestibility with pepsin and pancreatin due to its protease-resistant property(Reference Sasaki, Yamada and Kato33). Undigested sericin functions as a dietary fibre in the intestine, thereby improving constipation(Reference Sasaki, Yamada and Kato33) or increasing the faecal excretion of cholesterol and TAG, ultimately altering systemic lipid profiles for the anti-hyperlipidaemic effect(Reference Seo, Um and Rico34). In a similar manner, undigested sericin in group S could alter epidermal barrier lipids in which ceramides are composed of a multilayered lamellar structure together with cholesterol and fatty acids(Reference Rawlings and Harding1). In our previous study(Reference Park, Choi and Kim14), the epidermal level of total ceramides, which was significantly decreased in group CA, was significantly increased in group S (1 % sericin and 22 % casein in the diet of group S). However, addition of a low percentage of sericin in casein does not inhibit pepsin and pancreatin digestion(Reference Sasaki, Yamada and Kato33). Furthermore, when either peptide (prepared by proteases treatment in vitro: molecular weight confirmed to be about 5410(Reference Park, Choi and Kim14)) or AA types of sericin are supplemented in the diet as the same percentage of the protein type of sericin (group S), the epidermal level of total ceramides is far less than that in group CA(Reference Park, Choi and Kim14). These data, coupled with previous reports, in which although serum lipid levels were decreased, faecal excretion and digestibility of lipids with 4 % dietary sericin did not differ from that of the control group(Reference Okazaki, Kakehi and Xu35), suggest that the increased levels of ceramides(Reference Park, Choi and Kim14) and filaggrins, and/or the enhanced filaggrin degradation into free AA in group S could not be explained solely by the altered systemic lipid profile with the low digestibility of sericin. Alternatively, dietary sericin increases the serum level of adiponectin(Reference Okazaki, Kakehi and Xu35), which indirectly mediates a decrease in hepatic lipogenesis(Reference Awazawa, Ueki and Inabe36), and specifically increases the secretion of IgA from the colon without affecting the caecal profile of the microflora(Reference Okazaki, Tomotake and Tsujimoto37). In our previous study(Reference Rawlings and Harding1, Reference Holleran, Takagi and Uchida38), dietary sericin inhibited the protein expression of ceramidase, the degradative enzyme of ceramides, which was significantly increased in group CA(Reference Park, Choi and Kim14). Furthermore, in the present study(Reference Sandilands, Sutherland and Irvine5), dietary sericin increased the protein expressions of PPARγ, PAD3 and caspase-14, a regulator or proteases involved in filaggrin metabolism, suggesting that ingested sericin or partially digested sericin peptide fragments, of which the AA sequences might be distinct from those prepared by the proteases treatment in vitro (Reference Park, Choi and Kim14), could indirectly promote systemic responses to alter the protein expression and/or the activity of target enzymes or regulators in filaggrin or ceramide metabolism. The systemic responses of dietary sericin to filaggrin and/or ceramide metabolism could be due to complex mechanisms; these remain to be elucidated in future studies.
In contrast to the beneficial effects of sericin, fibroin, another silk protein examined in the present study, did not recover dry skin conditions despite increased levels of filaggrins and total free AA to levels similar to those found in group C. Compared with high serine content (30·4 % of the total AA) and the resemblance of the AA profiles of sericin to those of NMF(Reference Kato, Sato and Yamanaka13), glycine (44·4 %) and alanine (29·3 %) are the most prominent AA and serine accounts for only 12·1 % of the total AA in fibroin. Glycine and serine can be converted to each other by a reversible reaction in mammal tissue(Reference Fallon, Davis and Goyer39); dietary supplementation of the silk protein fibroin could possibly provide large quantities of serine, therefore increasing the synthesis of filaggrins and ceramides, the major lipid of the epidermal barrier which is synthesised de novo with the enzymatic condensation of serine and palmitoyl-CoA by serine palmitoyl transferase(Reference Rawlings and Harding1, Reference Holleran, Takagi and Uchida38), to the similar level of group C. However, epidermal levels of total filaggrins as well as total ceramides in group F(Reference Park, Choi and Kim14) were lower than those in group S, suggesting that serine supplied by glycine conversion in group F was less efficient for the synthesis of either filaggrin or ceramides in the epidermis than serine directly supplied in group S. Although epidermal levels of total free AA as well as of PPARγ and PAD3 proteins were increased to the similar level of group C and methionine, which is highly detected in group CA, was not detected, epidermal levels of the major AA of NMF such as glutamate and serine were not increased in group F. Notably, tyrosine, phenylalanine, cysteine and isoleucine, minor AA of NMF(Reference Scott, Harding and Barrett4), were highly detected and the protein expression of caspase-14, a protease of filaggrin degradation into free AA, was less than that in group C. These data, coupled with the identification of possible substrates other than filaggrins for transglutaminases, PAD or protease(Reference Tarcsa, Marekov and Mei40, Reference Fischer, Stichenwirth and Dockal41), suggest that dietary fibroin did not enhance filaggrin degradation into free AA, and/or possible substrates other than filaggrins could become proteolysed by these enzymes, resulting in the abnormal free AA contents and therefore not enough for recovering the dry skin condition in group F despite the increase in total filaggrins and free AA to the similar level of group C.
Although the systemic responses of dietary sericin to filaggrin metabolism, specifically modulation of the activities or expressions of regulators and proteases, require further elucidation in depth, the present in vivo study demonstrated a novel observation that dietary supplementation of the silk protein sericin improved epidermal hydration in NC/Nga mice, in parallel with the profilaggrin accumulation and its enhanced degradation into free AA that is coupled with elevated protein expressions of PPARγ, PAD3 and caspase-14, a regulator or proteases in filaggrin metabolism. Sericin may be a potential alternative therapeutic agent, or may serve as an adjunct to conventional therapies for dry skin conditions such as AD.
Acknowledgements
We wish to thank Dr Peter M. Elias in the Dermatology Service and Research Unit, Department of Veterans Affairs Medical Center, University of California-San Francisco, CA, USA for his personal communication and precious advice on epidermal hydration and filaggrin degradation. This study was supported by a grant of the National Research Foundation (2010-0016554), Republic of Korea. H. K. was responsible for the development of the study design, evaluation and the interpretation of the data and statistics, and was involved in the drafting and writing of the manuscript. Y.-J. L. and J.-H. P. were responsible for the immunofluorescence analysis. Y. C. was responsible for the supervision of the analytical work, writing and final approval of the manuscript. All authors read and approved the final manuscript. None of the authors has any conflicts of interest.
