Original Paper

Skin Pharmacol Physiol 2015;28:227–235 Received: September 15, 2014 DOI: 10.1159/000369830 Accepted after revision: November 8, 2014 Published online: February 14, 2015

Effects of Food-Derived Peptides on the Expression of and Keratin-Associated in the Mouse Skin

a a a a a, e Phuong Le Vu Ryo Takatori Taku Iwamoto Yutaka Akagi Hideo Satsu a b d a, c Mamoru Totsuka Kazuhiro Chida Kenji Sato Makoto Shimizu

a b c Departments of Applied Biological Chemistry and Animal Resource Sciences, The University of Tokyo, and Department d of Nutritional Science, Tokyo University of Agriculture, Tokyo , Division of Applied Life Sciences, Kyoto Prefectural e University, Kyoto , and Department of Biotechnology, Maebashi Institute of Technology, Maebashi , Japan

Key Words were cultured without fibroblasts, suggesting that the pres- Coculture · Collagen · Food-derived collagen peptides · ence of fibroblasts is essential for the effects of Pro-Hyp. Our expression · Keratinocytes · Skin study presents new insights into the effects of CP on the skin, which might link to the cycle. © 2015 S. Karger AG, Basel

Abstract Oral ingestion of collagen peptides (CP) has long been sug- gested to exert beneficial effects on the skin, but the mo- Introduction lecular events induced by CP on the skin remain unclear. Here, we investigated the effects of oral CP administration The skin is the foremost protective barrier of the body, on gene expression in hairless mouse skin and of prolyl-hy- which acts to prevent environmental harmful factors as droxyproline (Pro-Hyp), a collagen-derived dipeptide, on well as water loss. The outer compartment of the skin is gene expression in a coculture of mouse skin keratinocytes the epidermis, which comprises 4 layers of keratinocytes, and fibroblasts. Using microarray analysis, we found that oral which functions as a physical, biochemical and immuno- administration of CP to hairless mice for 6 weeks induced logical barrier of the skin. Underlying the epidermis is the increased expression of Krtap and Krt genes in the skin. An- dermis, which comprises mainly fibroblasts in an extra- notation analysis using DAVID revealed that a group of the cellular matrix [1]. The current literature indicates that up-regulated genes, Gprc5d, Sprr2a1, Krt27 and Krtap16-7, dermal extracellular matrix remodeling is involved in is associated with the development of the epidermis and the skin homeostasis, aging and wound healing [2], and that hair cycle. In addition, the presence of Pro-Hyp (200 μM ) in- dermis-epidermis interactions play an indispensable role duced an increase in the expression of Krtap16-7, Krtap15, in the regulation of these processes [3, 4] . Furthermore, Krtap14 and Krtap8-2 in keratinocytes in coculture, partially the molecular signals exchanged between these compart- resembling the in vivo result. The Pro-Hyp-induced up-reg- ments also mediate hair follicle formation and hair cy- ulation of these genes was not observed when keratinocytes cling [5] .

© 2015 S. Karger AG, Basel Makoto Shimizu 1660–5527/15/0285–0227$39.50/0 Department of Nutritional Science Tokyo University of Agriculture, Sakura-gaoka E-Mail [email protected] 1-1-1, Setagaya-ku, Tokyo (Japan) www.karger.com/spp E-Mail ms205346 @ nodai.ac.jp Downloaded by: Kyoto University 130.54.110.72 - 3/5/2015 2:07:27 AM Not only is the skin affected by extrinsic factors, it is animal model [13] , and intestinal absorption and tissue also altered by intrinsic aging. Attempts have been made distribution of gelatin hydrolysates, hydrolyzed products to search and analyze functional food compounds that of collagen, were assessed in a study using 14 C-labeled hy- can protect the skin from damage and aging. Among the drolysates [14, 15] . Recent clinical studies have also estab- most studied food substances are polyphenols from vari- lished that oral intake of collagen peptides (CP) reduces ous sources. Oral administration of baicalein (5,6,7-trihy- skin wrinkles and improves biophysical parameters in droxyflavone) and wogonin (5,7-dihydroxy-8-methoxy- aged skin [16, 17] . Furthermore, following the ingestion flavone) extracted from Scutellaria baicalensis roots, for of gelatin hydrolysates, prolyl-hydroxyproline (Pro-Hyp) example, was reported to suppress ultraviolet-induced was significantly increased in human blood, which skin damage [6] . Green tea polyphenols, especially EGCG reached up to 200 μ M in some individuals, among the [(–)-epigallocatechin-3-gallate], was found to have pro- collagen-derived di- and tripeptides and free amino acids tective effects against photocarcinogenesis when topical- [18] . Pro-Hyp was found to have a stimulatory effect on ly applied to hairless mice [7, 8]. In addition, beneficial cell growth of mouse skin fibroblasts, leading to the hy- effects of dietary ceramides, such as glucosylceramides pothesis that it is the key peptide in biological activities of and sphingolipids, in restoring skin functions have been dietary collagen [19] . suggested. Hasegawa et al. [9] reported improved corni- Although oral CP administration was proven to pro- fied envelope formation in the skin following oral admin- tect mouse skin from the UV-induced transepidermal istration of glucosylceramide. Another ceramide deriva- water loss and UV-induced matrix metalloproteinase 1 tive was found to have an inhibitory effect on skin inflam- protein expression [20] , little is known about the effects mation [10] . Apart from polyphenols and ceramides, the of CP on normal (nonirradiated) skin, especially their potent benefits of diet-derived collagen on the skin have effects on gene expression levels in the skin. We per- also been considered [11] . Long-term oral administration formed DNA microarrays to identify the effects of CP of marine collagen hydrolysates was found to exert pro- ingestion on gene expression levels in hairless mouse tective effects on skin aging, which was attributable to the skin under normal conditions (nonirradiated). In addi- antioxidant property of collagen [12] . tion, we addressed the effects of Pro-Hyp on gene ex- , along with fibrillins, and other pression in a coculture of mouse skin keratinocytes and macromolecules, are synthesized in fibroblasts and se- fibroblasts. creted into the matrix, where they are cross-linked to form a complex network. These characterize the structure of the dermis, and, most importantly, they are Materials and Methods involved in various signaling pathways to maintain skin homeostasis. The biosynthesis process of collagen is initi- Materials Porcine CP mixture (mean molecular weight: 5,000) was ob- ated via the transcription of collagen genes, followed by tained from Nitta Gelatin (Osaka, Japan) and dissolved in distilled the translation into precursor polypeptides. The structure water at 0.02 g/ml. Pro-Hyp was purchased from Bachem (Buben- of these polypeptides is featured by the repeat of Gly-X-Y dorf, Switzerland) and reconstituted in methanol according to the triplets, where X and Y are often proline and hydroxypro- manufacturer’s instructions. Reduced serum medium (MCDB) line. The precursor polypeptides, termed α-chains, are and fetal bovine serum (FBS) were from Sigma (St. Louis, Mo., USA). Collagenase was purchased from Wako (Osaka, Japan). then modified in the rough endoplasmic reticulum and Ham’s F12 medium was from Life Technologies (Tokyo, Japan). the Golgi apparatus. Hydroxylation of prolyl and lysyl Cell matrix type I-A was from Nitta Gelatin. The RNeasy mini kit residues is one of the important modifications which lead for RNA extraction in the in vivo experiment was from Qiagen to the assembly of three α-chains to form a triple helix (Tokyo, Japan). ISOGEN for RNA extraction in the in vitro ex- called procollagen. Once secreted into the extracellular periment was from Nippon Gene (Tokyo, Japan). Cell culture in- serts (12-well plate format) with a polyethylene terephthalate space, the amino and carboxyl extensions of procollagen (PET) membrane and companion 12-well plates were from Becton are removed, followed by the alignment of procollagens Dickinson (Franklin Lakes, N.J., USA). to generate collagen fibers. These fibers are cross-linked with other components of the extracellular matrix, form- Animals and Diets ing a bioactive network [2] . Six-week-old female hairless mice (Hos:HR-1) were purchased from Japan SLC (Shizuoka, Japan) and kept at 25 ° C under a Dietary collagen can be derived from sources such as 12-hour-light/12-hour-dark cycle during the experiment. They fish, porcine and chicken skin, and chicken tendon and were allowed free access to water and normal rodent diet (Labo MR cartilage. Oral intake of collagen was proven safe in an Stock in pellet form; Nosan, Kanagawa, Japan). Following 5 days

228 Skin Pharmacol Physiol 2015;28:227–235 Le Vu/Takatori/Iwamoto/Akagi/Satsu/ DOI: 10.1159/000369830 Totsuka/Chida/Sato/Shimizu Downloaded by: Kyoto University 130.54.110.72 - 3/5/2015 2:07:27 AM of acclimatization, mice were randomly divided into two groups: the control group and the collagen group (n = 5). The CP mixture was administered daily to the collagen group at 0.2 g/kg body Primary mouse weight p.o. whereas an equal volume of sterile water was adminis- skin keratinocytes tered to the control group. The amount of CP administered was Primary mouse based on a previous study by Tanaka et al. [20] . After 6 weeks, mice skin fibroblasts were sacrificed by cervical dislocation. Dorsal skin samples were obtained and snap-frozen in liquid nitrogen. The samples were stored at –80 ° C until analyzed.

DNA Microarray Analysis Total RNA was extracted from the dorsal skin samples using the RNeasy mini kit according to the instructions, and the quality was determined using an Agilent 2100 Bioanalyzer. Equal amounts of RNA from 5 mice/group were pooled to normalize individual Skin-like layer differences for microarray analysis. Sense-strand cDNA was syn- thesized using an Ambion wild-type (WT) expression kit follow- ing the manufacturer’s protocol. cDNA was fragmented and la- Pro-Hyp beled using an Affymetrix GeneChip WT terminal labeling kit.

The fragmented and labeled cDNA was hybridized at 45 ° C for 17 h to the Mouse Gene 1.0 ST Array (Affymetrix, Santa Clara, Calif., USA). The array was washed and stained using the Affymetrix GeneChip Fluidics Station 450 and then scanned using the Af- fymetrix GeneChip Scanner 3000 7G. Data were analyzed by means of ArrayStar software (DNASTAR Navigator) and normal- ized using the robust multi-array average procedure. Annotation analysis for differentially expressed genes was done by means of DAVID (Database for Annotation, Visualization and Integrated Discovery) following previously published instructions [21, 22] . Fig. 1. Coculture of primary mouse skin keratinocytes and fibro- blasts. Four days after air exposure, Pro-Hyp (200 μ M , reconsti- Coculture Method tuted in methanol) was added to the medium at the basal side of Primary mouse skin fibroblasts (from 2- or 3-day-old BALB/c the culture for 24 h, while methanol was used as control. The skin- mice; Nihon Clea, Tokyo, Japan) and keratinocytes (from 2-day- like layer and the dermis equivalent were then separated for gene old BALB/c mice) were cocultured in a three-dimensional model expression analysis using real-time RT-PCR. described previously [23] . Briefly, keratinocytes were overlaid on the surface of a collagen gel containing fibroblasts (dermis equiva- lent) and then cultured in the MCDB medium supplemented with 10% dialyzed FBS for 1 day. The medium was then replaced with fresh medium containing 200 μM of Pro-Hyp or an equivalent a mixture (1: 1) of MCDB and FAD medium (1 volume of Ham’s volume of methanol. Cells were incubated for 24 h in a humidi- F12, 3 volumes of DMEM and 7.5% FBS) for another day. The co- fied incubator. Total RNA was extracted using ISOGEN, and culture was then exposed to an air-liquid interface to allow strati- gene expression changes were determined using real-time fication of keratinocytes. The medium in the basal side was re- RT-PCR. placed daily with fresh FAD medium. Four days after exposure, fresh FAD medium containing 200 μ M of Pro-Hyp (reconstituted Real-Time RT-PCR in methanol) was added to the basal side of the coculture, whereas The 5× PrimeScriptTM RT Master Mix (Takara, Shiga, Japan)

FAD containing methanol was used as the control. The coculture was used for cDNA synthesis at 37 ° C for 15 min and at 85 ° C for ® was then incubated in a humidified incubator at 37 ° C for 24 h. The 5 s. Real-time RT-PCR was performed using the SYBR Premix Ex skin-like layer and the dermis equivalent were then separated TaqTM (Takara) on the LightCycler system (Roche Diagnostics). ( fig. 1 ). Total RNA was extracted from the skin-like layer and from The following thermal profile was used for all PCR reactions: pre- the fibroblasts using ISOGEN following the manual. Gene expres- heating at 95 ° C for 10 s, followed by 40 cycles that involved dena- sion changes in the two parts were determined using real-time turing at 95 ° C for 5 s, annealing at 56 ° C for 20 s and extension at

RT-PCR. 72 ° C for 20 s. Hprt (hypoxanthine phosphoribosyltransferase) was used as the reference gene, and the ΔΔC t formula was used to cal- Primary Keratinocyte Monolayer Culture culate gene expression levels. The following primer sequences Primary mouse skin keratinocytes were isolated from 2-day- were used: old BALB/c mice as described previously [23] . The cells were Hprt, 5′ -ctggtgaaaaggacctctcg-3′ and 5′ -tgaagtactcattatagtca then cultured in fibroblast-conditioned medium on a collagen- agggca-3′ ; coated plate according to the method described by Häkkinen et Gprc5d, 5′ -gtctcctccttactcatttgcag-3′ and 5′ -gaacggatgtctgg al. [24] . Four days after plating, the medium was replaced with tctcct-3′ ;

Effects of Collagen Peptides on Gene Skin Pharmacol Physiol 2015;28:227–235 229 Expression in the Skin DOI: 10.1159/000369830 Downloaded by: Kyoto University 130.54.110.72 - 3/5/2015 2:07:27 AM Table 1. Keratin-associated protein genes up-regulated in the col- Krt27, 5′ -gagatgaaacggactctccaa-3′ and 5′ -tcagggagcactccag lagen group aga-3′ ;

Sprr2a1, 5′ -ctgctccggagaacctgat-3′ and 5′ -acaaggctcaggg Gene symbol Gene name Fold change cacttc-3′ ;

Krtap16-7, 5′ -gctcaggctatggttgtgg-3′ and 5′ -gcagcaagaggaatag Krtap8-2 keratin associated protein 8-2 4.28 catttt-3′ ;

Krtap16-5 keratin associated protein 16-5 3.27 Krtap14, 5′ -gctgggctcttctcttcaca-3′ and 5 ′ -ataggctgtctggc Krtap15 keratin associated protein 15 3.23 agttgg-3′ ;

Krtap6-3 keratin associated protein 6-3 2.97 Krtap15, 5′ -ccccaccagcaatgtagtct-3′ and 5 ′ -ttcaagtggctcactga Krtap16-8 keratin associated protein 16-8 2.71 agg-3′ ;

Krtap6-1 keratin associated protein 6-1 2.67 Krtap8-2, 5′ -ctttgggaacctgggttatg-3′ and 5 ′ -tctgccatagccagatc Krtap16-2 keratin associated protein 16-2 2.66 cat-3′ ; Krtap16-1 keratin associated protein 16-1 2.59 Krt1, 5′ -tttgcctccttcatcgaca-3′ and 5 ′ -gttttgggtccgggttgt-3′ ;

Krtap6-1 keratin associated protein 6-1 2.40 Krt10, 5′ -cgtactgttcagggtctggag-3′ and 5′ -gcttccagcgattgt ttca- Krtap2-4 keratin associated protein 2-4 2.31 3 ′ ;

Krtap16-7 keratin associated protein 16-7 2.29 Krt6a, 5′ -gccaaggcagacagtctaaca-3′ and 5 ′ -caggctacggttgttg tcc- Krtap3-3 keratin associated protein 3-3 2.26 3 ′ , and Krtap14 keratin associated protein 14 2.24 Krt16, 5′ -tgagctgaccctgtccaga-3′ and 5′ -ctcaaggcaagcatctcctc-3′ . Krtap4-2 keratin associated protein 4-2 2.19 Krtap4-1 keratin associated protein 4-1 2.10 Statistical Analysis Krtap6-2 keratin associated protein 6-2 2.10 Real-time RT-PCR results were expressed as means ± SEM. Krtap5-5 keratin associated protein 5-5 2.03 Gene expression levels in treated samples were compared with those Krtap24-1 keratin associated protein 24-1 2.01 in untreated samples using one-way ANOVA, followed by Fisher’s Krtap3-1 keratin associated protein 3-1 1.96 LSD post hoc test to determine the level of significance (p < 0.05). Krtap5-1 keratin associated protein 5-1 1.96 Krtap4-7 keratin associated protein 4-7 1.95 Krtap9-1 keratin associated protein 9-1 1.90 Results Krtap1-3 keratin associated protein 1-3 1.88 Krtap4-16 keratin associated protein 4-14 1.85 Krtap5-4 keratin associated protein 5-4 1.85 Effect of Oral Administration of the CP Mixture on Krtap10-4 keratin associated protein 10-4 1.81 Gene Expression Levels in the Skin Krtap16-10b keratin associated protein 16-10b 1.80 DNA microarray analysis was carried out to examine Krtap4-8 keratin associated protein 4-8 1.80 Krtap7-1 keratin associated protein 7-1 1.79 the effect of oral CP administration on the skin. The ex- Krtap10-10 keratin associated protein 10-10 1.78 pression levels of 104 genes were shown to be up-regulat- Krtap5-2 keratin associated protein 5-2 1.77 ed by more than 1.5-fold in the CP group compared with Krtap4-13 keratin associated protein 4-13 1.77 the control group. Forty-one keratin-associated protein Krtap1-5 keratin associated protein 1-5 1.75 genes (table 1 ) and 6 keratin genes (table 2 ) were up-reg- Krtap5-3 keratin associated protein 5-3 1.71 Krtap4-6 keratin associated protein 4-6 1.66 ulated in the CP group. Sixteen genes of the olfactory re- Krtap13-1 keratin associated protein 13-1 1.66 ceptor family were also up-regulated (online suppl. ta- Krtap3-2 keratin associated protein 3-2 1.62 ble 1; for all online suppl. material, see www.karger.com/ Krtap16-3 keratin associated protein 16-3 1.57 doi/10.1159/000369830). Krtap16-4 keratin associated protein 16-4 1.56 Annotation analysis using DAVID revealed that a Krtap1-4 keratin associated protein 1-4 1.55 Krtap8-1 keratin associated protein 8-1 1.55 group of CP-induced up-regulated genes, Gprc5d, Krt27, Sprr2a1 and Krtap16-7 (1.61-, 1.53-, 1.88- and 2.29-fold, respectively), was associated with the development of the epidermis with a 5.3-fold enrichment (p < 0.05). Gprc5d, Table 2. Keratin genes up-regulated in the collagen group Krt27 and Krtap16-7 were also identified as a group re- lated to the hair cycle with a 9.56-fold enrichment (p < Gene symbol Gene name Fold change 0.05). Validation using real-time RT-PCR showed that Krt86 keratin 86 2.05 the expression levels of Gprc5d, Krt27, Sprr2a1 and Krt33a keratin 33A 1.77 Krtap16-7 in the CP group had a strong tendency to in- Krt34 1.68 duction compared with those in the control group, al- Krt31 keratin 31 1.61 though this was not statistically significant (fig. 2 ). Krt40 keratin 40 1.55 Oral administration of CP down-regulated the expres- Krt27 keratin 27 1.53 sion levels of 67 genes in the skin, including those coding

230 Skin Pharmacol Physiol 2015;28:227–235 Le Vu/Takatori/Iwamoto/Akagi/Satsu/ DOI: 10.1159/000369830 Totsuka/Chida/Sato/Shimizu Downloaded by: Kyoto University 130.54.110.72 - 3/5/2015 2:07:27 AM Gprc5d Sprr2a1 Krt27 Krtap16-7 5 5 3.5 10 3.0 4 4 8 2.5 3 3 2.0 6 2 2 1.5 4 1.0 1 1 2

Relative fold change Relative fold change Relative fold change Relative 0.5 fold change Relative 0 0 0 0 Control Collagen Control Collagen Control Collagen Control Collagen

Krtap15 Krtap8-2 Krtap14 4.0 14 6 3.5 12 5 3.0 10 2.5 4 8 2.0 3 6 1.5 2 1.0 4

Relative fold change Relative 1 0.5 fold change Relative 2 fold change Relative 0 0 0 Control Collagen Control Collagen Control Collagen

Fig. 2. Validation of the microarray result of CP-induced up-reg- reference gene. Values are means ± SEM (n = 5). Statistical analy- ulated genes related to the development of the epidermis and the sis using ANOVA, followed by Fisher’s LSD post hoc test, did not hair cycle. Total RNA from the dorsal skin sample of each mouse reveal any significant differences, although there was a strong in a given group was used for cDNA synthesis. Real-time RT-PCR trend to up-regulation. was used to determine mRNA expression. Hprt was used as the

for metabolic enzymes, such as Cyp17a1, St6galnac1, was added to the medium in the basal side 4 days after Ndufb6 and Idh3a (online suppl. table 2). Cyp17a1, in air exposure. particular, is involved in the synthesis of cortisol from Firstly, we assessed the impact of Pro-Hyp on the ex- cholesterol [25, 26]. Other genes involved in cholesterol pression levels of Gprc5d, Krt27, Krtap16-7, and Sprr2a1 metabolism in the skin (Cyp11a1, Cyp19, Hsd11b1 and in the coculture. Keratinocytes (the skin-like layer) and Hsd11b2), however, were not affected by oral CP admin- fibroblasts (the dermis equivalent) in coculture expressed istration in this experiment. In contrast with up-regulat- comparable levels of Gprc5d and Krtap16-7, whereas ed genes, the annotation analysis of the CP-induced Sprr2a1 and Krt27 were only detectable in keratinocytes down-regulated genes did not result in any gene clusters (data not shown). Only Krtap16-7 expression in kerati- that were specific for skin functions. nocytes was induced in the presence of Pro-Hyp (fig. 3 ). We sought to determine the effect of Pro-Hyp on the ex- Effect of the Collagen Dipeptide Pro-Hyp on the pression of other Krtap genes that were up-regulated Expression Levels of Krtap Genes in Keratinocytes in more than 2-fold in vivo. In cocultured keratinocytes, Mouse Skin Coculture Krtap15, Krtap14 and Krtap8-2 were increased ∼ 2-fold To explore whether the collagen-induced gene ex- in the presence of Pro-Hyp (fig. 3 ), resembling the move- pression changes observed in the above-mentioned in ment of these genes found in our in vivo experiment. vivo experiments occur in vitro, we used the coculture model of primary mouse skin keratinocytes and fibro- Effect of Pro-Hyp on the Expression of Keratinocyte blasts described by Ikuta et al. [23] . This coculture mod- Differentiation Markers in Mouse Skin Coculture el enables the investigation of two cell types in an archi- Although Pro-Hyp was reported to stimulate skin fi- tecture that resembles the skin in vivo. The coculture was broblast cell growth [19] , the effects of this peptide on exposed to an air-liquid interface, and Pro-Hyp (200 μM ) keratinocyte differentiation and proliferation have not

Effects of Collagen Peptides on Gene Skin Pharmacol Physiol 2015;28:227–235 231 Expression in the Skin DOI: 10.1159/000369830 Downloaded by: Kyoto University 130.54.110.72 - 3/5/2015 2:07:27 AM Krtap16-7 Krtap14 Krtap15 Krtap8-2 2.5 2.0 3.0 2.5 ** * 2.0 ** 2.5 * 2.0 1.5 2.0 1.5 1.5 1.0 1.5 1.0 1.0 1.0 0.5 0.5 Relative fold change Relative

Relative fold change Relative 0.5 Relative fold change Relative Relative fold change Relative 0.5

0 0 0 0 Control Pro-Hyp Control Pro-Hyp Control Pro-Hyp Control Pro-Hyp

Fig. 3. Pro-Hyp-induced up-regulation of Krtap genes in keratinocytes in coculture. Hprt was used as the reference gene. Values are representative of three independent coculture experiments and are expressed as means ± SEM (n = 3); * * p ≤ 0.01, * p ≤ 0.05 by ANOVA, followed by Fisher’s LSD post hoc test.

Krt1 Krt10 Krt6a Krt16 1.2 1.2 1.2 1.2

1.0 1.0 1.0 1.0 * 0.8 0.8 0.8 0.8 * ** 0.6 0.6 ** 0.6 0.6 0.4 0.4 0.4 0.4 Relative fold change Relative fold change Relative 0.2 0.2 fold change Relative 0.2 fold change Relative 0.2

0 0 0 0 Control Pro-Hyp Control Pro-Hyp Control Pro-Hyp Control Pro-Hyp

Fig. 4. Pro-Hyp-induced down-regulation of keratin genes in keratinocytes in coculture. Hprt was used as the reference gene. Values are representative of three independent coculture experiments and are expressed as means ± SEM (n = 3); * * p ≤ 0.01 * p ≤ 0.05 by ANOVA, followed by Fisher’s LSD post hoc test.

yet been elucidated. Therefore, we sought to determine when keratinocytes were cultured alone in fibroblast- whether Pro-Hyp can affect epidermal differentiation conditioned medium (online suppl. fig. 1), indicating that markers Krt1 and Krt10 using quantitative RT-PCR. The the presence of fibroblast cells and fibroblast-keratino- expression levels of these in keratinocytes in co- cyte interactions played a key role in the mechanism of culture were significantly reduced in the presence of Pro- the effects of Pro-Hyp. Hyp (fig. 4 ), which raised the question of whether the peptide acted to suppress epidermal differentiation. This prompted us to examine the effect of Pro-Hyp on the ex- Discussion pression of Krt6a and Krt16, i.e. markers of hyperprolif- erative keratinocytes. Interestingly, Krt6a and Krt16 ex- Oral administration of the CP mixture up-regulated pression was also reduced after adding Pro-Hyp to the the expression levels of keratin-associated protein genes coculture ( fig. 4 ). Thus, the effects of Pro-Hyp might not and keratin genes in the skin. The CP-induced up-regu- simply involve the stimulation of proliferation and sup- lation of a large number of Krtap and Krt genes is inter- pression of differentiation. esting, since these genes code for structural components More importantly, the addition of Pro-Hyp (200 μM ) of the hair and hair follicles [27] . As an important ap- did not affect the expression levels of Krtap and Krt genes pendage of the skin, hair functions to retain heat and the

232 Skin Pharmacol Physiol 2015;28:227–235 Le Vu/Takatori/Iwamoto/Akagi/Satsu/ DOI: 10.1159/000369830 Totsuka/Chida/Sato/Shimizu Downloaded by: Kyoto University 130.54.110.72 - 3/5/2015 2:07:27 AM hair follicles possess a stem cell pool that can be activated peptide on gene expression in keratinocytes and fibro- to regenerate both follicular and interfollicular cells [28] . blasts, however, have not been studied. The interactions between follicular keratinocytes and To elucidate the effects of a certain specific peptide on dermal papilla cells (specialized fibroblasts of the hair fol- cellular functions of cultured cells in vitro is essential. licle) govern the formation of hair and hair cycling [5] . Although monolayer cultures of keratinocytes and fi- Several plant-derived compounds have been reported to broblasts and reconstituted human epidermis have been induce hair growth, a promising therapeutic treatment used for this purpose [38] , a culture model which resem- for diseases such as alopecia [29, 30]. In contrast, the ef- bles the three-dimensional structure of the skin in vivo fects of animal-derived dietary compounds such as col- is needed. We therefore introduced a sophisticated co- lagen on hair biology have not been explored. culture system of skin keratinocytes and fibroblasts in Given that keratin-associated proteins are produced this study. Keratinocytes differentiated and formed a during the anagen stage (growth stage) of the hair cycle skin-like layer when cocultured with fibroblasts. The ad- [31, 32], the up-regulation of these genes might be related dition of Pro-Hyp to the basal medium of the coculture to an induction of this stage. The increased expression of after air exposure induced increased expression of Krtap and Krt genes collectively might also play a yet un- Krtap16-7, Krtap15, Krtap8-2 and Krtap14, and de- determined role in the skin. A recent study revealed that creased expression of Krt1, Krt10, Krt6a and Krt16 in the the skin was more susceptible to environmental ultravio- skin-like layer of the coculture. The presence of fibro- let radiation during the telogen stage (quiescent stage) blasts appeared to be essential for these effects of the than during the anagen stage of the hair cycle [33]. It will peptide, since the expression of these genes was un- be interesting to determine whether the effect of CP in- changed when Pro-Hyp was added to the keratinocyte take on these genes contributes to skin protection under monoculture. UV irradiation. It has been reported that the expression of Krtap genes Our DNA microarray analysis has suggested that cu- is mediated by several factors, including Hoxc13, Bmp4 taneous Gprc5d is up-regulated by oral administration of and Wnt10b [39–42] . Therefore, we assessed the expres- CP. The expression of Gprc5d, a member of the retinoic sion levels of these factors in keratinocytes cocultured acid-inducible gene-1 family (RAIG-1), was found to be with or without Pro-Hyp, but no significant changes were associated with the expression of some genes found (data not shown), indicating that other factors in the hair cortex in the mid- and late anagen stage of the might be involved in the Pro-Hyp-induced up-regulation hair cycle. In contrast, its expression was undetectable of Krtap genes. during the telogen stage [34] . Therefore, Gprc5d up-reg- Pro-Hyp down-regulated the expression levels of dif- ulation was in accord with the up-regulation of Krt and ferentiation markers in keratinocytes in coculture with Krtap genes in our study. It is unclear, however, whether fibroblasts. We hypothesized that Pro-Hyp can be recog- Gprc5d affects the synthesis of hair keratins themselves nized by fibroblasts which then induce the secretion of or the proliferation and differentiation of keratinocytes in factors that will in turn stimulate changes in keratino- hair follicles. The CP-induced increases in the expression cytes. mRNA levels of Tgfb1 and Itga1, the two crucial levels of Krt27 and Krtap16-7 were perhaps also related regulators of epidermis-dermis interactions [3] , however, to induction of the anagen stage, since these genes code were not affected by the addition of Pro-Hyp to the co- for a hair-specific type 1 keratin and a keratin-associated culture (data not shown). It should be noted that Pro-Hyp protein. On the other hand, Sprr2a1 codes for one of the was added to the culture for 24 h, which might not be the small proline-rich proteins that constitute the cornified optimal time to detect changes in these factors. Therefore, envelope of the skin. Recent studies have found that these time dependency of the effects of Pro-Hyp could be the small proline-rich proteins function not only as compo- subject of future studies. nents of the physical barrier of the skin, but also as mod- In summary, this study reports for the first time the ulators in the defense against oxidative stress [35–37] . regulatory effect of oral CP administration on gene ex- Thus, the up-regulation of this gene by oral intake of CP pression in the skin. The induction of genes related to might also present a novel aspect in the effects of dietary epidermis development, Gprc5d, Sprr2a1, Krt27 and collagen on the skin. Krtap16-7, was highlighted. The majority of the CP-in- Pro-Hyp was ascribed a major role in the biological duced up-regulated genes appeared to be related to the activity of diet-derived collagen based on its stimulatory hair cycle, indicating a novel effect of dietary collagen on effect on fibroblast cell growth [19]. The effects of this the skin. Given that the mice used in our study were hair-

Effects of Collagen Peptides on Gene Skin Pharmacol Physiol 2015;28:227–235 233 Expression in the Skin DOI: 10.1159/000369830 Downloaded by: Kyoto University 130.54.110.72 - 3/5/2015 2:07:27 AM less mice, further investigations of the mechanism of this Acknowledgment effect may be of interest. In addition, the up-regulation of certain Krtap genes was also induced by Pro-Hyp in This work was supported by a Grant-in-Aid for Scientific Re- search (B23380072) from the Japan Society for the Promotion of mouse skin coculture, confirming that Pro-Hyp is criti- Science. We are extremely grateful to Drs. Asami Yukio and Kano cally involved in the regulatory effects of dietary collagen Hiroshi, and Mr. Takabayashi Takuya (R & D Division, Meiji Co., on the skin. Ltd.) for support and advice during this study.

References

1 Proksch E, Brandner JM, Jensen JM: The skin: Sprague-Dawley male rats. J Food Sci 2010; 23 Ikuta S, Sekino N, Hara T, Saito Y, Chida K:

an indispensable barrier. Exp Dermatol 2008; 75:H230–H238. Mouse epidermal keratinocytes in three-di-

17: 1063–1072. 13 Wu J, Fujioka M, Sugimoto K, Mu G, Ishimi mensional organotypic coculture with dermal 2 Krieg T, Aumailley M: The extracellular ma- Y: Assessment of effectiveness of oral admin- fibroblasts form a stratified sheet resembling

trix of the dermis: flexible structures with dy- istration of collagen peptide on bone metabo- skin. Biosci Biotechnol Biochem 2006; 70:

namic functions. Exp Dermatol 2011; 20: 689– lism in growing and mature rats. J Bone Min- 2669–2675.

695. er Metab 2004; 22: 547–553. 24 Häkkinen L, Koivisto L, Larjava H: An im- 3 Werner S, Krieg T, Smola H: Keratinocyte- 14 Oesser S, Adam M, Babel W, Seifert J: Oral proved method for culture of epidermal kera- fibroblast interactions in wound healing. J In- administration of (14)C labeled gelatin hy- tinocytes from newborn mouse skin. Meth-

vest Dermatol 2007; 127: 998–1008. drolysate leads to an accumulation of radioac- ods Cell Sci 2001; 23: 189–196. 4 Yamaguchi Y, Hearing VJ, Itami S, Yoshikawa tivity in cartilage of mice (C57/BL). J Nutr 25 Hannen RF, Michael AE, Jaulim A, Bhogal R,

K, Katayama I: Mesenchymal-epithelial inter- 1999; 129: 1891–1895. Burrin JM, Philpott MP: Steroid synthesis by actions in the skin: aiming for site-specific tis- 15 Kawaguchi T, Nanbu PN, Kurokawa M: Dis- primary human keratinocytes; implications

sue regeneration. J Dermatol Sci 2005; 40: 1–9. tribution of prolylhydroxyproline and its me- for skin disease. Biochem Biophys Res Com-

5 Sennett R, Rendl M: Mesenchymal-epithelial tabolites after oral administration in rats. Biol mun 2011; 404: 62–67.

interactions during hair follicle morphogen- Pharm Bull 2012; 35: 422–427. 26 Slominski A, Zbytek B, Nikolakis G, Manna

esis and cycling. Semin Cell Dev Biol 2012; 23: 16 Proksch E, Schunck M, Zague V, Segger D, PR, Skobowiat C, Zmijewski M, Li W, Janje- 917–927. Degwert J, Oesser S: Oral intake of specific tovic Z, Postlethwaite A, Zouboulis CC, 6 Kimura Y, Sumiyoshi M: Effects of baicalein bioactive collagen peptides reduces skin wrin- Tuckey RC: Steroidogenesis in the skin: im- and wogonin isolated from Scutellaria ba- kles and increases dermal matrix synthesis. plications for local immune functions. J Ste-

icalensis roots on skin damage in acute UVB- Skin Pharmacol Physiol 2014; 27: 113–119. roid Biochem Mol Biol 2013; 137: 107–123. irradiated hairless mice. Eur J Pharmacol 17 Proksch E, Segger D, Degwert J, Schunck M, 27 Wu DD, Irwin DM, Zhang YP: Molecular

2011; 661: 124–132. Zague V, Oesser S: Oral supplementation of evolution of the keratin associated protein 7 Afaq F, Katiyar SK: Polyphenols: skin photo- specific collagen peptides has beneficial ef- gene family in mammals, role in the evolution

protection and inhibition of photocarcino- fects on human skin physiology: a double- of mammalian hair. BMC Evol Biol 2008; 8:

genesis. Mini Rev Med Chem 2011; 11: 1200– blind, placebo-controlled study. Skin Phar- 241.

1215. macol Physiol 2014; 27: 47–55. 28 Alonso L, Fuchs E: Stem cells in the skin:

8 Wang ZY, Agarwal R, Bickers DR, Mukhtar 18 Iwai K, Hasegawa T, Taguchi Y, Morimatsu F, waste not, Wnt not. Gene Dev 2003; 17: 1189– H: Protection against ultraviolet B radiation- Sato K, Nakamura Y, Higashi A, Kido Y, Nak- 1200. induced photocarcinogenesis in hairless mice abo Y, Ohtsuki K: Identification of food-de- 29 Li Z, Li JJ, Gui LJ, Begum S, Wang YB, Sun BS, by green tea polyphenols. Carcinogenesis rived collagen peptides in human blood after Lee M, Sung C: Chrysanthemum zawadskii

1991; 12: 1527–1530. oral ingestion of gelatin hydrolysates. J Agric extract induces hair growth by stimulating the

9 Hasegawa T, Shimada H, Uchiyama T, Ueda Food Chem 2005; 53: 6531–6536. proliferation and differentiation of hair ma-

O, Nakashima M, Matsuoka Y: Dietary gluco- 19 Shigemura Y, Iwai K, Morimatsu F, Iwamoto trix. Int J Mol Med 2014; 34: 130–136. sylceramide enhances cornified envelope for- T, Mori T, Oda C, Taira T, Park EY, Naka- 30 Shin HS, Park SY, Hwang ES, Lee DG, Song mation via transglutaminase expression and mura Y, Sato K: Effect of prolyl-hydroxypro- HG, Mavlonov GT, Yi TH: The inductive ef-

involucrin production. Lipids 2011; 46: 529– line (Pro-Hyp), a food-derived collagen pep- fect of ginsenoside F2 on hair growth by alter- 535. tide in human blood, on growth of fibroblasts ing the WNT signal pathway in telogen mouse

10 Kang JS, Yoon WK, Youm JK, Jeong SK, Park from mouse skin. J Agric Food Chem 2009; 57: skin. Eur J Pharmacol 2014; 730: 82–89. BD, Han MH, Lee H, Moon EY, Han SB, Lee 444–449. 31 Rogers MA, Langbein L, Praetzel-Wunder S, CW, Lee K, Park SK, Yang KH, Kim HM: In- 20 Tanaka M, Koyama Y, Nomura Y: Effects of Winter H, Schweizer J: Human hair keratin- hibition of atopic dermatitis-like skin lesions collagen peptide ingestion on UV-B-induced associated proteins (KAPs). Int Rev Cytol

by topical application of a novel ceramide de- skin damage. Biosci Biotechnol Biochem 2006; 251: 209–263.

rivative, K6PC-9p, in NC/Nga mice. Exp Der- 2009; 73: 930–932. 32 Rogers GE: Hair follicle differentiation and

matol 2008; 17: 958–964. 21 Huang da W, Sherman BT, Lempicki RA: Sys- regulation. Int J Dev Biol 2004; 48: 163–170. 11 Zague V: A new view concerning the effects of tematic and integrative analysis of large gene 33 Geyfman M, Kumar V, Liu Q, Ruiz R, Gordon collagen hydrolysate intake on skin proper- lists using DAVID bioinformatics resources. W, Espitia F, Cam E, Millar SE, Smyth P, Ihler

ties. Arch Dermatol Res 2008; 300: 479–483. Nat Protoc 2009; 4: 44–57. A, Takahashi JS, Andersen B: Brain and mus- 12 Liang J, Pei X, Zhang Z, Wang N, Wang J, Li 22 Huang da W, Sherman BT, Lempicki RA: Bio- cle Arnt-like protein-1 (BMAL1) controls cir- Y: The protective effects of long-term oral ad- informatics enrichment tools: paths toward cadian cell proliferation and susceptibility to ministration of marine collagen hydrolysate the comprehensive functional analysis of UVB-induced DNA damage in the epidermis.

from chum salmon on collagen matrix ho- large gene lists. Nucleic Acids Res 2009; 37: Proc Natl Acad Sci U S A 2012; 109: 11758– meostasis in the chronological aged skin of 1–13. 11763.

234 Skin Pharmacol Physiol 2015;28:227–235 Le Vu/Takatori/Iwamoto/Akagi/Satsu/ DOI: 10.1159/000369830 Totsuka/Chida/Sato/Shimizu Downloaded by: Kyoto University 130.54.110.72 - 3/5/2015 2:07:27 AM 34 Inoue S, Nambu T, Shimomura T: The RAIG 37 Mizuguchi Y, Specht S, Lunz JG 3rd, Isse K, 40 Botchkarev VA, Sharov AA: BMP signaling in family member, GPRC5D, is associated with Corbitt N, Takizawa T, Demetris AJ: SPRR2A the control of skin development and hair fol-

hard-keratinized structures. J Invest Derma- enhances p53 deacetylation through HDAC1 licle growth. Differentiation 2004; 72: 512–526.

tol 2004; 122: 565–573. and down regulates p21 promoter activity. 41 Saitoh A, Hansen LA, Vogel JC, Udey MC:

35 Vermeij WP, Backendorf C: Skin cornifica- BMC Mol Biol 2012; 13: 20. Characterization of Wnt gene expression in tion proteins provide global link between 38 Farwick M, Köhler T, Schild J, Mentel M, murine skin: possible involvement of epider- ROS detoxification and cell migration during Maczkiewitz U, Pagani V, Bonfigli A, Rigano mis-derived Wnt-4 in cutaneous epithelial-

wound healing. PLoS One 2010; 5:e11957. L, Bureik D, Gauglitz GG: Pentacyclic triter- mesenchymal interactions. Exp Cell Res 1998;

36 Vermeij WP, Alia A, Backendorf C: ROS penes from Terminalia arjuna show multiple 243: 150–160. quenching potential of the epidermal corni- benefits on aged and dry skin. Skin Pharmacol 42 Lim X, Nusse R: Wnt signaling in skin devel-

fied cell envelope. J Invest Dermatol 2011; 131: Physiol 2014; 27: 71–81. opment, homeostasis, and disease. Cold

1435–1441. 39 Awgulewitsch A: Hox in hair growth and de- Spring Harb Perspect Biol 2013; 5:a008029.

velopment. Naturwissenschaften 2003; 90: 193–211.

Effects of Collagen Peptides on Gene Skin Pharmacol Physiol 2015;28:227–235 235 Expression in the Skin DOI: 10.1159/000369830 Downloaded by: Kyoto University 130.54.110.72 - 3/5/2015 2:07:27 AM