ABSTRACT
It was recently revealed that epidermal growth following topical treatment with all-trans retinoic acid (atRA) was at least partly induced by heparin-binding EGF-like growth factor (HB-EGF) released from suprabasal keratinocytes. Since proliferation of keratinocytes appears to be one of the critical roles of atRA in depigmentation treatment and promotion of wound healing, HB-EGF is supposed to be suitable for assessing the therapeutic value of topical retinoids. In this study, HB-EGF mRNA expression in normal human keratinocytes after atRA treatment was examined, and the effects of a variety of natural and synthetic retinoids were compared to each other.
The present results of RT-PCR suggested that induction of differentiation increased HB-EGF mRNA expression in cultured keratinocytes. Real-time PCR analyses revealed that HB-EGF mRNA expression was elevated dose-dependently with atRA, most at 12 hours. This elevation was more prominent in keratinocytes of confluent state than those of subconfluent state, suggesting differentiated keratinocytes are more subject to be stimulated by atRA with regard to HB-EGF expression than proliferating keratinocytes.
HB-EGF mRNA was upregulated in differentiation-induced keratinocytes by all retinoids used in this study at 1 μM, and marked upregulation was seen when treated with three isotypes of retinoic acid (atRA, 9-cis and 13-cis retinoic acid). RARα-selective agonists (Am80, Am580, ER-38925, and TAC-101) and a panagonist of RARs (Re80) showed relatively low elevation of HB-EGF transcripts as well as all-trans retinol (Rol) and all-trans retinal (Ral). Although another panagonist (Ch55) showed the highest elevation of HB-EGF mRNA, it was relatively cytotoxic at the concentration employed. Ral and Rol was found to upregulate HB-EGF when used at 100 μM to 1 mM, to a similar extent of atRA at 1 to 10 μM. The capacity of retinoids to upregulate HB-EGF may be an important index to develop and seek an ideal synthetic retinoid, which has maximum benefits and minimal side effects.

INTRODUCTION
Topical tretinoin (all-trans retinoic acid; atRA) has been widely used for several skin diseases such as acne vulgaris and photoaged skin with remarkable success since 1970's. Retinoids have a variety of biological effects on epidermis and dermis including appendices, and those are mediated by specific nuclear receptors; RARs (retinoic acid receptors) and RXRs (retinoid X receptors) (1, 2). Epidermal hyperplasia is one of the most prominent histological changes in skin seen after treatment with atRA (1, 3). This phenomenon was commonly observed not only in vivo in several mammals, but also in skin equivalents using normal human keratinocytes (NHK) and fibroblasts (data not shown). However, since NHK proliferation was not consistently observed in monolayer-cultured NHK (2, 4), the mechanism of keratinocyte proliferation induced by retinoids remained unknown for a long time.
Recently, a study using cultured keratinocytes, organ culture and skin biopsies, revealed that transcripts of heparin-binding EGF-like growth factor (HB-EGF), a member of the EGF family of growth factors, are induced by treatment of retinoids, suggesting that epidermal hyperplasia after atRA treatment may be mediated at least in part by keratinocyte-derived HB-EGF (5). HB-EGF was shown to be upregulated in actual wound healing much more than other growth factors that accelerate epidermal growth (6). Thereafter, a paracrine action of HB-EGF released from suprabasal keratinocytes was found to be a key mechanism of epidermal growth following atRA treatment by a study using transgenic mice (7). Thus, it is suggested that atRA accelerates keratinocyte differentiation in a direct manner, and promotes keratinocyte proliferation in an indirect manner.
Combination topical therapies with atRA and hydroquine for pigmented skin lesions have been successfully performed since Kligman and Willis proposed their depigmenting formula in 1975 (8). The authors modified the protocol and demonstrated the depigmenting potential of atRA (9, 10). Since atRA appeared not to have a suppressive effect on melanogenesis, keratinocyte proliferation and acceleration of epidermal turnover (keratinocytes differentiation) appear to be the two critical roles of atRA in the depigmenting therapies (11). Since the former of the two is mediated by a paracrine action of HB-EGF released from suprabasal keratinocytes, HB-EGF mRNA is supposed to be suitable for assessing the therapeutic value of topical retinoids as far as they are used for treating hyperpigmentation or promoting wound healing.
The purposes of this study are to reconfirm the effect of atRA to promote the expression of HB-EGF mRNA in normal human keratinocytes, and to compare the HB-EGF-promoting abilities of a variety of natural and synthetic retinoids. The capacity of retinoids to upregulate HB-EGF may be an important index to develop and seek an ideal synthetic retinoid, which has maximum beneficial effects and minimal side effects.

METHODS
Cell isolation and cell culture
Human epidermal keratinocytes isolated from biopsies of healthy skin obtained from young Japanese patients during plastic surgery were used in this study. Keratinocytes were isolated using a modification of the method reported previously (12). Briefly, the specimens were washed 3 times in phosphate buffered saline (PBS) and finely shredded with scissors, and incubated with 0.25 % trypsin and 0.02 % EDTA in PBS for 16-24 hours at 4 ?C. The epithelium was separated from the dermis with forceps, and keratinocytes were isolated from the subepithelial side. Keratinocytes were grown in a modified serum-free Keratinocyte growth medium (KGM; Kyokuto Seiyaku, Tokyo), which consists of MCDB153 with high concentrations of amino acids, transferrin (final concentration 10 μg/ml), insulin (5 μg/ml), hydrocortisone (0.5 μg/ml), phosphorylethanolamine (14.1 μg/ml), and bovine pituitary extract (40 μg /ml). The final concentration of Ca2+ in the medium was 0.03 mM.
Keratinocytes in subconfluent state (30 % confluency) or those in confluent state (cultured for 48 hours after they showed 100 % confluency) were used in this study, representing growing keratinocytes and differentiated ones, respectively.

Reagents
The influences of various kinds of retinoids on HB-EGF mRNA expression of keratinocytes were compared in this study. Five kinds of natural retinoids and 6 kinds of synthetic retinoids were used (Fig. 1). For natural retinoids, tretinoin (all-trans retinoic acid; atRA), 13-cis retinoic acid (13cRA), 9-cis retinoic acid (9cRA), all-trans retinol (Rol), and all-trans retinal (Ral) were used. All of the natural retinoids were purchased from Sigma (St Louis, MO). 9cRA is known to be a ligand for RXRs and bind also to RARs. For synthetic retinoids, AM80 (4-[(5,6,7,8-tetrahydro-5,5,8,8-tetramethyl-2-naphthyl)-carbamoyl] benzoic acid), Am580 (4-[(5,6,7,8-tetrahydro-5,5,8,8-tetramethyl-2-naphthalenyl)carboxamido] benzoic acid), Ch55 ((E)-4-[3-(3,5-di-tert-butylphenyl)-3-oxo-1-propenyl] benzoic acid), and Re80 (4-[1-hydroxy-3-oxo-3-(5,6,7,8-tetrahydro-3-hydroxy-5,5,8,8-tetramethyl-2-
naphthalenyl)-1-propenyl] benzoic acid) were generous gifts from Dr. Kagechika (University of Tokyo, Tokyo), TAC101 (4-[3,5-bis(trimethylsilyl) benzamide] benzoic acid) and ER-38925 (4-[5-(4,7-dimethylbenzofuran-2-yl)pyrrol-2-yl] benzoic acid) were also generous gifts from Taiho Parmaceuticals Co, Ltd (Tokyo, Japan) and Eisai Co, Ltd. (Tokyo, Japan), respectively. Am80, Am580, TAC-101, and ER-38925 are RARα-selective agonists (13-16), while Ch55 and Re80 are panagonists for all three RAR subtypes (12, 17). All reagents were dissolved in ethanol at 1 mM as stock solutions (for atRA, Rol, and Ral; other stock solutions were also prepared), and 10 μl of each stock solution was added to 10 ml of the culture medium to get the designated final concentrations (for comparison of all retinoids; 1 μM). As a control, 10 μl of ethanol alone was added to 10 ml of the culture medium.

RNA isolation
After removing the culture media and washing twice with PBS, total RNA was obtained with RNeasyR Mini Kit (QIAGEN, Hilden, Germany). In order to eliminate residual genomic DNA, RNase-Free DNase Set (QIAGEN, Hilden, Germany) was also used after. The concentration of each RNA sample was measured with Spectrophotometer V-530 UV/VIS (JASCO, Tokyo, Japan).

Reverse transcription-PCR (RT-PCR) analysis
The amount of isolated total RNA was spectrophotomecally measured. A reverse transcriptase reaction was performed using RNA PCR Kit (AMV) Ver.2.1 (TaKaRa, Tokyo, Japan) according to the instruction manual; 5 μg of total RNA in a 100 μM of reaction mixture (final concentrations: 5 mM MgCl2, 1 mM dNTP Mixture, 1 U/μl RNase Inhibitor, 0.125 μM Oligo dT-Adaptor Primer, 10 mM Tris-HCl, 50 mM KCl, pH 8.3) containing 25 units of AMV Reverse Transcriptase XL at 42?C for 30 minutes, followed by inactivation of the enzyme at 99 ?C for 5 minutes with Program Temp Control System PC-700 (ASTEC, Fukuoka, Japan). The control reaction was performed simultaneously under identical conditions, but without reverse transcriptase.
For PCR amplification, 0.5 μl of each cDNA reaction mixture was added to 49.5 μl PCR mixture containing 5 μl MgCl2 (25 mM), 5 μl 10X PCR buffer, 1 μl deoxynucleotide mixture (10 mM), 0.5 μl each of the 3' and 5' primers (50 μM each), and 0.25 μl Taq polymerase, and 37.25 μl DEPC-treated water. Reaction mixtures were amplified using Microplate Gradient Thermal Cycler PC-960G (Corbett Research, Australia). The PCR cycle conditions were: melting for 30 seconds at 94 ?C, annealing for 30 seconds at 59 ?C, and extension for 1.5 minutes at 72 ?C. The oligonucleotide primers used for RT-PCR were as follows; human HB-EGF (forward) 5'- CACACCAAACAAGGAGGAGC -3' and (reverse) 5'- CATGAGAAGCCCCACGATGA -3' (PCR-product size: 279bp); human glyceraldehydes-3-phosphate dehydrogenase (GAPDH), (forward) 5'- GAAGGTGAAGGTCGGAGTC -3' and (reverse) 5'- GAAGATGGTGATGGGATTTC -3' (PCR-product size: 226bp). All reverse transcription-PCR products were separated on 2 % agarose gels, visualized by ultraviolet B using ethidium bromide staining.

Real time RT-PCT analysis
Expression of HB-EGF transcripts by keratinocytes was quantitatively measured using real-time quantitative PCR System (Sequence Detection System ABI PRISM 7700, PE Applied Biosystems, Foster City, CA). Real-time PCR was performed on ABI PRISMTM 96-Well Optical Reaction Plates (PE Biosystems, Foster City, CA). All PCR reaction mixtures per well contained 25 μl of TaqMan SYBRR Green PCR Master Mix (PE Biosystems, Foster City, CA), 0.25 μl of forward primer (10 pmol/μl), 0.25 μl of reverse primer (10 pmol/μl), 4 μl of each diluted sample (RT products), 20.5 μl of distilled water. PCR amplification of each sample was performed with both specific primer pairs of target human HB-EGF gene and human GAPDH gene on the same reaction plate. The oligonucleotide primers used for real-time PCR were as follows; human HB-EGF (forward) 5'- CAGATCTGGACCTTTTGAGAGTCA -3' and (reverse) 5'- TCCCGTGCTCCTCCTTGTT -3' (PCR-product size: 78bp); human GAPDH, (forward) 5'- GAAGGTGAAGGTCGGAGTC -3' and (reverse) 5'- GAAGATGGTGATGGGATTTC -3' (PCR-product size: 226bp). The PCR reaction was comprised of 40 cycles, consisting of denaturing at 95 ?C (15 sec.), annealing/extension at 60 ?C (1 min.). For comparison of the HB-EGF mRNA expression, the value of HB-EGF of each sample was normalized by deduction by the value of GAPDH of the same sample. In order to eliminate the possibility of contamination of genomic DNA during extraction of total RNA, a control reaction with each primer pair was performed at the same time under identical condtions without reverse transcription, and no amplification was detected. The normalized data were collected by at least 4 separate experiments.

Statistical analysis
In each experiment, the values of normalized HB-EGF mRNA expression of each sample were devided by the value of the control to obtain data for analysis. Statistical analysis was performed using Student t-test.

RESULTS
I. RT-PCR analysis
In preliminary experiments, HB-EGF mRNA expression was examined with RT-PCR. HB-EGF mRNA expression was elevated in keratinocytes, differentiation-induced by incubation for 48 hours with high concentration (1.8mM) of Ca2+ or 2% serum, or both, relative to the control (Fig. 2). This finding suggested that normal human keratinocytes increase HB-EGF mRNA expression when they differentiate.
The investigation of sequential changes of HB-EGF mRNA elevation by atRA (1 μM) revealed that HB-EGF mRNA expression increased most at 12 hours after atRA administration (Fig. 3).
In addition, our results also suggest that HB-EGF transcripts were dose-dependently elevated by incubation with atRA for 15 hours in the concentration range of 0.01 μM to 1 μM (Fig. 4).

II. HB-EGF mRNA elevation by atRA in keratinocytes in subconfluent or confluent condition
The sequential changes of HB-EGF mRNA elevation by atRA (1 μM) was quantitatively measured with real-time PCR system using normal human keratinocytes in subconfluent (30% confluence) and confluent (100% confluence) condition (Fig. 5). For the confluent condition, keratinocytes cultured for 48 hours after they were first observed confluent in culture dishes were used, representing differentiated keratinocytes as a model of suprabasal keratinocytes. For the subconfluent condition, keratinocytes showing 30% confluency were used immediately, representing proliferating keratinocytes as a model of basal keratinocytes. The normalized HB-EGF mRNA level was highest at 12 hours, and thereafter decreased in both cases. Throughout the time investigated (0-48 hours), the average value of HB-EGF mRNA level was higher in the confluent condition than the subconfluent condition. Statistical significance between the subconfluent condition and the confluent one was found only at 12 hours.

III. Comparison of HB-EGF mRNA upregulation by various kinds of natural and synthetic retinoids
The HB-EGF mRNA levels were quantitatively measured with real-time PCR in normal human confluent-state keratinocytes cultured for 15 hours with 11 kinds of retinoids (0.01 μM, 0.1 μM, and 1 μM for atRA; 1 μM for the other retinoids). The results demonstrated that all of the 11 kinds of retinoids studied showed significant upregulation of HB-EGF mRNA, yet to different extents (Fig. 6).　Three isotypes of retinoic acid (atRA, 9cRA, and 13cRA) showed higher degrees of upregulation of HB-EGF transcripts than any other natural or synthetic retinoids except Ch55, although there was no apparent difference among the three retinoic acids. Ch55, a panagonist of RARs, demonstrated a much higher degree of upregulation of HB-EGF transcripts than the three retinoic acids, but the amount of total RNA isolated from keratinocytes treated with Ch55 was very small compared to that from the other samples, suggesting Ch55 was cytotoxic to keratinocytes in the concentration employed here (1 μM). RARα-selective synthetic agonists (Am80, Am580, ER-38925, and TAC-101) and another panagonist of RARs (Re80) showed a lesser degree of upregulation of HB-EGF transcripts than the three retinoic acids, as well as the other two natural retinoids, Rol and Ral.

IV. Comparison of upregulations of HB-EGF mRNA by atRA, Rol, and Ral.
HB-EGF transcripts expression of normal human confluent-state keratinocytes treated for 15 hours with several concentrations of atRA, Rol and Ral were quantitatively measured using real-time PCR. Compared at 1 μM and 10 μM, HB-EGF was upregulated significantly more by atRA than by Rol or Ral (Fig. 7). When using the concentration of 100 μM to 1 mM, atRA was cytotoxic, whereas Rol and Ral markedly upregulated HB-EGF to a similar degree to atRA at 1 μM to 10 μM. This means that even Rol and Ral could promote keratinocyte proliferation as well as atRA when used at higher concentrations (presumably 40-100 fold). It remained unknown whether Ral and Rol upregulated HB-EGF through nuclear receptors directly or after oxidative conversion to atRA. An alternative pathway can not be fully excluded.

DISCUSSION
It was recently revealed that HB-EGF mRNA was induced by treatment of retinoids in human keratinocytes and organ-cultured skin, suggesting that epidermal hyperplasia after atRA treatment may be mediated at least in part by keratinocyte-derived HB-EGF(5). Thereafter, a paracrine action of HB-EGF released from suprabasal keratinocytes was suggested to be a key mechanism of epidermal growth following atRA treatment by the study, in which dominant-negative RARα was overexpressed in suprabasal layers of mice and the response of skin to atRA was examined (7). The present study confirmed that HB-EGF mRNA was markedly induced by atRA in human normal keratinocytes, and suggested that even cell differentiation of keratinocytes alone might increase keratinocyte-derived HB-EGF without retinoid treatment. Although the previous study (5) did not detect a significant difference in HB-EGF upregulation between confluent and subconfluent keratinocytes at 6, 24, 48 hours, our results revealed that differentiated keratinocytes in confluent condition upregulated HB-EGF significantly more than growing keratinocytes in subconfluent condition after 12 hours of atRA treatment, supporting the hypothesis of Xiao et al. (7) that HB-EGF released from differentiated suprabasal keratinocytes stimulate proliferation of basal keratinocytes. Whether the induction of HB-EGF mRNA is due to ligand-dependent transcription activation of the HB-EGF gene or not is to be explored, because the presence of retinoic acid response elements (RARE) has not yet been identified in the promoter region of the HB-EGF gene (5, 7).
As far as depigmenting treatment is concerned, promotion of keratinocyte proliferation and acceleration of kerationcyte differentiation appeared to be the two main roles of atRA (11). Assumed that HB-EGF released from suprabasal keratinocytes is the main reason for epidermal growth following retinoid treatment in vivo, the promotive ability of retinoids on HB-EGF expression can be used as an index in evaluating individual retinoid and developing the synthetic retinoids to promote epidermal growth. However, any reliable index indicating the ability of retinoids to differentiate keratinocytes has not been found.
All natural and synthetic retinoids employed in this study significantly upregulated HB-EGF, suggesting that epidermal growth can be induced by a wide range of topical retinoids. However, the extent of HB-EGF upregulation was distinct among those agents. All four RARα-specific synthetic reinoids, Am80, Am580, ER-38925, and TAC-101, showed relatively low promotion of HB-EGF mRNA to all three isotypes of retinoic acids, whereas those synthetic retinoids demonstrated a higher affinity to RARα and higher activity than atRA in other actions such as growth suppression of neoplasm (13-16). The reason for the variance in promotion of HB-EGF mRNA among retinoids remains unknown, but it may be partly due to differential binding affinity to RARγ. Ninety percent of RARs expressed in epidermis is composed of RARγ, and the other of RARα, while RXRα is a major RXRs in skin (2, 18). Both RARs and RXRs were reported to be expressed much higher in the spinous and granular layers in comparison to the basal layer in normal skin (19). Since RAR-β, or RAR-γ specific ligands were not evaluated in this study, the relationship between retinoid specificity to subtypes of RARs and HB-EGF expression requires further investigation.
In addition to mediating RARE- or RXRE(retinoid X receptor responsive elements)- dependent transactivation, retinoid receptors can also affect gene expression by inhibiting the action of other transactivation factors, including AP-1 (20). Furthermore, there are some RAR/RXR-independent pathways, such as activation of the mannose-6-phosphate (M6P)/insulin-like growth factor-II (IGF-II) (21).
Human keratinocytes are known to transform Rol into Ral, and then into atRA by two enzymatic steps involving dehydrogenases, and the conversion rate of Rol into atRA depends on the state of keratinocytes differentiation; differentiated keratinocytes can convert at a higher rate than non-differentiated keratinocytes (22). The binding affinity of Rol or Ral to retinoid nuclear receptors is quite low (23), so that their biological activity should result from their oxidative transformation into retinoic acid by epidermal keratinocytes. However, there have been some reports suggesting the existence of other pathways than that mediated by nuclear receptors (24).
Rol is constitutively present in human plasma at 1-2 μM (25) and the upper limit of extracellular Rol concentration in epidermis is thought to be 0.7 μM (26), whereas atRA is present at 4-14 nM in human plasma (27, 28). The present results revealed that all of atRA, Rol and Ral induce markedly HB-EGF mRNA at pharmacological concentration which is 100-1000 folds higher than physiological concentration. Even Rol and Ral induced high level of HB-EGF mRNA when used at much higher (40-100 fold) concentration than atRA. This means that Rol and Ral when used at high concentration could promote epidermal growth in vivo as well as atRA, although Rol and Ral have been reported to show much less side effects than atRA when used at the similar concentration. Rol and Ral are though to be more tolerable because of less side effects, and may be of great value in clinical use.
The results indicated the potentiality of use of Rol or Ral at high concentration in vivo, especially in depigmenting treatment which needs promotion of epidermal growth and turnover. Indeed, our preliminary clinical study with 5-10% Rol aqueous gel suggested similar depigmenting effects to 0.1% atRA aqueous gel (in preparation). Thus, promotion of HB-EGF expression may be used as one of promising indexes to evaluate the depigmenting ability of topical retinoids.

ACKNOWLEDGEMENTS
We thank Dr. Hiroyuki Kagechika (University of Tokyo, Tokyo, Japan) for his generous provision of Am80, Am 580, Ch55, and Re80. We also acknowledge Yuka Kuwahara for her technical assistance.

REFERENCES

1) Kligman A M, Grove G L, Hirose R, Leyden J J. Topical tretinoin for photoaged skin. J Am Acad Dermatol 1986: 15: 836-859.

2) Fisher G J, Voorhees J J. Molecular mechanisms of retinoid actions in skin. FASEB J 1996:10: 1002-1013.

3) Tur E, Hohl D, Jetten A, Panizzon R, Frenk E. Modification of late epidermal differentiation in photoaged skin treated with topical retinoic acid cream. Dermatology 1995:191: 124-128.

4) Gibson D F, Bikle D D, Harris J. All-trans retinoic acid blocks the antiproliferative prodifferentiating actions of 1,25-dihydroxyvitamin D3 in normal human keratinocytes. J Cell Physiol 1998: 174: 1-8.

5) Stoll S W, Elder J T. Retinoid regulation of heparin-binding EGF-like growth factor gene expression in human keratinocytes and skin. Exp Dermatol 1998: 7: 391-397.
6) Stoll S, Garner W, Elder J. Heparin-binding ligands mediate autocrine epidermal growth factor receptor activation In skin organ culture. J Clin Invest 1997: 100:1271-1281.

7) Xiao J H, Feng X, Di W, et al. Identification of heparin-binding EGF-like growth factor as a target in intercellular regulation of epidermal basal cell growth by suprabasal retinoic acid receptors. EMBO J 1999: 18:1539-1548.

8) Kligman A M, Willis I. A new formula for depigmenting human skin.
Arch Dermatol 1975: 111: 40-48.

9) Yoshimura K, Harii K, Aoyama T, Shibuya F, Iga T. A new bleaching protocol for hyperpigmented skin lesions with a high concentration of all-trans retinoic acid aqueous gel. Aesthetic Plast Surg 1999: 23: 285-291.

10) Yoshimura K, Harii K, Aoyama T, Iga T. Experience with a strong bleaching treatment for skin hyperpigmentation in Orientals. Plast Reconstr Surg 2000: 105:1097-1108.

11) Yoshimura K, Tsukamoto K, Okazaki M, et al. Effects of all-trans retinoic acid on melanogenesis in pigmented skin equivalents and monolayer culture of melanocytes. J Dermatol Sci 2001: 27: S68-75.

12）Tsunenaga M, Kohno Y, Horii I, et al. Growth and differentiation properties of normal and transformed keratinocytes in organotypic culture. Jpn J Cancer Res 1994: 85: 238-244.

13) Hashimoto Y, Kagechika H, Kawachi E, Shudo K. Specific uptake of retinoids into human promyelocytic leukemia cells HL-60 by retinoid-specific binding protein: possibly the true retinoid receptor. Jpn J Cancer Res 1988: 79: 473-483.

14) Kagechika H, Kawachi E, Hashimoto Y, Himi T, Shudo K. Retinobenzoic acids. 1. Structure-activity relationships of aromatic amides with retinoidal activity. J Med Chem 1988: 31: 2182-2192.

15) Murakami K, Wierzba K, Sano M, Shibata J, Yonekura K, Hashimoto A, Sato K, Yamada Y. TAC-101, a benzoic acid derivative, inhibits liver metastasis of human gastrointestinal cancer and prolongs the life-span. Clin Exp Metastasis 1998: 16: 323-331.

16) Yoshimura H, Kikuchi K, Hibi S, et al. Discovery of novel and potent retinoic acid receptor alpha agonists: syntheses and evaluation of benzofuranyl-pyrrole and benzothiophenyl-pyrrole derivatives. J Med Chem 2000: 43: 2929-2937.

17) Kagechika H, Kawachi E, Hashimoto Y, Shudo K. Retinobenzoic acids. 2. Structure-activity relationships of chalcone-4-carboxylic acids and flavone-4'-carboxylic acids. J Med Chem 1989: 32: 834-840.

18) Fisher G J, Talwar H S, Xiao J H, et al. Immunological identification and functional quantitation of retinoic acid and retinoid X receptor proteins in human skin. J Biol Chem 1994: 269: 20629-20635.

19) Xu X C, Lotan R. Aberrant expression and function of retinloid receptors in cancer. In: Nau H, Blaner W S, ed. Retinoids. Berlin: Springer, 1999: 335-336.

20) Schule R, Rangarajan P, Yang N, et al. Retinoic acid is a negative regulator of AP-1-responsive genes. Proc Natl Acad Sci USA: 1991: 88: 6092-6096.

21) Kang JX, Li Y, Leaf A. Mannose-6-phosphate/insulin-like growth factor-II receptor is a receptor for retinoic acid. Proc Natl Acad Sci USA 1997: 94:13671-13676.

22) Siegenthaler G, Saurat J H, Ponec M. Retinol and retinal metabolism. Relationship to the state of differentiation of cultured human keratinocytes.
Biochem J 1990: 268: 371-378.

23) Crettaz M, Baron A, Siegenthaler G, Hunziker W. Ligand specificities of recombinant retinoic acid receptors RAR alpha and RAR beta. Biochem J 1990: 272: 391-397.

24) Saurat J H, Sorg O, Didierjean L. New concepts for delivery of topical retinoid activity to human skin. In: Nau H, Blaner W S, ed. Retinoids. Berlin: Springer, 1999: 521-538.

25) Soprano D, Blaner W S. Plasma retinol-binding protein. In: Sporn MB, Roberts AB, Goodman DS, ed. The retinoids: biology, chemistry, and medicine, 2nd edn. New York: Raven, 1994: 257.

26) Randolph R K, Siegenthaler G. Vitamin A homeostasis in human epidermis: native retinoid composition and metabolism. In: Nau H, Blaner W S, ed. Retinoids. Berlin: Springer, 1999: 499-500.

27) DeLeenHeer A P, Lambert W E, Claeys I. All-trans retinoic acid: Measurement of reference values I human serum by high performance liquid chromatography. J Lipid Res 1982: 23: 1362-1367.

28) Eckhoff C, Nau H. Identification and quantitation of all-trans- and 13-cis retinoic acid and 13-cis-4-oxo-retinoic acid in human plasma. J Lipid Res 1990: 31: 1445-1454.

LEGENDS

Fig.1. Chemical structures of retinoids used in this study. (left) 5 kinds of natural retinoids; atRA: all-trans retinoic acid, 13cRA: 13-cis retinoic acid, 9cRA: 9-cis retinoic acid, Rol: all-trans retinol, Ral: all-trans retinal. (Right) 6 kinds of synthetic retinoids.

Fig.2. HB-EGF mRNA expression in keratinocytes differentiation-induced by 48 hours incubation with the medium containing 2 % serum or/and 1.8 mM Ca2+. It was amplified by RT-PCR and showed along with control incubated with normal medium. CTL: control, S: the medium contained 2 % serum, Ca: the medium contained 1.8mM Ca2+, S+Ca; the medium contained both 2% serum and 1.8mM Ca2+.

Fig.3. HB-EGF mRNA expression in keratinocytes cultured with medium containing 1 μM atRA for 0, 6, 12, 24, and 48 hours. The expression was amplified by RT-PCR.

Fig.4. HB-EGF mRNA expression in keratinocytes cultured with medium containing 3 kinds of concentration (1 μM, 0.1 μM, and 0.01 μM) atRA for 15 hours. The expression was amplified by RT-PCR.

Fig.5. HB-EGF mRNA expression in keratinocytes in subconfluent (30 %) or confluent (100 %) condition cultured with medium containing 1 μM atRA for 0, 6, 12, 24, and 48 hours. HB-EGF mRNA was measured quantitatively with real-time PCR system and the normalized value was calculated by deducting by the GAPDH value of each sample. The values are demonstrated as relative values to the value obtained from the subconfluent keratinocytes at 0 hour, and indicated as mean + standard error. : p<0.05.

Fig.6. HB-EGF mRNA expression in normal human keratinocytes in confluent condition cultured for 15 hours with 1 μM each retinoid (0.01, 0.1, 1μM for atRA). HB-EGF mRNA was measured quantitatively with real-time PCR system. The values are demonstrated as relative values to the value obtained from the control in each measurement, and indicated as mean + standard error. CTL: control, atRA(-6): 1 μM atRA, atRA(-7); 0.1 μM atRA, atRA(-8): 0.01 μM atRA, Er: ER-38925, TAC: TAC-101. The data of all retinoids are significantly higher than that of control.

Fig.7. HB-EGF mRNA expression in normal human keratinocytes in confluent condition cultured for 15 hours with one of the three natural retinoids (atRA, Rol, and Ral). HB-EGF mRNA was measured quantitatively with real-time PCR system. The values are demonstrated as relative values to the value obtained from the control (normal medium) in each measurement. Four kinds of concentration (0.01 μM, 0.1 μM, 1 μM, and 10 μM for atRA; 1 μM, 10 μM, 100 μM, and 1 mM for Rol and Ral) were employed for each measurement. The values are indicated as mean + or - standard error. At the concentration of 1 μM and 10 μM, HB-EGF mRNA expression was significantly higher in keratinocytes treated with atRA than those treated with Rol or Ral. : p<0.05.