Biochemical Characterization of Endothelin-converting Enzyme-1α in Cultured Skin-derived Cells and Its Postulated Role in the Stimulation of Melanogenesis in Human Epidermis*

The vasoconstrictive peptide endothelin-1 (ET-1) is expressed in human epidermis at the gene and protein levels and plays an important role in stimulating pigmentation via its increased secretion by keratinocytes following ultraviolet B (UVB) irradiation. However, one or more biological mechanisms underlying the secretion of ET-1 by keratinocytes in human skin have never been evaluated. In mammalian endothelial cells, a membrane-bound neutral metalloproteinase, termed endothelin-converting enzyme (ECE), catalyzes the specific cleavage of the inactive precursor Big ET to produce mature active ET, which leads in turn to the secretion of ET by those cells. To clarify the potential involvement of ECE in the processing and secretion of ET-1 by human keratinocytes, we synthesized the N-terminal peptide of human ECE-1α and generated a rabbit polyclonal antibody (αPEPT6) that specifically recognizes ECE-1α. Reverse transcription PCR and Western blotting analysis revealed that significant expression of ECE-1 transcripts and ECE-1α protein occurs in human keratinocytes. When ECE activity was assayed in extracts of human keratinocytes at pHs ranging from 5.0 to 8.0, the enzymatic profile had an optimal neutral pH of 7.0 and was sharply pH-dependent. Furthermore, when extracts of human keratinocytes were treated with αPEPT6, ECE activity was significantly reduced compared with extracts treated with the prebleed serum of αPEPT6, which supports the notion that ECE-1α is a major metalloproteinase with ECE activity in human keratinocytes. The exogenous addition of the pro-inflammatory cytokine interleukin-1α significantly increased expression of ECE-1 transcripts in cultured human keratinocytes, which suggests an association with post-inflammatory hyperpigmentation. Collectively, our results demonstrate for the first time that ECE-1α is expressed at significant levels in various types of human skin cells (including keratinocytes) and that it plays a constitutive role in the processing and UVB-inducible secretion of ET-1 by human keratinocytes, which leads to the stimulation of pigmentation in the epidermis.

Three major steps are involved in epidermal hyperpigmen-tation, i.e. the proliferation of melanocytes (1), the synthesis and activation of tyrosinase (2,3) and the transfer of melanosomes to keratinocytes (4). With respect to proliferation and melanogenesis of melanocytes, keratinocyte-derived cytokines, including basic fibroblast growth factor (5), ET-1 1 (6 -10), ␣-melanocyte-stimulating hormone (11)(12)(13)(14)(15), stem cell factor (16,17), and nitric oxide (18) have been documented to be up-regulated in their production and secretion/release following UVB irradiation and can act as mitogens and/or melanogens for human melanocytes. Among them, ET-1 is a unique cytokine that acts as a mitogen and a melanogen for human melanocytes via the G protein-coupled endothelin B receptor (ET B R) (16,19). In addition to its function in melanocytes, ET-1 has also been reported to act as an autocrine growth factor for normal human keratinocytes that express both ET A R and the ET B R (20,21). However, little is known about the biochemical mechanism(s) underlying the processing and secretion of ET by human keratinocytes.
ET isopeptides are first expressed as corresponding ϳ200residue inactive prepropolypeptides (preproendothelins) that are encoded by distinct genes (22). After removal of their signal peptides during their early biosynthesis processing, the propeptides are cleaved at pairs of basic amino acids to yield the intermediate Big ETs. This early processing step is presumably involved with furin, a prohormone convertase of the constitutive secretory pathway (23). Big ETs are then further cleaved by an endopeptidase termed endothelin-converting enzyme (ECE) at Trp 21 -(Val 22 /Ile 22 ) to produce biologically active ETs, which are mature 21-residue peptides (24). All evidence to date regarding the processing and secretion of ETs has been restricted to endothelial cells. Until now, no reports have described the biological features of ECE in human keratinocytes, even though ET-1 is known to play an important role in the stimulation of pigmentation in the epidermis, which consists mainly of keratinocytes. It is therefore of considerable interest to identify and characterize ECE expressed by human keratinocytes; such a study would provide a new insight into the regulatory mechanisms involved in the secretion of ETs and their functional role in the epidermis. We now report for the first time the constitutive expression of ECE-1 at the gene and protein levels and that the ECE-1␣ isoform is dominant within human keratinocytes. This epidermal ECE-1␣ is characterized by high sensitivity to a metalloproteinase inhibitor, phosphoramidon, has an optimal enzymatic activity at neutral pH, and can be induced at the transcriptional and translational levels by IL-1␣. Thus, it is likely that ECE-1␣ plays a constitutive role in the processing and secretion of ETs by human keratinocytes and that it is involved in the stimulation of pigmentation in the epidermis.

EXPERIMENTAL PROCEDURES
Materials-Normal human endothelial cells, fibroblasts, keratinocytes, and E300 culture medium were obtained from Kyokutou (Tokyo, Japan). ET derivatives and endothelin enzyme-linked immunosorbent assay (ELISA) kits were purchased from Immuno-Biological Laboratories (Gunma, Japan). Serum-free keratinocyte medium (SFM), bovine pituitary extract (BPE), epidermal growth factor (EGF), and Dulbecco's modified Eagle's medium were purchased from Invitrogen (Rockville, MD). Normal human melanocytes and serum-free melanocyte growth medium was purchased from Sankou Pure Chemicals (Tokyo, Japan). Other chemicals were of reagent grade.
Cell Culture-All cells used in this study were derived from male newborns, and third-passage normal human cells were used for all experiments. Cells were maintained at 37°C with 5% CO 2 . Human umbilical vein endothelial cells, fibroblasts, keratinocytes, and melanocytes were maintained in E300 medium, Dulbecco's modified Eagle's medium (containing 10% fetal bovine serum), SFM (supplemented with 5 ng/ml EGF and 50 g/ml BPE), and melanocyte growth medium (supplemented with 1 ng/ml recombinant basic fibroblast growth factor, 5 g/ml insulin, 0.5 g/ml hydrocortisone, 10 ng/ml phorbol 12-myristate 13-acetate, 50 g/ml streptomycin, and 0.2% (v/v) BPE), respectively. In experiments using cultured keratinocytes, the medium was exchanged for a fresh one without supplements (EGF and BPE), and cells were cultured for 24 h, then followed by each assay. In experiments using cultured melanocytes, the medium was exchanged for a fresh one without phorbol 12-myristate 13-acetate, and cells were cultured for 24 h, then followed by each assay.
Peptide Synthesis and Antibody Production-Peptides named PEPT6 (NH 2 -MSTYKRATLDEEDLC) and PEPT7 (PPGSPMNPPHKCEVW-COOH), which correspond to the N and C termini of ECE-1␣ encoded by the human ECE-1 gene, respectively (25), were synthesized and coupled to bovine serum albumin SuperCarrier (Pierce Chemical Co., Rockford, IL). Immunization and serum collection from rabbits, and the titration and specificity of antibody production as determined by ELISA, were as described elsewhere (26,27).
Measurement of Endothelin Derivative Concentrations-For the ET ELISA assay, cells were seeded in six-well plates (Falcon, Franklin Lakes, NJ) at a density of 5 ϫ 10 4 to 10 5 cells/ml and were cultured for 24 -96 h. The medium in each well was then aspirated and exchanged for fresh medium as described above, and the cells were cultured for another 24 h. The conditioned media were then collected and quantified at 100 l/well for levels of ET-1 and Big ET-1 by ELISA. The ET ELISA kit is a solid phase enzyme immunoassay that uses the multiple antibody sandwich principle. A human purified polyclonal antibody specific for human ET-1 or Big ET-1 was attached to wells in the 96-well microtiter plates. ET-1 and Big ET-1, present in known standards and in the conditioned media, were captured by the solid phase antibody. Horseradish peroxidase-labeled rabbit anti-human ET-1 or Big ET-1 IgG was added, which bound to multiple epitopes on ET-1 or Big ET-1 attached to the solid phase. Levels of immunoreactive ET-1 and Big ET-1 were then measured by absorbance at 450 nm using an ELISA plate reader (Bio-Rad, model 3550). The amount of ET-1 and Big ET-1 in the conditioned media was determined by comparing their absorbance to that produced by standards. The standard curve was linear from 78.1 to 5000 pg/ml.
Western Blotting-ECE-1␣ expression in human endothelial cells, fibroblasts, keratinocytes, and melanocytes was also investigated using Western blotting analysis. Proteins from Nonidet P-40/SDS-solubilized cells were separated on 7.5% SDS gels, transferred to polyvinylidene difluoride membranes (Immobilon-P, Millipore Corp., Bedford, MA), and then incubated with a purified polyclonal rabbit antibody specific for ECE-1␣ at a 1/1000 dilution. Subsequent visualization of antibody binding was carried out with Enhanced Chemiluminescence (Amersham Biosciences, Inc., Arlington Heights, IL) according to the manufacturer's instructions.
Measurement of ECE Activity-These assays were performed according to Xu et al. (28). Briefly, the reaction mixture for ECE contained 0.1 M sodium phosphate buffer (pH 6.8), 0.5 M NaCl, 0.1 M human big ET-1 (1-38), and the enzyme fraction in a total of 100 l. Proteins from cells solubilized in 25 mM sodium phosphate buffer (pH 6.8) and 0.1% Triton X-100 were used as the enzyme fraction. All reactions were preincu-bated at 37°C for 15 min prior to the addition of Big ET-1. Five micrograms of the endothelial cell and fibroblast extract, 10 g of the keratinocyte extract, or 25 g from the melanocyte extract were assayed at 37°C for 2-6 h. Reactions were terminated by adding 100 l of 5 mM EDTA. The mixture was then directly assayed for mature ET-1 (1-21) using the ELISA assay. For the pH profiling studies, the buffer solution (0.1 M) was substituted as designated.
Immunoprecipitation-The cells were washed with phosphate-buffered saline (PBS) without CaCl 2 and MgCl 2 and were solubilized at 4°C in 25 mM sodium phosphate buffer (pH 6.8) containing 1% Triton X-100. The resulting 200 l (2 g/l) of each extract was then incubated with 4 l of ␣PEPT6, the prebleed of PEPT6 or ␣PEPT7, or the prebleed of PEPT7 for 1 h at 4°C, and then complexed with 50 l of Protein G-Sepharose for 30 min at 4°C. After a brief centrifugation, each supernatant was incubated with antibodies and complexed with Protein G-Sepharose an additional two times. The resulting supernatants were used for Western blotting analysis and measurement of ECE activity, as detailed above.
Amplification was performed using Taq polymerase (Invitrogen) over 29 cycles for prepro-ET-1, 29 cycles for ECE-1, and 18 cycles for G3PDH, with an automated thermal cycler (MJ Research, Inc.). Each cycle consisted of the following steps: denaturation at 94°C, annealing at 62°C, and extension for 1.5 min at 72°C. Polymerase chain reaction products were analyzed by agarose gel electrophoresis. To quantitate the expression of the transcripts, the intensities of PCR bands were measured by densitometry and are expressed as relative intensities to G3PDH.
Irradiation with Ultraviolet Light-UVB irradiation was administered with a bank of nine unfiltered FL20S-E sun lumps (Toshiba, Tokyo, Japan). The irradiance was measured with a UV radiometer (Tokyo Optical Co., Ltd., Tokyo, Japan). Keratinocytes were seeded in T-75 flasks (Falcon) at a density of 5 ϫ 10 4 to 10 5 cells/ml. After cultivation for 24 -96 h, the culture medium in each flask was replaced with PBS. The cells were washed with 10 ml of PBS three times and were then kept in 10 ml of PBS. The cells were irradiated twice at a 48-h interval with doses of UVB light ranging from 10 to 80 mJ/cm 2 , because double irradiations were more effective than a single irradiation in secreting cytokines as described in our previous studies (7). UVB light emitted most of its energy within the UVB range (295-315 nm) with a peak of 305 nm. Immediately after each UVB treatment, the PBS was removed and replaced with fresh medium. After cultivation for another 24 h, the cells were rinsed with PBS and collected.
Statistics-The level of significance of the difference was calculated by the Student t test.

Preparation and Specificity of Antibodies to ECE-1␣-Pep-
tides named PEPT6 and PEPT7, which correspond to the 15amino acid sequences at the N and C termini of ECE-1␣, respectively, were synthesized. The distinct reactivity of the anti-serum raised against PEPT6 (designated here as ␣PEPT6) with PEPT6 and the lack of its cross-reactivity with any other peptides (e.g. PEPT7 or PEP13 (the silver locus protein (32)) was demonstrated by ELISA ( Fig. 1), showing ␣PEPT6 to be immunizing peptide-specific. It should be noted that ␣PEPT7 (which recognizes the C terminus of ECE-1␣), the prebleed of ␣PEPT7, and the prebleed of ␣PEPT6 did not react with PEPT6, nor did ␣PEPT6 recognize peptides PEPT7 or PEP13.
Western blotting analysis of endothelial cell extracts, which are known to predominantly express ECE-1␣ (33), revealed that reactivity of ␣PEPT6 or ␣PEPT7 with the 130-kDa band was inhibited in a dose-dependent manner by the peptides PEPT6 or PEPT7, respectively, but not by PEPT7 or PEPT6, respectively, which demonstrates the high specificity of ␣PEPT6 and ␣PEPT7 for ECE-1␣ (Fig. 2).

Secretion of ETs by Human Keratinocytes in Comparison with Human Endothelial Cells and Other Types of Skin Cells-
ELISA of Big ET-1 and ET-1 in conditioned medium revealed that human endothelial cells, fibroblasts, and keratinocytes, but not melanocytes, have the capacity to secret significant amounts of Big ET-1 and ET-1, with the highest conversion ratio from Big ET-1 to ET-1 in fibroblasts (Fig. 3).
Detection of ECE-1 Gene Transcripts in Human Keratinocytes-RT-PCR analysis of ECE-1 transcript levels using various types of skin cells revealed that ECE-1 was distinctively expressed in cultured human keratinocytes. There was also a marked expression of the ECE-1 transcript in other types of cultured human skin cells, including fibroblasts and melanocytes, with the highest level being in endothelial cells (Fig. 4).
Detection of ECE-1␣ Protein in Human Keratinocytes-Western blotting with ␣PEPT6 revealed that ECE-1␣ protein was readily detectable in cultured human keratinocytes. Fibroblasts also produced ECE-1␣ protein, although at undetectable levels in melanocytes (Fig. 5).
Effect of Phosphoramidon on ET-1 Secretion-When human keratinocytes were treated for 24 h with phosphoramidon, an inhibitor of metalloproteinases, ET-1 secretion into the medium was remarkably suppressed at concentrations of 1-100 M, in a dose-dependent manner, and this was accompanied by a concomitant increase in the release of Big ET-1 (Fig. 6).
Detection of ECE Activity in Human Keratinocytes-We compared ECE activity among various types of skin cells and endothelial cells. When cell extracts were incubated with Big ET-1 at pH 6.8, the highest activity of ECE was observed with endothelial cells. Lesser levels of ECE activity were found in fibroblasts and keratinocytes, and no detectable activity was found in melanocytes (Fig. 7).
pH Profile of ECE Activity in Human Keratinocytes-Analysis of the optimal pH for ECE activity revealed that there was a sharp pH dependence in its activity with a peak (at pH 7.0 -7.2) in endothelial cells and in keratinocytes (Fig. 8). Although the mean value of ECE activity at pH 6.8 in endothelial cells is different from that in Fig. 7, this difference may be due to substantial inter-donor variability. When 100 M phosphoramidon was added to the assay mixture to exclude phosphoramidon-insensitive activity, the activity of phosphoramidonsensitive ECE was found to exhibit only single peak at pH 6.6 and 7.0 in human endothelial cells and human keratinocytes, respectively (Fig. 9). Immunoprecipitation-To clarify to what extent ECE-1␣ contributes to the detectable ECE activity of human keratinocytes, we measured ECE activity following depletion of ECE-1␣ by immunoprecipitation with antibodies to ECE-1␣ (␣PEPT6 and ␣PEPT7). Western blotting following the immunoprecipitation of ECE-1␣ demonstrated that the supernatant obtained after incubation of cell extracts contains a detectable band of ECE-1␣ following immunoprecipitation with ␣PEPT7 but not after immunoprecipitation with ␣PEPT6 (Fig. 10A). In con-trast, the precipitates from the same assays have distinct or weak bands of ECE-1␣ following immunoprecipitation with ␣PEPT6 or ␣PEPT7, respectively (Fig. 10B). These results indicate that ␣PEPT6, but not ␣PEPT7, was able to immunoprecipitate ECE-1␣. Assays of ECE activity in supernatants following immunoprecipitation with ␣PEPT6 demonstrated that endothelial cells and human keratinocytes have detectable activities of ECE, which correlate well with the ECE-1␣ immunoprecipitated with ␣PEPT 6 (Fig. 11).
Effect of UVB Exposure on ECE-1␣ Expression-UVB irradiation is a potent stimulator of ET-1 secretion by human keratinocytes (34). Therefore, we tested whether UVB exposure, or its inducible primary cytokines, IL-1␣ and TNF␣, stimulates the expression of ECE-1 in human keratinocytes. When human keratinocytes were exposed twice at a 48-h interval to UVB irradiation, the expression of transcripts encoding ECE-1 (as FIG. 4. RT-PCR analysis of RNA isolated from human endothelial cells, fibroblasts, keratinocytes, and melanocytes. Total RNA was extracted and reverse-transcribed with Moloney murine leukemia virus reverse transcriptase using oligo(dT) 16 as a primer. The cDNA synthesized was PCR-amplified with the ECE-1-specific primer set, the prepro-ET-1-specific primer set, or the G3PDH primer set. Semiquantitative RT-PCR was carried out for 29 cycles for ECE-1 and prepro-ET-1, and for 18 cycles for G3PDH. PCR products were analyzed by agarose gel electrophoresis. Relative intensity was calculated by expressing the intensity in endothelial cells as 100%. assessed by RT-PCR) (Fig. 12) or of ECE-1␣ protein (as assessed by Western blotting analysis) (Fig. 13) was unchanged. This was in contrast to a marked dose-dependent increase in the expression of the transcript encoding prepro-ET-1. On the other hand, the addition of the UVB-inducible cytokine, IL-1␣, at concentrations ranging from 0.05 to 0.5 nM stimulated the expression of ECE-1 transcripts (Fig. 14) and ECE-1␣ protein (Fig. 15) in a dose-dependent manner, although TNF␣ had no effect.
Effect of a High Calcium Concentration on the Expression of ET-1 and ECE-When human keratinocytes were cultured at a high calcium concentration (1.5 mM), the expression of prepro-ET-1 transcripts (as assessed by RT-PCR) (Fig. 16) and ET-1 secretion (as measured by ELISA) were significantly increased (Fig. 17). In contrast, the expression of ECE-1 transcripts (as assessed by RT-PCR) (Fig. 16) and ECE activity (Fig. 18) remained unchanged at an increased calcium concentration. DISCUSSION ECE is membrane-bound metalloproteinase that plays a pivotal role in the final step of ET biosynthesis (28,35). Two isoforms of ECE, ECE-1 and ECE-2, have been identified to date (28,35,36). ECE-1 is expressed abundantly in endothelial cells, and it was first purified to homogeneity from rat lung microsomes (37). ECE-2, which is encoded by a distinct gene, was cloned from bovine adrenal cortex in the course of cloning of ECE-1 (36). The deduced amino acid sequence of ECE-2 is quite similar to ECE-1, with an overall identity of 59%. The main structural features of ECE-1 are completely conserved in ECE-2, both being type II integral membrane proteins with a typical zinc binding consensus sequence. On immunoblot analysis, both are expressed as proteins of ϳ130 kDa under reducing conditions. However, ECE-2 is strikingly different from ECE-1 in several respects. The sensitivity of ECE-2 to phosphoramidon is 250-fold higher than ECE-1, whereas FR901533 (WS79089B; 1,6,9,14-tetrahydroxy-3-(2-hydroxypropyl)-7-methoxy-8,13-dioxo-5,6,8,13-tetrahydrobenzo-[a]naphthacene-2-carboxylate⅐Na) inhibits both enzymes at similar concentrations. Yet another distinction between ECE-1 and ECE-2 is their optimal pH; ECE-2 has an acidic pH optimum at pH 5.5, which is in sharp contrast to the neutral pH optimum of ECE-1. Human ECE-1 has recently been reported to exist as four distinct isoforms encoded by the same gene (38) but using different promoters. Those four isoforms have distinct differences in their tissue distribution as well as in their subcellular localization, although such evidence is mainly limited to endothelial cells and smooth muscle cells.
We report here, for the first time, that ECE-1 transcripts and ECE-1␣ protein are constitutively expressed in normal human keratinocytes as well as in human fibroblasts. Although ECE-1 transcript was detected in human melanocytes, protein expression and catalytic function were not observed in this experiment, which suggests that keratinocytes are a major epidermal cell associated with ECE activity. ECE-1 expression in human keratinocytes is approximately one order of magnitude less than in human endothelial cells, whereas the level of ECE-1 in human fibroblasts was slightly lower than in endothelial cells. Consistent with those expression levels, phosphoramidon-sensitive ECE activity in human keratinocytes was approximately one order of magnitude less than in human endothelial cells and was 5-fold lower than in human fibroblasts. In contrast, the conversion ratio from Big ET-1 to ET-1 in these three different types of cells was similar, which indicates that the biosynthesis of ET-1 may be associated with the expression level of ECE-1 even in different types of cells by virtue of unknown mechanisms.
As for the biochemical characteristics of ECE-1␣ expressed by human keratinocytes, ET-1 secretion was found to be abrogated by a metalloproteinase inhibitor, phosphoramidon, in a dose-dependent manner, and this was accompanied by a concomitant increase in the secretion of Big ET-1. This is the same inhibitory profile as reported for bovine aortic endothelial cells (39). Analysis of the pH dependence for the enzymatic action of ECE-1 revealed an optimal activity at a neutral pH, 6.6 -7.0, in human keratinocytes and in endothelial cells, with the former showing a sharp pH dependence. Collectively, these results indicate that human keratinocyte-derived ECE-1 is almost identical in enzymatic properties to that derived from human endothelial cells. Our immunodepletion study using antibodies to ECE-1␣ demonstrated that ECE-1␣ is a major isoform in human keratinocytes and that it plays an essential role in the secretion of ETs by human keratinocytes.
We have succeeded in preparing two rabbit polyclonal antibodies against the N and C termini of ECE-1␣. Both antibodies specifically recognized ECE-1␣ protein in Western blotting, but ␣PEPT7 was unable to immunoprecipitate ECE-1␣ in the keratinocyte extracts. ECE-1 has been reported to form a disulfidelinked homodimer via Cys 412 with a certain three-dimensional conformation in which the N terminus tends to face toward the cytoplasmic domain, whereas the C terminus is located outside the plasma membrane. Such a conformational feature might explain the distinct immunological specificities observed between ␣PEPT6 and ␣PEPT7.
Previous studies in vitro with human epidermal cells indicated that keratinocytes regulate melanocyte activity in response to several stimuli, modulating the proliferation of melanocytes via ETs, production of which are markedly increased by UVB light or by IL-1␣ and IL-1␤ (7). Because UVB light stimulates keratinocytes to release IL-1 (40) and ET secretion is increased by UVB light (the level of which is very comparable with that induced by IL-1), and because this can be abrogated by the addition of antibodies to IL-1␣, it has been suggested that UVB light indirectly increases ET production through IL-1␣ (7). One important issue to be resolved in mechanism of ET-1 stimulation should be whether increased ECE-1␣ expression contributes to the increased ET-1 production during UVB stimulation. It is thus intriguing to determine whether ECE-1␣ expression is also stimulated by UVB irradiation or by its inducible primary cytokines. In human keratinocytes exposed to UVB irradiation, ECE-1 transcript and ECE-1␣ protein expression were not increased, despite the fact that prepro-ET-1 transcript expression was enhanced in a dose-dependent manner. This is consistent with our previous report describing that prepro-ET-1 transcript expression was enhanced by two UVB doses of 50 mJ/cm 2 (7). Therefore, it is likely that up-regulation of prepro-ET-1, but not of ECE-1␣, is mainly responsible for the UVB-induced stimulation of ET-1 production and secretion.
To clarify the regulatory mechanism(s) involved in the production and function of ECE-1 in the epidermis (which consists mainly of keratinocytes), we have found that the exogenous application of IL-1␣, but not of TNF␣, at concentrations rang- ing from 0.05 to 0.5 nM, to cultured human keratinocytes significantly up-regulates the expression of ECE-1 transcript within 48 h. This suggests that the functional role of ECE-1 is strictly regulated by the primary inflammatory cytokine IL-1␣, the secretion of which is generally stimulated by several factors, including UV irradiation. The lack of a stimulatory effect by UVB irradiation in the present study might result from an insufficiency of such cytokines in the culture medium (determined to be about 0.005 nM) in a concentration sufficient to increase expression of ECE-1.
In human keratinocytes, protein expression and differentiation are regulated by calcium concentrations (41). Thus, it is intriguing to know whether genetic and enzymatic profiles involved in ECE-1␣ and ET-1 in human keratinocytes are affected by calcium. We found that ECE activity and ECE-1 transcripts expression were not substantially affected by a high calcium concentration (1.5 mM) in cultured human keratinocytes, whereas the secretion of ET-1 was significantly increased at a high calcium concentration accompanied by an increased expression of prepro-ET-1. These results suggest that calciuminduced ET-1 secretion is not attributable to the increased ECE FIG. 13. Western blotting of ECE-1␣ in human keratinocytes after UVB irradiation. Human keratinocytes were irradiated twice at a 48-h interval with doses of UVB light ranging from 10 to 80 mJ/cm 2 and were collected after cultivation for another 24 h, as described under "Experimental Procedures." Human keratinocytes were solubilized in Nonidet P-40/SDS buffer; 10 g protein from each extract were electrophoresed and analyzed by Western blotting, as described under "Experimental Procedures."  12. RT-PCR analysis of ECE-1 transcript levels in human keratinocytes after UVB irradiation. Human keratinocytes were irradiated twice at a 48-h interval with doses of UVB light ranging from 10 to 80 mJ/cm 2 and were collected after cultivation for another 24 h, as described under "Experimental Procedures." Total RNAs were extracted and were reverse-transcribed with Moloney murine leukemia virus reverse transcriptase using oligo(dT) 16 as a primer. The cDNAs synthesized were PCR-amplified with the prepro-ET-1-specific primer set, the ECE-1-specific primer set or the G3PDH primer set. Semiquantitative RT-PCR was carried out for 29 cycles for prepro-ET-1 and ECE-1, and for 18 cycles for G3PDH. PCR products were analyzed by agarose gel electrophoresis. Separate experiments were repeated three times. Relative intensity was calculated by expressing the intensity in un-irradiated cells as 100%.

FIG. 14. RT-PCR analysis of ECE-1 transcript levels in human keratinocytes after exposure to IL-1␣ (A) or TNF␣ (B).
Human keratinocytes were cultured with IL-1␣ or TNF␣ at concentrations of 0.05-0.5 nM as noted. After cultivation for 48 h, human keratinocytes were collected and total RNAs were extracted. Total RNAs were reverse-transcribed as described under "Experimental Procedures." The cDNAs synthesized were PCR-amplified with the ECE-1-specific primer set or with the G3PDH primer set. Semiquantitative RT-PCR was carried out for 29 cycles for ECE-1, and for 18 cycles for G3PDH. PCR products were analyzed by agarose gel electrophoresis. Separate experiments were repeated three times. Relative intensity was calculated by expressing the intensity in untreated cells as 100%. activity but to the increased expression of prepro-ET-1 as seen for UVB-induced stimulation of ET-1 production.
In conclusion, our results have provided a new insight into the constitutive existence of ECE-1 as an important metalloproteinase expressed by human keratinocytes and into the essential role of ECE-1␣ in epidermal pigmentation via its control of the secretion of ET-1 by keratinocytes. ET-1 acts as a potent activator for melanocytes via the G i -coupled transmembrane receptor, ET B R, which leads to a variety of epidermal hyperpigmentary disorders (42). Importantly, the present findings facilitate a fundamental understanding of mechanisms involved in regulating melanogenesis as well as the homeostatic maintenance of skin color. FIG. 16. RT-PCR analysis of the expression of ECE-1 and prepro-ET-1 transcripts by human keratinocytes at a high calcium concentration. Before assay, the medium was exchanged for a fresh one without supplements, and cells were cultured for 24 h, then followed by incubation at a low calcium concentration (0.09 mM), which was originally included in SFM, or a high calcium concentration (1.5 mM) for 24 h. After the cultivation, total RNA was extracted and reverse-transcribed with Moloney murine leukemia virus reverse transcriptase using oligo(dT) 16 as a primer. The cDNA synthesized was PCR-amplified with the ECE-1-specific primer set, the prepro-ET-1specific primer set, or the G3PDH primer set. Semiquantitative RT-PCR was carried out for 29 cycles for ECE-1 and prepro-ET-1 and for 18 cycles for G3PDH. PCR products were analyzed by agarose gel electrophoresis. Separate experiments were repeated three times. Relative intensity was calculated by expressing the intensity in cells treated at a low calcium concentration as 100%.
FIG. 18. ECE activity in human keratinocytes at high calcium concentration. Before assay, the medium was exchanged as described in the legend of the Fig. 16. After the cultivation with low (0.09 mM) or high (1.5 mM) calcium concentration for 24 h, enzyme fractions from the cells were incubated with 0.1 M Big ET-1 and 0.5 M NaCl in 0.1 M sodium phosphate buffer at pH 6.8, and the mixture was then directly assayed for mature ET-1 as described under "Experimental Procedures." The values reported represent means Ϯ S.D. from three independent experiments.