Search

To identify the cis-acting element that is responsible for the pigment cell-specific expression of the human tyrosinase gene, we analyzed the promoter activity of its 5'-flanking region by transient expression assays. The fusion genes were constructed by inserting the 5'-flanking region of the human tyrosinase gene upstream from the firefly luciferase gene and were introduced into human melanoma cells and HeLa cells. We thus found the element, located between 2.7 and 1.8 kilobase pairs upstream from the transcription initiation site, that enhances the transient expression of the luciferase reporter gene in melanoma cells, but not in HeLa cells, the tyrosinase gene expression of which is not detectable. Using the fusion genes containing putative enhancer elements under the control of the heterologous simian virus 40 promoter, we identified the pigment cell-specific enhancer of approximately 200 base pairs (bp) between -2.0 and -1.8 kilobase pairs and localized the core sequence to a 39-bp region. This 39-bp core element was then confirmed to direct the melanoma cell-specific expression of the reporter gene under the tyrosinase gene promoter. We thus propose that this core element is responsible for the pigment cell-specific expression of the human tyrosinase gene.

The pigment cell-specific gene, located at the brown (b) locus in mouse, has been cloned and characterized. Its gene product is required for the formation of black melanin rather than brown, although its exact function remains to be elucidated. We thus tentatively named it b-locus protein in this report. The b-locus protein gene is about 18 kilobase pairs long and organized into 8 exons and 7 introns. Functional analysis of its promoter region suggests that the nucleotide residues -38/154 is sufficient to direct the pigment cell-specific transcription in melanoma whole cell extracts. On the other hand, we were unable to detect its transcripts in HeLa whole cell extracts. Sequence comparison with the promoter region of the tyrosinase gene, another pigment cell-specific gene, reveals that two elements of the b-locus protein gene (-33/-24 and 18/28) are also conserved in the tyrosinase gene at equivalent positions, suggesting that these two elements may be involved in their pigment cell-specific transcription. Furthermore, we have cloned a cDNA, pMT3, coding for an isoform of b-locus protein from a cDNA library of mouse B16 melanoma cells. Sequence analysis of pMT3 reveals a deletion of 103 base pairs, which corresponds to the 5'-end of the exon 8 of the b-locus protein gene, indicating that pMT3 represents a mRNA species generated by alternative splicing. Since this deletion changes the reading frame and eliminates the transmembrane domain of b-locus protein, the pMT3-type mRNA may code for a soluble isoform. Such an isoform, consisting of 553 amino acids, differs only in its carboxyl terminus and is larger than b-locus protein by 16 amino acids. Using transient expression assays, we confirmed that such an isoform is able to react with anti-b-locus protein monoclonal antibody, TMH-1, suggesting that a b-locus protein isoform may have some function in pigmentation.

In this study we explored the possible application of MAT-1, which has been established as a monoclonal antibody against human tyrosinase, for detection of mouse tyrosinase. The MAT-1 reacted with B16 mouse melanoma cells, but not with tyrosinase-negative NIH-3T3 mouse fibroblasts. In western blot analysis of the large granule fraction (LGF) of B16 cells, MAT-1 detected a single protein of 80 kDa, whose size was close to that of human tyrosinase detected with MAT-1 in extracts of human melanocytes. Furthermore, the 80 kDa band that was detected with MAT-1 in the LGF of B16 cells was also detected by DOPA reaction. In order to confirm that the protein detected with MAT-1 is tyrosinase, a transient expression assay was carried out. When mouse tyrosinase or mouse tyrosinase-related protein 1, which shares high homology with human tyrosinase, was transiently expressed in tyrosinase-negative K1735 mouse melanoma cells by cDNA transfection, MAT-1 reacted only with the cells expressing mouse tyrosinase. These results indicate that MAT-1 specifically reacts with mouse tyrosinase.

Immunohistochemical localization of tyrosinase was examined with a monoclonal antibody (MoAb MAT-1) against human tyrosinase on routine formalin-fixed paraffin-embedded sections of 3 normal skin specimens, 15 melanocytic tumors (6 pigmented nevi, 3 juvenile melanomas and 6 malignant melanomas) and 3 non-melanocytic tumors. In the melanotic melanomas, almost all tumor cells were clearly stained with the antibody. In the nevocytic nevi, the nevus cells in lower epidermis and upper dermis were positive for MoAb MAT-1, but negative in middle and lower dermis. All three juvenile melanomas, one amelanotic melanoma, and three non-melanocytic tumors were entirely negative for MoAb MAT-1. Thus, MoAb MAT-1 could recognize the cells with melanogenic activity on routine formalin-fixed paraffin-embedded sections. However, the staining quality was not adequate for normal epidermal melanocytes, indicating that small technical innovations in the immunostaining process such as formalin fixation after PBS washing are required. Nevertheless, MoAb MAT-1 can be expected to be very useful for identifying melanogenic cells on paraffin-embedded sections, because we have to date no other antibody available for it.

Tyrosinase is a rate-limiting enzyme in melanin biosynthesis and is specifically expressed in differentiated melanocytes. We have identified the enhancer element in the 5'-flanking region of the human tyrosinase gene that is responsible for its pigment cell-specific transcription and have termed it tyrosinase distal element (TDE) (positions -1861 to -1842). Transient expression assays showed that TDE confers efficient expression of a firefly luciferase reporter gene linked to the tyrosinase gene promoter in MeWo pigmented melanoma cells but not in HeLa cells, which do not express tyrosinase. TDE was specifically bound by nuclear proteins of MeWo and HeLa cells, the binding properties of which were indistinguishable in gel mobility shift assays. TDE contains the CATGTG motif in its center, and mutation analysis indicates that the CA dinucleotides of this motif are crucial for protein binding and pigment cell-specific enhancer function. The CATGTG motif is consistent with the consensus sequence recognized by a large family of transcription factors with a basic helix-loop-helix structure, which prompted us to examine the possible involvement of a ubiquitous transcription factor, USF, and a novel factor, microphthalmia-associated transcription factor (MITF), recently cloned as the human homolog of the mouse microphthalmia (mi) gene product. The mi phenotype is associated with a mutant mi locus and characterized by small eyes and loss of melanin pigments. Both USF and MITF are predicted to contain a basic helix-loop-helix structure and a leucine zipper structure. We provide evidence that USF binds to TDE, whereas we were unable to detect the DNA-binding activity of MITF. Transient coexpression assays showed that MITF specifically transactivates the promoter activity of the tyrosinase gene through the CATGTG motif of TDE but not the promoter of the ubiquitously expressed heme oxygenase gene, while USF is able to activate both promoters. These results indicate that MITF is a cell-type-specific factor that is capable of activating transcription of the tyrosinase gene.

We have cloned and sequenced the human genomic DNA segments encoding the 5'-flanking region and the first two exons of the tyrosinase-related protein (TRP) gene, a pigment cell-specific gene. Functional analysis of its promoter suggests that the downstream region of the TRP gene, including the first intron, enhances the transient expression of the luciferase gene under control of the TRP gene promoter about 16- to 20-fold. This enhancer-like activity is detected not only in melanoma cells but also in HeLa cells whose TRP gene expression is assumed to be repressed. We suggest a possibility that the downstream region is not sufficient to confer pigment cell-specific expression, but is required for efficient transcription of the TRP gene in pigment cells.

The pigment cell-specific gene, located at the brown (b)-locus in mouse, encodes the protein that determines the type of melanin synthesized. This protein is known as tyrosinase-related protein, but here we tentatively term it b-locus protein to avoid confusions with the related sequence cross-hybridizing to the tyrosinase gene. In order to identify the mutation at the b-locus, we have cloned and characterized the b-locus protein gene of BALB/c mouse (b/b, c/c). The gene is about 18 kb long and organized into 8 exons and 7 introns. Sequence analysis of the b-locus protein gene reveals four base changes within the protein-coding regions: two missense mutations and two silent mutations. Two missense mutations result in the Cys to Tyr substitution at position 86 (codon 110) and the Arg to His substitution at position 302 (codon 326) of a b-locus protein molecule. Using allele-specific amplification, we confirmed that these missense mutations are actually present in the genomic DNA of two b-mutant strains examined, BALB/c and DBA/2 (b/b, C/C) mice, suggesting that these mutations are specific for the mutant mice at the b-locus. Moreover, we are able to show that the b-locus protein containing Tyr 86 is not reactive with the anti-b-locus protein monoclonal antibody, TMH-1, in transient expression assays.

There is no available monoclonal antibody which reacts specifically recognizes human tyrosinase. Employing a synthetic peptide, MEKEDYHSLYQSHL, corresponding to the carboxyl terminus of human tyrosinase as an immunogen, we produced a mouse monoclonal antibody MAT-1 of the IgG1 isotype. The epitope for MAT-1 was determined to be EDYH, the sequence of which is not present in human tyro-sinase-related protein-1 (TRP-1) or tyrosinase-related protein-2 (TRP-2). By transient expression assays and immunofluorescence technique, we show that MAT-1 reacts specifically with cells expressing human tyrosinase cDNA but not with cells expressing TRP-1 or TRP-2 cDNA. The results of immunohisto-chemical staining also confirmed that MAT-1 reacts specifically with epidermal melanocytes in human skin sections. MAT-1 should be invaluable for studying the interaction between tyrosinase and TRPs and for detecting the changes in the levels of tyrosinase expression. In addition, MAT-1 should be useful as a sensitive immunohistochemical tool for investigation of various pigmentary disorders and possibly for the diagnosis of melanoma.