Functional Analysis of Microphthalmia-associated Transcription Factor in Pigment Cell-specific Transcription of the Human Tyrosinase Family Genes*

(Received for publication, July 9, 1996, and in revised form, August 30, 1996)

Ken-ichi Yasumoto Dagger , Kouji Yokoyama Dagger , Kazuhiro Takahashi Dagger , Yasushi Tomita § and Shigeki Shibahara Dagger

From the Dagger  Department of Applied Physiology and Molecular Biology, Tohoku University School of Medicine, Aoba-ku, Sendai, Miyagi 980-77 and the § Department of Dermatology, Akita University School of Medicine, Akita 010, Japan

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
Acknowledgment
REFERENCES


ABSTRACT

Tyrosinase, tyrosinase-related protein-1 (TRP-1), and TRP-2 are the enzymes involved in melanin biosynthesis and are preferentially expressed in pigment cells. Their human gene promoters share the 11-base pair M box containing a CATGTG motif, which was shown here to be bound in vitro by microphthalmia-associated transcription factor (MITF). Transient cotransfection analysis showed that MITF overexpression increased the expression of a reporter gene under the control of the human tyrosinase or TRP-1 gene promoter but not the TRP-2 promoter. The promoter activation caused by MITF is dependent on each CATGTG motif of the distal enhancer element, the M box, and the initiator E box of the tyrosinase gene and the TRP-1 M box. Furthermore, a truncated MITF lacking the carboxyl-terminal 125 amino acid residues transactivated the tyrosinase promoter less efficiently than did MITF, suggesting that MITF's carboxyl terminus contains a transcriptional activation domain, but unexpectedly such a truncated MITF remarkably transactivated the TRP-2 gene promoter. These results suggest that MITF is sufficient to direct pigment cell-specific transcription of the tyrosinase and TRP-1 genes but not the TRP-2 gene.


INTRODUCTION

Melanin production is specifically observed in the differentiated melanocytes of neural crest origin and in the retinal pigment epithelium derived from the optic cup of the brain. Tyrosinase (EC 1.14.18.1) is a rate-limiting enzyme in melanin biosynthesis that catalyzes the conversion of tyrosine to 3,4-dihydroxyphenylalanine (DOPA),1 DOPA to DOPA-quinone, and possibly 5,6-dihydroxyindole to indole-5,6-quinone (1). Two other enzymes were recently shown to be involved in melanin production: tyrosinase-related protein-1 (TRP-1) (2, 3) and TRP-2 (4, 5). There is about 40% amino acid identity among human tyrosinase (6, 7, 8), TRP-1 (9, 10), and TRP-2 (11, 12), all of which constitute the tyrosinase family and are specifically expressed in melanin-producing cells. Thus, the tyrosinase family genes provide a useful experimental model for the study of a network of transcription factor interactions involved in cell type-specific gene expression.

The tyrosinase family gene promoters share an M box of 11 bp, containing a CATGTG motif in its center whose motif is a consensus sequence of the binding site for a large family of transcription factors with a basic helix loop helix (bHLH) structure (13, 14). The bHLH structure is required for DNA binding of transcription factors (15). Microphthalmia-associated transcription factor (MITF) is the human homolog of the mouse microphthalmia (mi) gene product, which was recently identified as a novel factor with a bHLH-leucine zipper structure (16, 17, 18). The mi phenotype is associated with the mutant mi locus and is characterized by small nonpigmented eyes, a lack of melanocytes in the skin and inner ear, a deficiency in mast cells, and osteopetrosis caused by dysfunction of osteoclasts. Mi protein (mi gene product) shares the significant amino acid sequence similarity with several transcription factors, such as TFE3, TFEB, and TFEC (19, 20, 21), and was shown to bind in vitro to DNA as a heterodimer formed with TFE3, TFEB, or TFEC (22). Recently, the carboxyl-terminal region of TFE3, sharing a significant amino acid similarity with that of Mi or TFEB, was shown to contain an activation domain (23).

The structural gene coding for MITF is organized into nine exons (24), although at least two types of exon 1 of the mouse mi gene have been reported (25). In this context, we have recently identified a melanocyte-type promoter of the MITF gene, which was able to direct preferential expression of a reporter gene in melanoma cells (26). The mutations in the MITF gene were identified in two families affected with Waardenburg syndrome type 2, which is a dominantly inherited syndrome of sensorineural hearing loss and patchy abnormal pigmentation of the eyes, hairs, and skin (24). It is therefore conceivable that MITF is involved in the differentiation and maintenance of melanocytes. Recently, we reported that MITF is able to transactivate the human tyrosinase gene promoter through the tyrosinase distal element (TDE) containing a CATGTG motif, located 1842 bp upstream from its transcriptional initiation site (27). TDE is sufficient for melanocyte-specific expression of a reporter gene linked to either the homologous or heterologous promoter (27, 28). In contrast, melanocyte-specific expression of the human TRP-2 gene is directed by the two separate regulatory regions, the 32-bp element and the proximal region containing an M box (29). On the other hand, melanocyte-specific promoter function was not detectable within the 3.5-kilobase pair 5'-flanking region of the human TRP-1 gene by transient expression assays (29), although the TRP-1 promoter contains an M box (30).

Here we explore the possible involvement of MITF in directing pigment cell-specific transcription of the human tyrosinase family genes and show the functional heterogeneity of the M box shared by these genes. We also show that the 133-bp TRP-1 promoter region, containing an M box, is able to direct preferential expression of a reporter gene in pigment cells. In addition, we suggest that the carboxyl terminus of MITF contains a transcriptional activation domain and also plays an important role in defining the target gene.


EXPERIMENTAL PROCEDURES

Construction of Fusion Genes

All constructs used contain the firefly luciferase gene as a reporter linked to each 5'-flanking region of the human tyrosinase family genes. The luciferase gene was derived from pSV2/L, containing the luciferase gene under the control of SV40 early promoter (31), kindly provided by Dr. S. Subramani (University of California at San Diego, La Jolla, CA). A promoterless plasmid pL1 and the tyrosinase-luciferase fusion plasmids pHTL6, pHTL10, and pHTL12 were constructed as described previously (28). The constructs pHTRPL16 (29) and pHTRPL4 (30) contain the human TRP-1 gene segments spanning from positions about -3600 to +82 and from positions -839 to +82, respectively. The constructs pHTRPL18, pHTRPL19, and pHTRPL20 contain the SpeI-Fnu4HI fragment (positions -133 to +82), the BlnI-Fnu4HI fragment (positions -484 to +82), and the BglII-Fnu4HI fragment (-291 to +82), derived from pHTRPL4, respectively. The TRP-2-luciferase fusion plasmids, pHDTL1, pHDTL2, pHDTL8, pHDTL10, and pHDTL12, were described previously (29). pHHOSVL4 (27) contains the MscI-XhoI fragment (positions -176 to +19) of the human heme oxygenase-1 gene (32).

Site-directed Mutagenesis

The nucleotide substitutions were introduced into each CATGTG motif of the tyrosinase family gene by using a TransformerTM site-directed mutagenesis kit (Clontech) according to the method of Deng and Nickoloff (33). The mutagenic primers used were the oligonucleotide M01 (5'-GATCAGCACA<UNL>GT</UNL>ATTCCCTTCC-3', complementary to positions -55 to -34) for the TRP-1 M box (30), M02 (5'-CATTAGCACA<UNL>GT</UNL>ACCCAAAGTG-3', complementary to positions -124 to -145) for the TRP-2 M box (29), M03 (5'-AGTCATCACA<UNL>GT</UNL>ATCTCCACAA-3', complementary to positions -1844 to -1865) for TDE (27), M04 (5'-TGAAAAGCACA<UNL>GT</UNL>ACTGATTTTC-3', complementary to positions -115 to -92) for the tyrosinase proximal element (TPE) (27), and M05 (5'-GTGATTATCACA<UNL>GT</UNL>TCTTGGCTGAG-3', complementary to positions -23 to +2) for the initiator E box (27). The native TG dinucleotides were converted to the GT dinucleotides (underlined). The selection primer used for these constructions was a Trans Oligo NdeI/NcoI (Clontech), and experimental procedures were according to the manufacturer's instructions. The introduced mutations were confirmed by determining the nucleotide sequences. The constructs, pHTRPL4M and pHDTL12M, thus obtained contain the ACTGTG motif instead of the CATGTG motif at each M box carried by the original constructs, pHTRPL4 and pHDTL12. Each CATGTG motif of TDE, TPE, and initiator E box of pHTL12 was changed to the ACTGTG motif, yielding pHTL12M, pHTL12M4, and pHTL12M5, respectively.

Cell Culture and Transient Expression Analysis

MeWo human melanoma cells and HeLa human cervical cancer cells were transfected with each fusion plasmid and a beta -galactosidase expression plasmid by the calcium phosphate-precipitation method as described previously (28, 34). HMV-II human melanoma cells (35, 36) were maintained and transfected as described previously (27, 34). Luciferase activity was assayed by using a PicaGene luciferase assay system (Toyo Ink) and was measured with a Lumat LB9501 (Berthold). The obtained luciferase activity was normalized with each beta -galactosidase activity that represents an internal control. An expression plasmid for a human MITF cDNA, pRc/CMV-MITF, was constructed as described previously (27).

Gel Mobility Shift Assay

The entire region of MITF was fused to the carboxyl terminus of glutathione S-transferase (GST), and the recombinant MITF was produced in Escherichia coli as described previously (34). GST-MITF fusion protein was prepared by affinity chromatography with glutathione-sepharose 4B column (Pharmacia Biotech Inc.). Synthetic TDE was end-labeled with [gamma -32P]ATP (NEN Research) (specific radioactivity, 2 × 108 cpm/µg) and was used for the gel mobility shift assay as described previously (34). The oligonucleotides used as a probe or competitors are shown in the figures (see Figs. 3 and 5).


Fig. 3. Binding of GST-MITF fusion protein to the CATGTG motifs of the human tyrosinase gene. Gel mobility shift assay was performed with the synthetic TDE as a binding probe and GST-MITF fusion protein. Three major bands representing protein-DNA complexes and unbound probes are indicated by arrows. The sequences of the synthetic TDE, TPE, the oligonucleotides initiator E (Inr-E) (positions -20 to -1), and their mutated elements are indicated at the right. The CATGTG motifs are underlined, and the base changes are indicated by double underlining. Each competitor was added to the reaction mixture at a 70- or 700-fold molar excess to the input probe as indicated with open triangles. Uppercase letters indicate the nucleotide sequence from the tyrosinase gene, and lowercase letters indicate the linker sequence. Lane 1 contained no GST-MITF fusion protein (buffer control), indicated as Mock; lane 2, protein-DNA complexes formed in the absence of competitor.
[View Larger Version of this Image (39K GIF file)]



Fig. 5. Binding of GST-MITF fusion protein to the TRP-1 gene promoter. Binding reaction was performed with the synthetic TDE as a probe and GST-MITF fusion protein in the presence of the unlabeled competitor, the oligonucleotide TRP1E containing the M box of the TRP-1 gene (27), or each of three oligonucleotides (DDE 1, 2, and 3) for the 32-bp element of the TRP-2 gene. The oligonucleotide DDE1 represents the 32-bp element (-447 to -416). These oligonucleotide sequences are indicated to the right, and the E-box motif is underlined. Each competitor was added to the reaction mixture at a 70- or 700-fold molar excess to the input probe as indicated with open triangles. Uppercase letters indicate the nucleotide sequence from the TRP-1 or TRP-2 gene, and lowercase letters indicate the linker sequence.
[View Larger Version of this Image (31K GIF file)]


Construction of Expression Plasmids for an MITF Isoform and a Mutant MITF

An HMV-II human melanoma cDNA library was screened with an MITF cDNA probe as described previously (27). The cDNA insert of an isolated clone was subcloned into the pRc/CMV expression vector, yielding pRc/CMV-MITFDelta . A deletion mutant MITF cDNA, pRc/CMV-MITFDelta 2, was constructed as follows. The HindIII-XhoI fragment coding for the amino-terminal 294 amino acids of MITF was isolated from a subclone phMI9 containing a full-length MITF cDNA (27) and was inserted between the HindIII and XhoI sites of pRc/CMV expression vector (Invitrogen). Both restriction sites are located in the polylinker sequence. pRc/CMV-MITFDelta 2 thus codes for a mutant MITF lacking the carboxyl-terminal 125 amino acids but instead containing the Cys-Ile dipeptide, encoded by the polylinker sequence.


RESULTS

MITF Transactivates the Tyrosinase and TRP-1 Promoters but Not the TRP-2 Promoter

The constructs, containing each 5'-flanking region of the human tyrosinase family genes, are schematically shown in Fig. 1. There are three cis-regulatory elements in the human tyrosinase gene, TDE (positions -1861 to -1842), the M box (positions -104 to -94) located in TPE (positions -112 to -93), and the initiator E box (CATGTG, positions -12 to -7) (27). The two regulatory regions of the TRP-2 gene, conferring pigment cell-specific transcription, are the 32-bp element (positions -447 to -416) and the proximal region (positions -268 to -56) containing an M box (positions -138 to -128) (29). The TRP-1 promoter also contains an M box (positions -48 to -38), although its pigment cell-specific promoter function was not detectable in our previous work (29, 30).


Fig. 1. Effects of overexpression of an MITF cDNA on the promoter activities of the human tyrosinase family genes. The fusion plasmids, containing each human tyrosinase family promoter upstream of the luciferase gene, are schematically shown to the left. The regulatory regions indicated are TDE, TPE containing the M box, and the initiator E box (Inr-E) of the tyrosinase gene, the TRP-1 M box, and the 32-bp element and the M box of the TRP-2 gene. MeWo melanoma cells (A) or HeLa cervical cancer cells (B) were cotransfected with each promoter-luciferase construct and MITF expression plasmid (pRc/CMV-MITF) or vector DNA (pRc/CMV). The magnitude of activation is presented as the ratio of the normalized luciferase activity obtained with MITF cDNA to that with vector DNA (induction ratio). The data shown are means ± standard deviations (narrow bars) for three independent experiments.
[View Larger Version of this Image (37K GIF file)]


To investigate the effects of MITF overexpression on the tyrosinase family gene promoters, we performed transient cotransfection assays in MeWo melanoma cells and nonpigmented HeLa cells, in which MITF mRNA expression was not detectable (26). pSV2/L was highly expressed in both cell lines and was included as a control for the cotransfection assay. As evident in Fig. 1, MITF transactivated the TRP-1 gene promoter as well as the tyrosinase promoter, although the magnitude of activation of the TRP-1 gene promoter is smaller than that of the tyrosinase gene promoter. In contrast, the effect of MITF seen on the TRP-2 promoter was similar to the effect on the expression of pL1 or pSV2/L. We were thus unable to detect the transactivation by MITF of the TRP-2 gene promoter in both cell lines.

Functional Analysis of the CATGTG Motifs of the Tyrosinase Gene Promoter

To explore the functional significance of each CATGTG motif in the pigment cell-specific expression of the human tyrosinase gene, we constructed pHTL12M, pHTL12M4, and pHTL12M5, each containing an ACTGTG motif instead of the CATGTG motif at TDE, the M box, and the initiator E box, respectively (Fig. 2A). Transfection experiments demonstrated that the introduced modification at TDE reduced the relative luciferase activity by about 4-fold in MeWo melanoma cells, and the relative luciferase activity of pHTL12M was comparable with that of pHTL12M5, containing the substitution at the initiator E box (Fig. 2A). In contrast, the luciferase activity of pHTL12M4 was slightly decreased (about 82% of the luciferase activity of pHTL12), but this difference is not statistically significant. Expression levels of pHTL6 or pHTL10, containing the M box and the initiator E box, are comparable with those of pHTL12M containing the mutation at TDE. The expression levels of all the constructs used were similarly lower in HeLa cells. These results suggest that TDE and initiator E box are required for the efficient expression of the human tyrosinase gene in pigment cells, and TPE may have only weak promoter activity. The similar results of the transfection experiments were also obtained in another human melanoma cell line, HMV-II (data not shown).


Fig. 2. Functional significance of the CATGTG motifs of the human tyrosinase gene. A, effects of mutations at TDE, TPE, or initiator E box (Inr-E) on basal promoter activity. MeWo or HeLa cells were transfected with tyrosinase-luciferase fusion plasmids shown to the left. Constructs pHTL12M, pHTL12M4, and pHTL12M5 each contain the ACTGTG motif instead of the CATGTG motif at TDE, TPE, and initiator E, indicated with crossed lines. The normalized luciferase activity for each plasmid was divided by the normalized value obtained with promoterless construction, pL1, and shown as relative luciferase activity. B, effects of mutations on the transactivation by MITF. HeLa cells were cotransfected with each tyrosinase-luciferase construct and MITF expression plasmid or vector DNA. The relative luciferase activity shown is the ratio of each normalized luciferase activity to that obtained with pRc/CMV-MITF and vector plasmid, pRc/CMV. The values shown on the right are means ± standard deviations (narrow bars) for at least three independent experiments.
[View Larger Version of this Image (30K GIF file)]


Subsequently, we performed cotransfection assays with an MITF expression plasmid in HeLa cells not expressing the endogenous MITF. The expression levels of any fusion genes were low in HeLa cells (Fig. 2A) but were remarkably increased by MITF overexpression (Fig. 2B). The induction ratio of pHTL12 is about four times higher than that of pHTL12M, pHTL12M4, pHTL6, or pHTL10. The magnitude of transactivation of pHTL12M5 was lowest, suggesting that initiator E box may play an important role in the transactivation by MITF. These results were reproducible in MeWo melanoma cells (data not shown), except that the magnitude of transactivation by MITF was much lower in MeWo cells than in HeLa cells as shown in Fig. 1.

Binding of MITF to the CATGTG Motifs of the Human Tyrosinase Gene

Our initial attempt to detect the DNA binding activity of MITF produced by in vitro translation was unsuccessful (27). Consequently, in the present study, using GST-MITF fusion protein and synthetic TDE as a probe, we analyzed the DNA binding properties of MITF by gel mobility shift assay (Fig. 3). The specific DNA-protein complexes were detected as at least three bands, suggesting that the fusion protein may be partially degraded during experimental procedures. The formation of these bands was inhibited by excess amounts of synthetic TDE (Fig. 3, lanes 2-4) but not by the oligonucleotide TDEM containing an ACTGTG motif (Fig. 3, lanes 5 and 6), indicating that GST-MITF fusion protein specifically binds to the CATGTG motif of TDE but not to the ACTGTG motif. The formation of these specific bands was also inhibited by either synthetic TPE containing the M box or the oligonucleotide Inr-E containing the initiator E box (Fig. 3, lanes 7 and 8 and lanes 13 and 14, respectively). The oligonucleotide TPEM1 also competed for the formation of the protein-DNA complexes (Fig. 3, lanes 9 and 10). It should be noted that TPEM1 contains a base change at the 5' end of the tyrosinase M box and thus TPEM1 contains the sequence identical to the TRP-2 M box, GGTCATGTGCT (29). In contrast, any competitors containing the base changes at each CATGTG motif did not affect the specific protein bindings (Fig. 3, lanes 5 and 6, lanes 11 and 12, and lanes 15 and 16). These results suggest that MITF is able to bind each CATGTG motif of TDE, the tyrosinase M box (TPE), the initiator E box, and the TRP-2 M box.

The M Box of the Human TRP-1 Gene Promoter Is Able to Confer the Transactivation by MITF

The apparent activation of the TRP-1 promoter by MITF (see Fig. 1) prompted us to re-evaluate its pigment cell-specific promoter function. Indeed, in the course of cotransfection assays (Fig. 1), we noticed that expression levels of the TRP-1-reporter construct, pHTRPL16, are slightly higher in MeWo cells than those in HeLa cells. Transfection experiments in MeWo melanoma cells revealed the similar relative luciferase activity obtained with either pHTRPL16, pHTRPL4, pHTRPL19, pHTRPL20, or pHTRPL18 (Fig. 4A). These expression levels were always higher in MeWo cells than those in HeLa cells by about 2-3-fold, but the differences were smaller than those seen for the tyrosinase or TRP-2 gene by more than 20-fold (see Fig. 2; data not shown), as reported previously (29). However, the differences in the expression levels of pHTRPL16 between the two cell lines were not statistically significant. It is also noteworthy that expression levels of a construct pHTRPL4M containing a mutated M box were lower than those of pHTRPL4 by about 5-fold in MeWo cells, but expression of both constructs was similar in HeLa cells (Fig. 4A). These results suggest that the segment (positions -133 to +82) containing the M box, carried by pHTRPL18, is able to direct preferential expression of the TRP-1 gene in pigment cells.


Fig. 4. Functional analysis of the human TRP-1 gene promoter. A, functional significance of the M box. MeWo or HeLa cells were transfected with the fusion plasmids containing TRP-1 promoter as indicated on the left of the figure. pHTRPL4M contains the ACTGTG motif instead of the CATGTG motif at the M box, indicated with crossed lines. Relative luciferase activity shown represents the ratio of each normalized luciferase activity to the value obtained with pL1. The data shown are means ± standard deviations for at least three independent experiments. B, identification of the TRP-1 M box as a cis-acting element responsible for transactivation by MITF. MeWo melanoma cells were cotransfected with TRP-1 promoter-luciferase construct together with MITF cDNA or vector DNA. The magnitude of activation is presented as the ratio of normalized luciferase activity obtained with MITF cDNA to that with vector DNA (induction ratio). The data shown are means ± standard deviations (narrow bars) for three independent experiments.
[View Larger Version of this Image (31K GIF file)]


We then searched for the cis-acting element of the TRP-1 gene, which is responsible for the transactivation by MITF. The relative luciferase activities of any TRP-1 constructs were increased about 3-fold by coexpression with MITF cDNA in MeWo melanoma cells, whereas no activation was detected with pHTRPL4M (Fig. 4B). Because all of the constructs used in this experiment contain an intact M box, except for pHTRPL4M, the M box seems to be responsible for the transactivation of the TRP-1 promoter by MITF.

MITF Does Bind to the TRP-1 M Box but Not to the 32-bp Element of the TRP-2 Gene

To examine whether MITF binds to the regulatory elements of the TRP-1 or TRP-2 gene, we performed gel mobility shift assay using the synthetic TDE as a binding probe. As evident in Fig. 5, the binding of GST-MITF fusion protein to TDE was completely inhibited by the oligonucleotide TRP1E containing the TRP-1 M box (Fig. 5, lanes 2 and 3) but was not by the 32-bp element (-447 to -416) containing a CAATTG motif of the TRP-2 gene (Fig. 5, lanes 4-9) (29). These results suggest that MITF binds the TRP-1 M box but not the 32-bp element of the TRP-2 gene, indicating the significance of a CATGTG motif for the DNA binding by MITF.

The TRP-2 M Box Is Not Sufficient for the Transactivation by MITF

To confirm the observations that the TRP-2 gene promoter is not activated by MITF as shown in Fig. 1, we explored the effect of MITF on the TRP-2 promoter activity using several other constructs (Fig. 6). pHDTL12, containing the 32-bp element and the proximal region, was preferentially expressed in MeWo cells (29). The expression level of pHDTL12M, containing the base changes at the M box, was similar to that of pHDTL1 (Fig. 6), suggesting that the TRP-2 M box is essential for efficient transcription in pigment cells and thus represents the function of the proximal region (29). On the other hand, we could not detect a significant increase in the luciferase activities by the cotransfection with MITF cDNA as compared with the transactivation seen in the tyrosinase or TRP-1 construct (see Fig. 1). The relative luciferase activities of pHDTL12M, pHDTL10, pHDTL2, and pHDTL1 were not noticeably influenced by MITF. The induction ratio obtained with pHDTL12 was slightly higher than that with pHDTL10, but the difference was not statistically significant.


Fig. 6. Effects of overexpression of an MITF cDNA on the human TRP-2 gene promoter. MeWo cells were cotransfected with each TRP-2 promoter-luciferase construct together with MITF cDNA or vector DNA. The structure of the fusion plasmids used are schematically shown to the left. Two regulatory regions required for pigment cell-specific expression of the TRP-2 gene (29), the 32-bp element and the proximal region, are indicated. The M box is located in the proximal region, and the mutated M box is shown with crossed lines. The relative luciferase activity shown is the ratio of each normalized luciferase activity to the value obtained with pHDTL1 and vector DNA. The magnitude of activation is also presented as the ratio of normalized luciferase activity obtained with MITF cDNA to that with vector DNA (induction ratio). The data shown are means ± standard deviations (narrow bars) for three independent experiments.
[View Larger Version of this Image (27K GIF file)]


Functional Analysis of an MITF Isoform and a Truncated MITF

In the course of cDNA cloning for MITF (27), one type of cDNA lacking the 84-bp DNA segment representing the exon 3 was isolated (Fig. 7). This type of mRNA may be generated by alternative splicing and codes for an MITF isoform containing an internal in-frame deletion of 28 amino acids, termed MITFDelta . Expression of this mRNA species in melanoma cells was confirmed by reverse transcription and polymerase chain reaction (data not shown). In addition, we constructed an expression cDNA coding for a mutant MITF lacking the carboxyl-terminal 125 amino acids, termed MITFDelta 2. This deleted region is enriched in serine residues and shares a significant amino acid similarity with the carboxyl-terminal region of TFE3, which was shown to contain an activation domain (23). The effects of MITFDelta or MITFDelta 2 on the tyrosinase and TRP-2 gene promoters were examined, because both 5'-flanking regions possess the strong enhancer(s) that directs melanocyte-specific expression on the reporter gene. MITFDelta transactivated the tyrosinase promoter as efficiently as MITF but not the TRP-2 promoter. Thus, no functional differences were detected between MITF and MITFDelta . On the other hand, MITFDelta 2 transactivated the tyrosinase promoter 2-fold, but this activation level is 2-3-fold less than that obtained with MITF. The reduced activation of the tyrosinase promoter by MITFDelta 2 suggests that the carboxyl terminus of MITF may contain an activation domain. Unexpectedly, MITFDelta 2 transactivated the TRP-2 promoter by about 4-fold, whereas MITFDelta 2 did not transactivate the heme oxygenase-1 gene promoter containing the two cis-acting elements, each containing an E box motif (37, 38).


Fig. 7. Functional analysis of MITF variants. MeWo cells were cotransfected with each promoter-luciferase construct together with an indicated MITF cDNA. The structures of MITF and its variants used are schematically shown at the top. MITFDelta contains an internal deletion of 28 amino acids, encoded by exon 3, and MITFDelta 2 lacks the carboxyl-terminal 125 amino acids. Both deleted regions are shown with broken lines. A leucine-zipper structure and a serine-rich region are indicated with LZ and S, respectively. The magnitude of activation is presented as the ratio of normalized luciferase activity obtained with each MITF cDNA to that with vector DNA. The data shown are means ± standard deviations (narrow bars) for three independent experiments.
[View Larger Version of this Image (31K GIF file)]



DISCUSSION

The present study indicates that MITF transactivates the tyrosinase promoter mainly by binding to the CATGTG motifs of TDE and the initiator E box. It should be noted that even a single copy of TDE is able to confer the pigment cell-specific expression on a tyrosinase promoter (27) and on a heterologous SV40 promoter (28) and that TDE deviates from the M box sequence except for a CATGTG motif (Fig. 8). By introducing base changes into a CATGTG motif, we have established the significance of this motif of TDE in its enhancer function (Fig. 2) and in the MITF binding (Fig. 3). Furthermore, we suggested that the initiator E box rather than the M box plays an important role in transcription of the tyrosinase gene (Fig. 2), although MITF overexpression transactivated the promoter activity through the M box (Figs. 2B and 3). These results suggest that the affinity of MITF to the M box is lower in vivo than to TDE and the initiator E box. Other investigators also reported by using a construct containing the 300-bp human tyrosinase promoter that the initiator E box is responsible for the transactivation of the human tyrosinase promoter by Mi protein, a mouse homolog of MITF (39). Considering all these results, we suggest that MITF is sufficient to direct the pigment cell-specific transcription of the tyrosinase gene by interacting with TDE and the initiator E box.


Fig. 8. Comparison of the cis-acting elements of the human tyrosinase family genes. Functional properties of each cis-acting element containing a CATGTG motif are summarized. The negative number indicates the position of the 5' end of each sequence (27, 30, 29). The M box (underlined) and its flanking sequences, the sequence of which is also termed TPE in the case of the tyrosinase gene, are shown. Inr-E includes the initiator E box and its flanking sequences. Basal Promoter Activity, shown in the first column, indicates whether involvement of each element in efficient expression of a reporter gene is detectable in pigment cells by transient expression assays. Binding, shown in the column for MITF, indicates that each element was bound in vitro by GST-MITF fusion protein. Pigment Cell-Specific Expression, shown in the last column, indicates whether pigment cell-specific promoter function is detectable by transient expression assays of each 5'-flanking region. The symbols + and ± indicate that statistically significant effects were detectable and undetectable, respectively, under the conditions used. The symbol ++ represents that the observed effects are especially remarkable.
[View Larger Version of this Image (15K GIF file)]


The present study also indicates that the M box, shared by the human tyrosinase family genes, shows the functional heterogeneity as summarized in Fig. 8. For example, in contrast to the tyrosinase M box, each of the TRP-1 and TRP-2 M boxes is essential for the respective basal promoter activity in pigment cells (Figs. 4 and 6) but shows different responses to MITF overexpression. MITF transactivated the TRP-1 gene promoter by binding to the CATGTG motif of the M box (Fig. 4) but not the TRP-2 gene promoter (Fig. 6), even though MITF bound in vitro to the TRP-2 M box (Fig. 3). In this context, it is noteworthy that the sequences flanking each M box are entirely different among the tyrosinase family genes, and even the M box is not completely identical (10 of 11 are identical to each other) (Fig. 8). It is therefore conceivable that each M box or its overlapping sequence may be bound in vivo by certain other factors, which are involved in the basal promoter function for the TRP-1 and TRP-2 genes but not for the tyrosinase gene.

Using a sensitive assay method for luciferase activity that was not available in our previous studies (29, 30), we could detect the noticeable promoter function of the TRP-1 gene, directing preferential expression of a luciferase activity in pigment cells. However, such an activity is lower than the pigment cell-specific promoter activity of the tyrosinase or TRP-2 gene by more than 20-fold (Fig. 2 and data not shown) as reported previously (29). These results suggest that additional regulatory element(s) equivalent to TDE or the initiator E box for the tyrosinase gene may be present either in a region further upstream or somewhere downstream of the human TRP-1 gene, which is however not represented in the fusion genes used in this study. It is therefore conceivable that MITF may regulate pigment cell-specific transcription of the TRP-1 gene by binding to such an unrecognized element and to the M box. Alternatively, MITF bound to a hitherto unrecognized element may interact with a certain factor bound to the M box or its overlapping sequence, leading to pigment cell-specific expression of the TRP-1 gene. In this context, only 133-bp 5'-flanking region of the mouse TRP-1 gene is sufficient for pigment cell-specific promoter activity, and its M box may be responsible for the transactivation by MITF (34). A repressor sequence, MSE, has been reported in the mouse TRP-1 gene (40), but the sequence homologous to MSE is not present within the 839-bp 5'-flanking region of the human TRP-1 gene.

The pigment cell-specific transcription of the TRP-2 promoter is directed by the two separate regulatory elements, the 32-bp element and the proximal region containing an M box (29). Here we show that the M box is essential for efficient expression of the TRP-2 promoter in pigment cells (Fig. 6), but this M box by itself is not sufficient for the transactivation by MITF. The presence of the 32-bp element containing a CAATTG motif in addition to the TRP-2 M box caused no significant transactivation by MITF (Fig. 6), which is consistent with the results that the 32-bp element was not bound by MITF (Fig. 5). These results suggest that MITF is not indispensable for the melanocyte-specific transcription of the TRP-2 gene, and thus the mechanism for melanocyte-specific transcription of the TRP-2 gene is different from that of the tyrosinase gene. Such a difference may reflect a character of TRP-2, whose expression is detected in melanoblasts of developing mouse embryo and precedes the expression of the tyrosinase and TRP-1 genes (41).

We also identified an MITF isoform lacking the region encoded by exon 3, whose function was, however, indistinguishable from that of MITF. Thus, the deleted region of 28 amino acid residues exerted no noticeable effects on the function of MITF. On the other hand, MITFDelta 2, which lacks the carboxyl-terminal 125 residues of MITF, showed remarkable functional differences compared with MITF. As expected from the functional analysis of TFE3 (23), the activation levels of the tyrosinase promoter by MITFDelta 2 were noticeably reduced, suggesting that MITF may contain an activation domain in its carboxyl-terminal region. On the other hand, MITFDelta 2 unexpectedly transactivated the TRP-2 gene promoter,but not the heme oxygenase-1 and SV40 early promoters. Aberrant activation of the TRP-2 promoter by MITFDelta 2 suggests that the carboxyl-terminal portion of MITF may be involved in the specific MITF-protein interactions, which in turn define the binding sequence in the tyrosinase family genes. It should be noted that the heme oxygenase-1 gene promoter, carried by pHHOSVL4, contains the CATATG and CACGTG motifs, each of which is bound by USF and other factors (37, 38). Furthermore, expression of pHDTL12M containing the mutated M box was not increased by MITFDelta 2 (data not shown). These results suggest that MITFDelta 2 still retains the ability to recognize the CATGTG motif. The carboxyl terminus deleted in MITFDelta 2 is encoded by the 5' portion of exon 9, which was shown to represent the 3' end of the MITF gene (24). Thus, it remains to be investigated whether a naturally occurring MITF isoform, functionally equivalent to MITFDelta 2, is present in pigment cells. Alternatively, it is a hitherto unknown factor rather than an MITF isoform that may function as a transcription factor for the TRP-2 gene, which seems to be consistent in part with the unique expression patterns of the TRP-2 gene among the tyrosinase family genes (41).

In addition to transcription factors, such as TFE3, TFEB, and TFEC (22), the retinoblastoma protein was recently shown to interact with Mi protein in vitro (42). The retinoblastoma protein has been characterized as a regulator for gene transcription or cell cycles (43). Thus, MITF may play an important role in cell differentiation and maintenance of not only melanocytes but also certain cell types of other lineage, as evident from the phenotype of mi mice. These observations together with a function of MITFDelta 2 are of particular interest in view of the pathogenesis of a dominantly inherited Waardenburg syndrome type 2. It is conceivable that certain mutations in or near the MITF's carboxyl-terminal portion may lead to produce a mutant MITF with an aberrant function.

The regulation of the tyrosinase family gene expression will provide a model system to study a regulatory network of multiple transcription factors involved in the cell-specific gene transcription. The present study will facilitate the research on searching for the factors involved in the regulation of pigment cell-specific transcription of each human tyrosinase family gene as well as on the pathogenesis of Waardenburg syndrome type 2.


FOOTNOTES

*   This work was supported in part by Grants-in-Aid for Encouragement of Young Scientists (to K.-i. Y.), for Developmental Scientific Research (to S. S.), and for Scientific Research (to S. S. and Y. T.) from the Ministry of Education, Science, and Culture of Japan. The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) D10243[GenBank].


   To whom correspondence should be addressed. Tel.: 81-22-717-8117; Fax: 81-22-717-8118.
1    The abbreviations used are: DOPA, 3,4-dihydroxyphenylalanine; TRP-1, tyrosinase-related protein-1; TRP-2, tyrosinase-related protein-2; bHLH, basic helix loop helix; MITF, microphthalmia-associated transcription factor; TDE, tyrosinase distal element; TPE, tyrosinase proximal element; GST, glutathione S-transferase; bp, base pair(s).

Acknowledgment

We thank S. Subramani for providing pSV2/L.


REFERENCES

  1. Hearing, V. J., and Tsukamoto, K. (1991) FASEB J. 5, 2902-2909 [Abstract/Free Full Text]
  2. Jackson, I. J. (1988) Proc. Natl. Acad. Sci. U. S. A. 85, 4392-4396 [Abstract]
  3. Shibahara, S., Tomita, Y., Sakakura, T., Nager, C., Chaudhuri, B., and Müller, R. M. (1986) Nucleic Acids Res. 14, 2413-2427 [Abstract]
  4. Jackson, I. J., Chambers, D. M., Tsukamoto, K., Copeland, N. G., Gilbert, D. J., Jenkins, N. A., and Hearing, V. J. (1992) EMBO J. 11, 527-535 [Abstract]
  5. Tsukamoto, K., Jackson, I. J., Urabe, K., Montague, P. M., and Hearing, V. J. (1992) EMBO J. 11, 519-526 [Abstract]
  6. Kwon, B. S., Haq, A. K., Pomerantz, S. H., and Halaban, R. (1987) Proc. Natl. Acad. Sci. U. S. A. 84, 7473-7477 [Abstract]
  7. Shibahara, S., Tomita, Y., Tagami, H., Müller, R. M., and Cohen, T. (1988) Tohoku J. Exp. Med. 156, 403-414 [Medline] [Order article via Infotrieve]
  8. Takeda, A., Tomita, Y., Okinaga, S., Tagami, H., and Shibahara, S. (1989) Biochem. Biophys. Res. Commun. 162, 984-990 [Medline] [Order article via Infotrieve]
  9. Cohen, T., Müller, R. M., Tomita, Y., and Shibahara, S. (1990) Nucleic Acids Res. 18, 2807-2808 [Medline] [Order article via Infotrieve]
  10. Vijayasaradhi, S., Bouchard, B., and Houghton, A. N. (1990) J. Exp. Med. 171, 1375-1380 [Abstract]
  11. Bouchard, B., Del Marmol, V., Jackson, I. J., Cherif, D., and Dubertret, L. (1994) Eur. J. Biochem. 219, 127-134 [Abstract]
  12. Yokoyama, K., Suzuki, H., Yasumoto, K., Tomita, Y., and Shibahara, S. (1994) Biochim. Biophys. Acta 1217, 317-321 [Medline] [Order article via Infotrieve]
  13. Murre, C., McCaw, P. S., and Baltimore, D. (1989) Cell 56, 777-783 [Medline] [Order article via Infotrieve]
  14. Murre, C., McCaw, P. S., Vaessin, H., Caudy, M., Jan, L. Y., Jan, Y. N., Cabrera, C. V., Buskin, J. N., Hauschka, S. D., Lassar, A. B., Weintraub, H., and Baltimore, D. (1989) Cell 58, 537-544 [Medline] [Order article via Infotrieve]
  15. Kadesch, T. (1993) Cell Growth & Differ. 4, 49-55 [Medline] [Order article via Infotrieve]
  16. Hodgkinson, C. A., Moore, K. J., Nakayama, A., Steingrímsson, E., Copeland, N. G., Jenkins, N. A., and Arnheiter, H. (1993) Cell 74, 395-404 [Medline] [Order article via Infotrieve]
  17. Hughes, M. J., Lingrel, J. B, Krakowsky, J. M., and Anderson, K. P. (1993) J. Biol. Chem. 268, 20687-20690 [Abstract/Free Full Text]
  18. Tachibana, M., Perez-Jurado, L. A., Nakayama, A., Hodgkinson, C. A., Li, X., Schneider, M., Miki, T., Fex, J., Francke, U., and Arnheiter, H. (1994) Hum. Mol. Genet. 3, 553-557 [Abstract]
  19. Beckmann, H., Su, L.-K., and Kadesch, T. (1990) Genes & Dev. 4, 167-179 [Abstract]
  20. Fisher, D. E., Carr, C. S., Parent, L. A., and Sharp, P. A. (1991) Genes & Dev. 5, 2342-2352 [Abstract]
  21. Zhao, G.-Q., Zhao, Q., Zhou, X., Mattei, M.-G., and De Crombrugghe, B. (1993) Mol. Cell. Biol. 13, 4505-4512 [Abstract]
  22. Hemesath, T. J., Steingrímsson, E., McGill, G., Hansen, M. J., Vaught, J., Hodgkinson, C. A., Arnheiter, H., Copeland, N. G., Jenkins, N. A., and Fisher, D. E. (1994) Genes & Dev. 8, 2770-2780 [Abstract]
  23. Artandi, S. E., Merrell, K., Avitahl, N., Wong, K.-K., and Calame, K. (1995) Nucleic Acids Res. 23, 3865-3871 [Abstract]
  24. Tassabehji, M., Newton, V. E., and Read, A. P. (1994) Nat. Genet. 8, 251-255 [Medline] [Order article via Infotrieve]
  25. Steingrímsson, E., Moore, K. J., Lamoreux, M. L., Ferré-D'Amaré, A. R., Burley, S. K., Sanders Zimring, D. C., Skow, L. C., Hodgkinson, C. A., Arnheiter, H., Copeland, N. G., and Jenkins, N. A. (1994) Nat. Genet. 8, 256-263 [Medline] [Order article via Infotrieve]
  26. Fuse, N., Yasumoto, K., Suzuki, H., Takahashi, K., and Shibahara, S. (1996) Biochem. Biophys. Res. Commun. 219, 702-707 [CrossRef][Medline] [Order article via Infotrieve]
  27. Yasumoto, K., Yokoyama, K., Shibata, K., Tomita, Y., and Shibahara, S. (1994) Mol. Cell. Biol. 14, 8058-8070 [Abstract]
  28. Shibata, K., Muraosa, Y., Tomita, Y., Tagami, H., and Shibahara, S. (1992) J. Biol. Chem. 267, 20584-20588 [Abstract/Free Full Text]
  29. Yokoyama, K., Yasumoto, K., Suzuki, H., and Shibahara, S. (1994) J. Biol. Chem. 269, 27080-27087 [Abstract/Free Full Text]
  30. Shibata, K., Takeda, K., Tomita, Y., Tagami, H., and Shibahara, S. (1992) Biochem. Biophys. Res. Commun. 184, 568-575 [Medline] [Order article via Infotrieve]
  31. de Wet, J. R., Wood, K. V., DeLuca, M., Helinski, D. R., and Subramani, S. (1987) Mol. Cell. Biol. 7, 725-737 [Medline] [Order article via Infotrieve]
  32. Shibahara, S., Sato, M., Muller, R. M., and Yoshida, T. (1989) Eur. J. Biochem. 179, 557-563 [Abstract]
  33. Deng, W. P., and Nickoloff, J. A. (1992) Anal. Biochem. 200, 81-88 [Medline] [Order article via Infotrieve]
  34. Yasumoto, K., Mahalingam, H., Suzuki, H., Yoshizawa, M., Yokoyama, K., and Shibahara, S. (1995) J. Biochem. (Tokyo) 118, 874-881 [Abstract]
  35. Hasumi, K., Sakamoto, G., Sugano, H., Kasuga, T., and Masubuchi, K. (1978) Cancer (Phila.) 42, 2675-2686 [Medline] [Order article via Infotrieve]
  36. Inoue, S., Ito, S., Imai, Y., Kasuga, T., and Fujita, K. (1987) Biochem. Pharmacol. 36, 3537-3540 [Medline] [Order article via Infotrieve]
  37. Sato, M., Ishizawa, S., Yoshida, T., and Shibahara, S. (1990) Eur. J. Biochem. 188, 231-237 [Abstract]
  38. Muraosa, Y., and Shibahara, S. (1993) Mol. Cell. Biol. 13, 7881-7891 [Abstract]
  39. Bentley, N. J., Eisen, T., and Goding, C. R. (1994) Mol. Cell. Biol. 14, 7996-8006 [Abstract]
  40. Yavuzer, U., and Goding, C. R. (1994) Mol. Cell. Biol. 14, 3494-3503 [Abstract]
  41. Steel, K. P., Davidson, D. R., and Jackson, I. J. (1992) Development 115, 1111-1119 [Abstract/Free Full Text]
  42. Yavuzer, U., Keenan, E., Lowings, P., Vachtenheim, J., Currie, G., and Goding, C. R. (1995) Oncogene 10, 123-134 [Medline] [Order article via Infotrieve]
  43. Kouzarides, T. (1995) Semin. Cancer Biol. 6, 91-98 [CrossRef][Medline] [Order article via Infotrieve]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.