Molecular biologv of melanogenic proteins
Vincent J. Hearing
Laboratory of Cell Biology, National Cancer Institute
National Institutes of Health, Bethesda, MD 20892 USA
The 14th meeting of the IPCC held recently in Kobe, Japan (chaired by Dr Y Mishima) gave heavy emphasis to topics concerning the molecular biology of pigment cells; this was reinforced by a Molecular Biology Symposium held immediately thereafter in Sendai (chaired by Dr T Takeuchi). It is the purpose of this review to provide an overview of research in this field, and perhaps supply some insight as to where such studies might be heading. The field of molecular biology encompasses many different disciplines and approaches, but at this time primarily revolves in pigment cell research around two broad classes of genes, i.e. oncogenes and pigment-related genes. Studies on both have examined the expression, structure, and function of these genes, the effects of various mutations on the properties of their encoded proteins, and their effects on cells following transfection. In this review, I will confine my summaries and comments to the pigment-related genes.
As we all know, tyrosinase represents a critical regulatory point in the pathway of melanin formation, and is the key lesion in many types of albinism. Because of its prominent role in the modulation of visible pigmentation, and its potential application as a genetic marker, there has been tremendous interest in the cloning of its gene. The genes for tyrosinases from lower species [1-3] had been available for some time, and have served as useful tools for structural and functional studies; however, since primary sequence data for mammalian tyrosinase was not available, the direct approach of synthesizing probes by which the mammalian gene could be identified and isolated could not be used. Various laboratories used different alternative and indirect approaches for cloning the tyrosinase gene, and several candidate clones were obtained from murine and human systems [4-8]. Interestingly and unexpectedly, several putative but distinct clones were identified which shared significant sequence homology. More importantly, each of those clones encoded proteins that had many properties predicted for tyrosinase, including melanocyte specificity, the correct size (i.e. ~ 60-70 kD), a transmembrane region, two potential copper binding domains, multiple glycosylation sites, and so on [ reviewed in 9]. Much of the research currently being pursued revolves around identifying mutations which affect the functions of those gene products, and how those encoded proteins normally interact to regulate melanin production.
The Albino Locus Gene
Structure: The albino locus has historically been proposed as the structural locus for tyrosinase. This interpretation results from the dramatic lack of pigment produced by mutations at that locus, but is not consistent with studies demonstrating significant tyrosinase activity in several of those albino mutants [cf 9 for review]. cDNAs which mapped to the albino locus were originally cloned from murine  and human  melanocytes; the original sequence reported for the murine gene was a prematurely truncated version of the full length transcript - the full sequence has been subsequently published [10-12]. The albino gene in mice (and its human homologue) is composed of 5 exons and 4 introns, and is predominantly untranslated material. In mice the gene is ~ 70 kb  and in humans it is >35 kb , while the processed mRNAs are ~ 2.4 kb in length; thus the precursor mRNA synthesized must be processed to delete the introns. This process does not always occur accurately, and misspliced mRNAs (10 to 40% of the total) are generated which translate into altered proteins [13-15]. This may represent an important regulatory step since those altered proteins are apparently not competent catalytically as tyrosinase, although whether they fulfill some other role in the melanocyte has yet to be determined. Although the albino gene is present as a single copy (chromosome 7 in mice, chromosome 11 in humans), it has been shown by in situ hybridization  that there is a second site (on human chromosome 11) that is partially homologous. This second site has now been shown to be a pseudogene  composed of exons 4 and 5 of the authentic sequence. The fact that the protein encoded by the albino locus is expressed in melanocytes as a single chain with the amino and carboxy termini predicted by the nucleic acid sequence has been confirmed using antibodies generated against synthetic peptides which correspond to those sequences .
Function: The ability of the protein encoded by the albino gene to function as tyrosinase has now been shown by several approaches. Initially, transfection of the gene into tyrosinase deficient cells resulted in the expression of tyrosinase activity [10,19]; not only was it shown that the full length and correctly spliced mRNA encoded active tyrosinase, but that several aberrantly spliced mRNAs encoded proteins that were catalytically incompetent . In subsequent studies, minigenes (which contain the tyrosinase coding sequence as well as a competent regulatory sequence) were used to produce transgenic mice [20-22]. Although the transgenic animals were not fully and normally pigmented, there were dramatic increases in pigmentation and it was shown that the pigmentation was restricted to melanocytes. Since the regulatory sequence used to construct the minigene was derived from the authentic 5' flanking region of the tyrosinase gene, that regulatory sequence is thought to be critical and sufficient for the tissue specific expression of tyrosinase. The ability of the aIbino-locus encoded protein to function catalytically as tyrosinase has also been shown by immunopurification protocols [18,23].
At this time, I believe that all laboratories would agree that the albino locus encoded protein is tyrosinase, and that its function is crucial to the production of pigment; whether it is the sole protein which can perform as a tyrosinase, and exactly how it interacts with other melanogenic proteins is still under discussion. It is also interesting to note that while expression of albino locus encoded tyrosinase seems to be necessary for pigment production, it is not the only determininant, since many melanocyte lines with normal expression of the tyrosinase mRNA and/or protein are unpigmented [10,19,24]. Similarly, while melanogenic agents which stimulate pigmentation (such as Ñ MSH) typically increase the transcription and translation of the albino gene 2 to 3-fold [11,25], such stimulations are not sufficient to explain the much greater increases in tvrosinase activities noted, suggesting that other control points in the pathway are also critical to the regulation of melanogenesis.
Mutations: The characterization of mutations, be they natural or mutagen induced, on the structure and function of proteins encoded by the mutated gene can provide important insights into the mechanisms of biologic processes. This is most obvious with mutations at the albino locus, which lead to dramatic changes in pigmentation; initial studies in this field depended on animal models, but recently, melanocyte cell lines derived from coat color mutants have become available for study [26,27]. Studies with these cells have demonstrated that heterokaryons of brown and albino mutant cell lines produce normal black pigment, showing that these genes are complementary in culture, thus opening the door for future studies on their interactions in vitrb . Other studies on mutations at the albino locus (including Himalayan and chinchilla) have shown that such mutations result in the production of tyrosinases with significantly altered properties. With the albino mutation, catalytic activity is virtually quantitatively lost, while with the Himalayan mutation, there is an alteration in glycosylation which results in a temperature sensitive phenotype; with the chinchilla mutation there is an increased sensitivity to proteolytic inactivation, and thus a decrease in enzyme function . The mutations that elicit these changes are now being detailed in mice [20,28-30] and should he invaluable in the future for understanding the functional properties of the enzyme. Perhaps the most dramatic recent revelation has been the characterization of the exact amino acid change which is responsible for the albino phenotype; it is a point mutation which results in the change of a conserved cysteine to a serine in the first cysteine-rich domain of tyrosinase . This single change in primary structure causes the virtually quantitative loss of catalytic function, although the exact reason (i.e. alteration in secondary structure, loss of metal binding activity, etc.) has not yet been determined. However, in light of the virtually identical nature of the point mutation causing the brown phenotype (cf below), there can be no doubt that future studies will be intensely directed at this domain of tyrosinase which appears to be so critical to its activity. All strains of albino mice examined had this same mutation suggesting that all were derived from the same initial mutation . Naturally there has been tremendous interest in defining mutations which are critical to tyrosinase activity in humans. To date, many different mutations in oculocutaneous albinos have been described, all of which appear to result in loss of tyrosinase activity and thus in melanin production. Space considerations don't allow a thorough discussion of the mutations described to date, other than to say that they now number in the dozens, occur at various areas on the enzyme (i.e. they are not confined to a single region or domain), and result from a variety of mechanisms, including point mutations and insertions, which occur in the structural or the promoter region of the gene [32-36]. In both humans and mice, polymorphisms have been described (these are nondestructive point mutations which do not demonstrably alter enzyme function) as well as the mutations noted above. It is obvious from the meetings held in Japan that many groups are currently working on defining: (1) the exon-intron structures of the genes, (2) mutations which lead to various forms of coat color mutations, (3) the structure of the regulatory regions (functional promoters and enhancer-like elements), and (4) mechanisms of splicing and aberrant splicing. Of particular interest, a minigene has been created  with only 1 kb of human tyrosinase 5' flanking sequence, which was fused with murine structural locus and used to create transgenic mice. This resulted in pigmentation of the mice only in melanocytes, suggesting that there is a common or highly similar mechanism between gene regulation in mice and humans; this sequence is located in the 5' flanking region. Several groups [38,39] have described more than 12 different genetic mutations in the tyrosinase gene; these cluster in a non-random fashion, but are not localized to any distinct area. Two new mutations were described  for type 1B OCA (yellow) which are novel; the defect is either within an intervening sequence of tyrosinase gene, or distant to it. A point mutation in the human gene was also described  which results in a temperature sensitive tyrosinase which is analogous to the Himalayan mutation in mice.
TRPI: The Brown Locus Gene
Structure: The first cloned pigment related gene proposed to be the structural locus for tyrosinase  was quickly mapped to the brown locus on chromosome 4 in mice [8,42]. The structure and organization of this gene is similar to the albino gene, and the predicted protein has all of the features noted above which would be predicted for tyrosinase [cf 9 for review]. There is only a single copy of this gene in the genome, and the analogous gene has now also been found in humans and cloned [43,44]. In general there is about 90% sequence identity between the brown protein expressed by murine and human melanocytes (this is similar to the conservation found between murine and human tyrosinase), and there is about 55% nucleic acid identity overall between the brown and albino genes (and about 43% primaly sequence identity). The conservation of residues is much higher in several areas of the proteins which are thought to be important to their structure and/or function (such as the copper binding and cysteine rich domains) and is almost completely lost elsewhere, such as in the transmembrane region and cytoplasmic determinant at the carboxyl termini of the proteins. Interestingly, although the brown locus gene (~ 18 kb in mice) is broken up into 8 exons and 7 introns , no data has yet been reported which suggests that alternative processing occurs with this gene, as it does with the albino gene.
Function: The specific function of the brown locus encoded protein is not altogether clear as yet. Phenotypically, we can see that a mutation at this locus causes the production of brown rather than black melanin, but what this means chemically andior enzymatically is as yet undefined. Nevertheless, it is obvious that whatever the role of the brown protein, it must somehow elicit the production of black (vs. brown) melanin in wild-type animals. The brown protein has been postulated by various groups to be: (a) DOPAchrome tautomerase (nee DOPAchrome conversion factor or oxidoreductase) , (b) Dihydroxyindole conversion factor ; (c) a melanosomal specific catalase ; or (d) another tyrosinase [18,46]. Although my laboratory favors the latter possibility (in fact we presented evidence at these meetings that this protein acts synergistically to stimulate the albino locus encoded tyrosinase), the substrates and products of reactions catalyzed by the brown protein have not yet been identified conclusively, and the question remains open at this time. It has become evident that the brown protein is present in a higher quantity (typically ~ 10-fold) in melanocytes as compared to the albino protein; it is also expressed in similar quantities by human melanocytes [44,47]. As far as I know, a human mutation of the brown locus has not yet been identified, but surely will be before long, now that the human gene has been cloned and sequenced. It is also an indication of the important function of this protein that there is typically a better correlation of its expression with visible pigmentation than there is with the expression of the albino protein. As noted above, the brown and the albino loci complement each other .
Mutations: There are multiple mutations at the brown locus in mice, and several of them are currently under active study (including cordovan and light) [48,49]. As with the albino mutation, the critical molecular lesion resulting in the brown phenotype has been identified , and it results from a point mutation leading to the replacement of a conserved cysteine residue with a tyrosine. This critical replacement occurs in the first cysteine-rich domain of the protein, at a residue only 3 amino acids away from the critical cysteine mutation that occurs in albino mice. This could perhaps be coincidence, but in light of the recent proposal  that these cysteines may be involved in an iron binding site, the importance of this domain to the structure and function of these pigment-related proteins is sure to draw much interest in the coming years. It is an important feature to note that, as with the basic albino mutation, all brown mice from numerous different strains have the same mutation, suggesting that all brown mice derive from the same original mutation .
TRP2: The Slaty (?) Locus Gene
The clone originally identified by Jackson (termed 5A  which was initially thought to be identical to the brown locus gene, has now been identified as a distinct gene, but one which shares significant sequence homology to tyrosinase . This gene has been mapped to chromosome 14 and tentatively assigned to the slaty locus, and has common features with the genes discussed above, including a transmembrane region, highly conserved putative copper binding sites, two conserved cysteine-rich domains, potential glycosylation sites and a signal peptide. The size of the protein encoded by this gene (~ 75-80 kD) is somewhat larger than the products of the albino and brown loci (unpublished). The function of this gene is currently unknown.
Pmel 17-1: The Silver (?) Locus Gene
This clone was originally termed Pmell7-l [6j; the expression of its mRNA was specific for melanocytes, could be induced with MSH (again ~ 2-fold) or IBMX, and its abundance correlated well with pigmentation . It was known that the protein encoded by this gene was homologous to the albino locus, and it has been reported that the encoded protein has a predicted molecular weight similar to tyrosinase (~ 70 kD), and also had putative glycosylation sites and a transmembrane region . There is approximately 95% identity between the sequence of the human and the murine protein. This gene has been mapped to chromosome 12 in humans and to chromosome 10 in mice, and has been tentatively assigned to the silver locus of mice. The function of this protein is also unknown.
The molecular biology of pigmentation is becoming both more interesting and complex every minute. The plethora of genes belonging to a tyrosinase family was totally unexpected and has underscored the complexity of the regulation of mammalian melanogenesis -a process once thought to be a simple one enzyme:one product system. The next several years will see the characterization of even more pigment related genes, since clones have now been derived from the agouti, dilute and pink-eyed dilution loci, and more information should soon be forthcoming from those gene sequences. I would expect the following types of studies to provide interesting new approaches to elucidating the regulatory controls of mammalian melanogenesis in the immediate future:
A) What are the critical regulatory elements within the tyrosinase gene which control the tissue specific and timely expression of these pigment related genes? Recent studies have begun to describe the promoters and enhancer-like elements which may be critical to the regulation of the expression of these tyrosinase-related genes [52-54].
B) Does alternative splicing play a role in the regulation of tyrosinase activity? We now know that responses of melanocytes to environmental stimuli involve increased levels of transcription and translation of the albino and brown loci, but not enough to explain the dramatic increases in melanin production by those cells. This suggests either that post-tyrosinase factors are important to controlling melanin production and/or that the accuracy of processing of precursor mRNA might improve, leading to increases in functionally competent enzyme(s).
C) What are the catalytic functions of these cloned gene products and how do they interact to determine melanogenic function? Such studies are underway using many different approaches in a variety of laboratories.
D) What are the intracellular processing and delivery pathways involved in the delivery of these gene products to the melanosome? Are all of these melanogenic proteins transported to the melanosome en masse within the vesicles, or are they segregated in different vesicles and combined only following their arrival at the melanosome? The latter pathway would provide a potential mechanism whereby melanogenesis might be delayed until all melanogenic factors are in place in the melanin granule; preliminaIy evidence supporting this was presented at the meeting by Drs. Jimbow and Boissy.
E) What roles do melanogenic inhibitors and post-tyrosinase factors play? These points of regulation have gained added significance in light of what we now know about the regulation of gene expression as noted above, and the variety of post-tyrosinase 'factors' which play a role in the determination of melanin formation. The next International meeting in London (chaired by Dr. Riley) in 1993, and indeed the intervening Regional meetings, should provide us with exciting new developments in these areas in the future.
1. Bernan V, Filpula D, Herber, W, Bibb M, Katz E: Gene, 37, 101-110, 1985
2. Huber M, Hinterman G, Lerch K: Biochem, 24, 6038-6044, 1985
3. Kupper U, Niedermann DM, Travaglini G, Lerch K: J Biol Chem, 264, 17250-17258, 1989
4. Shibahara S, Tomita Y, Sakakura T, Nager C, Chaudhuri B, Muller R: Nucl Acids Res, 14, 2413-2427, 1986
5. Kwon BS, Halaban R, Kim GS, Usack L, Pomerantz SH, Haq AK: Mol Biol Med, 4, 339-355, 1987
6. Kwon BS, Haq AK, Pomerantz SH, Halaban R: Proc Natl Acad Sci USA, 84, 7473-7477, 1987
7. Yamamoto H, Takeuchi S, Kudo T, Makino K, Nakata A, Shinoda T, Takeuchi T:
Jpn J Gen, 62, 271-274, 1987
8. Jackson IJ: Proc Natl Acad Sci USA, 85, 4392-4396, 1988
9. Hearing VJ, Jiménez M: Pigment Cell Res, 2, 75-85, 1989
10. Kwon BS, Wakulchik M, Haq AK, Halaban R, Kestler D: Biochem Biophys Res
Commun, 153, 1301-1309, 1988
11. Müller G, Ruppert S, Schmid E, Schütz G: EMBO J, 7, 2723-2730, 1988
12. Terao M, Tabe L, Garattini, Sartori D, Mintz B: Biochem Biophys Res Commun, 159, 848-853, 1989
13. Ruppert S, Muller G, Kwon BS, Schiitz G: EMBO J, 7, 2715-2722, 1988
14. Shibahara S, Tomita Y, Tagami H, Muller RM, Cohen T: Tohoku J Exp Med, 156, 403-414, 1988
15. Porter S, Mintz B: Gene, (in press) Jan, 1991
16. Barton DE, Kwon BS, Francke U: Genomics, 2, 17-24, 1988
17. Takeda A, Matsunaga J, Tomita Y, Tagami H, Shibahara S: Tohoku J Exp Med,
159, 333-340, 1989
18. Jiménez M, Tsukamoto K, Hearing VJ: J Biol Chem, 266, 1147-1156, 1991
19. Bouchard B, Fuller BB, Vijayasaradhi S, Houghton AN: J Exp Med, 169, 2029-2042, 1989.
20. Yamamoto H, Takeuchi S, Kudo T, Sate C, Takeuchi T: Jpn J Gen, 64, 121-135, 1989
21. Beermann F, Ruppert S, Hummler E, Bosch FX, Müller G, Ruther U, Schütz G:
EMBO J, 9, 2819-2826, 1990
22. Tanaka S, Yamamoto H, Takeuchi S, Takeuchi T: Development, 108. 223-227, 1990
23. Halaban R, Moellmann G: Proc Natl Acad Sci USA, 87, 4809-4813, 1990
24. Tamate HB, Hirobe T, Wakamatsu K, Ito S, Shibahara S, Ishikawa K: J Exp Zool,
250, 304-311, 1989
25. Eller M, Naeyaert JM, Gordon PR, Gilchrest BA: J Invest Dermatol 94, 521, 1990
26. Halaban R, Moellman G, Tamura A, Kwon BS, Kuklinska E, Pomerantz SH, Lerner
AB: Proc Natl Acad Sci USA, 85, 7241-7245, 1988
27. Bennett DC, Cooper PJ, Dexter TJ, Devlin LM, Heasman J, Nester B:
Development, 105, 379-385, 1989
28. Kwon BS, Halaban R, Chintamaneni: Biochem Biophys Res Commun, 161,
29. Shibahara S, Okinaga S, Tomita Y, Takeda A, Yamamoto H, Sato M, Takeuchi
T: Eur J Biochem, 189. 455-461, 1990
30. Takeuchi S, Yamamoto H, Takeuchi T: Biochem Biophys Res Commun, 155,
31. Jackson IJ, Bennett DC: Proc Natl Acad Sci USA, 87, 7010-7014, 1990
32. Spritz RA, Strunk K, Oetting W, King R: Nucl Acids Res, 16, 9890, 1988
33. Tomita Y, Takeda A, Okinaga S, Tagami H, Shibahara S: Biochem Biophys Res
Commun, 164, 990-996, 1989
34. Giebel LB, Strunk KM, King RA, Hanifin JM, Spritz RA: Proc Natl Acad Sci USA,
87, 3255-3258, 1990
35. Spritz RA, Strunk KM, Giebel LB, King RA: New Eng J Med, 322, 1724-1728,
36. Takeda A, Tomita Y, Matsunaga J, Tagami H, Shibahara S: J Biol Chem, 265,
37. Tanaka M, Tanaka S, Miura H, Yamamoto H, Kikuchi H, Takeuchi T: Proc XIV
IPCC, p113, 1990
38. King RA Handoko HY, Mentink MM, Getting WS: Proc XIV IPCC, p115, 1990
39. Spritz RA: Proc XIV IPCC, p66, 1990
40. Giebel LB, Tripathi RK, Strunk KM, Spritz RA: Proc XIV IPCC, p116, 1990
41. Spritz RA, Giebel LB, Tripathi RK, Strunk KM, King RA: Proc XIV IPCC, p116,
42. Bennett DC, Huszar D, Laipis PJ, Jaenisch R, Jackson IJ: Development, 110,
43. Cohen T, Muller RM, Tomita Y, Shibahara S: Nucl Acids Res, 18, 2807-2808,
44. Vijayasaradhi S, Bouchard B, Houghton AN: J Exp Med, 171, 1375-1380, 1990
45. Shibahara 8: Prog Clin Biol Res, 256, 263-271, 1988
46. Jiménez M, Maloy WL, Hearing VJ: J Biol Chem, 264, 3397-3403, 1989
47. Vijayasaradhi S, Houghton AN: Int J Cancer, (in press), 1991
48. Jackson IJ, Chambers D, Rinchik EM, Bennett DC: Genetics, 126, 451-460, 1990
49. Zdarsky E, Favor J, Jackson IJ: Genetics, 126, 443-450, 1990
50. Jackson lJ: personal communication, 1990
51. Kwon BS, Chintameneni C: Proc XIV IPCC, p89, 1990
52. Takeuchi T: Proc XIV IPCC, p65, 1990
53. Takeuchi S, Yamamoto H, Takeuchi T: Proc XIV IPCC, p114, 1990
54. Ruppert 8, Beermann F, Kluppel M, von Langsdorff M, Hummler E, Schmid E,
Schütz G: Proc XIV IPCC, p119, 1990