C.R.C.-B.A.C.R.-A.I.C.R. International Workshop

Melanogenesis : its chemistry as a therapeutic strategy in melanoma

E.J. Land

CR.C. Department of Biophysical Chemistry, Paterson Institute
for Cancer Research, Christie Hospital and Holt Radium
Institute, Manchester M20 9BX, U.K.

This Workshop brought together 60 scientists from 13 countries investigating mechanisms of melanogenesis and rational approaches towards using the melanogenic pathway as a means of targeting the cytotoxic therapy of malignant melanoma. The organisers were : Prof. D.G. Harnden and Dr E.J. Land (Paterson Institute for Cancer Research, Manchester), Prof. P.A. Riley (University College and Middlesex School of Medicine, London), Dr N. Thatcher (Christie Hospital, Manchester) and Prof T.G. Truscott (University of Keele).
The unique biochemical characteristic of melanocytes is the propensity to produce melanin. The meeting was convened specifically to examine recent advances in knowledge of melanogenesis and the possibility that there might be some way of exploiting the melanin-forming property of malignant melanocytes as a means treating melanoma. An improved understanding of the chemistry of melanogenesis might lead to the ability to manipulate such chemistry which could lead to a treatment of melanoma based on subverting the chemistry of melanogenesis.
The problem of melanoma was put in perspective by Prof. Rona MacKie (Glasgow, U.K.) who discussed the epidemiology and pathogenesls of melanoma and emphasized the importance of early detection, with the prognosis worsening rapidly for increasingly thick lesions due to disseminated disease. Epidemiological and case-controlled studies strongly implicate ultra-violet radiation, in particular its B component (280-315 nm), as a major aetiological factor in cutaneous melanoma. Those at greatest risk appear to be white-skinned subjects who have an indoor occupation but who indulge in intense sun exposure for short periods of time, e.g. on vacation. Additional risk factors include fair complexion and large numbers of benign melanocytic naevi. The unsatisfactory nature of current therapy for disseminated melanoma was reviewed by Dr N. Thatcher (Manchester, U.K.) who gave an overview of the metastatic behaviour of melanoma and the relatively poor results of treatment with a range of cytotoxic agents, immunotherapy, and adjuvant treatments. Although in most clinical series there were a proportion of responses, the efficacy of current treatments for metastasizing melanoma is currently highly unsatisfactory and new strategies are urgently needed.
The classical Raper-Mason scheme for melanogenesis postulates a pathway involving, successively, tyrosine, dopa, dopaquinone, dopachrome, dihydroxyindoles, indolequinones... leading to eumelanin, or phaeomelanin with the added involvement of cysteine. Enzymes are essential to some reactions in this process. Dr F. Solano (Murcia, Spain) described how the regulation of mammalian eumelanogenesis is mainly carried out not only by the well-known enzyme tyrosinase, but also by the more recently discovered enzyme dopachrome tautomerase. This enzyme is able to catalyze dopachrome tautomerisation into 5,6-dihydroxyindole-2-carboxylic acid (DHICA), thus preventing dopachrome decarboxylation taking place in the spontaneous rearrangement of dopachrome into 5,6-dihydroxyindole (DHI) at neutral pH. As o-diphenols, both DHI and DHICA are possible substrates of tyrosinase leading to the corresponding o-quinones. The relative concentrations of these dihydroxyindoles. and hence their o-quinones and the composition of the resulting eumelanin polymer, depends crucially upon the activity of both enzymes. This controlled polymerisation may be a natural mechanism to protect melanocytes against the known cytotoxicity of the decarboxylated indoles, which are more reactive than their 2-carboxylated counterparts. The possibility that peroxidase could also be involved in later stages of the biosynthesis of eumelanins was discussed by Prof G. Prota (Naples, Italy). Although peroxidase cannot convert the monophenol tyrosine to the diphenol, dopa, and hence lead to melanin, evidence was presented showing that peroxidase is much more effective than tyrosinase in catalysing the oxidative conversion of DHI and DHICA to melanin pigments.
The stimulation of tyrosinase in melanoma cells by adrenoceptor agonists and by catecholic compounds was described by Prof H. Rorsman (Lund, Sweden), the aim being to try to develop ways of influencing the metabolism of catechols. An increase in tyrosinase activity could thus lead to the enhanced production of cytotoxic quinones. The stimulating and cytotoxic effects of isoprenaline, theophylline, terbutaline and DOPAC on IGR1 melanoma cells was discussed in terms of their abilities to penetrate the cells, modify cell proliferation and generate active oxygen species, including H2O2. Active oxygen species appear to increase tyrosinase activity in IGR1 cells. Moreover, both the activation of tyrosinase and the cytotoxic effects of catechols were found to be eliminated by catalase.
The melanocytes of normal humans and melanoma patients also contain the enzyme calechol-O-methyl transferase (COMT). This enzyme methylates the indolic melanin precursors DHI and DHICA. Dr S. Pavel (Amsterdam, The Netherlands) described the detection of such indoles in media from human melanoma cell cultures.
Most of these possess a methoxy group in positions 5 or 6 which prevents their oxidation to simple o-quinones. The methylation, which can be considered as a protective mechanism in melanocytes against intrinsically generated toxic o-hydroxy, products, is accomplished by intracellularly localised COMT.
Several contributions dealt with detailed aspects of the chemistry of early stages, melanogenesis and related processes. An overview of quinone reactivity with respect to melanin formation was provided by Dr J.M. Bruce (Manchester, U.K.). Although much of the chemistry of the melanogenesis pathway up to the dihydroxyindoles DHI and DHICA is fairly well understood, the subsequent stages, even those leading to species with comparatively low molecular weight, are not. The problem is compounded by the complexity of the oxidation of DHI (and DHICA), which can, in principle, lead not only to the corresponding ortho-quinone, but also to both a quinone-imine and a quinone-methide, all three of which may be in tautomeric equilibrium. Home- and hetero-coupling reactions involving one or more of these tautomers, leading in the first place to a multiplicity of isomeric dimers, could be important in the polymerisation ultimately resulting in melanin.
Quinones and quinone-methides are also involved in the molecular mechanisms for cuticular sclerotization as described by Prof M. Sugumaran (Boston, USA). The exoskeleton of insects and other arthropods are hardened to protect their soft bodies by a process called sclerotization. During hardening, soluble structural proteins and chitin fibres are rendered insoluble by reaction with reactive species derived from enzymatic activation of catecholamines, such as N-acetyldopamine and
N-Ò -alanyldopamine. Based on the reactive species formed, two different mechanisms have been identified to acount for sclerotization reactions. They are quinone tanning and quinone-methide sclerotization. Although initially these two mechanisms were considered to be dependent of each other, the recent discovery of two new enzymes, quinone isomerase and quinone methide isomerase, led to a unification of these mechanisms.
The initial steps involved in sclerotization closely resemble the initial reactions observed during melanization. Both processes involve the initial enzymic oxidation of catechols to quinones, and quinone tautomerisation to cluinone methides. The introduction of a double bond into the side-chain and the oxidation of the side-chain desaturated catecholamine parallels the aromatisation of dopachrome to dihydroxyindole, and the oxidation of dihydroxyindole to the corresponding quinone.
Details of the chemistry of the transition : dopachrome --> dihydroxyindole were described by Prof H. Wyler (Lausanne, Switzerland). Four protonated forms of dopachrome were identified, the 6-OH group deprotonating with a pK of 0.8, the carboxy group with a pK of 3.1 and the nitrogen with a pK of 9.1. At neutral pH, the only product of dopachrome decay is dihyroxyindole, whereas at high pH (> 10) it is  exclusively the anion of dihydroxyindole-2-carboxylic acid.
The maturity of the technique of pulse racliolysis is such that one can now be confident enough to tackle problems as complex as the process of melanogenesis. Although one-electron oxidatinn of DHI and DHICA is not thought to be a part of the melanogenic pathway, the rapid disproportionation of such radicals is an excellent means of preparing high concentrations of the subsequent metastable intermediates, e.g. indolequinones, which are thought to be important components of the melanogenic pathway. Dr P. O'Neill (Chiltonl U.K.) described the one-electron oxidation of a series of hydroxy- and methoxy-indoles using pulse radiolysis. One-electron oxidation of dihydroxyindole in the pH range 5-10 yields the corresponding oxygen-centred indole semiquinone radical (pK 6.8). With hydroxylated mono-methoxyindoles, the corresponding methoxyindoloxyl radical is formed. Further methylation of the hydroxy substituents, as in dimethoxyindole, results in the stabilisation of the corresponding cation which, depending on the pH, deprotonates at N(1) to yield the nitrogen-centred indolyl radical, with a pK of 6.0.
With the exception of dihydroxyindole, the radicals were all found to decay bimolecularly to yield semi-permanent products which eventually decayed  unimolecularly. In the case of dihydroxyindole, using very low pulse doses the radical was found to decay unimolecularly. The latter decay was assigned to a reaction of the radical with the parent dihydroxyindole, rate constant ~ 106 M-1 s-1. This could be a component of the route for melanin polymerisation.
Prof T.G. Truscott (Keele, U.K.) described a closely similar study of the species resulting from pulse radiolytic one-electron oxidation of dihydroxvindoles and their methoxylated metabolites. Whereas there was good agreement with the results reported by the previous speaker on the initially formed one-electron oxidised radicals, differences were apparent in the nature of the reactions of some of the radicals and subsequent metastable intermediates. A possible scheme for polymerisation to melanin was proposed involving a reaction of a quinone-methide with water to produce a trihydroxyindole, which may itself then react successively with the methide to produce dimers, then trimers, etc.
Prof B. Kalyanaraman (Milwaukee, USA) described the application of the technique of electron spin resonance spectroscopy to the identification and characterisation of free radicals derived from melanin precursors. Complexed with diamagnetic ions, especially Mg²+ and ZnL²+, catechol(amine) semiquinones live 104 times as long as in the absence of metal, allowing their protonation-deprotonation reactions to be readily studied. The addition reactions of quinones with nucleophiles e.g. amino acids (proline and methionine), peptides and proteins were also studied via the radicals derived from such adducts.
It was the discovery some 20 years ago that tyrosinase is capable of oxidising substrates that are structural analogues of tyrosine which suggested the possibility of using the melanogenic pathway as a targeting strategy for melanoma chemotherapy. Prof P.A. Riley (London, U.K.) described a series of investigations on the mechanism of action of substituted phenols of which the lead compound is the well-known depigmenting agent 4-hydroxyanisole (4HA) which is oxidised to the corresponding orthoquinone. The mechanism of the cytotoxlcity of the orthoquinone probably depends predominantly on the formation of adducts with important cellular thiol-containing proteins. The possibility that the action depends on the generation of active oxygen species by redox cycling has been excluded and there is little evidence that semiquinones are involved in the toxic process (see below). 4HA also exhibits direct toxic actions, including inhibition of ribonucleotide reductase, inhibition of mitochondrial electron transport, and other effects on cell physiology. Separation of these direct actions from the tyrosinase-dependent cytotoxicity would serve to amplify the therapeutic index against melanogenic cells, and the results of tests on a series of derivatives of 4HA including a range of oxy-ethers and thio-ethers of differing chain length were presented. In discussion, Dr K. Schwabe (Berlin, Germany) reported that his group had synthesized over 100 analogues of 4HA and shown that the propyl oxy-ethers were therapeutically effective against tumour volume in melanoma-bearing animals, but had shown no effect on survival of the animals employed in these tests.
The results of studies on the mechanism of cytotoxicity and analogues were described by Dr E.J. Land (Manchester, U.K.). Using pulse radiolysis, semiquinones were generated from the corresponding hydroquinones chemically synthesized by Mr C.J. Cooksey. The semiquinones neither reacted with oxygen nor with trans-butanoic acid, a water-soluble model for unsaturated fatty acids. Consequently, it appears unlikely that redox cycling or the initiation of lipid peroxidation via semiquinones comprise the cytotoxic mechanism. The 3,4-quinones which form rapidly by disproportionation of the corresponding semiquinone were found to react rapidly with several thiols and with ascorbic acid. Nucleophilic addition of protein thiols to the quinone is thus a more probable mediator of cytotoxicity. 4-(n-propoxy)phenol, which possesses five times the in vitro tyrosinase-dependent cytotoxicity towards rat cells compared with 4HA, was found to behave almost identically as far as the reactivity of semiquinones and quinones was concerned.
A closely similar approach to targeted cytotoxicity was described by Prof K. Jimbow (Edmonton, Canada) who showed that 4-S-cysteaminylphenol and some analogues are selectively incorporated into murine melanoma tissue and into actively melanising hair follicles with selective destruction of melanocytes resulting in depigmentation of black hair follicles in mice, N-acetyl-4-S-cysteaminylphenol being the most potent inducer of selective destruction of follicular melanocytes. This compound may prove to be a useful anti-tumour agent judged by the data on subcutaneously inoculated B16 F10 melanoma cells in mice.
Dr P.G. Parsons (Queensland, Australia) reported that despite many promising in vitro demonstrations of antimelanoma activity by redox active agents such as catechols, and by drugs such as buthionine sulphoximine which depress cellular defences against oxygen radicals, there remain many problems regarding the successful application of this approach to melanomas in vivo. It may be possible to overcome the lack of potency and selectivity of such agents by using combinations of drugs. Such combination therapy, together with measures for minimising mutations rates, may also combat phenotypic instability leading to the development of drug resistance.
Dr B.S. Larsson (Uppsala, Sweden) described an approach towards the therapy of melanoma using aminothiol compounds which are selectively incorporated into newly-synthesized melanin by reacting with orthoquinones generated during melanogenesis. Various compounds of this class, including thiourea and 2-thiouracil, have been shown by autoradiography to be specifically taken up into melanising tissue including murine melanoma. Radioiodinated 5-iodo-2-uracil has been used in patients for melanoma screening and pilot studies on treatment with 35S-thioracil have been carried out on melanoma-bearing mice. However, the radiation doses required for substantial effects using this radionuclide are extremely high, making therapeutic clinical application hazardous. As an alternative, boron neutron capture therapy has been attempted, using boronated thioureas which are selectively incorporated into newly-synthesized melanin and the 10B can then be activated to undergo nuclear fission by irradiatin of tumours with thermal neutrons from an external source.
Dr A.J. Winder (Oxford, U.K.) reported on the effects of L-tyrosine phosphate and cytochalasin D on induction of pigmentation in cells, with a view to the possibility of developing differenciation therapy for melanoma. Both agents produced a small increase in the amount of tyrosinase messenger RNA, although the amount observed was insufficient to account for the increase in enzyme activity.
The relationship between abnormal melanosome structure and cytotoxic phenomena was discussed by Dr J. Borovansky (Prague, Czechoslovakia). Electron microscope investigations have confirmed the presence of abnormal and incomplete melanosomes in human melanomas from epidermal and mucosal sites, in melanoma metastases and in the B16 mouse melanoma. For example, 90% of melanosomes in cutaneous melanomas had membrane defects. Evidence that this leads to significant leakage of reactive melanin precursors was furnished by raised free radical-mediated lipid peroxidation in the liver of B16 melanoma-bearing mice.
Prof T. Sarna (Krakow, Poland) discussed thephotoactivation of melanin and the possible generation of cytotoxic products resulting from it. Direct photo-oxidation of melanin was accompanied by oxygen consumption and the production of hydrogen peroxide. This process was shown to be strongly wavelength- and pH-dependent and was also influenced by the presence of metal ions. Photo-oxidation of melanin in the presence of photosensitizers was also reported, and the possibility of covalent binding of photosensitizers to melanin as an approach to photodynamic therapy was discussed. Dr A.R. Young (London, U.K.) discussed the apparently paradoxical photoprotection of skin to UV-induced damage by prior exposure to photosensitizers such as psoralens. Tans induced by 5-methoxypsoralen (5-MOP) contained in a UVB sun-screen with solar stimulating radiation (SsR) were found to reduce unscheduled DNA svnthesis following minimal erythema doses of SSR. Despite the fact that 5-MOP is weak photocarcinogen, judicious use of 5-MOP-containing sun-screens may offer benefits with regard to protection against various forms of light-induced skin cancer which could outweigh the risk involved. This may be of importance in protecting future generations from the risk of developing malignant melanoma.
The Workshop generated much detailed discussion which it is hoped will be reflected in the stimulation of new initiatives towards the introduction of a successful treatment for disseminated melanoma. Clearly one potentially powerful strategy is targeted cytotoxicity utilising the melanogenic pathway to activate an otherwise innocuous pro-drug.

Acknowledgements
As well as the Cancer Research Campaign, the British Association for Cancer
Research and the Association for International Cancer Research, the following are thanked for providing financial support : Amgen Ltd., Asta Pharma AG, Beckman
Instruments Inc., Boehringer Ingelheim, Bristol-Myers Pharmaceuticals, Ciba Geigy Pharmaceuticals, Farillon Ltd., Institut de Recherches Internationales Servier, Laboratoires Pharmaceutiques Bergaderm SA, Lilly Industries Ltd., I'Oréal, Sterling Drug Inc., US Air Force Office of Aerospace Research and Development, and Windsor Phamaceuticals Ltd.