Encapsulation of a reactive core in neuromelanin

Melanin pigments are found widely in animals and plants (1). Melanin's most conspicuous presence is on the surface of the body, where it serves to protect underlying tissues from harmful UV radiation and to camouflage the organism from enemies. Two chemically distinct types of melanin pigments are produced in the melanocytes of mammals and birds: the black-to-brown eumelanin and the yellow-to-reddish-brown pheomelanin. Biochemical pathways to these melanin pigments are now well clarified (1, 2). Both pigments are derived from a common precursor, dopaquinone (DQ), which is formed from the amino acid tyrosine upon oxidation with a specialized enzyme, tyrosinase. DQ is a highly reactive intermediate that, after cyclization, undergoes a complex series of redox reactions leading to the production of eumelanin, a highly heterogeneous polymer consisting of 5,6-dihydroxyindole and 5,6-dihydroxyindole-2-carboxylic acid. On the other hand, when cysteine intervenes in this pathway, it gives rise to cysteinyldopa isomers that, upon oxidation, lead to pheomelanin production via benzothiazine intermediates (see ref. 2 for the pathway of mixed melanogenesis). A brown, insoluble, melanin-like pigment also is found in the central nervous system of humans and primates (3). This pigment, termed neuromelanin (NM), is present in highest concentration in catecholaminergic neurons of the substantia nigra and locus ceruleus regions of the midbrain. In recent years, research on NM has attracted much attention because of its possible role in the pathogenesis of Parkinson's disease. In contrast to the cutaneous melanin, however, much less is known about the structure and function of NM because of difficulty with isolation and lack of adequate biosynthetic models (3). In this issue of PNAS, Bush et al. (4) have shown by employing sophisticated physical methods, in particular photoelectron emission microscopy (PEEM) coupled to a free-electron laser (FEL) (5), that NM is composed of granules with ≈30-nm diameters consisting of pheomelanin at the core and eumelanin at the surface.

Casing Model of Mixed Melanogenesis

In a previous work (6), the same group used the FEL–PEEM technique to establish that human eumelanosomes from black hair have a surface oxidation potential of −0.2 V vs. the normal hydrogen electrode. On the other hand, human melanosomes from red hair contain a mixture of eumelanin and pheomelanin, which is reflected by oxidation potentials of −0.2 and +0.5 V, respectively. Because FEL–PEEM is a surface technique, the results indicate that melanosomes from red hair have both types of melanin on or near their surface. Chemical degradation of NM shows that this pigment also is composed of both eumelanic and pheomelanic components in a ratio of 3–4 to 1 (7). But the application of FEL–PEEM to NM granules isolated from human brains (4) reveals a single oxidation potential (−0.1 V), indicating the presence of only eumelanin on the surface; no pheomelanic species are detected on or near the surface of the NM granules. Therefore, an explanation is needed for how the eumelanin surface is built up on the pheomelanin core.

Kinetic studies on the fate of DQ in the presence or absence of cysteine have clearly indicated that pheomelanin production is favored over eumelanin production as long as cysteine concentration is >1 μM (2, 8). This favored production of pheomelanin suggests that, in the process of mixed melanogenesis, pheomelanin is always formed first, and then eumelanin is deposited on the preformed pheomelanin (Fig. 1). This casing model was originally suggested by Agrup et al. (9), based on biochemical findings. Now, the work by Bush et al. (4) provides biophysical evidence supporting this model. In NM production, this casing process is further favored by the fact that dopaminequinone (DAQ) cyclizes ≈100-fold more slowly than DQ (2), leading to a preferential production of cysteinyldopamine followed by oxidation to pheomelanic pigment.

Fig. 1.

Casing model of mixed melanogenesis. In the process of mixed melanogenesis, pheomelanic pigment is produced first, followed by deposit of eumelanic pigment. In the granule with the eumelanin surface, the side is intentionally cut away to reveal the inner pheomelanin core.

Implications of Architecture of NM

The physiological and pathological roles of NM in the dopaminergic neurons are not well understood (3). Nevertheless, the present study by Bush et al. (4) makes several intriguing interpretations possible for the roles of NM. First, the neurotoxicity of dopamine may be explained by the casing model (Fig. 1). Dopamine is known to be highly cytotoxic to neuronal cells, where it is oxidized under oxidative stress to DAQ, the ultimate toxic intermediate (10). As the kinetic data (2, 8) indicate, in neurons, DAQ rapidly conjugates with cysteine to form cysteinyldopamine isomers. In fact, the highest levels of 5-S-cysteinyldopamine among various brain tissues, although in trace concentrations (<100 nM), were detected in the substantia nigra (11). This mechanism could serve to detoxify the harmful DAQ intermediate. If DAQ is not scavenged by reaction with cysteine, then it may bind to neuroprotective proteins, such as parkin, through the cysteine residues to form cysteinyldopamine–protein adducts, thereby exerting neurotoxicity (10).

The casing model (Fig. 1) of NM architecture indicates that, in dopaminergic neurons of the substantia nigra, NM synthesis is kinetically regulated in such a manner that DAQ is conjugated first with cysteine to form cysteinyldopamine isomers, which are oxidized by DAQ through a redox reaction to form a pheomelanic core. This reaction is followed by the production of a eumelanic surface by a series of spontaneous reactions of DAQ, leading to the detoxification of DAQ. The extremely low content of cysteine in substantia nigra neurons is consistent with this view (12).

Second, the casing model may be relevant to protection against the development of Parkinson's disease, a common neurodegenerative disorder with clinical features that include tremor, slowness of movement, and stiffness. In Parkinson's disease, a selective loss of highly pigmented dopaminergic neurons in the substantia nigra occurs, whereas less-pigmented neurons are spared (13). Although exact mechanisms underlying this phenomenon are still under extensive investigation, NM is believed to play a neuroprotective role by binding toxic organic molecules and redox-active metal ions, especially iron (3). The interaction between iron and NM has been a focus of intensive research (14), because a marked accumulation of iron in parallel with disease severity is reported in the parkinsonian substantia nigra. NM is only partially saturated with iron in a healthy brain, thus playing a physiological role in intraneuronal iron homeostasis. The casing model (Fig. 1) is relevant in this respect in that pheomelanin is less efficient than eumelanin in binding drugs and metal ions (15). If pheomelanin were at the surface, the neuroprotective role of NM would not be expected.

Another possible implication of the casing model (Fig. 1) is related to the prooxidant property of pheomelanin. Ye et al. (16) have shown that, even in the absence of light, pheomelanin is able to reduce molecular oxygen at an appreciable rate, which is a property not observed with eumelanin. As one possible mechanism for the selective loss of dopaminergic neurons in Parkinson's disease, it has been proposed that the saturation of iron-binding sites caused by high iron levels leads to an increase in cytosolic, redox-active iron and subsequent cellular damage (17). Increased oxidative stress under such conditions could result in degradation of the eumelanic surface of NM by reactive oxygen species, thus exposing the pheomelanic core that is not only ineffective in iron binding but also behaves as a prooxidant itself. This exposure of pheomelanic core would lead to a vicious circle of neurodegenerative events in the parkinsonian substantia nigra. In fact, the iron-binding ability of NM in the parkinsonian brain is reduced compared with that in the healthy brain (17).

The casing model (Fig. 1) has not attracted much attention among pigment cell researchers. The past two decades have witnessed great progress in melanin research, especially regarding the biosynthesis and functions of eumelanin and pheomelanin (1, 2). The work by Bush et al. (4) now points out the significance of looking at the surface properties of melanosomes, i.e., whether they are eumelanic, pheomelanic, or both. Interest in NM has been increasing steadily over the years but has been hampered by the difficulty with isolation of NM and the scanty information on its structure. The architecture of NM granules proposed by Bush et al. (4) not only would make interpretation of the existing vast amounts of information on NM possible but also should stimulate renewed interest in the roles of NM in the pathogenesis of Parkinson's disease, the second-most-prevalent neurodegenerative disorder.