SCIENTIFIC PERSPECTIVE Neuromelanin, Aging, and Neuronal Vulnerability in Parkinson’s Disease Miquel Vila, MD, PhD1,2,3* 1 Neurodegenerative Diseases Research Group, Vall d’Hebron Research Institute–Center for Networked Biomedical Research on Neurodegenerative Diseases, Barcelona, Spain 2 Department of Biochemistry and Molecular Biology, Autonomous University of Barcelona, Barcelona, Spain 3 Catalan Institution for Research and Advanced Studies, Barcelona, Spain A B S T R A C T : Neuromelanin, a dark brown intracellular pig- allowed, for the first time, for the consequences of progres- ment, has long been associated with Parkinson’s disease sive neuromelanin accumulation—up to levels reached in (PD). In PD, neuromelanin-containing neurons preferentially elderly human brains—to be assessed in vivo. In these ani- degenerate, tell-tale neuropathological inclusions form in mals, intracellular neuromelanin accumulation above a spe- close association with this pigment, and neuroinflammation is cific threshold compromises neuronal function and triggers a restricted to neuromelanin-containing areas. In humans, PD-like pathology. As neuromelanin levels reach this thresh- neuromelanin accumulates with age, which in turn is the main old in PD patients and presymptomatic PD patients, the risk factor for PD. The potential contribution of neuromelanin modulation of neuromelanin accumulation could provide to PD pathogenesis remains unknown because, in contrast a therapeutic benefit for PD patients and delay brain aging. to humans, common laboratory animals lack neuromelanin. © 2019 The Authors. Movement Disorders published by The recent introduction of a rodent model exhibiting an age- Wiley Periodicals, Inc. on behalf of International Parkinson dependent production of human-like neuromelanin has and Movement Disorder Society. Neuromelanin and PD: this finding failed to gain him any significant recognition during his lifetime, his observation remains to this day the A 100-Year-Old Relationship cardinal pathologic diagnostic criterion for PD.1 The loss A century ago, in 1919, Konstantin Tretiakoff reported of nigral pigmented neurons, which we now know pro- for the first time in his remarkable doctorate thesis the duce the neurotransmitter dopamine, leads to the classical presence of a marked loss of pigmented neurons in motor symptoms of PD and constitutes the only robust the substantia nigra (SN), visible with the naked eye, in clinico-pathological correlation associated with the dis- the brains of Parkinson’s disease (PD) patients.1 Although ease.1 The pigment contained within these neurons, ter- med neuromelanin because of its similar appearance to cutaneous melanin, is so abundant in the SN of the human brain that this structure can be seen macroscopi- -This - - - is- -an- - open - - - - -access - - - - - - article - - - - - under - - - - - -the- - terms - - - - - of- - the - - - Creative - - - - - - - Com- -------- cally as a darkened area (hence the origin of the name mons Attribution License, which permits use, distribution and reproduc- given to this brain region).2 Later, other neuromelanin- tion in any medium, provided the original work is properly cited. containing neurons in different brain regions were also *Correspondence to: Dr. Miquel Vila, Vall d’Hebron Research Institute Neurodegenerative Diseases Pg. Vall d’Hebron 119-12908035, found to consistently degenerate in PD; these include the Barcelona, Spain; E-mail: miquel.vila@vhir.org noradrenergic neurons of the locus coeruleus and dorsal Relevant conflicts of interests/financial disclosures: Vall d’Hebron motor nucleus of the vagus, leading to characteristic non- Research Institute has applied for a use patent for the modulation of neuromelanin levels in the treatment of Parkinson’s disease and/or brain motor symptoms of the disease.3,4 In contrast, neuronal aging. loss in nonmelanized brain regions appears either incon- Received: 29 March 2019; Revised: 29 May 2019; Accepted: 30 sistent, not specific to PD, or secondary to the loss of May 2019 interconnected neuromelanin-containing neurons.5-7 Published online 28 June 2019 in Wiley Online Library In the normal human midbrain, dopamine-producing (wileyonlinelibrary.com). DOI: 10.1002/mds.27776 cell groups markedly differ from each other in terms of 1440 Movement Disorders, Vol. 34, No. 10, 2019 N E U R O M E L A N I N, A G I N G, N E U R O N A L V U L N E R A B I L I T Y I N P D the percentage of neuromelanin-pigmented neurons they brain area is not apparently observed in other animal spe- contain.8-10 In PD, the estimated cell loss in these groups cies at the macroscopic level.20 In fact, in the mammalian correlates directly with the percentage of neuromelanin- brain, neuromelanin accumulation increases progressively pigmented neurons normally present in them.8-10 Like- as the evolutionary relation to man becomes closer,20 and wise, within each neuromelanin-containing cell group, humans are the only species that naturally develops PD.24 there is greater relative sparing of weakly pigmented than Consequently, a factor so intimately linked to PD such as of strongly melanized neurons.8 For instance, the loss of neuromelanin has been surprisingly neglected to date in catecholaminergic neurons in PD is severe in the SN pars experimental in vivo paradigms of the disease. compacta (SNpc), in which virtually all neurons are pigmented, and almost undetectable in the central gray substance, in which most catecholaminergic neurons are Synthesis of Neuromelanin: not pigmented.9 In the peri- and retro-rubral regions of the Mechanisms and Significance midbrain, which contain pigmented and nonpigmented catecholaminergic neurons in similar proportions, the In contrast to the widespread distribution of other population of melanized catecholaminergic neurons is sig- brain pigments such as lipofuscin, neuromelanin is nificantly decreased, whereas the total population of cate- restricted to catecholamine-producing regions of the cholaminergic neurons devoid of neuromelanin is not brain and forms only in neurons. Neuromelanin first affected.9 Similarly, dopaminergic neurons in the ventral becomes observable in the human SNpc at approxi- tegmental area, which are largely spared in PD, produce mately 3 years of age and progressively accumulates over minimal neuromelanin over a lifetime.8,11 time within the cells in which it has been produced, as Classical Lewy bodies (LB), that is, α-synuclein-containing neurons apparently lack the mechanisms for degrading intracytoplasmic inclusions that constitute the pathological or eliminating this pigment.25,26 As a consequence, intra- hallmark of PD besides cell loss, and their presumed precur- cellular neuromelanin builds up with age until occupying sor structures, termed pale bodies, typically appear within most of the neuronal cytoplasm.25 Although aging is the the intracellular areas of the cytoplasm in which main risk factor for developing PD,27 the molecular sub- neuromelanin accumulates and form in close physical asso- strate linking PD with aging is currently unknown. ciation with this pigment.12 Consistent with these observa- Melanins (from the Greek word melanos [“black”]) are tions, studies on human brains have shown that α-synuclein a group of complex, brown-black pigmented biopolymers redistributes to the neuromelanin pigment in early stages of that include eumelanin (the pigment associated with dark PD and becomes entrapped within neuromelanin gran- hair and skin), pheomelanin (characteristic of red and ules.13-15 Further linking the PD neuropathology with blond hair), and neuromelanin (ie, brain melanin, which is neuromelanin, neuroinflammatory changes occurring in PD thought to contain a pheomelanin core and eumelanin brains, such as activation of innate (ie, microgliosis) and surface28). Melanins derive from a complex biosynthetic adaptive (ie, lymphocyte infiltration) immune responses, are pathway initiated with the hydroxylation of L-tyrosine to highly localized within neuromelanin-containing areas and L-dihydroxyphenylalanine (L-dopa). After this common only sparingly observed in nonmelanized regions.16 For step, L-dopa serves as a precursor to both melanins and cat- instance, extracellular neuromelanin released from dying echolamines, acting along separate pathways. Although neurons activates microglia in PD brains to phagocytize the synthesis of peripheral melanin pigments (eg, in skin and degrade/eliminate this pigment.17 Also, neuromelanin- and hair) is relatively well understood, the mechanism lead- containing SNpc neurons from PD brains exhibit ing to neuromelanogenesis is still a matter of speculation. Immunoglobulin G accumulation and antigenic major histo- Both neuromelanin and peripheral melanins are produced compatibility complex class I expression in a manner directly as oxidative products downstream from L-dopa. In fact, correlated to neuronal loss.18 In addition, antimelanin anti- neuromelanin synthesis is regarded as a protective antioxi- bodies are increased in the sera of PD patients.19 dant mechanism to remove potentially toxic oxidized The PD pathogenesis thus appears to be inextricably dopamine species, such as quinones and semiquinones, by linked to the presence of neuromelanin. However, despite their conversion into neuromelanin.29 However, although the close and long-established association between neuro- peripheral melanogenesis occurring within specialized cells melanin and PD, the physiological significance of neuro- (ie, melanocytes) is known to result from an enzymatically melanin and its potential contribution to PD pathogenesis driven biosynthetic pathway in which tyrosinase is the remain unknown because, in contrast to humans, labora- key, rate-limiting enzyme,30 it is widely assumed that tory animal species commonly used in experimental neuromelanin is instead produced by spontaneous non- research, such as rodents, lack neuromelanin. Although enzymatic dopamine auto-oxidation.31 Several observa- neuromelanin is actually present in some other species as tions argue, however, against the latter concept. For varied as monkeys,20,21 dolphins,22 and frogs,23 the highly instance, the distribution of neuromelanin does not fully abundant quantity of neuromelanin in the brain stem match that of the dopamine-producing enzyme tyrosine seems unique to humans, as a dark pigmentation of this hydroxylase (TH), with many dopamine and noradrenergic Movement Disorders, Vol. 34, No. 10, 2019 1441 V I L A neurons completely lacking neuromelanin.8,32-34 In addi- human brain. Although it is currently unknown whether tion, neuromelanin is not observed in most animal species tyrosinase may contribute to neuromelanin synthesis, this despite the presence of dopamine and other catecholamines possibility is theoretically conceivable because tyrosinase in these animals.20 Although this could be related to differ- not only mediates the hydroxylation of tyrosine and the ences in lifespan or rate of catecholamine synthesis between oxidation of L-dopa necessary for the formation of periph- species, experimentally induced increases in dopamine eral melanins but also is able to oxidize dopamine’s cate- and/or oxidized dopamine in mice and rats, achieved either chol ring, which is an essential event required for with chronic L-dopa treatment35,36 or by genetically neuromelanin synthesis.49 In addition, a rare loss-of- enhancing TH activity,37 are not sufficient by themselves to function mutation in tyrosinase linked to albinism has produce neuromelanin in these animals, as might be been recently associated with an increased risk for PD, expected if neuromelanin represents a mere process of auto- which has been attributed to a possible inability of this oxidized dopamine. Along this line, it is worth noting that tyrosinase variant to synthesize neuromelanin in these George C. Cotzias’s initial use of L-dopa in PD patients was patients and the subsequent accumulation of potentially apparently aimed at restoring neuromelanin levels, and not toxic dopamine-derived species that cannot be detoxified dopamine, as he thought that PD might result from the loss by conversion into neuromelanin.50 On the other hand, of the neuromelanin pigment in the SNpc.38 This approach, arguing against a potential role of tyrosinase in however, failed to reinstate the missing pigment in PD neuromelanin synthesis, the presence of neuromelanin was brains (but it provided instead a dramatic antiparkinsonian reported in the brains of 2 subjects with albinism,51 which effect by replacing the depleted dopamine, thus establishing are usually assumed to lack tyrosinase activity. However, a revolutionary new treatment for PD, although apparently that report was published in the pregenetics diagnostic era for the wrong reason).38 and, in absence of specific information about the genetic The identification of specific phases and changes in the type of albinism suffered by these cases, it cannot be rate of neuromelanin production over time in humans25,26 excluded that residual tyrosinase activity was actually pre- suggests the regulation of neuromelanin production and sent in these subjects.52 Indeed, because of the clinical over- turnover, possibly through enzymatic processes,26 as is the lap between oculocutaneous albinism (OCA) subtypes, case for all other melanin pigments.39 Interestingly, tyrosi- molecular diagnosis is necessary to establish the specific nase expression might not be restricted to melanocytes but gene defect and thus the OCA subtype. Among the most could also be present, although at very low levels, in the common types of albinism, only the OCA1 subtype is actu- brain, including the human SNpc.40-43 For instance, brain ally caused by mutations in the tyrosinase gene.53 In turn, tyrosinase messenger ribonucleic acid (mRNA) expression within OCA1 subjects, tyrosinase activity is completely has been reported in human SN by reverse transcription abolished only in the OCA1A subtype while in OCA1B polymerase chain reaction/real-time polymerase chain some enzymatic activity still remains, allowing some accu- reaction, although at barely detectable levels.41-43 Simi- mulation of melanin pigment over time.53 In this context, larly, 3 different microarray experiments in human brain at least one of the subjects examined by Foley and Baxter tissue (Entrez_id: 7299; Allen Brain Institute Seattle, was diagnosed with “partial albinism,”51 implying that Washington), nonhuman primate brain tissue (Entrez_id: some residual tyrosinase activity could have been present 705792; National Institutes of Health [NIH] Blueprint in these brains. Therefore, whether tyrosinase or other Non-human Primate [NHP] Atlas), and developing human related enzymes may contribute to neuromelanin synthesis prenatal tissue (Entrez_id: 7299; BrainSpan Atlas of the remains an open question. developing human brain) detected the presence of some tyrosinase transcripts in the brain. Consistent with the very low levels at which tyrosinase mRNA might be expressed, Modeling Neuromelanin Production if at all, in the brain, the actual protein has not been detected in the human brain when using immunohisto- In Vivo: Age-Dependent Intracellular chemistry or Western blot.44,45 Similarly, mass spectrome- Neuromelanin Accumulation Drives try techniques failed to detect significant amounts of PD Pathology typical enzymes and proteins involved in melanogenesis, including tyrosinase, in isolated neuromelanin-containing To assess the potential role of neuromelanin in PD path- organelles from human SN.15,46,47 In contrast, tyrosinase ogenesis, it would first be necessary to develop an in vivo enzymatic activity has been reported in brain extracts from model of human-like neuromelanin production, up to the human SN,41 although at 100,000 times lower than in abundant levels reached in elder humans, as no such ani- melanocytes.48 These inconsistent results could be poten- mal model currently exists, experimentally or naturally. tially attributed to the very low levels at which tyrosinase Considering the failure by previous studies to produce might be actually expressed, if at all, in the brain and fur- neuromelanin in vivo solely by increasing the levels of ther experiments are thus warranted to unambiguously dopamine and oxidized dopamine,35-37 we recently opted demonstrate or refute the existence of tyrosinase in the as a potential alternative strategy for overexpressing 1442 Movement Disorders, Vol. 34, No. 10, 2019 N E U R O M E L A N I N, A G I N G, N E U R O N A L V U L N E R A B I L I T Y I N P D (A) (B) (C) (D) (E) Human * AAV-hTyr Rat * FIG. 1. Human-like neuromelanin production in tyrosinase-overexpressing rats. Representative images from elderly human control brains (top) and brains from rats unilaterally injected with an adeno-associated viral (AAV) vector expressing human tyrosinase (hTyr) above the right substantia nigra pars compacta (SNpc; bottom). Humans and AAV-hTyr-injected rats exhibit analogous neuromelanin production from a macroscopic (A,B), microscopic (C,D), and ultrastructural (E) point of view. No neuromelanin is observed in the contralateral (noninjected) hemisphere of AAV-hTyr-injected rats, as rodents normally lack this pigment. (A) Unstained midbrain (dashed outline, melanized SNpc). (B) Neuromelanin-sensitive high-resolution T1-weighted magnetic resonance imaging (dashed outline, melanized SNpc visualized as an hyperintense area). (C) Hematoxylin-eosin stained 5-μm-thick SNpc sections (unstained neuromelanin is seen in brown). (D) Masson-Fontana melanin staining in 5-μm-thick SNpc sections (neuromelanin is seen in dark brown). (E) Electron microscopy (neuromelanin, electron dense matrix; asterisks, associated lipid droplets). Scale bars, 12.5 μm (C,D), 500 nm (E). Adapted from ref. 43. human tyrosinase (hTyr) via a viral vector in the SNpc of accumulation of human-like neuromelanin within PD- both rats and mice, which normally lack neuromelanin.43 vulnerable nigral neurons at levels up to those reached in Strikingly, hTyr-overexpressing rodents exhibited an age- elderly humans. The model thus constitutes a unique dependent production and accumulation of a neuro- experimental tool enabling assessment of the possible con- melanin-like pigment within nigral dopaminergic neurons sequences of neuromelanin production/accumulation on that was virtually analogous to human neuromelanin43 neuronal function and viability. (henceforth referred to as neuromelanin; see Fig. 1). Simi- In this context, one of the main questions to be addressed lar to humans, neuromelanin from hTyr-overexpressing concerns how a neuron containing very high levels of intra- rodents (1) could be visualized with the naked eye as a cellular neuromelanin, to the point where the neuromelanin darkened SNpc area43 (Fig. 1A); (2) was detected as an occupies most of the neuron’s cytoplasm, can function prop- hyperintense area by neuromelanin-sensitive high- erly. The answer is, in fact, that it cannot. Indeed, it was resolution T1-weighted magnetic resonance imaging found in hTyr-overexpressing rodents that the progressive (MRI)43 (Fig. 1B); (3) accumulated progressively over time intracellular build-up of neuromelanin ultimately com- within SNpc dopaminergic neurons until occupying most promised neuronal function when allowed to accumulate of their neuronal cytoplasm, reaching levels consistent above a specific threshold, eventually triggering in an age- with those seen in elderly humans43 (Fig. 1C); (4) stained dependent manner the main pathological features of PD, prominently with the melanin marker Masson-Fontana, including hypokinesia, LB-like inclusion formation and which reflects the ability of neuromelanin to chelate nigrostriatal neurodegeneration.43 By ~4 months, nigro- metals43 (Fig. 1D); and (5) was encapsulated within auto- striatal denervation in hTyr-overexpressing rodents was phagic structures exhibiting, at the ultrastructural level, an equivalent to that seen in early PD patients (<10 years’ evo- electron-dense matrix associated with characteristic lipid lution), and by ~2 years the extent of denervation reached droplets43 (Fig. 1E). The striking resemblance between levels comparable to those seen in advanced PD (>20 years’ human neuromelanin and neuromelanin from hTyr- evolution).43,54 In parallel to dopaminergic cell death, these overexpressing rodents, combined with the fact that hTyr animals exhibited other neuropathological features typical overexpression is sufficient in itself to produce neuro- of aged and PD human brains, including (1) extracellular melanin in these animals (as opposed to the lack of neuromelanin, released from dying neurons43; (2) neuro- neuromelanin by solely increasing dopamine/oxidized nophagia (ie, extracellular neuromelanin surrounded by, or dopamine levels), supports the possibility that brain tyrosi- within, activated microglia), indicative of an active, ongoing nase or other related enzymes might potentially contribute neurodegenerative process43; and (3) perivascular to neuromelanin formation in humans. Whatever the case neuromelanin, resulting from the migration of neuro- may be, these animals represent the first experimental melanin-filled microglia to blood vessels to remove extracel- in vivo model of the age-dependent production and lular neuromelanin from the brain.43 Movement Disorders, Vol. 34, No. 10, 2019 1443 V I L A Prior to degeneration, neuromelanin-filled neurons from hTyr-overexpressing rodents exhibited early signs of neu- ronal dysfunction similar to those occurring in PD patients, including impaired dopamine release, axonal swelling, decreased striatal levels of dopamine transporter, and phe- notypic loss of TH expression, all of which was accompa- nied by early motor deficits in these animals.43 The occurrence of neuronal dysfunction before overt cell death may have important therapeutic implications, as it provides a therapeutic window in which neuronal function could be potentially restored before the actual loss of the cell. Interest- ingly, PD-type inclusion body formation in hTyr- overexpressing rodents, comprising both pale bodies–like and LB-like structures exclusively observed within neuromelanin-filled neurons, peaked at the time of early neuronal dysfunction and was substantially reduced once neurodegeneration was established.43 This observation sug- gest that inclusion-containing neurons are those that become dysfunctional and preferentially degenerate in these animals. Consistent with this, inclusion-containing neurons in both humans55 and hTyr-overexpressing rodents43 often exhibit TH downregulation, which reflects neuronal dysfunction at FIG. 2. Putative pathogenic threshold of intracellular neuromelanin early stages of neurodegeneration.8 In addition, similar to accumulation. (Top) Age-dependent intracellular neuromelanin accumulation hTyr-overexpressing rodents, the number of neuronal inclu- in human tyrosinase (hTyr)–overexpressing rodents is associated with Parkinson’s disease–like neuronal dysfunction and degeneration when sions in PD brains at advanced stages of the disease is much reaching a specific threshold of accumulation (dotted red line)43. In aged lower than that observed in cases involving early PD.16 postmortem brains from healthy individuals intracellular neuromelanin levels Overall, the unprecedented use of neuromelanin- are maintained below this pathogenic threshold while in age-matched brains from Parkinson’s disease patients and pre–Parkinson’s disease (ie, inciden- producing hTyr-overexpressing rodents revealed that age- tal Lewy Body disease) subjects intracellular neuromelanin levels appear dependent neuromelanin production within SNpc dopami- above this pathogenic threshold.43 Please note that such a pathogenic nergic neurons is associated with neuronal dysfunction and threshold would be valid regardless of the actual mechanism of neuro- melanin synthesis (whether enzymatic or nonenzymatic). (Bottom) Neuronal progressive nigrostriatal neurodegeneration, equivalent to “rejuvenation” strategies aimed at reducing intracellular levels of neuro- that occurring in PD patients, once a certain threshold of melanin back to younger states (ie, below the pathogenic threshold of intracellular neuromelanin accumulation is reached (Fig. 2). neuromelanin accumulation) should prevent, halt, or delay neuronal dysfunc- tion and degeneration linked to both Parkinson’s disease and brain aging. Relevant to humans, intracellular neuromelanin levels reach Because of the putative protective role of neuromelanin synthesis (ie, to this threshold in both PD patients and subjects with inciden- remove potentially toxic dopamine-derived oxidative species from the tal LB disease (ie, clinically healthy individuals exhibiting LB cytosol), such strategies should probably favor the elimination of neuromelanin once it has already been produced (for instance, by pro- pathology at autopsy who are considered to represent early, moting the exocytosis of neuromelanin-filled lysosomal/autophagic presymptomatic stages of PD).43 In contrast, in healthy structures)43 instead of inhibiting neuromelanin synthesis per se. Photo- elderly individuals, intracellular neuromelanin levels remain micrographs correspond to hematoxylin-eosin stained human brain sec- tions illustrating dopaminergic nigral neurons with high or low levels of below this threshold.43 Although previous studies reported a intracellular neuromelanin. decrease in neuromelanin levels in dopaminergic neurons from PD patients,10 this was observed only in those neurons in the final stages of degeneration.13 In contrast, in adjacent, above a specific threshold within individual neurons, the vulnerable (but yet morphologically intact) dopaminergic lowering of intracellular neuromelanin to levels below this neurons, the neuromelanin density was augmented,13 fur- pathogenic threshold in hTyr-overexpressing rodents, in part ther supporting the notion that increased intracellular neuro- by promoting the release of intracellular neuromelanin to the melanin levels may precede dopaminergic cell death and outside of the cell (see the next section for details), dimin- predispose these neurons to degeneration in PD. Please note ished PD-linked inclusion formation, attenuated nigrostriatal that in these studies intracellular neuromelanin levels were neurodegeneration, and reversed hypokinesia in these ani- measured by optical densitometry within individual neuro- mals.43 These results demonstrate the feasibility and thera- melanin-containing neurons,10,13,43 as opposed to regional peutic potential of modulating intracellular neuromelanin quantifications of total SNpc neuromelanin,56 the latter levels in vivo and indicate that strategies to maintain or being dramatically decreased in PD brains because of the decrease intracellular neuromelanin to levels below its patho- massive loss of neuromelanin-containing dopaminergic neu- genic threshold may provide unprecedented therapeutic rons.56 Supporting a pathogenic role for intracellular opportunities to prevent, halt, or delay neuronal dysfunction neuromelanin accumulation when allowed to accumulate and degeneration linked to PD and brain aging (Fig. 2). 1444 Movement Disorders, Vol. 34, No. 10, 2019 N E U R O M E L A N I N, A G I N G, N E U R O N A L V U L N E R A B I L I T Y I N P D Mechanisms of Neuromelanin-Linked antioxidant role of neuromelanin synthesis.31 Further arguing against a major pathogenic contribution of Neurodegeneration dopamine-mediated toxicity in hTyr-overexpressing The oxidation of dopamine to form neuromelanin gener- rodents, it has been recently reported by 2 independent ates many o-quinones, which can be potentially toxic to groups that dopamine-induced toxicity in rodents, either neurons31 (Fig. 3). Indeed, dopamine oxidation has long by chronic L-dopa treatment35 or by enhancing TH been proposed as a leading pathogenic factor in PD (for a activity,37 is only observed in animals displaying additional review on this specific topic, see, for instance, ref. 31). PD-related alterations, such as DJ-1 deficiency35 or over- Therefore, the pathological PD phenotype observed in expression of PD-linked A53T mutant α-synuclein,37 but neuromelanin-producing hTyr-overexpressing rodents not in regular wild-type animals35,37 (as opposed to the PD could potentially result from the production of toxic phenotype observed in wild-type rodents overexpressing dopamine-derived intermediate species by overexpressed hTyr43). Indeed, although high concentrations of dopa- hTyr, with neuromelanin accumulation in these animals mine can be acutely toxic, mostly in vitro,11 the chronic simply representing an inert by-product of these reactions. enhancement of dopamine levels does not cause dopami- However, hTyr-overexpressing rodents showed no signifi- nergic nerve terminal or cell body loss in vivo,35-37 proba- cant accumulation of oxidized dopamine species, either bly because of compensatory protective mechanisms.57 In preceding or concomitant with neuronal dysfunction/ addition, dopamine-induced toxicity has been shown to be degeneration.43 In fact, the significant accumulation of dependent on α-synuclein expression,11,35,37,57 in contrast such potentially toxic species in these animals was actually to the nonessential role of α-synuclein for neuromelanin- prevented by their continuous conversion into neuro- linked pathology in hTyr-overexpressing animals43 (see later melanin,43 this being consistent with a putative protective for details). Although a potential pathogenic contribution of FIG. 3. Molecular mechanisms of neuromelanin-linked neurotoxicity. (1) Dopamine oxidation generates many o-quinones, which can be potentially toxic to neurons. However, the continuous conversion of these oxidized dopamine species into neuromelanin prevents their pathological accumulation in the cytoplasm,43 which is consistent with a protective antioxidant role of neuromelanin synthesis. (2) The continuous intracellular build-up of neuromelanin within undegraded lysosomal-autophagic structures, until occupying most of the neuronal cytoplasm, may physically interfere with intracellular trafficking/communication and ultimately exhaust the vesicular storage capacity of the cell, leading to a general failure of cellular proteostasis coursing with lysosomal/autophagy dysfunction, ubiquitin-proteasome (UPS) impairment, intracellular inclusion body formation, reduced mitochondrial respira- tion, increased production of reactive oxygen species, impaired neurotransmission, and neurodegeneration. All of these pathological changes are observed in neuromelanin-producing human tyrosinase–overexpressing animals and Parkinson’s disease patients.43 (3) Extracellular neuromelanin released from dying neurons activates microglia that phagocytize this pigment43 (ie, neuronophagia) and may trigger an antigenic response able to induce a selective T cell-mediated cytotoxic attack against neuromelanin-laden neurons,18 potentially contributing to the progression of the neurode- generative process. Movement Disorders, Vol. 34, No. 10, 2019 1445 V I L A oxidized dopamine species in neuromelanin-producing promoting cellular clearance involve the activation of lyso- hTyr-overexpressing rodents cannot be completely ruled somal exocytosis, a process in which lysosomes fuse with the out, because neuromelanin production is indissociable from plasma membrane and empty their contents to the outside dopamine oxidation, other explanations for the observed of the cell.69,72 Because neuromelanin is trapped within pathogenicity in these animals could be envisaged. lysosomal-autophagic structures, TFEB-induced reduction In both humans and hTyr-overexpressing rodents, of intracellular neuromelanin levels occurred, at least in part, intracellular neuromelanin accumulates within lysosomal- by the exocytosis of neuromelanin-filled lysosomes outside autophagic structures in an attempt by the cell to degrade the cell.43 This approach may thus represent a potential ther- this pigment15,58 (Fig. 3). However, because of its insoluble apeutic tool to reduce intracellular neuromelanin levels nature, neuromelanin cannot be degraded by lysosomes and below their pathogenic threshold of accumulation without remains trapped within undegraded or partly degraded interfering with neuromelanin synthesis. autophagic structures that accumulate with age.58 The con- tinuous build-up of neuromelanin within autophagic com- partments may ultimately exhaust the vesicular storage α-Synuclein in PD: A Key Factor capacity of the cell, interfere with lysosomal proteases and or a Misleading Clue? other degradative pathways, impair intracellular vesicular trafficking, and alter endocytic/secretory tasks.58,59 Sup- Given the formation of α-synuclein-positive inclusions porting this concept, L-dopa–induced toxicity in vitro is and oligomers in neuromelanin-producing hTyr-overex- only observed at doses associated with neuromelanin forma- pressing rodents and the pathogenic role attributed to tion11 and has been attributed to an interference by α-synuclein aggregation in PD, it is feasible that the patho- neuromelanin with intracellular neurotrophin signaling, and logical phenotype of these animals could be potentially not to an acute toxic effect of L-dopa per se.59 Neuromelanin mediated by α-synuclein. Strikingly, α-synuclein was could thus be toxic to aminergic neurons insofar as it found to be dispensable for PD-like inclusion formation physically interferes with intracellular communication,60,61 and nigrostriatal degeneration, as well as for neuro- causing a “macromolecular crowding” effect,62 thereby melanin production, in hTyr-overexpressing rodents.43 interfering with the synthesis and degradation of cellular Indeed, the induction of neuromelanin production by hTyr proteins.63 Consistent with this concept, the progressive overexpression in α-synuclein-deficient mice resulted in accumulation of neuromelanin-filled autophagic structures levels of pale bodies/LB-like inclusion formation and in hTyr-overexpressing cells was associated with parallel nigrostriatal degeneration comparable to those seen in decreases in both autophagic and ubiquitin-proteasome hTyr-overexpressing wild-type animals.43 However, (UPS) degradation system activities, leading to a general fail- although Lewy pathology in the latter animals displayed ure of cellular proteostasis and subsequent inclusion forma- LB markers such as p62, ubiquitin, or α-synuclein, in tion.43 Relevant to PD, both autophagy and UPS systems are hTyr-overexpressing α-synuclein-deficient mice Lewy impaired in neuromelanin-laden, but not in nonmelanized, pathology displayed p62 and ubiquitin but lacked α-syn- regions from postmortem PD brains.64,65 Impairment of uclein.43 These results indicate that α-synuclein, which is autophagy and UPS is indicative of a general, late-stage currently regarded as a major component of LB, might not proteostasis failure in which cell function and survival in fact be required for LB formation or for neu- become compromised.66 For instance, mitochondrial quality rodegeneration (whether in the absence of α-synuclein control is tightly linked to proteolytic cytosolic systems; if the these structures can still be called “LB” is a matter of latter are blocked, damaged mitochondria gradually accu- semantics). Consistent with a dispensable role of mulate.66 Consistent with this, neuromelanin-producing α-synuclein in the pathogenesis of PD, α-synuclein accu- cells also exhibit impaired mitochondrial respiration and mulation correlates poorly with clinical symptoms and increased production of reactive oxygen species, 2 patho- synaptic/neuronal loss in PD patients.73 Also, although genic events that have been consistently linked to PD.43 Con- α-synuclein-derived peptides have been proposed to act as firming a pivotal pathogenic role for impaired proteostasis in antigenic epitopes capable of inducing a selective T-cell- neuromelanin-laden cells, enhancement of lysosomal- mediated cytotoxic attack in PD patients,74 the activation mediated proteolysis in hTyr-overexpressing rodents by of both innate and adaptive immune responses in post- overexpression of transcription factor EB (TFEB) reduced mortem PD brains is localized within neuromelanin- intracellular neuromelanin to levels below the pathogenic containing areas and absent from nonmelanized regions threshold, attenuated PD-like inclusion formation, prevented such as the cortex despite the latter exhibiting abundant nigrostriatal neurodegeneration, and reversed motor impair- α-synuclein depositions.16,18 Furthermore, α-synuclein ment in these animals.43 TFEB is a master regulator of pathology is not mandatory for the development of PD, as autophagy that induces the biogenesis of lysosomes and shown by some genetic PD cases caused by Leucine-rich autophagosomes, boosts autophagic cellular clearance, and repeat kinase 2 and parkin mutations that course with modulates general proteostasis67-70 as part of a pleiotropic pure nigrostriatal degeneration without Lewy pathol- range of effects.71 Interestingly, some of TFEB’s effects at ogy.75 Moreover, phosphorylation of α-synuclein, which 1446 Movement Disorders, Vol. 34, No. 10, 2019 N E U R O M E L A N I N, A G I N G, N E U R O N A L V U L N E R A B I L I T Y I N P D is widely considered to be an index of PD pathology, alterations, such as mitochondrial dysfunction, oxidative (1) can also be seen in tissue from control subjects76; stress, autophagy/UPS deficits, inclusion body formation (2) may result from a nonspecific cross-reaction with other or neuroinflammation (all of which may occur secondarily phospho-proteins, such as phosphorylated neuro- to intracellular neuromelanin accumulation or extracellular filaments77; and (3) does not affect neuronal survival in neuromelanin release); (5) the uniqueness of PD to humans in vitro PD models.78 In addition, peripheral α-synuclein (as only humans accumulate such high levels of aggregates from PD patients, which are believed to spread neuromelanin); or (6) the supposed transmission of LB to the brain and initiate PD, lack the capacity to promote pathology from host PD patients to grafted embryonic mes- α-synuclein pathology, propagate between neuronal net- encephalic neurons (as LB are only observed within aged works, or induce neurodegeneration when inoculated into grafted neurons containing adult levels of neuromelanin). experimental animals.79 Finally, in the original studies In a broader context, these results imply that all humans reporting an apparent transmission of LB pathology from would potentially develop PD if they were to live long host PD patients to embryonic mesencephalic neurons enough to reach the pathogenic threshold of intracellular grafted into the striatum many years earlier,80,81 which neuromelanin accumulation. Supporting this concept, gave rise to the currently prevalent pathogenic hypothesis although only some individuals actually develop PD (1 to up of cell-to-cell transmission of α-synuclein, LB formation ~5% of the population older than 60, increasing with age),82 occurred exclusively within grafted neurons that con- 10% to 30% of apparently healthy people older than tained adult levels of neuromelanin and was not observed 60 years exhibit PD-type LB pathology in their melanized in nonmelanized grafted cells or intrinsic striatal neurons brain stem nuclei (ie, incidental LB disease),83 and these sub- of the host. Therefore, the latter studies may not in fact jects display intracellular neuromelanin levels above the reflect a host-to-graft propagation of α-synuclein as was pathogenic threshold defined in neuromelanin-producing widely interpreted but, rather, indicate that grafted cells hTyr-overexpressing rodents.43 In addition, parkinso- had reached pathogenic levels of intracellular neuromelanin nian signs such as bradykinesia, stooped posture, and accumulation. Overall, it is possible that α-synuclein gait disturbance are common in elderly individuals, in pathology in PD may merely represent an epiphenomenon the absence of overt PD, reaching a prevalence of up to within a more general PD-causing pathogenic process, such 50% in individuals older than 80 years of age.84,85 Also, as the one linked to neuromelanin accumulation.43 multiple reports have documented an age-related loss of Although this concept does not undermine the value of pigmented nigral neurons, estimated at about 10% per α-synuclein as a potential biomarker of PD, it raises the decade, in otherwise healthy individuals.17 Consistent importance of reassessing the value of α-synuclein as a with age-dependent alterations of neuromelanin- potential therapeutic target for the disease. containing neurons, brains from aged individuals com- monly exhibit a downregulation of dopaminergic phenotypic markers and α-synuclein accumulations Implications of Neuromelanin-Driven within neuromelanin-laden SNpc neurons54,86 as well Pathology for PD and Brain Aging as abundant extracellular neuromelanin associated with sustained microglial activation17 when compared with The recent development of a rodent model producing young adult brains. Also, in a recent retrospective study human-like neuromelanin has allowed for the consequences on healthy individuals of different ages (from 5-83 of age-dependent neuromelanin accumulation up to levels years), the neuromelanin-sensitive MRI SNpc signal reached in elderly humans to be experimentally assessed for peaked in middle age before declining in older ages, the first time in an in vivo setting. The use of this animal which was attributed to an age-related decrease of mela- model has revealed that the progressive intracellular neuro- nized nigral neurons.87 Overall, neuromelanin-filled melanin build-up that occurs with aging ultimately triggers neurons from apparently healthy aged individuals PD-like neuronal dysfunction and neurodegeneration once exhibit early signs of neuronal dysfunction/degenera- a specific threshold of neuromelanin accumulation is rea- tion. Whether these changes precede clinical PD or are ched. Therefore, although neuromelanin synthesis per se part of normal brain aging remains to be determined. In may be neuroprotective, its long-term accumulation may any case, based on the results discussed here, PD will have deleterious consequences. appear in those subjects that have reached earlier the From a pathogenic point of view, these results could pathogenic threshold of intracellular neuromelanin explain major features of PD, including (1) the selective accumulation. vulnerability of specific neuronal groups (ie, neurons From a therapeutic point of view, these results also containing neuromelanin); (2) the correlation of PD with imply that strategies to maintain or decrease intracellular age (as neuromelanin accumulates progressively with neuromelanin to levels below the pathogenic threshold age); (3) the progressive nature of the neurodegenerative should be able to prevent, halt, or delay neuronal dysfunc- process (as not all neurons accumulate neuromelanin at tion and degeneration linked to both PD and brain aging the same rate); (4) the known PD-linked cellular (Fig. 2). Because of the putative protective role of Movement Disorders, Vol. 34, No. 10, 2019 1447 V I L A neuromelanin synthesis itself (ie, to prevent the cytosolic Concluding Remarks accumulation of dopamine-derived toxic species), poten- tial therapies aimed at maintaining or reducing intracellu- and Future Directions lar neuromelanin to levels below the pathogenic threshold The development of the first rodent model producing should probably favor the elimination of neuromelanin human-like neuromelanin in PD-vulnerable dopaminergic from the cell once neuromelanin has already been pro- nigral neurons has allowed for the first time to experimen- duced instead of inhibiting neuromelanin synthesis per se, tally assess the consequences of age-dependent neuro- which may instead prove detrimental. It would be thus melanin accumulation up to levels reached in elderly important for any therapeutic strategy aimed at modulat- humans in an in vivo setting. The observation in this animal ing neuromelanin levels to tightly regulate neuromelanin model that age-dependent intracellular neuromelanin build- levels above its protective threshold and below its patho- up ultimately leads to PD-like neuronal dysfunction and genic threshold. neurodegeneration when reaching a certain pathogenic If PD were indeed triggered by an excessive accumula- threshold of accumulation may have far-reaching implica- tion of intracellular neuromelanin, above a certain patho- tions for both PD and brain aging. genic threshold, then neuromelanin levels might provide an Among the outstanding questions that remain to be indication of the risk of developing PD. For instance, answered is whether neuromelanin synthesis in humans neuromelanin’s binding of iron provides a paramagnetic source for its detection in living brain tissue by MRI.88 Cur- occurs enzymatically, nonenzymatically, or by a combina- rently, neuromelanin-sensitive MRI is mainly used to dif- tion of both. Although the pathogenic effects linked to ferentiate between PD patients and age-matched healthy neuromelanin accumulation discussed here appear to be controls based on the loss of neuromelanin-sensitive MRI independent of the actual mechanisms of neuromelanin signal in PD subjects secondary to SNpc neurodegene- synthesis, as they occur downstream of neuromelanin pro- ration.88 However, one could envisage the use of this tech- duction and subsequent accumulation, the elucidation of nique in subjects at risk for PD (eg, nonmanifesting carriers the mechanisms of neuromelanin synthesis could provide of PD-linked genetic mutations or subjects with idiopathic important clues about the significance of neuromelanin rapid eye movement sleep behavior disorder) or even in the production or identify potential molecular targets to mod- general aging population. This would allow to assess ulate neuromelanin levels. Understanding these mecha- whether these subjects might be approaching or reaching nisms may also help unravel why PD patients would reach the pathogenic threshold of neuromelanin accumulation the pathogenic threshold of neuromelanin accumulation and thus allow the prompt application of potential earlier than healthy subjects. This could be due, for neuromelanin modulatory therapies in these subjects to instance, to an increased or accelerated production of prevent disease. This could also be envisaged with the use neuromelanin in PD patients caused by an upregulation of of [18F]-AV-1451 (Flortaucipir), a novel Positron Emission hTyr or other related enzymes, such as tyrosinase-related Tomography (PET) tracer that was initially developed for protein-1 and tyrosinase-related protein-2. Consistent its high affinity to neurofibrillary tau in Alzheimer’s disease with this, the expression of tyrosinase-related protein-2 but was subsequently found to bind strongly and specifi- (ie, dopachrome tautomerase) is increased in dopaminer- cally to neuromelanin and melanin.89,90 Alternatively, it gic neurons derived from induced pluripotent stem cells could be also interesting to determine whether brain from PD patients.96 Alternatively, increased neuromelanin neuromelanin levels might correlate with, and thus be production in PD could arise from increased levels of free inferred from, the production and/or type of pigmentation cytosolic dopamine that oxidizes into neuromelanin, for in the skin or hair. Supporting this concept: (1) PD patients instance, by defects in vesicular monoamine transporter have an increased risk of developing cutaneous melanoma 2–mediated encapsulation of dopamine within synaptic and, reciprocally, patients with cutaneous melanoma have vesicles, as observed in early PD cases.97 Indeed, there is an increased risk of developing PD91; (2) similar to neurons, an inverse relationship between neuromelanin content and cutaneous melanocytes derive from pluripotent neural vesicular monoamine transporter 2 immunoreactivity in crest cells and their maturation is influenced by the same human midbrain dopaminergic neurons, with neurons signaling molecules that play a role in central and exhibiting the highest vesicular monoamine transporter peripheral nervous tissue92; (3) loss-of-function variants 2 levels and the lowest neuromelanin levels being the less of the melanocortin 1 receptor gene, which are associated vulnerable to PD-linked neurodegeneration.63 with red hair and fair skin, are linked to an increased risk of It would also be important to further uncover the exact both cutaneous melanoma and PD93,94; (4) genetically molecular mechanisms driving neuromelanin-linked modified mice carrying an inactivating mutation of mela- pathology. In addition of the potential detrimental effects nocortin 1 receptor mimicking the human redhead phe- of a physical cytosolic overcrowding by neuromelanin or notype have compromised nigrostriatal dopaminergic the formation of potentially toxic dopamine-derived oxi- neuronal integrity and are more susceptible to dopami- dized species during neuromelanin synthesis, other factors nergic parkinsonian neurotoxins.95 could also be considered. For instance, neuromelanin has 1448 Movement Disorders, Vol. 34, No. 10, 2019 N E U R O M E L A N I N, A G I N G, N E U R O N A L V U L N E R A B I L I T Y I N P D the ability to chelate various transition metal ions98 and 7. Szego EM, Gerhardt E, Outeiro TF, Kermer P. Dopamine-depletion and increased alpha-synuclein load induce degeneration of cortical neuromelanin is in fact the main iron storage molecule in cholinergic fibers in mice. J Neurol Sci 2011;310:90-95. SNpc dopaminergic neurons.99,100 The equilibrium 8. Hirsch E, Graybiel AM, Agid YA. Melanized dopaminergic neu- between iron, dopamine, and neuromelanin seems to be cru- rons are differentially susceptible to degeneration in Parkinson’s cial for cell homeostasis,99,101 and PD brains exhibit disease. Nature 1988;334:345-348. increased levels of iron when compared with healthy control 9. Hirsch EC, Graybiel AM, Agid Y. Selective vulnerability of pigmented dopaminergic neurons in Parkinson’s disease. Acta Neu- subjects.100 In this context, neuromelanin may represent a rol Scand Suppl 1989;126:19-22. potential source of toxic iron if the iron-binding capacity of 10. Kastner A. et al. Is the vulnerability of neurons in the substantia this pigment becomes saturated.101 Similarly, the capac- nigra of patients with Parkinson’s disease related to their ity of neuromelanin to bind environmental parkinso- neuromelanin content? J Neurochem 1992;59:1080-1089. nian neurotoxins, such as 1-methyl-4-phenylpyridinium 11. Mosharov EV. et al. Interplay between cytosolic dopamine, calcium, and alpha-synuclein causes selective death of substantia nigra neu- (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine MPTP’s rons. Neuron 2009;62:218-229. https://doi.org/10.1016/j.neuron. active metabolite) or the pesticide paraquat could also be 2009.01.033 considered.102 For instance, it has been observed in 12. Braak H, Del Tredici K. The pathological process underlying macaques that among dopaminergic neurons, those con- Alzheimer’s disease in individuals under thirty. Acta Neuropathol 2011;121:171-181. https://doi.org/10.1007/s00401-010-0789-4 taining neuromelanin are more susceptible to MPTP toxic- 13. Halliday G. M. et al. Alpha-synuclein redistributes to neuromelanin ity than nonmelanized neurons.103 lipid in the substantia nigra early in Parkinson’s disease. Brain These and other questions could now be addressed 2005;128:2654-2664. in vivo for the first time by means of the newly available 14. Fasano M, Giraudo S, Coha S, Bergamasco B, Lopiano L. Residual neuromelanin-producing rodent model.43 This model substantia nigra neuromelanin in Parkinson’s disease is cross-linked to alpha-synuclein. Neurochem Int 2003;42:603-606. could be used for (1) the study of the synthesis, biology, 15. Zucca FA. et al. Neuromelanin organelles are specialized and pathogenic mechanisms of neuromelanin in vivo in autolysosomes that accumulate undegraded proteins and lipids in both aging and disease, (2) the screening of novel thera- aging human brain and are likely involved in Parkinson’s disease. NPJ Parkinson Dis 2018;4:17. https://doi.org/10.1038/s41531-018- peutic strategies aimed at modulating neuromelanin levels 0050-8 within the aged or diseased brain, and (3) the search for 16. Braak H. et al. Staging of brain pathology related to sporadic potential neuromelanin-related biomarkers for the risk, Parkinson’s disease. Neurobiol Aging 2003;24:197-211 early detection, diagnosis, and/or progression of PD. The 17. Beach TG. et al. Marked microglial reaction in normal aging human introduction into experimental in vivo research of a factor substantia nigra: correlation with extraneuronal neuromelanin pig- ment deposits. Acta Neuropathol 2007;114:419-424. such as neuromelanin that is so intimately linked to PD 18. Cebrian C. et al. MHC-I expression renders catecholaminergic neu- and that has thus far been neglected in PD animal model- rons susceptible to T-cell-mediated degeneration. Nat Commun ing should open new research avenues and could lead to a 2014;5:3633. paradigm shift in the field of PD and, in a broader sense, 19. Double KL. et al. Anti-melanin antibodies are increased in sera in brain aging. Parkinson’s disease. Exp Neurol 2009;217:297-301. 20. Marsden CD. Pigmentation in the nucleus substantiae nigrae of mammals. J Anat 1961;95:256-261. Acknowledgments: The work described was funded by grants from the Ministry of Economy and Competitiveness (Spain), Institute of Health 21. Herrero MT. et al. Neuromelanin accumulation with age in cate- Carlos IIII–European Regional Development Fund, Parkinson’s U.K., The cholaminergic neurons from Macaca fascicularis brainstem. Michael J. Fox Foundation, “la Caixa” Foundation, and Center for Develop Neurosci 1993;15:37-48. https://doi.org/10.1159/ Networked Biomedical Research on Neurodegenerative Diseases 000111315 (CIBERNED). Figures 2 and 3 were drawn by Dr. Celine Perier. 22. Sacchini S. et al. Locus coeruleus complex of the family Del- phinidae. Sci Rep 2018;8:5486. https://doi.org/10.1038/s41598- 018-23827-z 23. Kemali M, Gioffre D. Anatomical localisation of neuromelanin in References the brains of the frog and tadpole. Ultrastructural comparison of 1. Lees AJ, Selikhova M, Andrade LA, Duyckaerts C. The black stuff neuromelanin with other melanins. J Anat 1985;142:73-83. and Konstantin Nikolaevich Tretiakoff. Mov Disord 2008;23:777- 24. Diederich NJ, James Surmeier D, Uchihara T, Grillner S, 783. https://doi.org/10.1002/mds.21855 Goetz CG. Parkinson’s disease: Is it a consequence of human brain 2. Fedorow H. et al. Neuromelanin in human dopamine neurons: evolution? Mov Disord 2019;34:453-459. https://doi.org/10.1002/ comparison with peripheral melanins and relevance to Parkinson’s mds.27628 disease. Prog Neurobiol 2005;75:109-124. 25. Halliday GM. et al. Evidence for specific phases in the development 3. Sara SJ. The locus coeruleus and noradrenergic modulation of cog- of human neuromelanin. J Neural Transm 2006;113:721-728. nition. Nat Rev Neurosci 2009;10:211-223. 26. Fedorow H. et al. Evidence for specific phases in the development 4. Halliday GM, Leverenz JB, Schneider JS, Adler CH. The neurobio- of human neuromelanin. Neurobiol Aging 2006;27:506-512. logical basis of cognitive impairment in Parkinson’s disease. Mov 27. Driver JA, Logroscino G, Gaziano JM, Kurth T. Incidence and Disord 2014;29:634-650. remaining lifetime risk of Parkinson disease in advanced age. Neurol- 5. Sulzer D, Surmeier DJ. Neuronal vulnerability, pathogenesis, and ogy 2009;72:432-438. https://doi.org/10.1212/01.wnl.0000341769. Parkinson’s disease. Mov Disord 2013;28:715-724. 50075.bb 6. Bensaid M, Michel PP, Clark SD, Hirsch EC, Francois C. Role of 28. Bush WD. et al. The surface oxidation potential of human pedunculopontine cholinergic neurons in the vulnerability of nigral neuromelanin reveals a spherical architecture with a pheomelanin dopaminergic neurons in Parkinson’s disease. Exp Neurol 2016; core and a eumelanin surface. Proc Natl Acad Sci U S A 2006;103: 275(Pt 1):209-219. 14785-14789. Movement Disorders, Vol. 34, No. 10, 2019 1449 V I L A 29. Zucca FA. et al. Neuromelanin of the human substantia nigra: an 51. Foley JM, Baxter D. On the nature of pigment granules in the cells update. Neurotox Res 2014;25:13-23. https://doi.org/10.1007/ of the locus coeruleus and substantia nigra. J Neuropathol Exp s12640-013-9435-y Neurol 1958;17:586-598. 30. Slominski A, Tobin DJ, Shibahara S, Wortsman J. Melanin pig- 52. Bjedov I. et al. Mechanisms of life span extension by rapamycin in mentation in mammalian skin and its hormonal regulation. Physiol the fruit fly Drosophila melanogaster. Cell Metab 2010;11:35-46. Rev 2004;84:1155-1228. 53. Gronskov K, Ek J, Brondum-Nielsen K. Oculocutaneous albinism. 31. Monzani E. et al. Dopamine, oxidative stress and protein-quinone Orphanet J Rare Dis 2007;2:43. https://doi.org/10.1186/1750- modifications in Parkinson’s and other neurodegenerative diseases. 1172-2-43 Angewandte Chemie 2018;58(20):6512-6527. https://doi.org/10.1002/ 54. Kordower JH. et al. Disease duration and the integrity of the anie.201811122 nigrostriatal system in Parkinson’s disease. Brain 2013;136:2419- 32. Saper CB, Petito CK. Correspondence of melanin-pigmented neu- 2431. rons in human brain with A1-A14 catecholamine cell groups. Brain 55. Beach TG. et al. Substantia nigra Marinesco bodies are associated 1982;105:87-101. with decreased striatal expression of dopaminergic markers. 33. Gaspar P. et al. Tyrosine hydroxylase and methionine-enkephalin J Neuropathol Exp Neurol 2004;63:329-337. in the human mesencephalon. Immunocytochemical localization 56. Zecca L. et al. The absolute concentration of nigral neuromelanin, and relationships. J Neurol Sci 1983;58:247-267. assayed by a new sensitive method, increases throughout the life 34. Halliday GM. et al. Distribution of monoamine-synthesizing neu- and is dramatically decreased in Parkinson’s disease. FEBS Lett rons in the human medulla oblongata. J Comp Neurol 1988;273: 2002;510:216-220. 301-317. https://doi.org/10.1002/cne.902730303 57. Mor DE, Daniels MJ, Ischiropoulo, H. The usual suspects, dopa- 35. Burbulla LF. et al. Dopamine oxidation mediates mitochondrial mine and alpha-synuclein, conspire to cause neurodegeneration. and lysosomal dysfunction in Parkinson’s disease. Science 2017; Mov Disord 2019;34:167-179. https://doi.org/10.1002/mds.27607 357:1255-1261. https://doi.org/10.1126/science.aam9080 58. Sulzer D. et al. Neuronal pigmented autophagic vacuoles: 36. Murer MG. et al. Chronic levodopa is not toxic for remaining lipofuscin, neuromelanin, and ceroid as macroautophagic responses dopamine neurons, but instead promotes their recovery, in rats during aging and disease. J Neurochem 2008;106:24-36. with moderate nigrostriatal lesions. Ann Neurol 1998;43:561-575. 59. Sulzer D. et al. Neuromelanin biosynthesis is driven by excess cyto- 37. Mor DE. et al. Dopamine induces soluble alpha-synuclein oligo- solic catecholamines not accumulated by synaptic vesicles. Proc mers and nigrostriatal degeneration. Nat Neurosci 2017;20:1560- Natl Acad Sci U S A 2000;97:11869-11874. 1568. https://doi.org/10.1038/nn.4641 60. Graham DG. Oxidative pathways for catecholamines in the genesis 38. Fahn S. The medical treatment of Parkinson disease from James of neuromelanin and cytotoxic quinones. Mol Pharmacol 1978;14: Parkinson to George Cotzias. Mov Disord 2015;30:4-18. https:// 633-643. doi.org/10.1002/mds.26102 61. Mann DM, Yates PO. Possible role of neuromelanin in the patho- 39. Mishima Y. Molecular and biological control of melanogenesis genesis of Parkinson’s disease. Mechanisms of ageing and develop- through tyrosinase genes and intrinsic and extrinsic regulatory fac- ment 1983;21:193-203. tors. Pigment cell Res 1994;7:376-387. 62. Ellis RJ. Macromolecular crowding: obvious but underappreciated. Trends Biochem Sci 2001;26:597-604. 40. Miranda M, Botti D, Bonfigli A, Ventura T, Arcadi A. Tyrosinase- like activity in normal human substantia nigra. Gen Pharmacol 63. Liang CL, Nelson O, Yazdani U, Pasbakhsh P, German DC. 1984;15:541-544. Inverse relationship between the contents of neuromelanin pigment and the vesicular monoamine transporter-2: human midbrain 41. Greggio E. et al. Tyrosinase exacerbates dopamine toxicity but is dopamine neurons. J Comp Neurol 2004;473:97-106. https://doi. not genetically associated with Parkinson’s disease. J Neurochem org/10.1002/cne.20098 2005;93:246-256. 64. Dehay B. et al. Pathogenic lysosomal depletion in Parkinson’s dis- 42. Xu Y. et al. Tyrosinase mRNA is expressed in human substantia ease. J Neurosci 2010;30:12535-12544. nigra. Brain Res Mol Brain Res 1997;45:159-162. 65. Cook C, Petrucelli L. A critical evaluation of the ubiquitin- 43. Carballo-Carbajal I. et al. Brain tyrosinase overexpression impli- proteasome system in Parkinson’s disease. Biochim Biophys Acta cates age-dependent neuromelanin production in Parkinson’s dis- 2009;1792:664-675. https://doi.org/10.1016/j.bbadis.2009.01.012 ease pathogenesis. Nat Commun 2019;10:973. https://doi.org/10. 1038/s41467-019-08858-y 66. Kaushik S, Cuervo AM. Proteostasis and aging. Nat Med 2015;21: 1406-1415. https://doi.org/10.1038/nm.4001 44. Tribl F, Arzberger T, Riederer P, Gerlach M. Tyrosinase is not detected in human catecholaminergic neurons by immunohisto- 67. Sardiello M. et al. A gene network regulating lysosomal biogenesis chemistry and Western blot analysis. J Neural Transm Suppl 2007; and function. Science 2009;325:473-477. (72):51-55. 68. Settembre C. et al. TFEB links autophagy to lysosomal biogenesis. 45. Ikemoto K. et al. Does tyrosinase exist in neuromelanin-pigmented Science 2011;332(6036):1429-1433. neurons in the human substantia nigra? Neurosci Lett 1998;253: 69. Medina DL. et al. Transcriptional activation of lysosomal exocyto- 198-200. sis promotes cellular clearance. Dev Cell 2011;21:421-430. 46. Tribl F. et al. Subcellular proteomics reveals neuromelanin granules 70. Song W. et al. TFEB regulates lysosomal proteostasis. Hum Mol to be a lysosome-related organelle. J Neural Transm 2006;113: Genet 2013;22:1994-2009. https://doi.org/10.1093/hmg/ddt052 741-749. 71. Torra A. et al. Overexpression of TFEB drives a pleiotropic neuro- 47. Plum S. et al. Proteomic characterization of neuromelanin granules trophic effect and prevents Parkinson’s disease-related neu- isolated from human substantia nigra by laser-microdissection. Sci rodegeneration. Mol Ther 2018;26:1552-1567. https://doi.org/10. Rep 2016;6:37139. https://doi.org/10.1038/srep37139 1016/j.ymthe.2018.02.022 48. Jimenez M, Kameyama K, Maloy WL, Tomita Y, Hearing VJ. 72. Magalhaes J, Gegg ME, Migdalska-Richards A, Schapira AH. Mammalian tyrosinase: biosynthesis, processing, and modulation Effects of ambroxol on the autophagy-lysosome pathway and mito- by melanocyte-stimulating hormone. Proc Natl Acad Sci U S A chondria in primary cortical neurons. Sci Rep 2018;8:1385. https:// 1988;85:3830-3834. https://doi.org/10.1073/pnas.85.11.3830 doi.org/10.1038/s41598-018-19479-8 49. Sanchez-Ferrer A, Rodriguez-Lopez JN, Garcia-Canovas F, Garcia- 73. Surmeier DJ, Obeso JA, Halliday GM. Selective neuronal vulnera- Carmona F. Tyrosinase: a comprehensive review of its mechanism. bility in Parkinson disease. Nat Rev Neurosci 2017;18:101-113. Biochim Biophys Acta 1995;1247:1-11. https://doi.org/10.1038/nrn.2016.178 50. Lubbe SJ. et al. Rare variants analysis of cutaneous malignant mel- 74. Sulzer D. et al. T cells from patients with Parkinson’s disease recog- anoma genes in Parkinson’s disease. Neurobiol Aging 2016;48: nize alpha-synuclein peptides. Nature 2017;546:656-661. https:// 222.e1-222.e7. doi.org/10.1038/nature22815 1450 Movement Disorders, Vol. 34, No. 10, 2019 N E U R O M E L A N I N, A G I N G, N E U R O N A L V U L N E R A B I L I T Y I N P D 75. Lang AE, Espay AJ. Disease modification in Parkinson’s disease: 89. Marquie M. et al. Lessons learned about [F-18]-AV-1451 off-target current approaches, challenges, and future considerations. Mov binding from an autopsy-confirmed Parkinson’s case. Acta Neuro- Disord 2018;33:660-677. https://doi.org/10.1002/mds.27360 pathol Commun 2017;5:75. https://doi.org/10.1186/s40478-017- 0482-0 76. Antunes L. et al. Similar alpha-synuclein staining in the colon mucosa in patients with Parkinson’s disease and controls. Mov Dis- 90. Marquie M. et al. Validating novel tau positron emission tomogra- ord 2016;31:1567-1570. https://doi.org/10.1002/mds.26702 phy tracer [F-18]-AV-1451 (T807) on postmortem brain tissue. Ann Neurol 2015;78:787-800. https://doi.org/10.1002/ana.24517 77. Sacino AN. et al. Amyloidogenic alpha-synuclein seeds do not invariably induce rapid, widespread pathology in mice. Acta Neu- 91. Liu R, Gao X, Lu Y, Chen H. Meta-analysis of the relationship ropathol 2014;127:645-665. https://doi.org/10.1007/s00401-014- between Parkinson disease and melanoma. Neurology 2011;76: 1268-0 2002-2009. 78. Inigo-Marco I. et al. E46K alpha-synuclein pathological mutation 92. Yaar M, Park HY. Melanocytes: a window into the nervous sys- causes cell-autonomous toxicity without altering protein turnover tem. J Invest Dermatol 2012;132:835-845. https://doi.org/10.1038/ or aggregation. Proc Natl Acad Sci U S A 2017;114:E8274-E8283. jid.2011.386 https://doi.org/10.1073/pnas.1703420114 93. Tell-Marti G. et al. The MC1R melanoma risk variant p.R160W is 79. Recasens A. et al. Lack of pathogenic potential of peripheral alpha- associated with Parkinson disease. Ann Neurol 2015;77:889-894. synuclein aggregates from Parkinson’s disease patients. Acta https://doi.org/10.1002/ana.24373 Neuropathol Commun 2018;6:8. https://doi.org/10.1186/s40478- 94. Gao X, Simon KC, Han J, Schwarzschild MA, Ascherio A. Genetic 018-0509-1 determinants of hair color and Parkinson’s disease risk. Ann Neu- 80. Kordower JH, Chu Y, Hauser RA, Freeman TB, Olanow CW. rol 2009;65:76-82. https://doi.org/10.1002/ana.21535 Lewy body-like pathology in long-term embryonic nigral trans- 95. Chen X. et al. The melanoma-linked "redhead" MC1R influences plants in Parkinson’s disease. Nat Med 2008;14:504-506. dopaminergic neuron survival. Ann Neurol 2017;81:395-406. 81. Li JY. et al. Lewy bodies in grafted neurons in subjects with https://doi.org/10.1002/ana.24852 Parkinson’s disease suggest host-to-graft disease propagation. Nat 96. Fernandez-Santiago R. et al. Aberrant epigenome in iPSC-derived Med 2008;14:501-503. dopaminergic neurons from Parkinson’s disease patients. EMBO 82. de Lau LM, Breteler MM. Epidemiology of Parkinson’s disease. Mol Med 2015;7:1529-1546. https://doi.org/10.15252/emmm. Lancet Neurol 2006;5:525-535. https://doi.org/10.1016/S1474- 201505439 4422(06)70471-9 97. Pifl C. et al. Is Parkinson’s disease a vesicular dopamine storage 83. DelleDonne A. et al. Incidental Lewy body disease and preclinical disorder? Evidence from a study in isolated synaptic vesicles of Parkinson disease. Arch Neurol 2008;65:1074-1080. human and nonhuman primate striatum. J Neurosci 2014;34: 8210-8218. 84. Bennett DA. et al. Prevalence of parkinsonian signs and associated mortality in a community population of older people. N Engl J 98. Bohic S. et al. Intracellular chemical imaging of the developmental Med 1996;334:71-76. phases of human neuromelanin using synchrotron X-ray micro- spectroscopy. Anal Chem 2008;80:9557-9566. 85. Prettyman R. Extrapyramidal signs in cognitively intact elderly people. Age Ageing 1998;27:557-560. 99. Zucca FA. et al. Interactions of iron, dopamine and neuromelanin pathways in brain aging and Parkinson’s disease. Prog Neurobiol 86. Chu Y, Kordower JH. Age-associated increases of alpha-synuclein 2015;155:96-119. https://doi.org/10.1016/j.pneurobio.2015.09.012 in monkeys and humans are associated with nigrostriatal dopamine depletion: Is this the target for Parkinson’s disease? Neurobiol Dis 100. Double KL. et al. Iron-binding characteristics of neuromelanin of 2007;25:134-149. the human substantia nigra. Biochem Pharmacol 2003;66:489-494. 101. Hare DJ, Double KL. Iron and dopamine: a toxic couple. Brain 87. Xing Y, Sapuan A, Dineen RA, Auer DP. Life span pigmentation 2016;139:1026-1035. https://doi.org/10.1093/brain/aww022 changes of the substantia nigra detected by neuromelanin-sensitive MRI. Mov Disord 2018;33:1792-1799. https://doi.org/10.1002/ 102. Karlsson O, Lindquist NG. Melanin and neuromelanin binding of mds.27502 drugs and chemicals: toxicological implications. Arch Toxicol 2016;90:1883-1891. 88. Sulzer D. et al. Neuromelanin detection by magnetic resonance imaging (MRI) and its promise as a biomarker for Parkinson’s dis- 103. Herrero MT. et al. Does neuromelanin contribute to the vulnerabil- ease. NPJ Parkinson Dis 2018;4:11. https://doi.org/10.1038/ ity of catecholaminergic neurons in monkeys intoxicated with s41531-018-0047-3. MPTP. Neuroscience 1993;56:499-511. Movement Disorders, Vol. 34, No. 10, 2019 1451