3,4-Dihydroxyphenylacetaldehyde is the toxic dopamine metabolite in vivo: implications for Parkinson’s disease pathogenesis

Brain Research 989 (2003) 205–213 www.elsevier.com / locate / brainres

Research report

3,4-Dihydroxyphenylacetaldehyde is the toxic dopamine metabolite in vivo: implications for Parkinson’s disease pathogenesis William J. Burke a,b,c,e , *, Shu Wen Li d , Evelyn A. Williams c , Randal Nonneman a,c , Daniel S. Zahm c a

Department of Neurology, Saint Louis University Medical Center, 3635 Vista at Grand, St. Louis, MO 63110, USA Department of Medicine, Saint Louis University Medical Center, 3635 Vista at Grand, St. Louis, MO 63110, USA c Department of Anatomy and Neurobiology, Saint Louis University School of Medicine, 1402 S. Grand Boulevard, St. Louis, MO 63104, USA d Department of Chemistry, Saint Louis University, 221 N. Grand Boulevard, St. Louis, MO 63103, USA e Veterans Affairs Medical Center, St. Louis, MO, USA b

Accepted 17 July 2003

Abstract In Parkinson’s disease (PD), there is a highly selective loss of dopamine (DA) neurons in the substantia nigra (SN) greater than in the ventral tegmental area (VTA). The simplest explanation for selective DA neuron loss in PD is that DA is toxic and, because only DA neurons contain significant amounts of DA, this highly localized synthesis of DOPAL accounts for selective vulnerability of DA neurons. However, the large concentrations of DA required to produce in vivo toxicity cast doubt on its role in PD pathogenesis. Alpha-synuclein (a-syn) is the major component of the Lewy body, the pathological marker of PD, and is genetically linked to the disease. Recent studies indicate that a-syn neurotoxicity is mediated by a free radical generating metabolite of DA. Here we test the hypothesis that 3,4-dihydroxyphenylacetaldehyde (DOPAL), the monamine oxidase metabolite of DA, mediates DA toxicity in vivo. We injected DOPAL, DA and its oxidative, reduced and methylated metabolites into rat SN and VTA. Five days post-surgery, the injection sites were evaluated in Nissl preparations and with tyrosine hydroxylase (for DA neurons), neuronal nuclear antigen (for neurons) and glial fibrillary acidic protein (for astrocytes) immunoreactivities. Lesion size in SN vs. VTA was compared using morphometry. DOPAL at concentrations as low as 100 ng was toxic to DA SN neurons.DA VTA neurons.glia. Neither DA nor its other metabolites showed evidence of neurotoxicity at fivefold higher doses. However, 20 mg of DA produced lesions in the SN and VTA. We conclude that DOPAL is the toxic DA metabolite in vivo. Implications for a unified hypothesis for PD pathogenesis are discussed.  2003 Elsevier B.V. All rights reserved. Theme: Disorders of the nervous system Topic: Neurotoxicity Keywords: Basal ganglia; Neuron death; Neurodegenerative disorder; Substantia nigra; Ventral tegmental area; 3,4-Dihydroxyphenylacetaldehyde

1. Introduction Parkinson’s disease (PD), the second most common neurodegenerative disease [1], affects 1 million Americans with an incidence of 50,000 / year [20]. PD is characterized by resting tremor, rigidity and bradykinesia associated with losses of more than 80% of dopamine (DA) neurons in the *Corresponding author. Department of Neurology, Saint Louis University Medical School, 3635 Vista at Grand, St. Louis, MO 63110, USA. Tel.: 11-314-577-8026; fax: 11-314-268-5101. E-mail address: [email protected] (W.J. Burke). 0006-8993 / 03 / $ – see front matter  2003 Elsevier B.V. All rights reserved. doi:10.1016 / S0006-8993(03)03354-7

pars compacta of the substantia nigra (SN) [15–17]. Substantially fewer DA neurons are lost in the ventral tegmental area (VTA) in PD, and most other neuronal subtypes are unaffected [14,31]. Because only DA neurons contain significant amounts of DA, the simplest explanation for selective DA neuron loss in PD is that DA itself is toxic. In fact, over 50 publications report that DA is neurotoxic in vitro and in vivo [8]. However, the very high concentrations required for DA toxicity in vitro (300 mM) [8] and in vivo (77 mg) [10] have led to skepticism about its toxicity in PD. Contrarily, a series of recent investigations by several laboratories

206

W. J. Burke et al. / Brain Research 989 (2003) 205–213

[3,19,25,40] now implicates DOPAL (3,4-dihydroxyphenylacetaldehyde), a DA metabolite as the critical endogenous toxin that triggers DA neuron loss in PD. Other major metabolites of DA include 3,4-dihydroxyphenylacetic acid (DOPAC) and 3,4-dihydroxyphenylethanol (DOPET). DOPAL, the product of monamine oxidase (MAO) action on DA, is oxidized by aldehyde dehydrogenase (ALDH) to DOPAC or reduced by aldehyde reductase (ALDR) to DOPET. DOPAC is methylated by catechol-O-methyltransferase (COMT) to homovanillic acid (HVA), the major brain metabolite of DA. Mutations in the gene for a-synuclein (a-syn) are found in autosomal dominant early onset PD [29] providing strong genetic evidence for a-syn involvement in nigral DA neuron loss in PD. Contrariwise, a-syn is widely distributed in brain neurons and glia [28] and therefore by itself cannot explain the highly selective loss of DA neurons in PD [14,31]. However, recent studies show that aggregation of a-syn is necessary for its toxicity [40]. Furthermore, Xu et al. [40] showed that aggregation of a-syn to a 53–83 kDa complex is toxic to cultured a-syn transfected DA neurons, but not non-DA neurons, which suggests that toxicity of a-syn aggregates depends on synthesis of DA. Inhibiting DA synthesis in these cultured DA neurons completely prevented a-syn induced neuron death. The a-syn transfected neurons also exhibited a marked increase in free radical generation, which, together with the accompanying cell death, was attenuated by antioxidants, leading to the conclusion that accumulation of a-syn aggregates renders endogenous levels of DA toxic by facilitating the production of a free radical generating DA metabolite [40]. In 1952 Blashko predicted that the aldehyde metabolites of amines would be toxic to cells in which they are formed [3]. Almost 50 years passed before it was shown that synthetic DOPAL [24], the monoamine oxidase-A (MAOA) metabolite of DA, but not DA itself or its oxidative or methylated metabolites, is neurotoxic in an in vitro model of DA neurons [19] and generates a free [–OH] radical under physiological conditions [25]. The experiments described here were done to determine if DOPAL is toxic in vivo.

2. Materials and methods

2.1. Materials Dopamine and most other biochemicals were purchased from Sigma (St. Louis, MO, USA). DOPAL was synthesized by our method [24].

2.2. Injections and tissue preparation The housing and nutrition of the rats used in this study

and all procedures performed on them conformed to standards set forth in the Guide for the Care and Use of Laboratory Animals of the National Research Council (National Academy Press, 1996). The experimental protocols reported here were reviewed and approved by the Animal Care Committee and monitored by the Department of Comparative Medicine of the Saint Louis University School of Medicine. The details of the surgical and immunohistochemical methods used have been published previously [32,41]. Briefly, male Sprague–Dawley rats (200–300 g; Harlan, Indianapolis, IN, USA) were deeply anesthetized with 0.16 ml / 100 g of a mixture of ketamine (100 mg / ml), xylazine (20 mg / ml) and saline (9:7:4, i.p.). The heads of the rats were fixed in a stereotaxic apparatus (David Kopf, Tujunga, CA, USA) and either DOPAL, DA, DOPAC, DOPET, HVA or the vehicle consisting of 1.0% benzyl alcohol in phosphate-buffered saline (pH 7.4), was injected unilaterally into the VTA and substantia nigra compacta (SNC) of each by pressure through a 1.5 mm pipette pulled to a tip diameter of 30–50 mm. Rhodamine microspheres (1.0 mm; Molecular Probes, Eugene, OR, USA) were injected with the compounds to reveal the injection sites. Thirty-one rats were used and all of these received multiple injections involving the SN and VTA on one or both sides of the brain. The numbers of injections involving the SN were as follows: DA—20 mg (2), 10 mg (2), 5 mg (2), 500 ng (4); DOPAL—500 ng (3), 250 ng (4), 100 ng (2), 50 ng (1); DOPAC—500 ng (2); DOPET—500 ng (2); HVA—500 ng (2); vehicle (1). The numbers of injections involving the VTA were as follows: DA—20 mg (2), 10 mg (2), 5 mg (2), 500 ng (4); DOPAL—750 ng (1), 500 ng (2), 250 ng (3), 100 ng (2), 50 ng (2); DA—500 ng (4); DOPAC—500 ng (2); DOPET—500 ng (2); HVA— 500 ng (6); vehicle (1). Injections were made in volumes of 0.2 ml.

2.3. Immunohistochemistry Eighteen hours (two rats only) or 5 days following the surgery the rats were reanesthetized and perfused through the left ventricle of the heart with 0.1 M phosphate buffer containing 4% paraformaldehyde. The brains were removed, post-fixed for at least 4 h, sunk in 25% sucrose, and sectioned frozen at 50 mm with a sliding microtome. Adjacent series of sections were subjected to a conventional immunoperoxidase protocol using antibodies against tyrosine hydroxylase (TH; Sigma, monoclonal, made in rat, used at a dilution of 1:6000), neuronal nuclear antigen (NeuN, Chemicon, Temecula, CA, USA, made in rabbit, used at 1:20,000) or glial fibrillary acidic protein (GFAP, made in rabbit, used at 1:5000). Briefly, the sections were immersed overnight in 0.1 M Sorenson’s phosphate buffer (SPB, pH 7.4) containing 0.2% Triton X-100 (SPB / Triton) and primary antibody with agitation. The following morn-

W. J. Burke et al. / Brain Research 989 (2003) 205–213

ing they were thrice rinsed in SPB / Triton and immersed for 1 h in SPB / Triton containing biotinylated secondary antibodies against the host species of the primary antibodies, used at a dilution of 1:200. The sections were again thrice rinsed in SPB / Triton and then immersed in SPB / Triton containing ABC reagents (Vector, Burlingame, CA, USA) used at a dilution of 1:200. After further rinsing in SPB the sections were reacted with 3,39-diaminobenzidine (DAB) and hydrogen peroxide to produce an insoluble brown reaction product that was further intensified with osmium and thiocarbohydrazide as has been described [32,41]. All immersions and rinses were done at room temperature. In addition, a series of sections from many of the brains was mounted and processed for Nissl staining using a standard cresyl violet staining procedure. Processing was concluded by coverslipping the sections under DPX (Fluka, Sigma–Aldrich, St. Louis, MO, USA). NeuN is a marker of neuronal differentiation [38,42] that allows neurons to be distinguished immunohistochemically and thus demonstrates neuronal loss more clearly than Nissl-staining by avoiding the confounding images of glia in the material. Antibodies against NeuN have been used as a convenient means to examine lesions following the injection of excitotoxins into the central nervous system [37].

207

2.4. Quantitation of lesion size Evaluations of the effects of injections were made based on the appearance of the tissue at injection sites as compared to adjacent parts of the same structures at some distance from the injection sites. Lesion sites were circumscribed using morphometry software (Northern Eclipse, Mississauga, Canada) and the areas were expressed as means and standard deviations and compared with an analysis of variance (ANOVA) followed by post-hoc evaluations using Fishers LSD test.

3. Results

3.1. DA neurotoxicity DA, at 20 mg, but not 10 mg, 5 mg, or 500 ng, consistently produced lesions in SN and less extensive lesions in the VTA. Lesions were characterized by loss of tyrosine hydroxylase (Fig. 1A and C) and NeuN immunoreactivities and marked gliosis (Fig. 1B), within which near complete loss of Nissl stained neurons was observed (Fig. 1D).

Fig. 1. Micrographs illustrating the effect of a microinjections of DA (20 mg in 0.2 ml) into substantia nigra compacta (SNC). Panels A and C show tyrosine hydroxylase immunoreactivity, with panel C simply an enlargement showing detail in the lesion site (arrows in A and B). Panels B and D illustrate the Nissl stained preparation of an adjacent section at lower (B) and higher (D) magnification. Note the marked gliosis at the site of the lesion. Note in D that neurons are not present in the gliotic lesion site. Additional abbreviations: SNR—substantia nigra reticulata; VTA—ventral tegmental area. Scale bar: 1.0 mm in A and B; 250 mm in C and D.

208

W. J. Burke et al. / Brain Research 989 (2003) 205–213

3.2. DOPAL neurotoxicity None of the immunohistochemical markers or Nissl preparations were detectably altered following any of the injections of DA at less than 20 mg or vehicle (Fig. 2A). On the contrary, DOPAL injections produced focal, generalized neuronal degeneration (Figs. 2 and 3). Such injections resulted in the loss of TH and NeuN immunoreactivity (ir) with accompanying gliosis at the site of injection in both the SNC and VTA as early as 18 h following the injections (not shown) and at the 5-day time point (Fig. 2B–C0). The SNC was consistently more severely affected than the VTA by equivalent injections (compare Fig. 2B9 with B0, and Fig. 2C9 with C0). In the same sections where 100 ng of DOPAL in identical volumes had been infused into the VTA and SN (e.g., Fig. 2B and C), morphometric comparisons of the area of the lesions in SN vs. VTA stained with NeuN revealed that comparable injections produced lesions that were larger in the SN by 3.2-fold (VTA: 0.05560.008 mm 2 vs. SN: 0.17560.010; P,0.01, randomized ANOVA with post-hoc Fisher’s LSD). A 250 ng DOPAL injection produced a similarly significant 3.7-fold larger lesion in SN compared to VTA. The largest injections of DOPAL were associated with frank tissue necrosis (Fig. 2D and E). Nissl stained preparations exhibited marked gliosis at lesion sites. Neuronal perikarya were absent within the patches of gliosis, although apparently normal numbers of neurons were present in the areas abutting the lesions. Interestingly, despite the Nissl evidence for gliosis, some sites exhibiting loss of TH and NeuN immunoreactivity following DOPAL injections also exhibited a loss of GFAP-ir within a restricted zone at the center of the injection. The area of loss of GFAP-ir in the SN was invariably less than that for TH and NeuN-ir (compare Fig. 2B9 and C9 with Fig. 3A). DOPAL injection sites typically were surrounded by a ‘halo’ of reactive GFAP-ir cells possessing a dense network of strikingly immunoreactive cellular processes, presumably indicative of astrocytic gliosis (Fig. 3A, C and D). A similar, but less robust, apparent gliosis occurred following DOPAL injections in the VTA (Fig. 3B and C), although none of the injections into the VTA produced a focal core of absent GFAP-ir as was seen after nigral injections of DOPAL (compare Fig. 2B9 and C9 with Fig. 3B). Interestingly, injections of 50 ng DOPAL into the SN that produced no detectable loss either of TH- or NeuN-ir were nevertheless associated with focal enhancement of GFAP staining (Fig. 3E) suggestive of a glial response. Altered GFAP-ir was not detected at sites of injection of 500 ng dopamine or vehicle.

3.3. Toxicity of other DA metabolites Neither DOPAC, DOPET or HVA at concentrations of 500 ng produced lesions in the SNC or VTA (Figs. 4A–H). In the absence of detectably altered tissue, rhodamine

microspheres demonstrated the precise locations of the injection.

4. Discussion These data are consistent with a dose-dependent toxic effect of DOPAL on neurons and glia in SN and VTA associated with tissue necrosis at the highest dose. Nissl stained preparations confirmed that the loss of the TH and NeuN immunohistochemical markers reflects loss of neurons as opposed to attenuated expression of the proteins. The toxicity of DOPAL was greater in SN than VTA, possibly due to a greater vulnerability of DA neurons in the SN. Interestingly, low concentrations of DOPAL that failed to produce any visible neuronal loss stimulated glial proliferation (Fig. 3E). The observation that DOPAL and DA affect SN neurons more than VTA neurons is consistent with previous findings that SN DA neurons are more vulnerable to 6-hydroxydopamine toxicity than VTA neurons [32]. We have previously reported that 3,4dihydroxyphenylglycolaldehyde (DOPEGAL), the MAO metabolite of norepinephrine, shows neuronal subtype selective toxicity [6]. The present data, however, do not demonstrate that intraparenchymally injected DOPAL and DA are selectively toxic to nigral as compared to VTA DA or any other type of neurons [6,32]. Indeed, it is apparent from the present results that DOPAL is highly toxic to all neurons. An apparent selective toxicity of externally applied free radical generating catecholamines (CAs) is unlikely to be due to any cell-type specificity in the mechanism by which they induce cell death [19], but rather can be explained by the extra- and intracellular characteristics of the respective DA neurons. For example, less vulnerable VTA neurons and glia [6,36] express greater endogenous levels of free radical scavengers than SN neurons [14]. SN DA neurons have a greater expression of the DA transporter than VTA neurons or glia [31], which would make them more susceptible to externally applied CA aldehydes that use this transporter to enhance entrance into CA neurons [6], although DOPAL obviously gets into neurons by means other than the dopamine transporter following injections of exogenous DOPAL at most of the concentrations used in this study. Ultimately, selective vulnerability is due to the fact that only DA neurons synthesize DA and its metabolites in amounts consistent with generation of toxic levels of DOPAL [11]. Numerous studies have shown that DA itself is toxic to neurons both in vitro and in vivo [8,10]. Filloux and Townsend reported that 77 mg of DA is toxic to DA terminals in striatum [10]. In the present experiments 20 mg, but not 10 mg, of DA was toxic to DA neurons in the SN, as compared to 50 ng, which was the smallest amount of DOPAL observed in the present study to exhibit histologically detectable toxicity (gliosis in Fig. 3E). This finding, along with the above data suggesting that DOPAL

W. J. Burke et al. / Brain Research 989 (2003) 205–213

209

Fig. 2. Micrographs illustrating the effects of microinjections of DOPAL into the ventral tegmental area (VTA) and substantia nigra compacta and reticulata (SNC and SNR). Panel A shows a case in which dopamine (500 ng in 0.2 ml) was injected into the VTA and substantia nigra compacta (SNC) on the right side of the brain and 0.2 ml of vehicle was injected in the same structures on the left. No lesions were observed. Panels B and C show a case in which 500 ng of DOPAL in 0.2 ml was injected into the SNR and SNC on the left side (arrow on the left) and into the VTA on the right side (arrow on right). B, B9 and B0 show sections processed to exhibit tyrosine hydroxylase immunoreactivity. C, C9 and C0 show an adjacent section processed to illustrate neuronal nuclear antigen (NeuN) immunoreactivity, a marker of neurons. B9 and C9 show the nigral injection site enlarged, while B0 and C0 show the VTA site enlarged. Note the complete loss of both immunohistochemical markers in the SNR / SNC (arrows in B9 and C9) and significantly smaller lesion in the VTA (surrounded by white arrows in B0 and indicated by an arrow in C0). Panels D and E show the results of injections of 250 (left arrow) and 750 ng (right arrow) of DOPAL in the VTA. Note the frank necrosis resulting from the larger injection. Scale bars: A, B and C—1.0 mm; B9, C9, B0 and C0—200 mm; D and E—0.5 mm.

210

W. J. Burke et al. / Brain Research 989 (2003) 205–213

Fig. 3. Micrographs illustrating GFAP immunostaining following DOPAL injections. Asterisks indicate sites of injection (0.2 ml). Panel A shows GFAP immunoreactivity in the substantia nigra compacta and reticulata (SNC and SNR) in a section serially adjacent to those illustrated in Fig. 1B9 and C9. Note the characteristic loss of GFAP in a central core surrounded by a zone in which GFAP expression is robustly enhanced (arrows). B shows GFAP immunoreactivity in the ventral tegmental area (VTA) of the same section (corresponds to panels B0 and C0 in Fig. 2). GFAP is enhanced in the vicinity of the asterisk. Panels C, D and E show GFAP immunoreactivity in a section adjacent to that shown in F, which illustrates tyrosine hydroxylase-stained material following injections of 100 ng DOPAL in the SNC and SNR (D) and VTA (C) on the right and 50 ng DOPAL in the same structures on the left. Note that the lesion in the VTA on the right (100 ng) is barely detectable as compared to that in the SNC and SNR. Note also that GFAP is enhanced in the SNC and SNR on the left (50 ng injection) despite an absence of any indication of lesion in the tyrosine hydroxylase-stained material. Additional abbreviation: ml—medial lemniscus. Scale bar: 1.0 mm in F; 200 mm in A–E.

is on the order of 400-fold more toxic than DA, indicates that DOPAL is the more probable mediator of DA toxicity in vivo. The precise concentration of DOPAL in neurons after injection into the SN cannot be determined, but clearly is not the same as the concentration in the pipette. First, the volume of solvent is immediately diluted by extra- and intracellular fluid in and surrounding glia and neurons. Second, DOPAL may be inactivated by enzymatic and nonenzymatic mechanisms prior to reaching neurons. Furthermore, it was observed that DOPAL, but neither DA nor its other metabolites, is toxic in vitro [19]. In addition,

a possible mechanism for this selective toxicity of DOPAL was described in that DOPAL, but neither DA nor its other metabolites, generates a free hydroxyl radical in the presence of H 2 O 2 [25]. Free hydroxyl radical induction of mitochondrial permeability transition has been implicated in DOPAL-mediated death of DA neurons [19,25]. Alternately, the aldehyde structure of DOPAL itself may mediate toxicity [18]. Here, it is shown that DOPAL, applied extraneuronally at high concentrations, can kill all cell types. However, Fornai et al. showed that in vivo, under normal basal conditions, DOPAL is synthesized

W. J. Burke et al. / Brain Research 989 (2003) 205–213

211

Fig. 4. Micrographs illustrating ventral tegmental area (VTA—A, C, E and G) and substantia nigra pars compacta (SNC—B, D, F and H) from rats that received 200 nl injections of buffer containing 500 ng of dopamine (DA—A and B) or its metabolites homovanillic acid (HVA—C and D), dihydroxyphenylacetic acid (DOPAC—E and F) and dihydroxyphenylethanol (DOPET—G and H). Sections were processed to exhibit tyrosine hydroxylase immunoreactivity. Inserts are epifluorescence images illustrating precise locations of infusion sites (arrows) revealed by co-injected rhodamine microspheres. Scale bar: 1.0 mm.

exclusively intraneuronally [11]. As noted above, this intraneuronal locus of synthesis can account for highly selective toxic effects in DA neurons in PD. Stated differently, under normal conditions, DOPAL is not syn-

thesized in, nor selectively taken up by, non DA neurons or glia and so would not reach toxic levels in these brain cells. DOPAL is a major metabolite of DA in human brain

212

W. J. Burke et al. / Brain Research 989 (2003) 205–213

[4,19]. Although the exact concentration required to cause death of human SN DA neurons is unknown, several findings suggest that DOPAL may rise to pathologically significant levels and play an important role in death of these neurons in PD. First, we previously examined DOPAL toxicity in the less complex in vitro model [19], observing that toxic levels of DOPAL in vitro are close to physiological levels found in human autopsy brain such that a threefold increase in DOPAL in human SN would produce a concentration that is toxic in vitro [4,19]. Injection of L-DOPA, a DA and DOPAL precursor commonly used to treat PD, elevates rat brain DOPAL levels by 18-fold [11]. In fact, a growing body of literature and recent imaging studies in PD patients suggests that LDOPA may accelerate loss of DA SN neurons in PD [27] and, thus, the long term progression of PD [8]. The question of whether therapeutic administration of L-DOPA is toxic in the long term remains to be answered [39]. Third, there is a deficit in mitochondrial complex I in PD SN [30]. Our studies show that cell death occurs at near physiological DOPAL concentrations in an in vitro model of DA neurons and that this toxicity is enhanced by inhibiting complex I with rotenone [19]. Fourth, Betarbet et al. [2] have developed a precise model for PD by inhibiting complex I with chronic rotenone injections, in which DA SN neurons are affected to a greater extent than DA VTA neurons [2]. Inhibition of complex I in PC12 cells decreases nicotinamide adenine dinucleotide (NAD) levels which results in inhibition of DOPAL catabolism by NAD-dependent aldehyde dehydrogenase and elevates DOPAL levels 12-fold [21,22]. The differential selective vulnerability for DA neuronal subtypes observed by Betarbet et al. [2] and others has been attributed to a differential expression of the dopamine transporter (DAT) with higher levels in SN compared to VTA neurons [31] and to higher levels of anti-oxidant in VTA neurons [14]. Furthermore, catecholamine-derived aldehydes are actively transported by the catecholamine transporter [6], which could contribute to the selective vulnerability of DA SN neurons to externally applied DOPAL. Interestingly, recent studies suggest that DAT blockers may slow degeneration of DA neurons in PD [9]. Fifth, the rotenone model for PD produces Lewy bodies, a specific pathological marker of PD, which contain a-syn. The accumulation of a-syn filaments was attributed to increased oxidative stress [2,12]. The free hydroxyl radical generated by DOPAL in the presence of H 2 O 2 could lead to oxidative modification of a-syn and its accumulation in DA SN neurons [25]. In fact, a self-regenerating cycle may obtain, insofar as the hydroxyl radical accelerates aggregation of a-syn fibrils into a toxic form [13,23,26,35,40]. Finally, a-syn has been genetically linked to PD pathogenesis [33] but, due to its widespread distribution [28], cannot, by itself, account for the highly selective DA neuron death in PD [31]. However, a-syn fibrils could enhance DOPAL neurotoxicity in several additional ways, including by binding to DAT and

thus enhancing DA uptake [23], permeabilizing storage vesicles [35], and catalyzing the formation of H 2 O 2 [34]. These pathologic functions of a-syn would be expected to increase cytosolic DOPAL levels and DOPAL generated free hydroxyl radicals [25]. As described above, several hypotheses exist to explain DA neuron death in PD. These involve DA [8,10], a-syn [23,33–35,40], free radicals [14,25]; DA transporter [31], complex I deficiency [30] and pesticides containing rotenone [2,12]. However, each of these hypotheses, in and of itself, provides an insufficient basis to precisely explain how neurodegeneration proceeds in PD. The present results, combined with previously reported findings [3,5,11,19,25,40], suggest that DOPAL may be the critical DA neuronal death messenger in PD that links these multiple mechanisms to provide for a unified neurochemical hypothesis of DA neuron loss in PD. This hypothesis may lead to new therapeutic approaches to delay progression of PD [7].

Acknowledgements This study was supported by Missouri Alzheimer’s and Related Disease Program (W.J.B., D.S.Z.), Souers Stroke Institute grant (W.J.B.), Saint Louis University Medical Center grant (W.J.B.), and NIH grant, NS 23805 (D.S.Z.).

References [1] D.A. Bennett, L.A. Beckett, A.M. Murray, K.M. Shannon, C. Goetz, D.M. Pilgrim, D.A. Evans, Prevalence of Parkinsonian signs and associated mortality in a community population of older people, New Engl. J. Med. 334 (1996) 71–76. [2] R. Betarbet, T.B. Sherer, G. MacKenzie, M. Garcia-Osona, A.V. Panov, J.T. Greenamyre, Chronic systemic pesticide exposure reproduces features of Parkinson’s disease, Nat. Neurosci. 3 (2000) 1302–1306. [3] H. Blaschko, Amine oxidase and amine metabolism, Pharmacol. Rev. 4 (1952) 415–453. [4] W.J. Burke, H.D. Chung, S.W. Li, Quantitation of 3,4-dihydroxyphenylacetaldehyde and 3,4-dihydroxyphenylglycolaldehyde, the monoamine oxidase metabolites of dopamine and noradrenaline, in human tissues by microcolumn high-performance liquid chromatography, Anal. Biochem. 273 (1999) 111–116. [5] W.J. Burke, S.W. Li, C.A. Schmitt, D.S. Zahm, H.D. Chung, A.D. Conway, P. Lampe, E.M. Johnson, T.-S. Lin, B.S. Kristal, J. Barg, M. Anwar, D.A. Ruggiero, Catecholamine-derived aldehyde neurotoxins, in: M.A. Collins, A. Storch (Eds.), Neurotoxic Factors in Parkinson’s Disease, Plenum Press, New York, 2000, pp. 167–180. [6] W.J. Burke, S.W. Li, D.S. Zahm, H. MacArthur, L.L. Kolo, T.C. Westfall, M. Anwar, S.B. Glickstein, D.A. Ruggiero, Catecholamine monoamine oxidase A metabolite in adrenergic neurons is cytotoxic in vivo, Brain Res. 891 (2001) 218–277. [7] W.J. Burke, 3.4-Dihydroxyphenylacetaldehyde: a potential target for neuroprotective therapy in Parkinson’s disease, Current Drug Targets—CNS Neurological Disorders, in press. [8] S. Fahn, Levodopa-induced neurotoxicity, CNS Drugs 8 (1997) 376–393.

W. J. Burke et al. / Brain Research 989 (2003) 205–213 [9] B.H. Falkenburger, K.L. Barstow, I.M. Mintz, Dendrodendritic inhibition through reversal of dopamine transport, Science 293 (2001) 2465–2470. [10] F. Filloux, J.J. Townsend, Pre- and postsynaptic neurotoxic effects of dopamine demonstrated by intrastitial injection, Exp. Neurol. 119 (1993) 79–88. [11] F. Fornai, F.S. Giorgi, L.L. Bassi, M. Ferrucci, M.G. Alessandri, G.U. Corsini, Modulation of dihydroxyphenylacetaldehyde extracellular levels in vivo in the rat striatum after different kinds of pharmacological treatment, Brain Res. 861 (2000) 126–134. [12] B.I. Giasson, V.M.-Y. Lee, A new link between pesticides and Parkinson’s disease, Nat. Neurosci. 3 (2000) 1227–1228. [13] M. Hashimoto, L.J. Hsu, Y. Xia, A. Takeda, A. Sisk, M. Sundsmo, M. Eliezer, Oxidative stress induces amyloid-like aggregate formation of NACP/ a-synuclein in vitro, Neuroreport 10 (1999) 717–721. [14] E.C. Hirsch, B. Facheux, P. Damier, A. Mouatt-Prigett, Y. Agid, Neuronal vulnerability in Parkinson’s disease, J. Neural Transm. 50 (1997) 79–88. [15] W.C. Koller, How accurately can Parkinson’s disease be diagnosed, Neurology 42 (Suppl. 1) (1992) 6–16. [16] W.C. Koller, When does Parkinson’s disease begin, Neurology 42 (Suppl. 4) (1992) 27–31. [17] W.C. Koller, An algorithm for the management of Parkinson’s disease, Neurology 42 (Suppl. 10) (1994) 1–52. [18] B.S. Kristal, B.K. Park, B.P. Yu, 4-Hydroxyenal is a potent inducer of the mitochondrial permeability transition, J. Biol. Chem. 271 (1996) 6033–6038. [19] B.S. Kristal, A.D. Conway, A.M. Brown, J.C. Jain, P.A. Ulluci, S.W. Li, W.J. Burke, Selective dopaminergic vulnerability: 3,4-dihydroxyphenylacetaldehyde targets mitochondria, Free Radic. Biol. Med. 30 (2000) 924–931. [20] L.T. Kurland, Epidemiology, incidence, geographic distribution and genetic considerations, in: W.S. Fields (Ed.), Pathogenesis and Treatment of Parkinsonism, Charles C. Thomas, Springfield, IL, 1958, pp. 5–49. [21] I. Lamensdorf, G. Eisenhofer, J. Harvey-White, Y. Hayakawa, K. Kirk, I.J. Kopin, Metabolic stress in PC 12 cells induces the formation of the endogenous dopaminergic neurotoxin, 3,4dihydroxyphenylacetaldehyde, J. Neurosci. Res. 60 (2000) 552–558. [22] I. Lamensdorf, G. Eisenhofer, J. Harvey-White, A. Neckustan, K. Kirk, I.J. Kopin, 3,4-Dihydroxyphenylacetaldehyde potentiates the toxic effects of metabolic stress in PC 12 cells, Brain Res. 868 (2000) 191–201. [23] F.J.S. Lee, F. Liu, Z.B. Pristupa, H.B. Niznik, Direct binding and functional coupling of a-synuclein to dopamine transporters accelerate dopamine-induced apoptosis, FASEB J. 15 (2001) 916–926. [24] S.W. Li, V.T. Spaziano, W.J. Burke, Synthesis of a biochemically important aldehyde, 3,4-dihydroxyphenylacetaldehyde, Bioorg. Chem. 26 (1998) 45–50. [25] S.W. Li, T.-S. Lin, S. Minteer, W.J. Burke, 3,4-Dihydroxyphenylacetaldehyde and hydrogen peroxide generate a hydroxyl radical: possible role in Parkinson’s disease pathogenesis, Mol. Brain Res. 93 (2001) 1–7. [26] N. Osterova-Golts, L. Petrucelli, J. Hardy, J.M. Lee, M. Farer, B. Wolozin, The A53T a-synuclein mutation increases iron-dependent aggregation and toxicity, J. Neurosci. 20 (2000) 6048–6054. [27] Parkinson Study Group, Dopamine transporter brain imaging to assess the effects of pramipexole vs. levodopa on Parkinson disease progression, J. Am. Med. Assoc. 287 (2002) 1653–1661.

213

[28] Y.S. Piao, K. Wakabayashi, S. Hayashi, M. Yoshimoto, H. Takahashi, Aggregation of a-synuclein / NACP in neuronal and glial cells in diffuse Lewy body disease: a survey of six patients, Clin. Neuropathol. 19 (2002) 163–169. [29] M.J. Polymeropoulos, C. Lavendan, E. Leroy, S.E. Ide, A. Dehaja, A. Dutra, B. Pike, H. Root, J. Rubenstein, E.S. Stenroos, S. Chandraskharappa, H. Athanassiadou, T. Papapetropoulos, W.G. Johnson, A.M. Lazzarini, R.C. Duvoisin, G. DiIorio, L.I. Golbe, R.L. Nussbaum, Mutation in the alpha-synuclein gene identified in families with Parkinson’s disease, Science 275 (1997) 2045–2047. [30] A.H.V. Schapira, J.M. Cooper, D. Dexter, J.B. Clark, P. Jenner, C.D. Marsden, Mitochondrial complex I deficiency in Parkinson’s disease, J. Neurochem. 54 (1990) 823–827. [31] A. Storch, J. Schwarz, The dopamine transporter: involvement in selective dopaminergic neurodegeneration, in: M.A. Collins, A. Storch (Eds.), Neurotoxic Factors in Parkinson’s Disease, Plenum Press, New York, 2000, pp. 17–40. [32] Y. Tan, E.A. Williams, A.J. Lancia, D.S. Zahm, On the altered expression of tyrosine hydroxylase and calbindin-D28KD immunoreactivities and viability of neurons in the ventral tegmental area of Tsai following injections of 6-hydroxydopamine in the medial forebrain bundle in the rat, Brain Res. 869 (2000) 56–68. [33] J.Q. Trojanowski, V.M.-Y. Lee, Parkinson’s disease and related neurodegenerative synucleinopathies linked to progressive accumulations of synuclein aggregates in brain, Parkinsonism Rel. Disord. 7 (2001) 247–251. [34] S. Turnbull, B.J. Tabner, O.M. El-Agnaf, S. Moore, Y. Davies, D. Allsop, Alpha-synuclein implicated Parkinson’s disease catalyzed the formation of hydrogen peroxide in vitro, Free Radic. Biol. Med. 30 (2001) 1163–1170. [35] M.J. Volles, S.-J. Lee, J.-C. Rochet, M.D. Shtilerman, T.J. Ding, J.C. Kessler, P.J. Lansbury, Vesicle permeabilization by protofibrillar a-synuclein: implications for the pathogenesis and treatment of Parkinson’s disease, Biochemistry 40 (2001) 7812–7819. [36] J.X. Wilson, Antioxidant defense of the brain: a role for astrocytes, Can. J. Physiol. Pharmacol. 75 (1997) 1149–1163. [37] P. Winn, H.L. Alderson, V.J. Brown, M.P. Latimer, K. Payne, G. Thiemann, E.A. Williams, D.S. Zahm, Examination of the effects of excitotoxic disconnection lesions of nucleus acucmbens core and shell, Soc. Neurosci. Abstr. 27 (2002) 422.11. [38] H.K. Wolf, R. Buslei, R. Schmidt-Kastner, T. Pietsch, D.D. Wiestler, I. Blumke, NeuN: a useful marker for diagnostic histopathology, J. Histochem. Cytochem. 44 (1996) 1167–1171. [39] F. Wooten, Agonist vs. levodopa in PD: the thrilla of witha, Neurology 60 (2003) 360–362. [40] J. Xu, S.-Y. Kao, F.J.S. Lee, W. Song, L.-W. Jin, B.A. Yankner, Dopamine-dependent neurotoxicity of a-synuclein: a mechanism for selective neurodegeneration in Parkinson’s disease, Nat. Med. 8 (2002) 600–606. [41] D.S. Zahm, S. Grosu, E.A. Williams, S. Qin, A. Berod, Neurons of origin of the neurotensinergic plexus enmeshing the ventral tegmental area in rat: retrograde labeling and in situ hybridization combined, Neuroscience 104 (2001) 841–851. [42] S.S. Zhou, S.M. Gospe, Double labeling of proliferating neurons with anti-Brdu and anti-NeuN: an improved immunohistochemical technique utilizing microwave irradiation, J. Histotechnol. 21 (1998) 201–204.