Preferential Extracellular Generation of the Active Parkinsonian Toxin MPP+ by Transporter-Independent Export of the Intermediate MPDP+

Article abstract of DOI:10.1089/ars.2015.6297

Aims: 1-Methyl-4-phenyl-tetrahydropyridine (MPTP) is among the most widely used neurotoxins for inducing experimental parkinsonism. MPTP causes parkinsonian symptoms in mice, primates, and humans by killing a subpopulation of dopaminergic neurons. Extrapolations of data obtained using MPTP-based parkinsonism models to human disease are common; however, the precise mechanism by which MPTP is converted into its active neurotoxic metabolite, 1-methyl-4-phenyl-pyridinium (MPP⁺), has not been fully elucidated. In this study, we aimed to address two unanswered questions related to MPTP toxicology: (1) Why are MPTP-converting astrocytes largely spared from toxicity? (2) How does MPP⁺ reach the extracellular space? Results: In MPTP-treated astrocytes, we discovered that the membrane-impermeable MPP⁺, which is generally assumed to be formed inside astrocytes, is almost exclusively detected outside of these cells. Instead of a transporter-mediated export, we found that the intermediate, 1-methyl-4-phenyl-2,3-dihydropyridinium (MPDP⁺), and/or its uncharged conjugate base passively diffused across cell membranes and that MPP⁺ was formed predominately by the extracellular oxidation of MPDP⁺ into MPP⁺. This nonenzymatic extracellular conversion of MPDP⁺ was promoted by O₂, a more alkaline pH, and dopamine autoxidation products. Innovation and Conclusion: Our data indicate that MPTP metabolism is compartmentalized between intracellular and extracellular environments, explain the absence of toxicity in MPTP-converting astrocytes, and provide a rationale for the preferential formation of MPP⁺ in the extracellular space. The mechanism of transporter-independent extracellular MPP⁺ formation described here indicates that extracellular genesis of MPP⁺ from MPDP is a necessary prerequisite for the selective uptake of this toxin by catecholaminergic neurons. Antioxid. Redox Signal. 23, 1001-1016.

Full text of DOI:10.1089/ars.2015.6297

1
Page 1 of 57
Original Research Communication
Preferential extracellular generation of the active parkinsonian
+
toxin MPP by transporter-independent export of the
+
intermediate MPDP
1
*
1*
2
1
Stefan Schildknecht, Regina Pape, Johannes Meiser, Christiaan Karreman, Tobias
3
3
3
3
4
Strittmatter, Meike Odermatt, Erica Cirri, Anke Friemel, Markus Ringwald, Noemi
5
5
6
3
7
Pasquarelli, Boris Ferger, Thomas Brunner, Andreas Marx, Heiko M. Möller,
2
1
Karsten Hiller, Marcel Leist
1
In vitro Toxicology and Biomedicine, Department of Biology, University of Konstanz, D-
8457 Konstanz, Germany;
Metabolomics Junior Research Group, Luxembourg Centre for Systems Biomedicine,
University of Luxembourg, L-4362 Esch-Belval, Luxembourg;
7
2
3
Department of Chemistry and Konstanz Research School Chemical Biology, University of
Konstanz, D-78457 Konstanz, Germany;
4
MCAT GmbH, Hermann-von-Vicari-Str. 23, D-78464 Konstanz, Germany;
5
CNS Disease Research, Boehringer Ingelheim Pharma GmbH  Co. KG,
D-88397 Biberach an der Riss, Germany;
6
Biochemical Pharmacology, Department of Biology, University of Konstanz, D-78457
Konstanz, Germany;
Institute of Chemistry, University of Potsdam, D-14476 Potsdam, Germany
7
*
These authors contributed equally
+
Running head: MPDP export from astrocytes
Word count:
Reference numbers: 56
Number of greyscale illustrations: 4
Number of color illustrations: 2
To whom correspondence should be addressed:
Stefan Schildknecht, PhD
University of Konstanz
Department of Biology
In vitro Toxicology and Biomedicine
PO Box M657
Universitätsstr. 10
7
8457 Konstanz, Germany
Fax: +49-7531-88-5039
Tel: +49 7531-88-5053
E-mail: Stefan.Schildknecht@uni-konstanz.de
+
+
Key words: astrocytes, autoxidation, parkinsons disease, MPDP , MPP
1

2
Page 2 of 57
Abstract
Aims: 1-methy-4-phenyl-tetrahydropyridine (MPTP) is among the most widely used
neurotoxins for inducing experimental Parkinsonism. MPTP causes Parkinsonian symptoms
in mice, primates, and humans by killing a subpopulation of dopaminergic neurons.
Extrapolations of data obtained using MPTP-based Parkinsonism models to human disease
are common; however, the precise mechanism by which MPTP is converted into its active
+
neurotoxic metabolite 1-methyl-4-phenyl-pyridinium (MPP ) has not been fully elucidated. In
this study, we aimed to address two unanswered questions related to MPTP toxicology: 1)
+
why are MPTP-converting astrocytes largely spared from toxicity; and 2) how does MPP
reach the extracellular space? Results: In MPTP treated astrocytes, we discovered that the
+
membrane-impermeable MPP , which is generally assumed to be formed inside astrocytes, is
almost exclusively detected outside of these cells. Instead of a transporter-mediated export,
+
we found that the intermediate MPDP and/or its uncharged conjugate base passively diffused
+
across cell membranes and that MPP was formed predominately by the extracellular
+
+
+
oxidation of MPDP into MPP . This non-enzymatic extracellular conversion of MPDP was
promoted by O , a more alkaline pH, and dopamine autoxidation products. Innovation and
2
Conclusion: Our data indicate that MPTP metabolism is compartmentalized between
intracellular and extracellular environments, explain the absence of toxicity in MPTP-
+
converting astrocytes, and provide a rationale for the preferential formation of MPP in the
+
extracellular space. The mechanism of transporter-independent extracellular MPP formation
+
described here indicates that extracellular genesis of MPP from MPDP is a necessary
prerequisite for the selective uptake of this toxin by catecholaminergic neurons.
2

3
Page 3 of 57
Introduction
The experimental study of dopamine (DA) neuron degeneration (a hallmark of Parkinson’s
disease) relies on toxicants that specifically induce pathological states reminiscent of human
Parkinsonism by inducing the selective death of DA neurons. Among the chemical tools
currently used to induce DA neuron degeneration, MPTP (1-methyl-4-phenyl-1,2,3,6-
tetrahydropyridine) most closely reproduces human Parkinsonian pathology in mice and
primates (3,31). MPTP is a lipophilic and non-cytotoxic prodrug that crosses the blood brain
+
barrier and is then converted into the active toxicant MPP (1-methyl-4-phenylpyridinium)
+
(
23,32,54). Once MPP reaches a critical concentration in the cerebrospinal fluid, it acts as a
molecular ‘Trojan Horse’ in DA neurons as result of its selective uptake by the DA
+
transporter (DAT) (26). Unlike DA, MPP accumulates in mitochondria, where it acts as an
inhibitor of complex I of the respiratory chain (38), and thereby kills its ‘host cell’ (10).
However, a vexing enigma pertaining to the neurotoxic mechanisms of MPTP remains
+
unsolved. MPTP is converted to MPP in astrocytes by the enzyme monoamine oxidase-B
(MAO-B), which is localized on the outer mitochondrial membrane of astrocytes (44). How
+
MPP exits astrocytes rather than accumulating in their mitochondria is unknown.
+
Due to its charge, the active toxicant MPP cannot cross membranes. Recent reports have
suggested that passive transporters, such as the family of organic cation transporters (OCT) or
+
the plasma membrane monoamine transporter (PMAT), are necessary for MPP -efflux from
astrocytes (9,37). OCT3 was observed by Cui et al. to be preferentially expressed in glial cells
in the vicinity of midbrain DA neurons that degenerated in response to MPTP treatment, and
+
it was hypothesized that a preferential export of intracellular astrocytic MPP via OCT3 might
contribute to the loss of adjacent DA neurons (9). However, a recent study clearly
+
demonstrated the global formation of MPP in the whole brain only minutes after MPTP
administration, and a subsequent rapid clearance from most regions (28). This indicates that
3

4
Page 4 of 57
+
glial cells devoid of OCT3 expression must also contribute to MPTP conversion and MPP
release. Moreover, conventions in membrane biophysics suggest that passive transporters like
+
OCT3 would contribute to an uptake of MPP from the cerebrospinal fluid into cells (49). A
plasma membrane potential of -70 mV would drive a 15-fold increase in the accumulation of
a positively charged molecule inside cells according to the Nernst equation. Hence, the
presence of OCT transporters in glial cells would rather lead to a reduction of extracellular
+
MPP and its toxicity in neurons as it was observed with monovalent paraquat (42).
+
MPTP can form several metabolites other than MPP in the brain (52), and their biochemical
and physical properties need to be taken into account in order to understand the possible
+
+
mechanisms of MPP transport across the membrane. The first metabolic step in MPP
formation is always an MAO-dependent two-electron oxidation yielding the unstable
+
+
intermediate MPDP (5). A non-enzymatic autoxidation of MPDP has been suggested as the
+
most likely second step of MPP formation (30). In addition, a disproportionation reaction
+
+
yielding one molecule of MPTP and MPP from two molecules of MPDP has been described
+
(
4,39). However, this reaction takes place only at concentrations of MPDP (millimolar range)
+
unlikely to occur in biological systems. The complex chemistry of MPDP involves further
+
spontaneous reactions; as an α,β-unsaturated iminium ion, MPDP can exist in tautomeric
forms and/or in its conjugate base form, 1-methy-4-phenyl-1,2-dihydropyridine (1,2-MPDP)
(4,5). The latter non-charged metabolite is capable of crossing biological membranes.
In the present work, we report two main findings: 1) the MAO-catalyzed MPTP intermediate
+
metabolite MPDP /1.2-MPDP undergoes transporter-independent export from the cell, and 2)
+
+
MPDP /1.2-MPDP is responsible for non-enzymatic MPP formation in the extracellular
+
space. These mechanisms explain why MPTP-converting glial cells are spared from MPP
+
toxicity and how extracellular MPP is generated, which is a necessary prerequisite for the
+
accumulation of MPP in catecholaminergic cells.
4

5
Page 5 of 57
Results
+
Extracellular MPP generation by MPTP-converting astrocytes predominates.
+
We studied the conversion of MPTP to MPP by astrocytes in the mouse astrocytic cell line
IMA 2.1, which displays MPTP metabolism properties comparable to primary astrocytes (Fig.
+
1
A,B (45)). We observed an accumulation of the biologically active toxicant MPP in the
IMA culture medium, but not within the cells (Fig. 1C, D). Addition of the OCT3 inhibitor
+
D22 to the culture medium had no influence on MPP release or uptake (data not shown),
suggesting that the previously suggested OCT3 transport mechanism (9) was not obligatory
+
for extracellular MPTP/MPP accumulation. However, the MPTP converting enzyme
monoamine oxidase-B (MAO-B) is located exclusively within the astrocytic cell body, and
+
MPP , as a charged molecule, cannot penetrate biological membranes. This raises the
+
question: how can MPP be efficiently generated extracellularly in the absence of a system for
its transport from the cell (Fig. 1C)?
To resolve this issue, we generated IMA sub-clones that stably express OCT3 (a passive
+
+
transporter of MPP ), DAT (accumulates MPP against a concentration gradient), or both
DAT and OCT3; these cells lines are referred to as OCT3-IMA, DAT-IMA, and OCT3/DAT-
IMA, respectively. Expression levels of OCT3 in the OCT3-IMA and OCT3/DAT-IMA lines
were identical (Suppl. Fig. S1A), and the transport capacity and pharmacological features
were comparable to those of the human kidney carcinoma cell line Caki-1 which
+
endogenously expresses OCT3 (22). We next assessed MPP transport in these cell lines. In
+
DAT-IMAs, uptake of MPP was prevented by the DAT-blocker GBR12909, while in OCT3-
+
IMAs MPP uptake was prevented by the OCT3 blocker D22 (Suppl. Fig. S1A-D).
+
We next assessed the response of wild type and the transporter-expressing cells lines to MPP
+
treatment. All cells were exposed to a fixed concentration (20 µM) of extracellular MPP
(Fig. 1E). The toxicant did not enter wild type IMAs, which was as expected given the
5

6
Page 6 of 57
+
absence of functional carriers in the plasma membrane in these cells and the inability of MPP
to cross lipid bilayers. Strong intracellular accumulation was however observed in DAT-
IMAs. Ectopic OCT-3 expression in OCT3-IMAs and OCT3/DAT-IMAs enabled influx of
+
+
MPP into cells. In OCT-3/DAT-IMAs, MPP efflux from cells that had accumulated very
+
high concentrations of MPP due to DAT activity was observed (Fig. 1E). Steady state levels
+
were reached within 2 h. At this time point, the absolute amount of accumulated MPP was
measured, and the cellular volume of IMA was determined (2.5 pl/cell) so that average
+
intracellular MPP concentrations could be calculated (Fig. 1F). The transporter-expressing
cells reached levels of around 180-680 µM, which is highly similar to tissue concentrations
found in mice exposed to MPTP (19). Within 24 h, these concentrations were sufficient to
diminish the cellular capacity to reduce resazurin or to produce ATP (Suppl. Fig. S2). By
+
contrast, IMAs neither accumulated MPP nor displayed impaired viability in response to up
+
+
to a 120 µM concentration of extracellular MPP . Indeed, the degree of MPP -induced
reductions in viability across the four cell lines correlated with their respective capacity to
+
accumulate MPP (Fig. 1G). These data support the notion that astrocytes are not resistant to
+
+
MPP toxicity, as reduced viability was observed as soon MPP accumulated within the
+
transporter-expressing cells. Our findings also confirm that MPP cannot cross the
+
membranes of IMAs unless they ectopically express MPP transporters. This led us to posit
+
that MPP crossed the IMA membrane through an OCT-3 independent mechanism during the
course of MPTP metabolism.
+
MPDP is a membrane-crossing intermediate generated from MPTP by MAO-B.
Since the pioneering work of Castagnoli and colleagues, it has been known that MAO-B
+
+
converts MPTP to intermediates such as MPDP , and that further oxidation to MPP proceeds
non-enzymatically (4,30,39). Some of these initial publications speculated that these
intermediates might play a role in transport processes, but further investigation of their
6

7
Page 7 of 57
biological formation and properties over the last 30 years has been highly limited. We thus
+
focused here on the transport of MPDP as a potential missing link that might explain
+
extracellular MPP formation. Analysis of the products formed in the conversion of MPTP by
+
purified MAO-B enzyme confirmed that MPDP is a major intermediate in the generation of
+
MPP (Fig. 2A). Similar observations were made in analysis of wildtype IMA and OCT3-
+
+
IMA cultures treated with MPTP, as MPDP was detected in the supernatant before MPP .
+
We further found that MPDP export was not dependent on OCT3 (Fig. 2B). To determine if
+
MPDP formation in this in vitro model also occurs in vivo, mice were injected with MPTP
and the striatum and cerebellum was collected from injected mice at different time intervals.
+
These experiments revealed a rapid (30 min) rise of MPDP in both brain areas that preceded
+
+
the formation of MPP . MPP remained elevated in the striatum, but not the cerebellum (Fig.
+
2
C). C-5 carbons of tetrahydropyridine structures such as MPDP are acidic (i.e. they possess
a α,β-unsaturated iminium ion structure). However, their conjugate free base form 1,2-MPDP
was also observed (4,30,39) (Fig. 2D) and this molecule is expected to be able to freely
diffuse across biological membranes. To more carefully examine the formation of
+
intermediates that are generated in the MAO-dependent conversion of MPTP to MPP , we
1
conducted time-resolved H-NMR analysis (Suppl. Fig. S3). These data confirmed a
+
significant time lag between the enzymatic formation of MPDP and its subsequent
+
autoxidation to MPP . Based on these findings, we hypothesized that the acid-base pair
+
+
MPDP /1.2-MPDP serves as membrane-permeable intermediate in MPP production that is
+
exported from astrocytes, and would thereby explain the presence of extracellular of MPP .
+
Free bidirectional flux of MPDP /1.2-MPDP across biological membranes.
+
In order to test the hypothesis that MPDP /1.2-MPDP possesses transporter-independent
+
+
membrane permeability, IMAs were treated with 50 µM fresh MPDP or decomposed MPDP
7

8
Page 8 of 57
(stored in an open tube at room temperature prior to use). Both the resazurin reduction assay
+
and measurements of intracellular ATP (Fig. 2E) indicated that only fresh MPDP led to cell
damage. For an appropriate interpretation of the data, it is important to note that the
extracellular volume in these experimental setups is substantially greater than the intracellular
volume. Treatment of the cells with MPTP hence results in a constant and steep concentration
+
gradient of freshly generated MPDP that supports its efflux into the extracellular space. In
+
contrast, extracellular addition of MPDP generates conditions in which the intracellular
volume is rapidly saturated (i.e. intracellular volume <<< extracellular volume).
+
To confirm a direct influence of intracellular MPP on mitochondria, metabolic flux
+
measurements were obtained as a more sensitive readout. MPP specifically inhibits complex
I of the electron transport chain. As a result, NADH cannot be oxidized via complex I and
+
accumulates in the cell. An increase in the NADH/NAD ratio results in an inhibition of
+
NAD -dependent reactions of the TCA cycle, such as pyruvate dehydrogenase activity. In the
case of complex I inhibition, glucose carbon contribution to the TCA cycle strongly
decreases, while pyruvate reduction to lactate increases. By increasing the rate of glycolysis,
the cell compensates for the lack of ATP, similar to the Warburg effect. To assess metabolic
1
3
flux, we labeled cells with the tracer [U- C ] glucose, which can be used to monitor relative
6
+
pyruvate oxidation that results from the NAD -dependent pyruvate dehydrogenase reaction
decarboxylating pyruvate in the mitochondrion during the production of acetyl-CoA, (Fig.
2
F). The activity of pyruvate dehydrogenase and of ATP dependent pyruvate carboxylase,
+
+
were both inhibited by MPDP , but not by MPTP or MPP (Fig. 2F, Suppl. Fig. S4).
+
+
+
To directly compare MPP and MPDP uptake, intracellular MPP levels in IMAs exposed to
+
+
either MPP or MPDP were investigated. To exclude any potential involvement of cellular
transporter systems, all experiments were conducted in parallel with artificial lipid vesicles
(Fig. 3A). As a third model system, purified human erythrocytes were used as they possess no
8

9
Page 9 of 57
+
+
+
MPP transporter activity, but do allow the conversion of MPDP to MPP due to the
presence of iron-containing hemoglobin. In all three model systems, extracellular
+
concentrations of MPP as high as 200 µM did not result in significant accumulation of
+
+
intracellular MPP , while the addition of MPDP immediately led to the formation of
+
+
+
intracellular/intravesicular MPP . Uptake of MPDP and MPP into erythrocytes was then
tested in vivo by intravenous injection of these compounds into mice and collection of blood
+
after 3 min. Similar to the ex vivo treatment of erythrocytes, the addition of MPDP resulted
+
in the intracellular accumulation of MPP , providing further evidence of transporter-
+
independent membrane passage of MPDP (Suppl. Fig. S5).
+
To further test the kinetics of cell loading and unloading of MPDP , IMAs were exposed to
+
MPDP (50 µM) for a 30 min loading phase. Analysis of the cellular compartment showed a
+
+
+
rapid increase of MPDP , followed by a rise in MPP concentrations. MPP was apparently
+
generated within these cells, as extracellular MPP cannot cross the cell membrane (Fig. 3B,
Left). After the loading phase, cells were washed, new medium was added, and metabolite
levels were detected over time in both the intracellular (Fig. 3B, Middle) and extracellular
+
compartments (Fig. 3B, Right). Accumulated MPDP vanished rapidly from the cells into the
+
extracellular media, whereas intracellular MPP was not able to exit the cells (Fig. 3B, Middle
+
and Right). Considering these data together, we concluded that extracellular MPP originated
+
from the autoxidation of extracellular MPDP . In addition, these findings indicate that the
+
MPDP concentration gradient is the predominant driving force underlying its transport across
+
membranes, which explains the apparent export of MPP . These observations indicate that
+
+
MPDP /1.2-MPDP is membrane permeable and suggest that MPP will remain in a given
biological compartment, unless transporters are present.
9

1
0
Page 10 of 57
+
The role of extracellular MPDP autoxidation in dopamine neuron toxicity.
+
Our data here-to-fore have suggested that upon MPTP intoxication, MPDP /1.2-MPDP would
have the potential to freely diffuse across the membranes of several cells, and eventually
+
+
convert to MPP in any cell type adjacent to MAO-containing glial cells. However, MPP
accumulation is largely limited to catecholaminergic neurons that take up the molecule via
their catecholamine transporter (e.g. DAT). To compare the respective contribution of DAT-
+
+
mediated MPP accumulation and passive MPDP /1.2-MPDP-dependent influx (Fig. 4A),
+
human dopaminergic neuronal cells (LUHMES), which can accumulate extracellular MPP
+
+
via DAT (46,47,48) were treated with MPP or MPDP (20 µM) in the presence or absence of
+
the DAT blocker GBR12909 (GBR). This compound blocks the uptake of MPP via DAT
+
inhibition, but not MPDP /1.2-MPDP as it is not a DAT substrate. Metabolic flux was
1
3
measured with labeled [U- C ] glucose, which provides a readout of the inhibition of
6
+
+
oxidative phosphorylation. Both MPP and fresh MPDP led to mitochondrial inhibition, and
GBR blocked toxicity in both cases (Fig. 4B). These observations were confirmed by cellular
staining with an anti-β-III-tubulin antibody, which demonstrated that pronounced neurite
+
+
degeneration was triggered by both MPP and fresh MPDP , and these effects were
completely prevented by GBR (Fig. 4C). Taken together these data suggest that at the
+
moderate MPDP /1.2-MPDP concentrations (low micromolar range) that exist in vivo,
+
intracellular autoxidation of MPDP would not be sufficient to generate damaging
+
intracellular MPP concentrations.
+
To explore whether toxic MPP can be formed within neurons from freely diffusing and
+
+
autoxidizing MPDP , LUHMES cells were exposed to high (200 µM) or low (20 µM) MPDP
concentrations in the presence or absence of the DAT blocker GBR. High extracellular
+
+
MPDP levels led to a DAT-independent accumulation of intracellular MPP doses sufficient
to inhibit mitochondria, whereas low concentrations did not (Fig. 4D). Thus, toxicity evoked
1
0

1
1
Page 11 of 57
+
by free diffusion of freshly generated MPDP in cells adjacent to astrocytes may play only a
marginal role. To test this assumption, a co-culture model containing IMA (astrocytes) and
LUHMES (neurons) was exposed to MPTP. This led to neurotoxicity via glial-dependent
+
generation of MPP . The addition of GBR completely inhibited neurotoxicity in this co-
+
culture system. These data confirm that the passive diffusion of freshly generated MPDP /1.2-
MPDP itself from glia immediately adjacent to LUHMES was not able to evoke significant
toxicity (Suppl. Fig. S6).
+
Identification of physiological factors that promote the autoxidation of MPDP in the
extracellular space
+
The findings described thus far indicate that MPP needs to be formed extracellularly in order
to become a substrate for DAT, and thus induce neurotoxicity. We next explored how a
+
preferential extracellular generation of MPP might proceed in structures such as the brain
with a high ratio between the intra- and the extracellular volume. As oxygen levels are likely
+
to play a role in MPDP autoxidation, we studied this environmental factor. High-resolution
measurements of oxygen concentrations in cells have provided evidence for steep oxygen
concentration gradients between the cytosol and mitochondria (27,36). This may result in
average intracellular oxygen concentrations that are significantly lower than the interstitial
+
levels in a tissue or cell culture media (17,51). We found that the speed of MPP formation
+
strongly depended on oxygen levels, and that MPDP is highly stable under low oxygen
conditions such as those that are typically found near the respiring mitochondria (27,36,51) of
astrocytes (Fig. 5A, B). Lower intracellular oxygen levels would hence stabilize intracellular
+
MPDP while at the higher extracellular oxygen tension, the non-enzymatic conversion into
+
MPP would be favored.
+
We posited that another important factor in determining the rate of MPDP autoxidation
might be the local pH. The cytosolic pH of most cells in the body is typically lower than in
1
1

1
2
Page 12 of 57
extracellular fluids and the blood. Local acidification is largely driven by respiratory chain
proton pump activity, particularly around mitochondria. We observed an accelerated
+
autoxidation of MPDP /1.2-MPDP under more alkaline conditions (Fig. 5C, D), while more
+
acidic conditions (e.g. close to mitochondria) stabilized MPDP . The combination of a more
acidic pH and lower oxygen tensions (e.g. the conditions typically encountered inside cells)
+
strongly prevented MPDP autoxidation (Fig. 5 E, F).
Based on these findings, we hypothesized that, in addition to DAT activity, the conditions in
brain regions enriched in dopaminergic neurons might promote greater autoxidation than in
other areas. The nigrostriatal system of the basal ganglia has a particularly high iron content,
and also contains various autoxidation products of dopamine (12,35). All of these compounds
+
promote the formation of radicals, which is required for autoxidation of MPDP . While free
radicals have very short half-lives inside cells due to several efficient free-radical buffering
mechanisms (e.g. glutathione system, iron buffering), they can persist for longer times
extracellularly. Taking this into consideration, several candidates were tested for their
+
+
+
influence on MPDP autoxidation. A striking acceleration of MPP genesis from MPDP was
observed in response to the treatment with autoxidized dopamine, melanin, iron containing
ferritin, or with free ferrous iron (Fig. 5G, H). Addition of the iron-chelator deferoxamine
+
prevented the iron-mediated acceleration of MPDP autoxidation (Suppl. Fig. S7). This is in
accordance with reports indicating a protective role of iron chelators in MPTP PD models
(11,29), although these reports did not demonstrate whether the conversion of MPTP was
affected by treatment with iron chelators or other oxidative stress related events. Enhanced
+
autoxidation in the extracellular milieu would not only explain preferential MPP generation
outside cells, but also ensures there is a constant, steep concentration gradient between
+
+
cytosolic MPDP and extracellular MPDP , which is the main driving force for its transporter-
independent efflux (Fig. 6).
1
2

1
3
Page 13 of 57
Discussion
Among the known human neurotoxicants, MPTP is exceptional in that it triggers relatively
selective death of nigrostriatal DA neurons upon systemic administration (3,25,31). A
+
necessary prerequisite for the selective DAT-dependent uptake of the active toxin MPP is the
+
presence of MPP in the extracellular space, as was elegantly confirmed by several
+
microdialysis studies (9,21,33). The mechanism by which MPP , a membrane-impermeable
molecule generated intracellularly by the enzymatic conversion of MPTP, reaches the
extracellular space has been elusive. The present work aimed at addressing this question, and
+
was inspired by the observation of efficient extracellular MPP generation in MPTP-
+
converting astrocytes in absence of significant intracellular MPP accumulation and toxicity.
+
We first determined that astrocytes (IMAs) were not immune to MPP , as these cells died
+
when cytosolic accumulation of extracellular MPP was enabled by the ectopic expression of
+
the MPP transporters DAT and/or OCT3. Although it can be concluded from these results
+
that astrocytes are not per se resistant to MPP , they do have the capacity to accelerate
glycolytic turnover and hence are able to at least partially compensate for inhibited
+
mitochondrial ATP synthesis caused by MPP . This conclusion is corroborated by several
experiments in the literature showing that any cell tested thus far, e.g. HEK293, hepatocytes,
+
COS, or HeLa cell, is killed by MPP if the toxicant is allowed to enter the cytosol.
Moreover, the findings of the present study provide a rationale for why cells that metabolize
MPTP with endogenous MAO, such as astrocytes, platelets, and hepatocytes, are not
+
+
necessarily killed by the final reaction product MPP . We found that the intermediate MPDP
or its neutral base 1.2-MPDP) exits cells along a concentration gradient via membrane
(
+
diffusion. MPP is mainly formed extracellularly, and in most cases it would be diluted and
carried away by body fluids or taken up by passive transporters in neighboring cells. In most
+
+
cells, MPP accumulation would therefore be transient and minimal, as MPDP /1.2-MPDP
1
3

1
4
Page 14 of 57
would exit the cell as soon as the extracellular concentration drops. Only cells that accumulate
+
MPP through an active intracellular transport mechanism, e.g. DA neurons in the central
nervous system, would therefore be able to retain the toxicant for a sufficient period of time
+
and at high enough concentrations to cause damage (Fig. 6). While accumulation of MPP in
cells is a necessary condition for the toxicity of this molecule, the degree of susceptibility of
+
particular cell types to intracellular MPP may vary due to the presence or absence of
resistance mechanisms, e.g. by sequestration as in adrenomedullary cells (43) or other more
poorly understood cellular factors (7,46). Alternatively, certain cells may possess factors that
+
promote MPP formation and toxicity; for instance, cytochrome P-450 2D6 was recently
suggested to directly contribute to the conversion of MPTP in neuronal cells (1). By contrast,
+
in LUHMES cells, MPTP treatment resulted in neither the formation of MPP nor cellular
degeneration (not shown).
+
The concentration of MPP required to cause toxicity in different cell types may therefore
+
vary. Studies of isolated mitochondria and submitochondrial particles have shown that MPP
concentrations of about 5-20 millimolar need to be reached in the mitochondrial matrix for
inhibition of complex I (40,41), which would in turn require cytosolic levels in the high
micromolar range (> 100 µM). This is corroborated by our measurements showing
mitochondrial effects and toxicity within a period of 24-48 h in IMAs expressing OCT3 or
+
DAT if intracellular MPP concentrations were larger than 100 µM (Fig. 1F). By contrast, our
measurements of wildtype IMAs showed that concentrations were approximately 10 µM (≈ 1
nmol/mg protein), which is in close accordance with the data of Di Monte et al. obtained from
primary mouse astrocyte cultures (1 nmol/mg, after 48 h exposure to 250 µM MPTP)
(15,16,55).
+
Here we have demonstrated that MPDP is the pivotal MPTP metabolite responsible for the
+
+
extracellular genesis of MPP , which minimizes its toxicity to most cell types. MPDP was
1
4

1
5
Page 15 of 57
described and synthesized by Neil Castagnoli and colleagues in the 1980s (4,5). Their
pioneering discovery of this metabolite and the seminal work of Di Monte et al. (13-16,55)
were quickly incorporated into all major toxicology textbooks, but experimental study of this
compound has been virtually non-existent for the past 25 years, despite its wide use in PD
research during this time. Here, we have advanced this early work by providing direct
+
+
evidence by NMR and MS of MPDP formation during MPTP to MPP conversion by MAO-
+
B. It was originally postulated that the disproportionation of two molecules of MPDP into
+
+
MPP and MPTP would be the dominating mechanism of MPP formation, however, our
+
previous studies on the autoxidation of resynthesized MPDP suggest that direct oxidation of
+
+
+
MPDP into MPP dominates (4,5). Although the mechanism of MPDP autoxidation has not
yet fully been elucidated, kinetic and thermodynamic principles observed in structurally
related compounds allow us to postulate that the reaction sequence is as follows: in the initial
step, a stabilized allyl radical is formed by the abstraction of a proton at C5 as a result of the
+
interaction of MPDP with molecular oxygen (O ), transition metals, or free radical species
2
(e.g. dopamine autoxidation products) (Suppl. Fig. S8); in a second reaction, the allyl radical
interacts with a second free radical species or O ; when there is an interaction with O , C-5
2
2
would carry a peroxide group, which would render the carbon 6 acidic and hence trigger the
release of a proton from C-6 (peroxometabolite); energetically, the elimination of the C-5
hydroperoxide substituent from the base form would strongly favor aromatization to the stable
+
toxin MPP (Suppl. Fig. S8); the final step of proton abstraction from the peroxometabolite is
favored under alkaline conditions. This sequence of events is strongly supported by our
+
experimental observations indicating accelerated MPDP autoxidation in the presence of high
oxygen levels and transition metals, and a relative stabilization under slightly acidic
conditions (Fig. 5). Importantly, autoxidation can efficiently proceed under conditions of
+
relatively low MPDP concentrations in contrast to disproportionation reactions.
1
5

1
6
Page 16 of 57
+
We have identified several factors that increase MPP formation in the extracellular space,
most notably higher oxygen tension and pH. Indeed, the low oxygen levels within cells
compared to the extracellular space may be reduced even further in astrocytes by MAO-B
metabolism of MPTP, which is an oxygen-consuming reaction (6). Thus, low oxygen tensions
+
and lowered pH around respiring astrocytes would prevent MPP formation within cells,
while this process would proceed faster in the extracellular space. Moreover, our data show
that MPTP conversion proceeds more than 10-times faster in the presence of biological
oxidation enhancers, like melanin, oxidized dopamine, or ferrous iron. Some of these factors
have particularly high levels not only in the nigrostriatal system in which DA
neurodegeneration prevails, but also in hemoglobin-carrying erythrocytes (Fig. 3A and Suppl.
Fig. S5) (50,53,56). These factors are neutralized and scavenged to a large extent inside cells
due to powerful redox buffers, lysosomal removal, or iron binding by ferritin, while they may
exert their full activity in the extracellular space. Apart from oxygen tension or pH, other
+
factors might indeed influence the autoxidation of MPDP . Our observation of accelerated
+
MPDP autoxidation, induced by ferrous iron and its inhibition by the iron chelator
deferoxamine (Suppl. Fig. S7), provides insight into the molecular mechanisms that might
underlie previous observations of protection against MPTP-toxicity in vivo by iron chelator
administration to model animals (11, 29). However, we tested ascorbic acid both in a cell-free
+
model of MPDP autoxidation, as well as in IMAs that were exposed to MPTP. In both cases,
+
+
the autoxidation of MPDP and the formation of MPP was not affected (Suppl. Fig. S7).
These findings indicate that a more detailed analysis of the molecular mechanisms involved is
required.
A recent study reported that there is selective expression of OCT3 in glial cells located in
close spatial vicinity to nigrostriatal neurons that were primarily affected in MPTP-challenged
+
brains. This suggested that preferential MPP release from astrocytes via OCT3 would be a
1
6

1
7
Page 17 of 57
major contributing factor in the selective loss of DA neurons in the substantia nigra (9).
+
However, we observed that MPDP generation in the brains of MPTP-treated mice preceded
+
the formation of MPP not only in the substantia nigra, but also in the cerebellum which is a
+
prerequisite for the transporter-independent efflux of MPDP into the extracellular space (Fig.
2
). A recent publication was in full accordance with our findings in demonstrating almost
+
uniform MPP generation in MPTP exposed brains only minutes after MPTP administration.
This was followed by rapid clearance from the brain, except in the basal ganglia, ventral
+
mesencephalon, and olfactory bulb (28). Given the findings that MPP is generated globally
in the extracellular space and taken up by specific cell types due to selective transporter
+
expression, the tissue distribution of MPP may be determined by complex dynamics that owe
+
to the differential expression of MPP transporters across cell types and brain regions. For
+
instance, MPP would be partially sequestered into cells expressing OCT transporters (Fig. 6),
as indicated by our experiments with OCT-3-expressing IMAs (Fig. 1). Indeed, such a
+
mechanism has been reported in vivo for the MPP -related compound paraquat, which
resulted in the protection of DA neurons in brains of paraquat-treated mice by lowering
extracellular concentrations of the toxin (42).
The basic principles identified in the present work are not limited to the experimental toxin
MPTP, but could also apply to endogenously formed small molecules such as quinolinic acid
(34), isoquinolinates, beta-carbolines (8), or dopamine autoxidation products (2), all of which
are suspected to play key roles in neurological and psychiatric disorders.
1
7

1
8
Page 18 of 57
Innovation
The well-studied Parkinsonian toxicant MPTP requires astrocytic metabolism to generate
+
+
MPP . Subsequently, MPP is transported into dopaminergic neurons by dopamine
+
transporters and it thereby accumulates and eventually kills these cells. Why MPP spares the
+
astrocytes in which it is supposedly generated and how MPP leaves these cells has been a
+
vexing enigma for decades. We found here that MPP is generated extracellularly, and that
+
the release of its membrane permeable precursor MPDP /1,2-MPDP from astrocytes is
transporter independent. Differences in O and pH between the intra- and extra-cellular
2
+
environment favor the stabilization of the labile MPTP intermediate MPDP /1,2-MPDP
within astrocytes and promote its nonenzymatic conversion in the extracellular space.
1
8

1
9
Page 19 of 57
Materials and Methods
Materials.
+
MPP (1-methyl-4-phenylpyridinium-iodide), purified monoamine oxidase-B (MAO-B),
GBR12909 (DAT blocker), dopamine, ferritin, and tyrosine-melanin were purchased from
Sigma (St. Louis, MO, USA).
+
Synthesis of MPTPHCI and MPDP bromide.
All reagents are commercially available and were used without further purification. Solvents
were dried over molecular sieves and used directly without further purification. All reactions
1
13
were conducted under exclusion of air and moisture. H- and C-NMR spectra were recorded
on an Avance III 400 MHz spectrometer from Bruker at room temperature. The time course
of enzymatic MPTP conversion was followed on a Bruker AV III 600 MHz spectrometer
equipped with a TCI-cytoprobe. Spectra were processed with the software MestReNova 6.1.1
1
13
from MestRelab Research and the H and C chemical shifts are reported relative to the
residual solvent peak. High resolution mass spectrometry (HRMS) was performed with
micrOTOF-Q II ESI-Qq-TOF from Bruker Daltonics.
1
9

2
0
Page 20 of 57
+
Synthesis of MPTP-HCI (4) and MPDP bromide (6)
Scheme 1. a) THF, -78°C→ rt, over night; (18) b) aq HCl, reflux, 5h, quantitative; (18) c) (i)
NaEtO, EtOH, 0°C (ii) aq H O , EtOH, 50°C, quantitative; (20) d) DCM, TFAA, aq HBr,
2
2
0
°C, 53%. (20)
Step a): 1-Methyl-4-phenyl-4-piperidinol 3 (18)
A solution of 24.6 ml phenyllithium 2 (1.8 M in dibuthylether) was added to a solution of 1-
metyl-4-piperidone 1 (5 g, 44.19 mmol) in 20 ml of dry tetrahydrofuran (THF) at -78°C. The
reaction was allowed to warm up to room temperature and was stirred over night. The
solvents were removed in vacuo to give crude 3 as yellowish solid.
Step b): MPTPHCI 4 (18)
Refluxing of crude 3 in 50 ml 10% aq HCl for 5 h furnished compound 4. The solvent was
removed in vacuo and crude 4 was re-crystallized from ethanol to give white crystalline 4 in a
quantitative yield.
1
H-NMR (400 MHz, CD OD) δ 7.53 – 7.47 (m, 2H), 7.44 – 7.31 (m, 3H), 6.15 (ddd, J = 5.1,
3
3
.4, 1.7 Hz, 1H), 4.08 (d, J = 15.3 Hz, 1H), 3.92 – 3.79 (m, 1H), 3.79 – 3.70 (m, 1H), 3.45 –
.35 (m, 1H), 3.03 (s, 3H), 3.00 – 2.92 (m, 1H), 2.87 (d, J = 21.4 Hz, 1H).
3
1
H-NMR (600 MHz, D O) δ 7.55 – 7.52 (m, 2H), 7.46 (dd, J = 10.2, 4.8 Hz, 2H), 7.43 – 7.39
2
(m, 1H), 6.15 – 6.13 (m, 1H), 4.05 (d, J = 16.4 Hz, 1H), 3.82 – 3.78 (m, 1H), 3.74 – 3.69 (m,
1
H), 3.43 – 3.34 (m, 1H), 3.00 (s, 3H), 2.96 – 2.90 (m, 1H), 2.86 (d, J = 19.0 Hz, 1H).
2
0

2
1
Page 21 of 57
1
3
C-NMR (101 MHz, CD OD) δ 139.9, 136.8, 129.7, 129.4, 126.2, 116.6, 53.5, 52.0, 42.9,
3
2
5.8.
1
3
C-NMR (150 MHz, D O) δ 138.3, 134.6, 128.7, 128.4, 125.0, 115.5, 52.0, 50.6, 41.7, 23.8.
2
+
HRMS (ESI, positive ion mode): [M+H] ; calculated: C H N: 174.1277 m/z; measured:
1
2
15
1
74.1288 m/z
Step c): 1-Methyl-4-phenyl-1,2,3,6-tetrahydropyridine N-Oxide 5 (20)
i) The solution of 4 (2.31 g, 11.01 mmol) in 60 ml ethanol was cooled to 0°C and 9.5 ml of a
(
sodium ethoxide solution (1.17 M in ethanol, 11.12 mmol) was added. The mixture was
filtered over celite and the filter cake was washed with 30 mL of ethanol.
(
ii) Next, 2.6 ml of 30% aq H O was added and the mixture was stirred at 50°C. After 2 h, an
2 2
additional portion of 2 mL 30% aq H O was added, and it was stirred over night. Afterwards
2
2
the excess of H O was quenched with 10% Pd/C. The resulting mixture was filtered over
2
2
celite and the filter cake was washed with 30 ml of ethanol. The solvent was removed in
vacuo to give 5 2 H O as yellow oil in a quantitative yield.
2
1
H-NMR (400 MHz, CDCl ) δ 7.42 – 7.26 (m, 5H), 5.96 (s, 1H), 4.32 – 4.10 (m, 2H), 3.76 –
3
3
.65 (m, 2H), 3.37 (s, 3H), 3.06 – 2.95 (m, 1H), 2.76 – 2.62 (m, 1H).
+
Step d): MPDP bromide 6 (20)
The suspension of 5 2 H O (2.51 g, 11.01 mmol) in 90 ml dichloromethane (DCM) was
2
cooled to 0°C and trifluoroacetic anhydride (TFAA) (4.49 g; 21.38 mmol) was added
dropwise within 20 min. It was stirred at 0°C for 2 h, followed by the addition of 48% aq HBr
(2.06 g, 12.23 mmol). The mixture was evaporated to dryness under reduced pressure at room
temperature. The resulting crude product was re-crystallized from acetone to yield 53 % 6
(1.48 g, 5.87 mmol) as a yellow hygroscopic solid.
1
H-NMR (400 MHz, CDCl ) δ 9.23 (d, J = 4.1 Hz, 1H), 7.69 – 7.60 (m, 2H), 7.52 – 7.40 (m,
3
3
H), 6.89 (d, J = 4.6 Hz, 1H), 4.14 (t, J = 9.4 Hz, 2H), 3.93 (s, 3H), 3.32 (t, J = 9.4 Hz, 2H).
2
1

2
2
Page 22 of 57
1
H-NMR (600 MHz, D O) δ 8.38 (d, J = 4.8 Hz, 1H), 7.82 (d, J = 7.9 Hz, 2H), 7.63 – 7.54
2
(m, 3H), 6.90 (d, J = 5.0 Hz, 1H), 4.04 (t, J = 10.1 Hz, 2H), 3.69 (s, 3H), 3.28 (t, J = 10.1 Hz,
2
H).
1
3
C-NMR (101 MHz, CDCl ) δ 164.1, 158.0, 134.5, 133.0, 129.5, 127.3, 113.9, 49.0, 47.7,
3
2
5.6.
1
3
C-NMR (150 MHz, D O) δ 163.5, 158.8, 134.6, 132.5, 129.0, 126.7, 112.9, 47.6, 46.0, 24.5.
2
+
HRMS (ESI, positive ion mode): [M]; calculated C H N : 172.1121 m/z; measured:
1
2
14
1
72.1124 m/z
+
-
Analysis of 1-Methyl-4-phenylpyridinium Iodide (MPP I ) (Sigma Aldrich)
1
H-NMR (400 MHz, DMSO) δ 9.02 (d, J = 6.9 Hz, 2H), 8.51 (d, J = 7.0 Hz, 2H), 8.11 – 8.04
(
1
(
1
m, 2H), 7.70 – 7.61 (m, 3H), 4.34 (s, 3H).
H-NMR (600 MHz, D O) δ 8.74 (d, J = 6.7 Hz, 2H), 8.28 (d, J = 6.7 Hz, 2H), 7.97 – 7.92
2
m, 2H), 7.70 – 7.62 (m, 3H), 4.36 (s, 3H).
3
C-NMR (101 MHz, DMSO) δ 154.3, 145.6, 133.5, 132.0, 129.7, 128.0, 124.1, 47.1.
1
3
C-NMR (150 MHz, D O) δ 156.3, 144.8, 133.8, 131.9, 129.5, 127.7, 124.6, 46.9.
2
+
HRMS (ESI, positive ion mode): [M]; calculated C H N : 170.0964 m/z; measured:
1
2
12
1
70.0968 m/z
LUHMES cell culture.
Methods for the generation, handling, and properties of LUHMES cells have been described
in detail previously (47,48). The cells were grown in standard conditions (serum-free) at a
density of 300.000 cells per well in 24-well plates. The cells were derived from human fetal
2
2

2
3
Page 23 of 57
mesencephalon and were conditionally immortalized by expression of v-myc under a tet-off
promoter. Upon addition of tetracycline, LUHMES cells differentiate within 5 days to fully
post-mitotic human dopaminergic neurons with neurites of up to 1000 µm length (48). At this
stage, the cells express highly active dopamine transporters and are susceptible to low
+
micromolar levels of MPP (47,48). For uptake experiments, cells were differentiated for 6
days and plated at a density of 300.000 cells/well in 24-well plates.
IMA 2.1 astrocytes (IMA).
Generation and properties of the IMA cell line have recently been described in detail (45).
These immortalized mouse astrocytes were cultured at 37 °C (5 % CO ) in DMEM (normal
2
glucose) (GIBCO) with 5 % FCS (fetal calf serum, from PAA) and 1%
Penicillin/Streptomycin (10.000 U/ml stock). The cells were passaged by trypsinization for 2
min with 0.5% trypsin/DMEM every 2-3 days and reseeded at a ratio of 1:5 or 1:10.
Experiments were performed 2 days after cells reached confluency in DMEM containing 2 %
FCS.
Overexpression of OCT3 and DAT.
Gene contracts for human solute carrier family 6, member 3 (dopamine transporter, DAT) and
sls22a - solute carrier family 22, member 3 (organic cation transporter, OCT3) were
purchased from SourceBioscience (Berlin, Germany). Both were obtained as inserts in the
pCR vector (InVitroGene, Life Technologies Corporation USA) as clone number
IRCBp5005O019Q (40147025 - IMAGE ID) and clone number IRCKp5014K038Q (8860680
-
IMAGE ID), respectively. The DAT gene was cloned into the phsCtR2AU-W expression
construct. This HIV-based vector was used as a plasmid to transfect the IMA cell line using
2
3

2
4
Page 24 of 57
Lipofectamine (InVitroGene, Life Technologies, Darmstadt, Germany). In this construct the
DAT gene was fused to the 3'-end of the tR2AU cassette (turbo red fluorescent protein from
Entacmaea quadricolor, the 2A sequence of Porcine teschovirus and human ubiquitin
sequences). Translation of this gene produces free tRFP protein by ribosomal skipping at the
2
A site and free DAT protein by the action of cellular hydrolases which cut at the 3'-end of
the ubiquitin peptide. Cells that were successfully transfected and expressed the DAT gene
could therefore by be identified by their expression of red fluorescence. Cells were assayed 3
days after transfection.
The gene for OCT3 was cloned into the pBIGFP vector. This construct contains a
bidirectional inducible promoter allowing the expression of any gene cloned therein and the
simultaneous expression of the linked marker gene eGFP (enhanced Green Fluorescent
Protein). Induction levels can be monitored by examination of eGFP intensities. Cells were
transfected using Lipofectamine (InVitroGene, Life Technologies Corporation, Darmstadt,
Germany) and subsequently selected with 200 µg/ml G418 for 10 days.
Immunostaining.
Cells were grown in 24-well plastic cell culture plates (Nunclon™). Following treatment,
cells were fixed with 4% PFA for 20 min at 37°C and washed with PBS. After blocking with
1
% BSA (Calbiochem, San Diego, CA) for 1 h, the primary antibody (anti--III-tubulin,
Convance, mouse 1:1000) was added in PBS-Tween (0.1%) and the cells were incubated at 4
C overnight. Secondary antibody (anti rat IgG-Alexa 488, Invitrogen, Darmstadt, Germany)
°
was added for 45 min at RT. For visualization, an Olympus IX 81 microscope (Hamburg,
Germany) equipped with an F-view CCD camera was used. Images were processed by Cell P
software (Olympus).
2
4

2
5
Page 25 of 57
Cell viability.
Resazurin (Sigma) was added to the cell culture medium at a final concentration of 5 µg/ml,
and was incubated at 37 °C for at least 30 min. The fluorescence intensity was measured at
5
30 nmex and 590 nmem with a Tecan Infinite M200 reader. Cell viability (in %) was
calculated by normalizing the fluorescence values to the values obtained from untreated
controls.
ATP assay.
Cells grown in 24-well plates were lysed in PBS-buffer containing 0.5% phosphatase inhibitor
cocktail 2 (Sigma) and boiled at 95 °C for 10 min. Following centrifugation at 10.000 g for 5
min to remove cell debris, protein content in the supernatant was measured and samples were
adjusted to have equal concentrations. Samples were then diluted 1:10 in PBS / 0.5%
phosphatase-inhibitor buffer. For the detection of ATP levels, a commercially available ATP
assay reaction mixture (Sigma) containing luciferin and luciferase was used (50 µl of adjusted
sample plus 100 µl of assay-mix were added to a black 96-well plate). Standards were
prepared by serial dilutions of ATP disodium salt hydrate (Sigma) to obtain concentrations
ranging from 1000 nM to 7.8 nM.
Vesicle uptake assay.
For vesicle preparation, stock solutions of 100 mg/ml DOPC (1,2-dioleoyl-sn-glycero-3-
phosphocholine) and 100 mg/ml DOPE (1,2-dioleoyl-sn-glycero-3-phosphoethanolamine)
(
Avanti) in CHCl were prepared. For the production of a 2 ml vesicle suspension, 102 µl of
3
DOPC and 98 µl of DOPE stock solutions were mixed in a glass tube and evaporated for 30
min under vacuum while rotating to remove CHCl . The lipid film was then resuspended in 1
3
2
5

2
6
Page 26 of 57
ml of 20% Triton X-100 in PBS, and air was removed by N -degassing. The mixture under N
2
2
atmosphere was then briefly sonicated in a water bath and allowed to rest at room temperature
until a clear solution was observed, indicating complete solubilization of the lipids. To
remove the detergent, 600 mg Bio-Beads SM-2 (BioRad) preequilibrated in PBS were added
to the clear lipid/detergent mixture diluted 1:1 in PBS, and incubated at 4 °C on a rotator for
1
6 h. For removal of the beads, the vesicle suspension was collected with capillary tips and
transferred to a separate tube. For uptake assays, the vesicle suspension (10 mg lipids/ml) in
+
+
+
+
PBS was incubated with MPP or MPDP for 2 h at 37 °C. To remove free MPP /MPDP ,
TM
TM
vesicles were separated by loading 100 µl of the sample onto a NAP -5 Sephadex G-25
column and a stepwise addition of 250 µl volumes of PBS for elution. As control, vesicle-free
+
solutions containing MPP or MPDP were added. The vesicle-containing fractions were
+
+
analyzed separately, MPP or MPDP values in the vesicle fractions were integrated.
Erythrocyte uptake assay
Human blood was collected in EDTA (1.6 mg/ml blood) containing tubes (Sarstedt S-
Monovette, Nuembrecht, Germany) and was centrifuged at 200 g for 4 min. The erythrocyte
fraction was resuspended in a buffer containing 21 mM Tris, 4.7 mM KCl, 2 mM CaCl , 140
2
mM NaCl, 1.2 mM MgSO , 5.5 mM glucose, pH 7.4. The cells were washed twice with the
4
+
buffer by centrifugation at 1000 g for 1 min. The cells were then treated with MPP or
+
+
MPDP as indicated. To assess intracellular MPP levels, the erythrocytes were again washed
+
by three consecutive centrifugation steps (1000 g, 1 min) to remove extracellular MPP or
+
MPDP . The erythrocytes were then lysed by resuspension in H O for 20 min, supported by
2
moderate sonication. Proteins were precipitated by the addition of 0.4 M perchloric acid, the
supernatant was filtered through a 0.22 µm membrane and analyzed by HPLC.
2
6

2
7
Page 27 of 57
MPTP conversion in the brain
C57BL/6JRj male mice were kept under 12h light/dark cycles at a temperature of 23°C and
5% humidity. The animals had ad libitum access to food and water. All animal studies were
5
approved by the appropriate institutional governmental agency (Regierungspraesidium
Tuebingen, Germany) and carried out in an Association for Assessment and Accreditation of
Laboratory Animal Care International-accredited facility in accordance with the European
Convention for Animal Care and Use of Laboratory Animals. The mice were intraperitoneally
injected with MPTP (30 mg/kg; free base in 0.9% NaCl), cared for over the time intervals
indicated, and then killed by cervical dislocation. The striatum and cerebellum of each mouse
was dissected after decapitation. The tissues were homogenized in 1 ml of 0.4 M perchloric
acid by sonication. Following a centrifugation step at 2.000 g for 10 min at 4°C, the pellet was
resuspended in 250 µM of 1 M NaOH. Total protein content was detected by a BCA assay kit.
The supernatant was filtered through a 0.22 µm membrane and analyzed by HPLC.
HPLC analysis.
+
+
Detection of MPTP, MPP , and MPDP was performed on a Kontron system (Goebel
Analytic, Au/Hallertau, Germany) comprising a model 520 pump, model 560 autosampler,
+
models 535 and 430 diode array detectors, set at 245 nm for MPTP, 295 nm for MPP , and
+
3
45 nm for MPDP . Samples were acidified with 9 µl perchloric acid (70%) per ml volume of
sample and centrifuged at 10,000 x g for 15 min. The supernatant was filtered through a
Chromaphil PET-20/15MS-filter with 0.2 µm pore size from Macherey Nagel (Düren,
Germany). Separation was carried out on a C18 nucleosil column (250 x 4.6 mm; 5 µm
particle size) from Macherey Nagel (Düren, Germany) at room temperature. The mobile
phase consisted of acetonitrile : distilled water : triethylamine : sulfuric acid (12.50 : 86.18 :
1
.04 : 0.28, v/v, pH 2.3). The mobile phase was degassed with an online vacuum degasser and
2
7

2
8
Page 28 of 57
delivered isocratically at a flow rate of 1 ml/min at an average pressure of 145 bar. Data
analysis was performed with Geminyx II software (Goebel Analytic). Peak areas of MPTP,
+
+
MPDP , and MPP were evaluated with the Geminyx II software by the application of
standard curves obtained for each of the three individual compounds.
13
Detection of C-labelled intracellular metabolites.
13
For stable isotope labeling, cells were incubated with [U- C ] glucose for at least 18 h. To
6
infer relative intracellular fluxes, intracellular metabolites were extracted and measured with
GC/MS. Cells grown in six-well plates were washed with 1 ml saline solution (0.9%) and
quenched with 0.4 ml of cold (- 20 °C) methanol. After adding an equal volume of cold water,
cells were collected with a cell scraper and transferred in tubes containing 0.4 ml of cold (-20
°
C) chloroform. The extracts were shaken at 1,400 rpm for 20 min at 4 °C and centrifuged at
6,000×g for 5 min at 4 °C. 0.3 ml of the upper aqueous phase was collected in specific GC
glass vials and evaporated under vacuum at -4 °C using a refrigerated CentriVap Concentrator
Labconco). Metabolite derivatization was performed using an Agilent Autosampler. Dried
polar metabolites were dissolved in 15 μl of 2% methoxyamine hydrochloride in pyridine at
5 °C. After 60 min, an equal volume of 2,2,2-trifluoro-N-methyl-N-trimethylsilyl-acetamide
1% chloro-trimethyl-silane was added and metabolites were incubated for 30 min at 45 °C.
1
(
4
+
GC/MS analysis was performed using an Agilent 6890GC equipped with a 30m DB-35MS
capillary column. The GC was connected to an Agilent 5975C MS operating under electron
impact ionization at 70 eV. The MS source was held at 230 °C and the quadrupole at 150 °C.
The detector was operated in scan mode and 1 μl of derivatized sample was injected in
splitless mode. Helium was used as carrier gas at a flow rate of 1 ml/min. The GC oven
temperature was held at 80 °C for 6 min and increased to 300°C at a rate of 6°C/min. After 10
min, the temperature was increased to 325°C at a rate of 10 °C/min for 4 min. The run time
2
8

2
9
Page 29 of 57
for each sample was 59 min. For determination of the mass isotopomer distribution (MID) of
citrate, the mass spectrum was corrected for natural isotope abundance. Data processing from
raw spectra to MID correction and determination, was performed using MetaboliteDetector
software (24).
Acknowledgments
This work was supported by RTG 1331, BMBF, the Doerenkamp-Zbinden-Foundation,
KoRS-Chemical Biology, and the Collaborative Research Center 969 “Chemical and
Biological Principles of Cellular Proteostasis”, funded by the Deutsche
Forschungsgemeinschaft (DFG).
Author Disclosure Statement
No competing financial interests exist.
List of Abbreviations
1
,2-MPDP = 1-methyl-4-phenyl-1,2-dihydropyridine
DA = dopamine
DAT = dopamine transporter
DOPC = 1,2-dioleoyl-sn-glycero-3-phosphatidylcholine
DOPE = 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine
IMA = immortalized mouse astrocytes
LUHMES = lund human mesencephalic neurons
MAO = monoamine oxidase
+
MPDP = 1-methyl-4-phenyl-2,3-dihydropyridinium
+
MPP = 1-methyl-4-phenyl-pyridinium
MPTP = 1-methyl-4-phenyl-tetrahydropyridine
OCT = organic cation transporter
PMAT = plasma membrane monoamine transporter
2
9

3
0
Page 30 of 57
References
1
. Bajpai P, Sangar MC, Singh S, Tang W, Bansal S, Chowdhury G, Cheng Q, Fang JK, Martin
MV, Guengerich FP, and Avadhani NG. Metabolism of 1-methyl-4-phenyl-1,2,3,6-
tetrahydropyridine by mitochondrion-targeted cytochrome P450 2D6: implications in
Parkinson disease. J Biol Chem. 288(6): 4436-51, 2013.
2
. Barzilai A, Daily D, Zilkha-Falb R, Ziv I, Offen D, Melamed E, and Shirvan A. The molecular
mechanisms of dopamine toxicity. Adv Neurol 91: 73-82, 2003.
3
4
. Beal MF. Experimental models of Parkinson's disease. Nat Rev Neurosci 2(5): 325-34, 2001.
. Castagnoli N, Chiba K, and Trevor AJ. Potential bioactivation pathways for the neurotoxin 1-
methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP). Life Sci 36(3): 225-30, 1985.
. Chiba K, Peterson LA, Castagnoli KP, Trevor AJ, and Castagnoli N. Studies on the molecular
mechanism of bioactivation of the selective nigrostriatal toxin 1-methyl-4-phenyl-1,2,3,6-
tetrahydropyridine. Drug Metab Dispos 13(3): 342-7, 1985.
5
6
7
. Chiba K, Trevor A, and Castagnoli N. Metabolism of the neurotoxic tertiary amine, MPTP, by
brain monoamine oxidase. Biochem Biophys Res Commun 120(2): 574-8, 1984.
. Choi WS, Kruse SE, Palmiter RD, and Xia Z. Mitochondrial complex I inhibition is not
required for dopaminergic neuron death induced by rotenone, MPP+, or paraquat. Proc Natl
Acad Sci USA 105(39): 15136-41, 2008.
8
9
. Cobuzzi RJ, Neafsey EJ, and Collins, MA. Differential cytotoxicities of N-methyl-beta-
carbolinium analogues of MPP+ in PC12 cells: insights into potential neurotoxicants in
Parkinson's disease. J Neurochem 62(4): 1503-10, 1994.
. Cui M, Aras R, Christian WV, Rappold PM, Hatwar M, Panza J, Jackson-Lewis V, Javitch
JA, Ballatori N, Przedborski S, and Tieu K. The organic cation transporter-3 is a pivotal
modulator of neurodegeneration in the nigrostriatal dopaminergic pathway. Proc Natl Acad
Sci USA 106(19): 8043-8, 2009.
10. Dauer W, and Przedborski S. Parkinson's disease: mechanisms and models. Neuron 39(6):
889-909, 2003.
3
0