© 1997 International Society for Neurochemistry
Neurotrophic Effects of L-DOPA in Postnatal Midbrain
Dopamine Neuron/Cortical Astrocyte Cocultures
*~M~j.ja
Angeles Mena, *Viviana Davila, and *David Suizer
*Departments of Neurology and Psychiatry, Columbia University and Department of Neuroscience, New York State Psychiatric Institute, New York, New York, U.S.A.; and ~Departamento de Investigación,
Hospital Ramón y Cajal, Madrid, Spain
Abstract: L-DOPA IS toxic to catecholamine neurons in
culture, but the toxicity is reduced by exposure to
astro-cytes. We tested the effect of L-DOPA on dopamine
neu-rons using postnatal ventral midbrain neuron/cortical
astrocyte cocultures in serum-free, glia-conditioned
me-dium. L-DOPA (50
tiM) protected against dopamine
neu-ronal cell death and increased the number and branching
of dopamine processes. In contrast to embryonically
de-rived glia-free cultures, where L-DOPA is toxic, postnatal
midbrain cultures did not show toxicity at 200
1iM
L-DOPA. The stereoisomer D-DOPA (50—400
btM)was not
neurotrophic. The aromatic amino acid decarboxylase
in-hibitor carbidopa (25
~zM)did not block the neurotrophic
effect. These data suggest that the neurotrophic effect of
L-DOPA is stereospecific but independent of the
produc-tion of dopamine. However, L-DOPA increased the level
of glutathione. Inhibition of glutathione peroxidase by
L-buthionine sulfoximine (3
1iM for 24 h) blocked the
neuro-trophic action of L-DOPA. N-Acetyl-L-cysteine (250
1iM
for 48 h), which promotes glutathione synthesis, had a
neurotrophic effect similar to that of L-DOPA. These data
suggest that the neurotrophic effect of L-DOPA may be
mediated, at least in part, by elevation of glutathione
con-tent. Key Words:
L-DOPA—Glia—Glutathione—Parkin-son‘s disease—Dopamine neurons.
J.
Neurochem. 69, 1398—1408 (1997).
The effects of L-DOPA on dopamine (DA) neurons
are quite different in vivo and in vitro. Whereas
rela-tively low levels
(25 jiM)of L-DOPA are toxic in
culture (Melamed, 1986; Mena et al., 1992, 1993;
My-tilineou et al., 1993; Pardo et al.,
1995),
the drug has
not been seen to damage DA neurons in healthy
ani-mals (Hefti et al., 1981; Perry et al., 1984; Yurek et
al., 1991; Blunt et al., 1993) or humans (Quinn et al.,
1986). As L-DOPA is the most effective compound
for symptomatic treatment of Parkinson‘ s disease
(PD), its influence on the progression of the disease
is particularly important. Although some reports have
indicated increased fluctuations and dyskinesia related
to the duration of the L-DOPA treatment (Lesser et
al., 1979), other studies have found no evidence of
negative impact (Markham and Diamond, 1986; Blint
et al., 1988; Uitti et al., 1993).
Previous studies on L-DOPA toxicity in culture have
used embryonically derived cells grown in the presence
of a very low level of glia
(<5%),
although in the
brain DA neurons are enveloped by glia. To investigate
the modulatory effects of glia on L-DOPA toxicity, we
have used a coculture system of postnatal midbrain
DA neurons and cortical astrocytes, a target area of
midbrain DA projections. The protective role of glia
has been demonstrated in experiments showing that
medium preconditioned by glia protects embryonically
derived midbrain cultures from L-DOPA toxicity
(Mena et al., 1996). The midbrain neurons/cortical
glia coculture system used in the present report is
ad-vantageous for this study because it produces a high
number of mature midbrain neurons
generally 30—
40% of the neurons are tyrosine hydroxylase (TH
)
-positive, as compared with <4% typical of
embryoni-cally derived cultures]. We suggest that this system,
which has previously been used to provide insights
into the mechanisms of methamphetamine
neurotoxic-ity (Cubells et al., 1994) and the role of superoxide
dismutase (SOD)
in protection from cell death
(Przedborski et al., 1996), is appropriate for studying
the effects on cell survival and the mechanisms of
action of L-DOPA.
MATERIALS AND METHODS
Materials
L- and o-DOPA, L-N-acetylcysteine (L-NAC),
L-buthio-nine- S,R-sulfoximine (BSO), glutathione [as its reduced
Received April 7, 1997; revised manuscript received May 16, 1997; accepted May 16, 1997.
Address correspondence and reprint requests to Dr. M. A. Mena at Departamento de Investigación, Hospital Ramón y Cajal, Can de Colmenar Km 9, Madrid 28034, Spain.
Abbreviations used: BSO, L-buthionjne-S,R-sulfoximine; DA, do-pamine; GSH, glutathione (reduced form); GSSG, glutathione (oxi-dized form); L-NAC, L-N-acetylcysteine; PD, Parkinson‘s disease; SOD, superoxide dismutase; TH, tyrosine hydroxylase.
L-DOPA AND POSTNATAL DOPAM1NE COCULTURE
1399
form (GSH)], 3,3 ‘-diaminobenzidine tetrahydrochloride, and 5,5 ‘-dithio-bis(2-nitrobenzoic acid) were obtained from Sigma Chemical (St. Louis, MO, U.S.A.). Biotinylated horse anti-mouse secondary antibody was from Vector Labora-tories (Burlingame, CA, U.S.A.).
Coculture of postnatal DA neurons/cortical
astrocytes
Cultures of ventral midbrain neurons from postnal day 1— 2 rats were prepared as in related studies (Rayport et al., 1992; Rayport and Suizer, 1995; Pothos et al., 1996; Przed-borski et al., 1996) except that serum-free medium was used as recently described (Mena et al., 1997). Astrocyte mono-layers were prepared from the rostral half of the cerebral cortex in ice-cold calcium- and magnesium-free phosphate-buffered saline on postnatal day 2. The chunks of tissue were dissociated with 20 U/ml papain in 1 mM cysteine, 0.001% phenol red, 500 jiM kynurenate, 116 mM NaCl, 5.4 mM KO, 26 mM NaHCO
5, 2 mM NaH2PO4, 1 mM magnesium phosphate, 500 jiM EDTA, and 25 mM glucose, pH 7.3. Cells were plated a density of 150,000 per well and fed with medium containing 90% minimal essential medium, 10% calf serum, 1.83 mM glucose, 5 jig/ml bovine pancreatic insulin, 500 jiM glutamine, 0.6 U/mI penicillin, and 60 jig! ml Streptomycin. 5-Flurodeoxyuridine (11 jiM with 55 jiM uridine) was added
4—5
days after plating, to prevent prolif-eration of oligodendrocytes.After 2 weeks in culture, midbrain neurons were plated onto glial monolayers. Twenty-four hours before plating, the glial medium was replaced by serum-free medium containing 47% minimal essential medium, 40% Dulbecco‘s modified Eagle‘s medium, 10% Ham‘s F-12 nutrient medium, 18.9 mM glucose, 0.25% albumin, 500 jiM glutamine, 100 jig! ml transferrin, 15
jiM
putrescine, 30 nM triiodothyronine, 25 jig/ml insulin, 200 nM progesterone, 125 nM corticoste-rone, 5 j.tg/ml SOD, 10 j.tg/ml catalase, and 500jiM
kynure-nate. On the next day, the tissue was dissociated, and 80,000 cells, resuspended in serum-free glia conditioned medium, were plated per well. During the following 24 h, cultures were exposed to 5-fluorodeoxyuridine and maintained in se-rum-free medium at 37°Cin 5%CO2.
Experimental designs
Dose-dependent effects of L-DOPA were studied in post-natal cocultures grown in serum-free medium and treated with L-DOPA (0, 25, 50, 100, 200, and 400 jiM) for 48 h at day 5 in vitro.
The effects of L-DOPA and D-DOPA were compared in cultures grown as above and treated with one of the two stereoisomers at concentrations of 50 and 400 jiM for 48 h. These two concentrations were chosen because we found that L-DOPA shows neurotrophic activity with the lower concentration and neurotoxic properties at the higher concen-tration (see below).
The neurotrophic properties of L-DOPA on cell survival were further investigated in cultures treated on the third day after plating with 50 jiM L-DOPA for either 24 h or 4 days. We investigated whether the neurotrophic activity of
L-DOPA is due to itself or to its metabolite DA, in cultures treated with 50 jiM L-DOPA or 50 ‚aM L-DOPA with 25
pM carbidopa. Previously, we found by HPLC with
electro-chemical detection that 25 ‚aM carbidopa blocked the conver-sion of L-DOPA to DA in our culture conditions (Mena et al., 1997).
The interaction betweenL-DOPA and GSH was examined.
Cells maintained in culture for 5 days were treated with GSH (200 and 400 ‚aM) and with the GSH precursor, L-NAC (250 and 500 ‚aM), for 2 days or with a GSH synthesis inhibitor, BSO (3 ‚aM), for 24 h. The effects of L-DOPA (50 ‚aM) on GSH intracellular levels were examined in co-cultures treated with L-DOPA for 24 h.
Measurement of levels of quinones
Extracellular quinones, which indicate extracellular
L-DOPA autooxidation, were evaluated according to the spec-trometric measurement of the optical density at 490 nm in the culture medium (Mena et al., 1992, 1993).
Cell culture immunocytochemistry
DA neurons in ventral midbrain cultures were identified by TH immunohistochemistry as previously described (Ray-port et al., 1992; Przedborski et al., 1996). In brief, cultures were fixed with 4% paraformaldehyde, washed in Tris-buf-fered saline (0.1 M Tris-HC1 in 9 g/L NaC1, pH 7.4), perme-abilized with 0.1% Triton X-100 (Sigma), and incubated at 4°C for 48 h in a 1:640 dilution of monoclonal anti-TH antiserum (Boehringer Mannheim) in Tris-buffered saline containing 0.1% Triton X-100 and 10% normal horse serum. The secondary antibody was a biotinylate-conjugated mono-clonal anti-mouse antibody (1:200) and a horseradish-conju-gated avidin/biotin complex (Vector Laboratories); immu-nostaining was revealed by 3,3 ‘-diaminobenzidine/H2O2. To assess the specificity of TH immunostaining, the primary or the secondary antibody was omitted.
Glutathione level measurements
Glutathione (GSH) levels were measured according to the method of Tietze (1969). In brief, 2 X 106 cells were washed twice with phosphate-buffered saline, lysed in 3% perchloric acid for 15 min at 4°C,and centrifuged, and supernatants were neutralized with 9 volumes of 5 mM EDTA in 0.1 M NaH2PO4, pH 7.5. Glutathione content was measured by addition of 5,5 ‘-dithio-bis (2-nitrobenzoïc acid), and the re-action was monitored at 412 nm. GSH content is expressed as a function of total proteins in the cell extract. Proteins were quantified in the pellet by the method of Bradford (1976).
Cell counts and statistics
Results are expressed as mean±SEM values for three to six independent experiments, and each data point corre-sponds to four or more cultures derived from the same litter. In each culture, eight to 10 consecutive fields (magnification of 100 or 200 X) were counted under phase optics; both the total number of neurons and TH-stained neurons were counted. The number of cells was corrected for area. Results were statistically evaluated for significance by ANOVA for multiple groups followed by Student‘ sttest. The difference was considered significant when p< 0.05.
RESULTS
Dose-dependent effect of L-DOPA on postnatal
cocultures
To observe the effects of L-DOPA, we incubated the cultures for 7 days in serum-free culture medium and added L-DOPA (0, 25,
50,
100, 200, and 400 ‚aM) for 48 h on day5.
In contrast to previous observations with embryonically derived cultures, there was no tox-icity or enhanced quinone formation at concentrationsFIG. 1. Effect of L-DOPA on TH-positive neurons. Micrographs show TH immunostaining of postnatal ventral midbrain cells treated
with L-DOPA. Cells were treated for 48 h after 5 days in culture with (A) vehicle, (B)L-DOPA (50pM), (C) L-DOPA (100pM), or (D) L-DOPA (200
pM).
Bar=30 jim.of L-DOPA up to 200
‚aM.
Moreover, L-DOPA at 50‚aM
increased the number of surviving TH-positive neurons by 166 ± 9% (Fig. 1 and Table 1).L-DOPA also markedly increased the proliferation of TH-positive neuntes. To compare these effects, we adapted semiquantitative methods for estimating the elaboration of processes in culture (Denis-Donini et
al., 1983). L-DOPA
(50 ‚aM)
increased the number of TH-positive processes that extended >80 ‚am from a point in the center of the cell body by 177 ± 14% and increased the number of TH-positive neurons that displayed more than three primary processes by 162 ± 12% (Fig. 2). This process outgrowth was attenu-ated at levels> 100 ‚aM, but there was no neurotoxicityTABLE 1.
Effect of L-DOPA (48 h) on neuronal survival and quinone formationCells/cm2 % TH-positive Quinones (ODX 10~) TH-positive TH-negative Controls 645 ±34 (100%) 1,251 ± 123 (100%) 34.2 ± 1.54 369 ± 19 (100%) L-DOPA 25 pM 949±516 (147%) 1,130 ±88 (90%) 45.3 ±3.6° 367 ± 15 (100%) 50pM 1,071 ±58~(166%) 1,076 ±75 (86%) 49.8 ± l.3l~ 328 ± II (89%) 100 pM 774 ±65 (120%) 1,451 ±10 (116%) 34.8 ±0.92 343 ± 11 (93%) 200 pM 710 ±71 (110%) 1,426 ±11 (114%) 33.2±1.3 373 ±11 (101%) 400 pM 534 ±49 (82.8%) 919 ±87° (73.4%) 36.6 ±3.1 635 ±26~(172%)
Dataare mean±SEM values. Numbers of TH-positive and -negative cells/cm2 are from two to six
experiments (n > 6for each data point). Cultures were treated with L-DOPA for 48 h after 5 days in
culture.
L-DOPA AND POSTNATAL DOPAMINE COCULTURE
1401
FIG. 2. Effects of 50 pML-DOPA on process outgrowth of TH-positive neurons. L-DOPA (50
pM)
increased the number ofTH-positive processes that extended >80 pm from a point in the center of the cell body and increased the number of TH-positive neurons that displayed more than three primary processes. Data are expressed as mean ±SEM (bars) percentages of TH-posi-tive neurons that displayed more than three primary processes versus total 1H-positive neurons (%TH + MP) and as mean±SEM (bars) percentages of TH-positive neurons with pro-cesses >80 pm versus total TH-positive neurons (% TH+ LP).
*p <0.01 for L-DOPA versus controls.
up to the highest concentration tested, and quinone formation was observed only at 400 ‚aM L-DOPA (Ta-blei and Fig. 1).
Effect of L-DOPA on cell survival of all neurons
and TH-positive cells and on quinone formation
From the previous experiment, we selected 50 ‚aM L-DOPA as the optimal concentration for neurotrophic effects. Cultures were treated after 3 days in vitro with 50
‚aM
L-DOPA or solvent for 24 h or 4 days and fixed at 2 h or 4 and 7 days after plating. In both cases, L-DOPA significantly increased the survival of TH-positive labeled neurons versus controls (Fig. 3). No changes in number of TH-unlabeled neurons or in qui-none formation were observed (data not shown).FIG. 3. Effect of 50
pM
L-DOPA on TH-positive neuronal sur-vival. Cultures were exposed at 3 days postplating to vehicle or 50pM
L-DOPA for 24 h or 4 days. The cells were fixed after 2 h or 4 or 7 days postplating. Data are expressed as mean±SEM(bars) values (n = 6 cultures) of TH-positive neurons. °p<0.05, **p< 0.01 for L-DOPA versus controls.
Comparison of
D-and L-DOPA effects
Two experiments were
performed
to study the dose-and time-dependent effects ofL-and D-DOPA on post-natal midbrain neuron/cortical astrocyte cocultures. To examine shorter-term effects, L- or D-DOPA (0, 50, and 400‚aM)
was added at day5
for 48 h. L-DOPA(50 ‚aM)
increased the number and percentage of TH-positive neurons without altering quinone formation or number of TH-unlabeled neurons. In contrast, D-DOPA (50‚aM)
decreased levels of TH-positive and -negative neurons to 83 and 64% of controls, respectively. Higher levels of L-DOPA (400 ‚aM) decreased levels of TH-positive and -negative neurons to 83 and 74%, whereas 400 ‚aM D-DOPA decreased levels of TH-positive and -negative cells to 60 and 49% of controls.Quinone formation was elevated with L-DOPA (400
‚aM)
to 173% of controls and to 498% of controls with D-DOPA (Fig. 4 and Table 2).TABLE
2. Effect of L-DOPA and D-DOPA (48 h) on neuronal survival and quinone formationCells/cm2 % TH-positive Quinones (OD X I 0~) 1H-positive TH-negative Controls 560 ±33 (100%) 1,086 ±51 (100%) 34.2 ±1.4 290 ± 1.2 (100%) L-DOPA 50 pM 846 ±78° (151%) 1,109 ±48 (102%) 43.1 ± 1.8° 300 ±19 (103%) 400pM 464±43 (83%) 804 ±82 (74%) 36.6 ±3 514±72~ (173%) n-DOPA 50pM 465 ±18°~(83%) 697 ±6.9°-°(64%) 40.1 ±0.9 305 ±3.3 (102%) 400pM 338 ±63~~(60%) 530 ±7‘~°(49%) 38.3 ±5.4 1,444 ±99‘° (498%) Data are mean ±SEM values. Numbers of 1H-positive and -negative cells /cm2 are from two to six experiments (n>6 for each data point). Cultures were treated withL- and D-DOPA (0, 50, and 400pM) for 48 h after5days in culture.
“p<0.05,b~<0.01, Cp< 0.001 versus controls.
FIG. 4. Comparison ofthe effect ofL-and D-DOPA onTH-positive neurons. TH immunostaining was done of postnatal ventral midbrain
cells treated withL- and D-DOPA. Cells were treated at5 days postplating for 48 h with (A) vehicle, (B) L-DOPA (50
pM),
(C) D-DOPAL-DOPA AND POSTNATAL DOPAMINE COCULTURE
1403
TABLE 3.
Effect of L-DOPA and D-DOPA (8 days) on neuronal survival and quinone formationCells/cm2 % 1H-positive Quinones (OD X 10~) TH-positive TH-negative Controls 302 ± 17 (100%) 510 ±15 (100%) 36.9 ± 1.7 352 ± 11 (100%) L-DOPA (200pM) 260 ±33 (86%) 501 ±27 (98%) 34 ±4 556 ±25‘ (160%) D-DOPA (200pM) 89 ±26~“’(30%) 192 ±33“ (38%) 26.6 ±2“ 1,144 ±25“ (325%) Data are mean ±SEM values. Numbers of 1H-positive and -negative cells/cm2 are from two experiments (n= 8 for each data point). Cultures were treated with L-and o-DOPA (200 pM)for 8 days after 5 days
in culture.
“p <0.05,b~<0.01,‘p<0.001 versus controls.
“p<0.001 for cultures treated with o-DOPA versus L-DOPA.
To examine longer-term effects, cocultures were
treated with L-DOPA or D-DOPA (200
‚aM)for 8
days.L-DOPA (200
‚aM)did not alter cell survival, although
it produced a 158% increase in levels of extracellular
quinones. D-DOPA (200
‚aM)for 8 days decreased the
level of TH-positive neurons to 28% of controls and that of TH-negative neurons to 38% of controls.
D-DOPA increased quinone levels to
325% of controls,
producing a distinct red-brown cast to the medium and
damaged cellular morphology (Fig. 5 and Table 3).
We conclude that D-DOPA has no neurotrophic effect
and is toxic at a dose where L-DOPA is
neuroprotec-tive.
Effect of carbidopa on L-DOPA response
To examine if the neurotrophic effect that we report
was dependent on the dopamine formation from
L-DOPA, we cotreated the cultures with L-DOPA with
carbidopa. Cocultures exposed to a combination of 25
‚aM
carbidopa [a level that blocks DA formation fromL-DOPA in culture (Basma et al., 1995; Mena et al.,
1997)] and 50
‚aML-DOPA had a neurotrophic effect
similar to the effect of L-DOPA alone. Therefore,
con-version of L-DOPA to DA is not required for the
neuro-trophic effect. However, carbidopa significantly
in-creased the levels of quinones produced (Table 4).
Effect of L-BSO, L-NAC, and glutathione on
L-DOPA response
A different mechanism by which L-DOPA could
ex-ert neurotrophic effects is by up-regulating oxygen
rad-ical scavenging systems. In particular, L-DOPA has
been shown to increase GSH formation in humans
(Tohgi et al., 1995) and in culture (Han et al., 1996).
We designed experiments to
test whether GSH
synthe-sis itself could provide the neurotrophic effects.
We added exogenous GSH (200 and 400
‚aMat 5
days in vitro for 48 h) and found an increased
elabora-tion of TH-positive processes identical to that by
L-DOPA (157 and 176%, respectively, of control
lev-els). The GSH precursor, L-NAC (250 and 500
‚aM),also induced process outgrowth identical to that with
L-DOPA or GSH
and increased the level of TH-positive neurons to 146 and 182%, respectively, of control lev-els (Fig. 6 and Table 5).If L-DOPA exerts a neurotrophic effect by potentiat-ing the GSH pathway, process outgrowth should be inhibited by GSH synthesis inhibition. We found that a GSH synthesis inhibitor, BSO, at a level that had no effect itself (3 ‚aM for 24 h) completely inhibited the neurotrophic effects of 50
‚aM
L-DOPA (Table 6).Effect of L-DOPA on glutathione levels
To determine whether L-DOPA increases intracellu-lar GSH content, we examined GSH intracelluintracellu-lar levels 24 h after treatment with 50 ‚aM L-DOPA using a procedure by Tietze (1969). The assay method does not distinguish between the reduced (GSH) and oxi-dized (GSSG) forms of glutathione. However, GSSG normally only represents a small fraction (0.2%) of
TABLE 4.
Effect of L-DOPA and carbidopa (CBD; 48 h) on neuronal survival and quinone formationCells/cm2 % 1H-positive Quinones (OD X I 0~) 1H-positive TH-negative Controls 509 ±25 (100%) 1,357 ± 100 (100%) 27.3 ±2.6 544 ±16 (100%) L-DOPA (50pM) 824 ±27‘ (162%) 1,485 ± 101 (109%) 35.7 ±1.4“ 582±14 (107%) CBD (25pM) 529 ±25 (104%) 1,194 ±48 (88%) 30.7 ±1.5 555 ±18 (102%) L-DOPA/CBD 794 ±38‘(156%) 1,397 ±68 (103%) 36.2 ± 1“ 674 ± 14“ (124%) Data are mean ±SEM values. Numbers of TH-positive and -negative cells/cm2 are from two experiments (n= 10 for each data point). Cultures were treated with L-DOPA(50 pM)and/or CBD (25pM) for48 h after 5 days in culture.
DISCUSSION
FIG. 5. Comparison of the effect of long-term L-and D-DOPA
on TH-positive neurons. Micrographs show TH immunostaining of postnatal ventral midbrain cells treated with L-DOPA (200pM)
or D-DOPA (200 pM). Cells were treated at 5 days postplating
for 8 days with (A) vehicle, (B) L-DOPA (200 pM), or (C)
o-DOPA (200pM). Bar= 30 pm.
the total glutathione (Pan and Perez-Polo, 1993; Ferrari
et al., 1995), and
it was previously found that
L-DOPA-induced elevated glutathione levels in midbrain cultures were due predominantly to an increased GSH content (Mytilineou et al., 1993). We found that 50
‚aM L-DOPA increased GSH content to 137% of
con-trol levels in midbrain neurons/cortical glia cocultures (from 20.8 ± 1.3 to 28.6 ±
2.2
nmol/mg of protein;n = 8, two experiments; p <
0.01).
This study demonstrates that L-DOPA has
neuro-trophic effects on DA neurons, stimulates elaboration
of neuntes, and protects DA neurons from cell death.
These results contrast with previous studies using
em-bryonic neurons cultured without an astrocyte layer,
where L-DOPA is toxic at relatively low levels (Mena et al., 1993; Mytilineou et al., 1993; Pardo et al., 1995). In a previous study, we found that glia-condi-tioned medium protects against this toxicity (Mena et al., 1996).
The present report points to a further
in-sight, that in a coculture system using postnatal neu-rons and glia, L-DOPA is neurotrophic. In untreated cells the death of TH-positive neurons was similar to that in previous studies (Brugg et al., 1996). This death rate was reduced by >50% by L-DOPA. This
may provide an explanation for discrepancies between
the results of L-DOPA in the various models. The neu-rotrophic actions of L-DOPA may underlie the long-term improved motor control after the cessation of drug administration.
Previous studies indicate that DA added to culture medium increases the expression of TH and aromatic amino acid decarboxylase in primary cultures of fetal neurons (De Vitry et al., 1991), and the
neurotransmit-ter itself has been postulated as a neurotrophic factor
(Leslie, 1993). Normally, L-DOPA in neuronal culture
is rapidly converted to DA and taken up into synaptic
vesicles (Pothos et al., 1996). If conversion of
L-DOPA to DA is required for its neurotrophic effects,
the response should be blocked by the aromatic acid
decarboxylase inhibitor carbidopa (Basma et al.,
1995). Our study reveals therefore that the mechanism
responsible for the neurotrophic effect of L-DOPA is
independent of its conversion to DA.
Astrocytes produce several neurotrophic and neu-rite-promoting agents
(Engele et al.,
1991; Nagata et al., 1993; Hoffen et al., 1994; Takeshima et al., 1994; Muller et al., 1995) that influence development, sur-vival, neunte extension, and neurotoxin resistance (Engele et al., 1991; Nagata et al., 1993; Takeshima et al., 1994). Among the systems identified are a gluta-mate uptake activity that protects against excitotoxicity (Rosenberg et al., 1991)and stimulation of GSH
syn-thesis and enzymes involved in GSH metabolism
(Nis-ticè et al.,
1992). GSH plays a major role in protection against oxidative stress (Meister, 1988; Makar et al.,1994) and functions to remove the H
202 generated by monoamine oxidase (Werner and Cohen, 1993). GSH is produced by GSH peroxidase in culture (Raps et al., 1989).
GSH peroxidase is mainly a mitochondrial
enzyme
(Victonica et al., 1984) exclusively located inthe human midbrain within glial cells (Slivka et al.,
1987; Damier et al., 1993). GSH levels decrease with
age in neurons but remain stable in astrocytes in culture
(Sagara et al., 1996). However, neurons can maintain
an intracellular glutathione poo1 by taking up cysteine
provided by glial cells (Sagara et al., 1993).
L-DOPA AND POSTNATAL DOPAMINE COCULTURE
1405FIG. 6. Effects of L-NAC and GSH on postnatal ventral midbrain cells. Cultures were treated at 5 days postplating for 48 h with (A)
vehicle, (B) L-NAC (0.25 mM), (C) L-NAC (0.5 mM), (D) L-NAC (1 mM), (E) GSH (200 pM),or (F) GSH (400 pM). Bar= 30 pm.
Protection from oxidation enhances the survival of
mesencephalic cultures (Mena et al., 1993, 1996;
Pardo et al., 1995), and the number of TH-positive
neurons is increased if SOD, USH peroxidase, or
L-NAC is added to the medium (Colton et al., 1995). We
have shown that TH-positive neurons that overexpress
SOD display increased neutite outgrowth and survival
(Przedborski et al., 1996) and are protected from
L-DOPA toxicity (Mena et al., 1997). Recently, Han et
al. (1996) have reported that L-DOPA raises GSH
lev-els in cultures of fetal rat mesencephalon, a mouse
neuroblastoma line (Neuro-2A), a human
neuro-blastoma (SKNMC), and glia from newborn rat brain
but not in midbrain neuronal cultures grown without glia.
In the present study, the effects of L-DOPA were not
associated with an elevation in content of extracellular
quinones, in contrast with
the observations in embry-onically derived, glial-free cultures (Mena et al., 1996), suggesting that astrocytes may inhibit L-DOPA quinone formation by providing an antioxidant system. D-DOPA was not neurotrophic, although it was more effective in inducing formation of extracellular qui-nones. Moreover, we found that the GSH synthesis promoter L-NAC and GSH itself reproducedL-DO-PA‘s effects, whereas the GSH synthase inhibitor BSO
blocked L-DOPA‘ s effect. Therefore,
the present study suggests that L-DOPA is neurotrophic owing to stimu-lation of GSH peroxidase. It is possible that L-DOPA is sequestered intracellularly in glia by L-amino acidTABLE
5. Effect of L-NAC and GSH (48 h) on neuronal survival Cells/cm2 % TH-positive TH-positive TH-negative Controls 342 ±17 (100%) 919 ±46 (100%) 27.3 ±2.1 L-NAC 1 mM 365 ±42 (107%) 1,030 ±97 (111%) 26.2 ±3.1 0.5 mM 500 ±46“ (146%) 1,119 ±203 (121%) 31.2 ±3.2 0.25 mM 623 ±49“ (182%) 1,061 ±97 (115%) 37.2 ±3.1“ GSH 200 pM 539 ±16“(157%) 1,026 ± 106 (112%) 34.1 ±1.9“ 400 pM 603 ±62‘ (176%) 946 ±92 (103%) 39.1 ±1.1‘ Data are mean±SEM values. Numbers of TH-positive and -negative cells/cm2 are from two experiments (n= 10 foreach data point). Cultures were treated with L-NAC or GSH for 48 h after 5 days in culture,“p<0.05, h~< 0.01, ‘p <0.001 versus controls.
transport systems and in turn stimulates astrocytic GSH
peroxidase.
It appears unlikely that L-DOPA provides its neuro-trophic effect by acting as an antioxidant itself because
high doses of L-DOPA do not provide neuroprotection
and increase quinone formation; however, a cautionary note is provided by the finding that L-NAC, previously
assumed to act as an antioxidant due to its role as a
substrate for GSH synthesis, may protect PC12 cells by acting directly as an oxygen radical scavenger (Yan et aI., 1995). L-NAC is a pluripotent protector that elevates the level of intracellular glutathione, prevents
against apoptotic death of neuronal cells, and enhances
trophic factor-mediated cell survival (Mayer and No-ble, 1994; Ferrari et al., 1995). Indeed, depletion of brain GSH with BSO in rats is accompanied by im-paired mitochondrial function and decreased complex
IV activity (Heales et al., 1995)
leading to energy crisis as a mechanism of nigral cell death (Jenner,1993).
In clinical studies, GSH content is decreased in the
substantia nigra of patients with idiopathic PD
(Ried-erer et al., 1989; Sofic et al., 1992; Sian et al., 1994).
L-DOPA treatment reverses GSH depletion (Tohgi et
al., 1995). In normal subjects, the nigrostriatal DA
neurons, the most important disease target associated
with PD, are normally surrounded by a low density of glia (Damier et al., 1993; Sagara et al., 1993; Makar et al., 1994), suggesting that these neurons are less protected from oxidative stress. It is striking that, in PD, the density of GSH peroxidase-positive glia
sur-rounding the
surviving DA neurons is increased, sug-gesting a compensatory proliferation of glia that pro-tects surviving neurons against pathological death (Da-mier et al., 1993).Although L-DOPA is toxic for catecholamine
neu-rons in other in vitro systems (Melamed, 1986; Mena
et aI., 1992, 1993; Pardo et al., 1995),
catecholamines are required for the development of the nervous system and survival. Inactivation of the TH gene produced the death of 90% of mutant mice fetuses, whereas adminis-tration of L-DOPA to pregnant females resulted in complete rescue of mutant mice (Zhou et al., 1995).Transgenic animals with mutations of the TH gene
have a poorer prognosis than those with mutations of the dopamine /3-hydroxylase gene (Thomas et al., 1995), suggesting that DA is essential for development of the nervous system. It is likely that L-DOPA and catecholamines are critical factors for promoting the survival and differentiation of DA neurons and that the effects of L-DOPA on DA cells in vivo might be neurotrophic or neurotoxic depending on the local en-vironment produced by glial cells.
In summary, conclusions from this study and that
of Mena et al. (1996) indicate that a long-term potenti-ation of DA release by L-DOPA may occur owing to promotion of neunitic arbonization and protection
against cell death. The neurotrophic effects are
inde-pendent of conversion to DA and deinde-pendent on the
presence of astrocytes and may result from antioxidant
mechanisms such as up-regulation of GSH. Significant
questions that need to be addressed include whether
the effect is due to L-DOPA itself acting as an
antioxi-TABLE 6.
Effect of L-DOPA and BSO (48 h) on neuronal survival and quinone formationCells/cm2 % TH-positive Quinones (OD X l0~) TH-positive TH-negative Controls 450 ±18 (100%) 1,303 ±60 (100%) 25.7 ±2.1 285 ±1.8 (100%) L-DOPA (50pM) 720 ±72“ (160%) 1,408 ±143 (108%) 33.9 ±0.8~‘ 284 ±7 (99.7%) BSO (3pM) 436 ±17 (96.8%) 1,160 ±72 (89%) 26.7 ±1.05 293 ±3 (103%) L-DOPA +BSO 441 ±40“ (98%) 1,329 ±11 (102%) 24.6 ±0.8 276 ±4 (96.7%)
Data are mean ±SEM values. Numbers of TH-positive and -negative cells/cm2 are from two experiments (n= 10 for each data point). Cultures were treated after 5 days in vitro with 50 pML-DOPA (for 48 h) or 3pML-BSO alone or in combination. Pretreatment with BSO for 24 h was followed by washout with neuronal medium, before L-DOPA or solvent was added.
b~<0.01 versus controls.
L-DOPA AND POSTNATAL DOPAMINE COCULTURE
1407dant or if it is due to up-regulation of a cellular system
such as the GSH peroxidase pathway, whether seques-tration of DA by synaptic vesicles plays a role in pro-tection from L-DOPA toxicity, and the precise path-ways of exchange of cysteine or other protective mech-anisms between glial cells and DA neurons.
Acknowledgment:
This work was funded by the Parkin-son‘sDisease Foundation, grants 10154 and 07418 from theNational Institute on DrugAbuse(to
D.S.), and grants PR
95-098 and SAF96-1099 from the Spanish MEC. Our thanksto Johanna Bogulavsky andIrma Ryjak for expert technical
assistance and to Mrs. Liselotte Gulliksen for editorial help.
REFERENCES
Abercrombie E. D., Bonatz A. E., and Zigmond M. J. (1990) Effects of L-DOPA on extracellular dopamine in striatum of normal and 6-hydroxydopamine-treated rats. Brain Res. 52, 536—544. Basma A. N., Morris E. J., Nicklas W. J., and Geiler H. M. (1995) L-DOPA cytotoxicity to PC 12 cells in culture is via its autoxida-tion. J. Neurochem. 64, 825—832.
Blint J., Bonnet A. M., and Agid Y. (1988) Does L-DOPA aggravate Parkinson‘s disease?Neurology38, 1410—1416.
Blunt S. B., Jenner P., and Marsden C. D. (1993) Suppressive effect ofL-DOPA on dopamine cells remaining in the ventral tegmen-tal area of rats previously exposed to the neurotoxin 6-hydroxy-dopamine. Mov. Disord. 8, 129—133.
Bradford M. M. (1976) A rapid and sensitive method for the quanti-tation of microgram quantities of protein utilizing the principle of protein—dye binding. Anal. Biochem. 72, 248—254. Brugg B., Michel P. P., Agid Y., and Ruberg M. (1996) Ceramide
induces apoptosis in cultured mesencephalic neurons. J.
Neuro-chem. 66, 733—739.
Colton C. A., Pagan F., Snell J., Colton J. S., Cummins A., and Gilbert D. L. (1995) Protection from oxidation enhances the survival of cultured mesencephalic neurons. Exp. Neurol. 132, 54—6 1.
Cubells J. F., Rayport S., Rajendran G., and Sulzer D. (1994)Meth-amphetamine neurotoxicity involves vacuolation of endocytic organelles and dopamine-dependent intracellular oxidative stress. J. Neurosci. 14, 2260—2271.
Damier P., Hirsch E. C., Javoy-Agid F., Zhang P., and Agid Y. (1993) Protective role of glutathione peroxidase against neu-ronal death in Parkinson‘s disease. Neuroscience52, 1—6.
Denis-Donini S., Glowinski J., and Prochiantz A. (1983) Specific influence of striatal target neurons on the in vitro outgrowth of mesencephalic dopaminergic neuntes: a morphological quanti-tative study. J.Neurosci.
3,
2292—2299.De Vitry F., Hillion J., Catelon J., Thibault J., Benoliel J., and Hamon M. (1991) Dopamine increases the expression of tyro-sine hydroxylase and aromatic amino acid decarboxylase in primary cultures of fetal neurons. Dey. Brain Res. 59, 123—
131.
Di Porzio U., Daguet M. C., Glowinski J., and Prochiantz A. (1980) Effect of striatal cells on in vitro maturation of mesencephalic dopaminergic neurons in serum-free conditions. Nature 288, 370—374.
Du X. and lacovitti L. (1995a) Synergy between growth factors and catecholamines (CA) required for novel expression of tyrosine hydroxylase in non-CA brain neurons. Soc.Neurosci.Abstr. 21,
1053.
Du X. and lacovitti L. (l995b) Synergy between growth factors and transmitter required for catecholamine differentiation in brain neurons. J. Neurosci. 15, 5420—5427.
Engele J., Schubert D., and Bohn M. C. (1991) Conditioned media derived from glial cell lines promote survival and differentiation of dopaminergic neurons in vitro: role of mesencephalic glia.
J. Neurosci. Res.30, 359—371.
Ferrari G., Irene C. Y., and Greene L. A. (1995) N-Acetylcysteine (D-and L-stereoisomers) prevents apoptotic death of neuronal cells. J.Neurosci. 15, 2857—2866.
Garns P. A., Ciolkowski E. L., Pastore P., and Wightman R. M. (1994) Efflux of dopamine from the synaptic cleft in the nucleus accumbens of the rat brain. J. Neurosci. 14, 6084—6093. Gasser T., Wszolek Z. K., Trofatter J., Ozelius B. S., Uitti R. J., Lee
C. S., Gusella J., Pfeiffer R. F., Calne D. B., and Breakefield X. 0. (1994) Genetic linkage studies in autosomal dominant parkinsonism: evaluation of seven candidate genes. Ann.
Neu-roi. 36, 387—396.
Han S-K., Mytilineou C., and Cohen G. (1996) L-DOPA up-regu-lates glutathione and protects mesencephalic cultures against oxidative stress. J. Neurochem. 66, 501—510.
Harden D. C. and Grace A. A. (1995) Activation of dopamine cell firing by repeated L-DOPA administration to dopamine-depleted rats: its potential role in mediating the therapeutic response to L-DOPA treatment. J.Neurosci. 15, 6157—6166.
Heales S. J. R., Davies S.E.C., Bates I.E., and Clark J. B. (1995) Depletion of brain glutathione is accompanied by impaired mi-tochondnial function and decreased N-acetyl aspartate concen-tration. Neurochem. Res. 20, 3 1—38.
Hefti F., Melamed E., Bhawan J., and Wurtman R. J. (1981) Long-term administration of L-DOPA does not damage dopaminergic neurons in the mouse.Neurology 31, 1194—1195.
Hirano T. and Kasano K. (1992) Spatial distribution of excitatory and inhibitory synapses on a Purkinje cell in a rat cerebellar culture. Soc. Neuro.sci. Abstr. 18, 1112.
Hoffen B. J., Hofman A., Bowenkamp K., Huettl P., Hudson J., Martin D., Lin L. F. H., and Gerhardt G. A. (1994) Glial cell line-derived neurotrophic factor reverses toxin-induced injury to midbrain dopaminergic neurons in vivo.Neurosci. Lert.182,
107—111.
lacovitti L. A. (1996) Factors regulating the induction of a dopamin-ergic phenotype in neurons grown in culture, inMesencephalic
Dopaminergic Geil Cultures Congress, p. 13. National Institutes
of Health, Bethesda, Maryland.
Jenner P. (1993) Altered mitochondrial function, iron metabolism and glutathione levels in Parkinson‘s disease. Acta Neurol.
Scand. 145, 6—13.
Leslie F. M. (1993) Neurotransmitters as neurotrophic factors, in
Neurotrophic Factors (Fallon J. H. and Loughlin S. E., eds),
pp. 565—598. Academic Press, San Diego.
Lesser R. P., Fahn S., Snider S. R., Cote L. J., Isgreen W. P., and Barret R. E. (1979) Analysis of the clinical problems in parkin-sonisms and the complications of long-term L-DOPA therapy. Neurology 29, 1253—1260.
Makar T. K., Nedergaard M., Preuss A., Gelbard A. S., Perumal A. S., and Cooper A. J. L. (1994) Vitamin E, ascorbate, gluta-thione, glutathione disulfide, and enzymes of glutathione metab-olism in cultures ofchick astrocytes and neurons: evidence that astrocytes play an important role in antioxidative processes in the brain. J. Neurochem. 62, 45—53.
Markham C. H. and Diamond S. G. (1986) Long term follow up of early DOPA treatment in Parkinson‘s disease. Ann. Neural. 19, 365 —372.
Mayer M. and Noble M. (1994) N-Acetyl-L-cysteine is a pluripotent protector against cell death and enhance of trophic factor-medi-ated cell survival in vitro. Proc. Natl. Acad. Sci. USA 91, 7496— 7500.
Meister A. (1988) Glutathione metabolism and its selective modifi-cation. J. Biol. C/scm. 263, 17205—17208.
Melamed E. (1986) Initiation of L-DOPA therapy in parkinsonian patients should be delayed until the advanced stages of the disease. Arch. Neurol. 43,402—405.
Mena M. A., Pardo B., Casarejos M. J., Fahn S., and Garcia de Yebenes J. (1992) Neurotoxicity of L-DOPA on catecholamine-rich neurons. Mov. Disord. 7, 23—31.
Mena M. A., Pardo B., Paino C. L., and Garciade Yebenes J. (1993) L-DOPA toxicity in foetal rat midbrain neurones in culture: modulation by ascorbic acid. Neuroreport 4, 438—440. Mena M. A., Casarejos M. J., Paino C. L., and Garcia de Yebenes
J. (1996) Glia conditioned medium protects fetal rat midbrain neurons in culture from L-DOPA toxicity. Neuroreport 7, 441— 445.
Mena M. A., Khan U., Togasaki D. M., Sulzer D., Epstein C. J., and Przedborski S. (1997) Effects of wild-type and mutated copper/ zinc superoxide dismutase on neuronal survival and L-DOPA-induced toxicity in postnatal midbrain culture. J. Neurochem. 69, 21—33.
Muller H. W., Junghans U., and Kappler J. (1995) Astroglial neuro-trophic and neunite-promoting factors. Pharmacol. Ther. 65, 1—
18.
Mytilineou C., Han S-K., and Cohen G. (1993) Toxic and protective effects of L-DOPA on mesencephalic cell cultures. J.
Neuro-c/scm. 61, 1470—1478.
Nagata K., Takei N., Nakajima K., Saito H., and Kohsaka S. (1993) Microglial conditioned medium promotes survival anddevelop-ment of cultured mesencephalic neurons from embryonic rat brain. J. Neurosci. Res.
34,
357—363.Nisticè G., Ciniolo M. R., Fiskin K., lannone M., De Martino A., and Rotillo G. (1992) NGF restores decrease in catalase activity
and increase superoxide dismutase and glutathione peroxidase activity in the brain of the aged rats. Free Radie. Biol. Med.
12,
177—181.Nutt J. G. (1995) Long-duration response to levodopa. Neurology 45, 1613—1616.
PanZ. and Perez-Polo R. (1993) Role of nerve growth factor in oxidant homeostasis: glutathione metabolism. J. Neurochem. 61, 1713—1721.
Pardo B., Mena M. A., Casarejos M. J., Paino C. L., and Garcia de Yebenes J. (1995) Toxic effects of L-DOPA on mesencephalic cell cultures: protection with antioxidants. Brain Res. 682, 133— 143.
Perry T. L., Wee Yong V.,ItoM., Foulks J. G., Wall R. A., Godin D. V., and Clavier R. M. (1984) Nigrostriatal dopamine neu-rons remain undamaged in rats given high doses of L-DOPA and carbidopa chronically. J. Neurochem.43,990—993. Pothos E., Desmond M., and Sulzer D. (1996)
L-3,4-Dihydroxyphe-nylalanine increases the quantal size of exocytotic dopamine release in vitro. J. Neurochem. 66, 629—636.
PrzedborskiS., Khan U., Kostic V., Carlson E., Epstein C. J., and
SulzerD. (1996) Increased superoxide dismutase activity im-proves survival of cultured postnatal midbrain neurons. J.
Neu-rochem. 67, 1383—1392.
Quinn N., Parker D., Janota I., and Marsden C. D. (1986) Preserva-tion of the substantia nigra and locus coeruleus in a patient receiving L-DOPA (2Kg) plus decarboxylase inhibitor over a four-year period. Mov. Disord. 1, 65—68.
Raps S. P., Lai J.,HertzI., and Cooper A. J. L. (1989) Glutathione is present in cultured astrocytes but flot in cultured neurons.
Brain Res. 493, 398—401.
Rayport S.and SulzerD. (1995) Visualization of antipsychotic drug binding to living mesolimbic neurons reveals D2 receptor, aci-dotropic, and lipophilic components. J. Neurochem.
65,
691— 703.Rayport S., Sulzer D., Shi W-X., Sawasdikosol S., Monaco J., Bat-son D., and Rajendran G. (1992) Identified postnatal mesolim-bic dopamine neurons in culture: morphology and electrophysi-ology. J. Neurosci.
12,
4264—4280.Riederer P., Sofic E., Rausch W-D., Schmidt B., Reynolds G. P., Jellinger K., and Youdim M. B. H. (1989) Transition metals,
fernitin, glutathione, and ascorbic acid in parkinsonian brains.
J. Neurochem. 52, 5 15—520.
Rosenberg P. A., Loning R., Xie Y., Zaleskas V., and Aizenman E. (1991) 2,4,5-Trihydroxyphenylalanine in solution forms a non— N-methyl-D-aspartate glutamatergic agonist and neurotoxin.
Proc. Nati. Acad. Sei. USA 88, 4865—4869.
Sagara J., Miura K., and Bannai S. (1993) Maintenance of neuronal glutathione by glial cell. J. Neurochem. 61, 1672—1676. Sagara J., Makino N., and Bannai S. (1996) Glutathione efflux from
cultured astrocytes. J. Neurochem. 66, 1876—1881.
Sian J., Dexter D. T., Lees A. J., Daniel S. E., Agid Y., Javoy-Agid F., Jenner P., and Marsden C. D. (1994) Alterations in glutathione levels in Parkinson‘s disease and other neurodegen-erative disorders affecting basal ganglia.Ann. Neurol. 36, 348— 355.
Slivka A., Mytilineou C., and Cohen G. (1987) Histochemical evalu-ation of glutathione in brain. Brain Res.
409,
275—284. Sofic E., Lange K. W., Jellinger K., and Riederer P. (1992) Reducedand oxidized glutathione in the substantia nigra of patients with Parkinson‘s disease. Neurosci. Lert. 17, 128—130.
Sundaresan M., Yu Z-X., Ferrans V. J., Irani K., and Finkel T. (1995) Requirement for generation of H
2O2 for platelet-derived growth factor signal transduction. Science 270, 296—299. Takeshima T., Shimoda K., Sauve Y., and Commissiong J. W.
(1994) Astrocyte-dependent and independent phases ofthe de-velopment and survival of rat embryonic day 14 mesencephalic neurons in culture. Neuroscience 60, 809—823.
Thomas S. A., Matsumoto A. M., and Palmiter R. D. (1995) Nor-adrenaline is essential for mouse fetal development.Nature374, 643—646.
Tietze F. (1969) Enzymic method for quantitative determination of nanogram amounts oftotal and oxidized glutathione: application to mammalian blood and other tissue. Anal. Biochem. 27, 502— 522.
Tohgi H., Abe T., Saheki M., Hamato F., Sasaki K., and Takahashi S. (1995) Reduced andoxidized forms of glutathione and alpha-tocopherol in the cerebrospinal fluid of parkinsonian patients: comparison between before and after L-dopa treatment.
Neu-rosci. Lett. 184, 21—24.
Uitti R. J., Ahlskog J. E., Maraganore D. M., Muentner M. D., At-kinson E.J., Cha RH., and O‘Brien P.C. (1993) L-DOPA therapy and survival in idiopathic Parkinson‘s disease: Olmsted County project. Neurology 43, 1918—1926.
Victorica J., Machado A., and Satndsteguï J. (1984) Age-dependent variations in peroxide-utilizing enzymes from rat brain mito-chondnia and cytoplasm. J. Neurochem. 42, 351—356. Werner P. and Cohen G. (1993) Glutathione disulfide (GSSG) as
a marker of oxidative injury to brain mitochondria. Ann. NY
Acad. Sei. 679, 364—369.
Yan C. Y. I., Ferrari G., and Greene L. A. (1995) N-Acetylcysteine-promoted survival of PC12 cells is glutathione-independent but transcription-dependent. J. Biol. Chem. 270, 26827—26832. Yurek D. M., Steece-Collier K., Collier T. J., and Sladek J. R.
(1991) Chronic L-DOPA impairs the recovery of dopamine agonist-induced rotational behavior following neural grafting.
Exp. Brain Res. 86, 97—l07.
Zhou
