Departments of
Neurology (M.A.M, V.D., J.B., D.S.) and
Psychiatry
(D.S.),
Columbia University and Department of Neuroscience (D.S.), New
York State Psychiatric Institute, New York, NY 10032
 |
Introduction |
The
catecholamine precursor L-DOPA elicits either neurotoxic or
neurotrophic responses depending on experimental conditions. Toxic
effects have been reported in neuroblastoma lines and catecholamine neurons cultured in the absence of astrocytes; this seems to be caused
by a promotion of oxygen radicals, because the toxicity is prevented by
the antioxidant ascorbic acid (Mena et al., 1993
; Pardo
et al., 1993; Pardo et al., 1995), overexpression
of superoxide dismutase (Mena et al., 1997c), and the
monoamine oxidase inhibitor deprenyl (Mena et al., 1992),
which inhibits production of hydrogen peroxide (Cohen and Spina, 1989).
Moreover, L-DOPA toxicity is correlated with high extracellular levels
of quinone oxyradical L-DOPA derivatives (Graham, 1978; Mena et
al., 1992; Basma et al., 1995).
Paradoxically, lower L-DOPA exposures (25-200 µM) are
selectively neurotrophic for DA neurons in ventral midbrain astrocyte cocultures, promoting cell survival (Mytilineou et al.,
1993; Mena et al., 1997b) and neurite outgrowth (Mena
et al., 1997b). The trophic effects may be caused by
factors synthesized in astrocytes that are upregulated by L-DOPA
exposure (Han et al., 1996; Mena et al.,
1997b). Indeed, astrocyte-conditioned medium alone protects against L-DOPA neurotoxicity in glial-free cultures of midbrain DA
neurons (Mena et al., 1997a), as would be expected if
soluble neuroprotective glial-derived factors were released by
astrocytes.
Recently, it has become evident that neurotrophic factors affect
neurotransmitter release as well as cell survival and development. The
striking ability of L-DOPA to induce neurite outgrowth in culture may
provide a basis for long-lasting changes in transmitter release, and
the often observed delay in full therapeutic benefit after L-DOPA
treatment for Parkinsonian disorders. Although target-derived growth
factors (including glial-derived neurotrophic factor) increase neurite
arborization of midbrain DA neurons (Lin et al., 1993; Pothos et al., 1998), there is at present no method to
trigger novel initiation of central DA neurite outgrowth. We therefore examined PC12 cells, a DA-secreting pheochromocytoma cell line that
produces neurites on exposure to NGF (Greene and Tischler, 1976). We
find that L-DOPA potentiates neurotrophin-elicited outgrowth and that
this response results in elevated quantal neurotransmission from the
neurite varicosities. An emerging hypothesis from these studies as well
as studies in other systems (Sundaresan et al., 1995) is
that oxidative metabolism strongly influences the presynaptic neurosecretory apparatus via a synergistic response with neurotrophic factors.
| |
Materials and Methods |
Cell culture.
PC12 cells were obtained from Dr. Lloyd Greene
(Department of Pathology Columbia University, New York, NY) and
cultured as described (Greene and Tischler, 1976) using RPMI 1640 medium supplemented with 10% heat-inactivated horse serum and 5%
fetal bovine serum (JRH Biosciences, Lenexa, KS). The cells were plated
onto plastic dishes or multiwell plates (Costar, Cambridge, MA) coated
with rat tail collagen (Vitrogen 100; Collagen Biomedical, Palo Alto, CA) or glass coverslips (Carolina Biological Supply #2, Burlington, NC)
coated overnight with 40 µg/ml poly-d-Lys (molecular
mass, 70-150 kDa; Sigma, St. Louis, MO) and then recoated with
10 µg/ml laminin for 2 hr (Collaborative Biomedical, Bedford, MA).
Cell density platings were as follows: for fluorescent observations, processes measurements, and electrophysiology, 40,000 cells per ml were
plated on 1-cm2 glass coverslips in a 50-mm
diameter Petri dish (total medium volume, 2.5 ml); for cell counts,
50,000 cells were plated per well in 24-well plates (total medium
volume, 1 ml); for GSH and protein measurements 400,000 cells were
plated in 6-well plates (total medium volume, 4 ml). NGF (human
recombinant) was donated by Genentech (South San Francisco, CA) and
used at a concentration of 50 ng/ml. Sulpiride was from Research
Biochemicals (Natick, MA); other compounds were obtained from Sigma
except where noted.
Cell measurements.
We adapted semiquantitative methods for
estimating the number and elaboration of processes in culture
(Denis-Donini et al., 1983). In each culture, 8-10
consecutive fields (magnification, 100× or 200×) were counted under
phase optics. Results are expressed as mean ± standard error
values for two to four independent experiments. Each data point
corresponds to four or more cultures. Images were produced using an
inverted microscope (Axiovert 135 TV; Carl Zeiss, Thornburg, NY), a
digital camera (Star I CCD; Photometrics, Tucson, AZ) and National
Institutes of Health Image software.
Electrochemical techniques.
Amperometric detection of DA
secretion in real time was as described (Pothos et al.,
1996; Pothos et al., 1998) except that electrodes were
placed on processes rather than cell bodies. The cultures were examined
1-2 weeks after plating. Data was analyzed using a locally written
program in the Superscope II environment (GW Instruments, Sommerville,
MA). The average background current in the vicinity of the spikes was
subtracted from the signal and spikes were identified if their
amplitude was 4.5 times greater than the root-mean-square background
current. The number of molecules (quantal size) oxidized at the
electrode face was determined by the relation N = Q/nF, where Q is the charge of the
spike, n is the number of electrons transferred [shown to
be two for catecholamines when used in a similar experimental
configuration (Ciolkowski et al., 1994)], N is
the number of moles and F is Faraday's constant (96,485 coulombs per
equivalent). Additional parameters that were measured for each quantal
event were: imax, the maximum
amplitude; t1/2 the width
at half imax; width, defined as
duration between the start of the upward slope and the point at which
the decay reaches baseline; and the interspike interval, a parameter
that indicates the release frequency.
Vital nuclear staining.
The membrane-permeant bisbenzimide
dye Hoechst 333342 (Molecular Probes, Eugene, OR) was used to stain
nuclei to identify apoptotic nuclei. A concentrated stock (2 mg/ml) was
prepared in water and sterilized by filtration. The stock was diluted
1:50 in sterile phosphate-buffered saline (22.8 mM
monobasic sodium phosphate, 76.8 mM dibasic sodium
phosphate, 154 mM sodium chloride, pH 7.3) 10 µl
was added directly to 100 µl of physiological medium (see below),
yielding a final concentration of 4 µg/ml.
Total cell counts.
For counts of living cells, plasma
membranes were lysed as described (Ferrari et al., 1995) and
intact nuclei with evident limiting membranes counted (10-16 cultures
per condition).
GSH measurements.
GSH levels were measured by the method of
Tietze (1969). Briefly, 2 × 106 cells were
washed twice with phosphate-buffered saline, lysed with 3% perchloric
acid for 15 min at 41/4, and centrifuged; supernatants were
neutralized with 9 volumes of 0.1 M
NaH2PO4, 5 mM
EDTA, pH 7.5. GSH content was measured by the addition of
5,5'-dithio-bis-(2-nitrobenzoic acid) and the reaction was monitored at
412 nm. GSH is expressed as a function of total protein of the cell
extract. In concordance with previous studies in PC12 cells (Ferrari
et al., 1995; Pan and Perez-Polo, 1996), no oxidized
glutathione was detectable in the cells by the method of Griffith
(1980). Protein was measured in the pellet by the method of
Bradford (1976) using bovine serum albumin as the standard.
Measurement of quinones.
Quinone formation, which can be
used to observe the rate of L-DOPA autoxidation, was evaluated
according to the spectrometric measurement of the absorbance at 490 nm
in the culture media (Mena et al., 1992; Mena et
al., 1993).
DCF label in culture.
Aliquots of 10 mM DCF were
prepared in dimethylsulfoxide and stored at
85o. Immediately before use, an aliquot was
diluted to 1 mM in physiological medium and the cells were
labeled with DCF (50 µM, 30 min) in physiological medium.
The cultures were then washed three times in Hanks' balanced saline
solution containing 10 mM HEPES buffer and 10 mM glucose.
Statistical analysis.
The results were statistically
evaluated for significance by ANOVA for multiple groups followed by
Tukey-Kramer post hoc testing for pairs where appropriate.
Differences were considered significant at p < 0.05. For comparison of quantal populations, the nonparametric Komolgorov-Smirnov statistic (Van der Kloot, 1991) was used (GB-Stat; Dynamic Microsystems, Silver Spring, MD).
| |
Results |
Induction of neurites by L-DOPA.
To test whether L-DOPA alone
might induce neurite outgrowth of undifferentiated (i.e., NGF-naive)
PC12 cells, we exposed the cultures to 12-200 µM L-DOPA
alone for 24-48 hr. In all cases, untreated cultures contained 
1% of cells with neurites. L-DOPA elevated the percentage of cells
that exhibited neuritic processes at 50, 100, and 200 µM
levels, although the affect was attenuated at the highest concentration
(Fig. 1, leftmost bars). These
neurites were relatively delicate in appearance and generally retracted about 3 days after the initial L-DOPA exposure, even in trials where
L-DOPA levels were maintained. Between different sets of cultures, the
percentage of cells that expressed temporary neurite induction varied
from 20-75%. Therefore, all experiments reported used sister cultures
derived from the same parent culture, which, as indicated in Fig. 1,
showed low variance.
View larger version (43K):
[in this window]
[in a new window]
|
Fig. 1.
Effect of L-DOPA and NGF on induction of neurites.
PC12 cultures at 2 days after plating were treated for 24 hr with
L-DOPA with or without NGF (50 ng/ml). The percentage of cells that
exhibited neurites are reported as mean ± standard error
(n = 10 cultures). L-DOPA increased neurite
expression both by itself (leftmost bars) and with NGF
(rightmost bars; p < 0.001 by ANOVA
for both cases).
|
|
L-DOPA (24 hr) in combination with NGF synergistically increased the
number of cells that exhibited processes compared with NGF treatment
alone (Fig. 1, rightmost bars) as well as the length and
arborization (see below). In contrast to treatments with L-DOPA alone
that produced temporary neurite outgrowth, there was little variance
between groups of sister cultures.
Because L-DOPA has been reported to produce apoptosis in PC12 cells
(Walkinshaw and Waters, 1995), we examined the occurrence of apoptosis
in undifferentiated PC12 cultures at 2-3 days after plating using the
nuclear stain Hoechst 33342 (Fig. 2,
inset). We found that 50 µM L-DOPA (48 hr) did
not increase apoptotic profiles, whereas 100 µM L-DOPA
increased apoptosis by nearly 30-fold (Fig. 2,
leftmost columns). NGF (50 ng/ml) effectively blunted L-DOPA-induced apoptosis (Fig. 2, rightmost
columns). In agreement with a previous study (Basma et
al., 1995), L-DOPA induced an overall dose-dependent decrease in
cell number (Fig. 3, leftmost
columns). This effect was also effectively blunted by NGF (Fig. 3,
rightmost columns). Note that the total number of cells in
groups treated with NGF was lower, because this factor differentiates
the cells and halts mitosis.
View larger version (43K):
[in this window]
[in a new window]
|
Fig. 2.
Effect of L-DOPA and NGF on apoptosis. To test
induction of apoptosis by L-DOPA, cultures at 2 days after lating were
treated for 24 hr with L-DOPA (0, 50 and 100 µM) with and
without NGF (50 ng/ml). The number of apoptotic cells labeled with
Hoechst 33342 are indicated. Results are expressed as mean ± standard error (n = 8 cultures). NGF protected the
cells from L-DOPA-induced apoptosis (p < 0.001, ANOVA). Inset, four stained nuclei;
arrow, apoptotic profile. Scale bar, 5 µm.
|
|
View larger version (48K):
[in this window]
[in a new window]
|
Fig. 3.
Effect of L-DOPA and NGF on cell number. PC12
cultures at 2-3 days after plating were treated with L-DOPA (0, 12, 25, 50, 100, and 200 µM) for 24 hr with and without NGF
(50 ng/ml). The number of surviving cells (as represented by the number
of intact nuclei) are indicated. Results are expressed as mean ± standard error (n = 10-16 cultures). L-DOPA
decreased the cell number both with and without NGF
(p < 0.001, ANOVA); however, this effect was
reduced in the presence of NGF. Note that the total numbers of cells
was reduced by NGF alone, as it induced differentiation and inhibited
cell division.
|
|
L-DOPA-mediated cell death could be caused by elevated intracellular
oxidative stress. To examine this, we adapted methods used previously
with cultured midbrain dopamine neurons to detect intracellular
oxyradicals caused by methamphetamine exposure (Cubells et
al., 1994). This approach uses DCF, a membrane-permeant
fluorogenic compound that is trapped in living cells after
de-esterification; in the presence of peroxides or hydroxy radicals,
this compound is converted to a fluorescent derivative. DCF has
recently been used to examine oxidative stress in PC12 cells using flow
cytometry (Aoshima et al., 1997). If oxyradicals are induced
within neurons by L-DOPA, we would expect to see an increase in
fluorescent cell bodies. Indeed, we found that a low percentage of the
cells were labeled by DCF, and that the number of labeled cells
increased 12-fold after high levels of L-DOPA (Table
1). The label generally appeared to have
a granular distribution (not shown), similar to the pattern of label in
cell bodies of midbrain dopamine neurons after methamphetamine
treatment (Cubells et al., 1994). Although NGF alone did not
reduce the number of DCF-positive cells not exposed to L-DOPA, NGF
inhibited intracellular oxidative stress observed after L-DOPA
administration (Table 1). Because of striking effects on cell
number, apoptosis, and intracellular oxidative stress observed with 200 µM L-DOPA, we chose to use 50 µM L-DOPA for
our experimental protocols.
View this table:
[in this window]
[in a new window]
|
TABLE 1
Effects of L-DOPA and NGF on intracellular DCF label
Five days after plating, cultures were exposed to 0, 50, or 200 µM L-DOPA with or without NGF (50 ng/ml) for 4 days. The
cultures were then examined for cells labeled by DCF. Data are
mean ± standard error (n = 6; at least 2000 cells
counted per condition).
|
|
Long term L-DOPA exposure potentiates quantal DA release from
neurites.
To test whether L-DOPA induced functional presynaptic
changes, we adapted recently developed techniques to measure quantal release of catecholamines directly from PC12 neurite varicosities (Zerby and Ewing, 1996). Preliminary observations indicated that quantal release from the NGF-naive processes was extremely rare (three
events observed from one site; Davila V, unpublished observations), probably related to the destability and gradual loss of processes that
occurred unless NGF was present.
After coexposure to NGF alone or NGF and L-DOPA, the electrodes were
placed over process varicosities (Fig. 4,
inset); quantal events were not observed along the neurites
between the varicosities. Depolarization with high
K+ promoted quantal release from the varicosities
(Fig. 4, A and B). Compared with cells exposed to NGF alone, cells
exposed to L-DOPA (50 µM for 48 hr at 2 days after
plating) and constant NGF released quanta composed of twice the number
of molecules, even > 1 week after exogenous L-DOPA was washed
out. The cumulative distribution of the cubed root of the quantal
sizes, which is normally distributed (Finnegan et al., 1996;
Pothos et al., 1996), demonstrates that the entire
population of quanta was shifted to larger values (Fig. 4C). The mean
interspike interval was ~2-fold shorter in duration in cells exposed
to both L-DOPA and NGF (Table 2; Fig. 4D). Therefore, the frequency of
release as well as quantal size was increased by a single 48 hr
exposure to L-DOPA. As the value of the standard deviation of the
interspike interval divided by the mean interval was unchanged, L-DOPA
did not elicit bursting release; rather, it reduced interspike
intervals to produce a greater number of release events per
stimulation. Together, the data indicate that L-DOPA potentiated
release from NGF-treated varicosities by increasing quantal size, the
number of quantal events (i.e., quantal content), and the release
frequency.
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 4.
Effect of L-DOPA on quantal release. A, An
amperometric recording of quantal release from a representative PC12
varicosity 1 week after NGF exposure (50 ng/ml). Secretion was
stimulated by 80 mM K+ administered for 3 sec
(arrow), evoking multiple quantal events.
Inset, example of a representative PC12 varicosity.
Scale bar; 2 µm. B, An amperometric recording from a
representative PC12 varicosity 1 week after exposure to both NGF and 50 µM L-DOPA. C, Cumulative distribution of the cubed root
of the quantal sizes of control and L-DOPA-treated cultures as reported
in Table 1. The cubed root of the number of DA molecules are displayed.
The population of quantal sizes in the L-DOPA and NGF-treated group is
shifted to higher values than NGF alone (p < 0.0001; Komolgorov-Smirnov test, KS-Z = 2.2258). D, The
cumulative distribution of intervals between quantal events reported in
Table 1. Cells treated with L-DOPA and NGF have a mean frequency of
release ~ 2-fold higher than NGF alone (Table 1) and the
population as a whole is shifted to shorter interspike intervals
(p < 0.005, KS-Z = 2.0442).
|
|
DA is not required for neurite induction.
The finding that
acidic fibroblast growth factor and catecholamines synergistically
up-regulate tyrosine hydroxylase activity in DA neurons in culture
(Stull and Iacovitti, 1996) suggests that an analogous action by NGF
and DA may occur after upregulation of DA levels because of exposure to
L-DOPA. To test whether DA synthesis was required for L-DOPA-induced
neurite outgrowth, we examined the effects of carbidopa (25 µM), which blocks DA synthesis at this exposure by
inhibiting aromatic acid decarboxylase (Basma et al., 1995;
Mena et al., 1997b).
Carbidopa itself did not induce neurite outgrowth (compare Fig.
5, A and C; Fig.
6) and did not block neurite outgrowth
induced by L-DOPA (compare Fig. 5, B and D; Fig. 6). Carbidopa slightly (20%) inhibited neurite outgrowth induced by the combination of L-DOPA
and NGF (p < 0.05; compare Fig. 5, F and H;
Fig. 6). Therefore, synthesis of DA from L-DOPA was not required to
induce neurite outgrowth, although there may be a small DA-mediated
potentiation of combined NGF/L-DOPA induction.
View larger version (95K):
[in this window]
[in a new window]
|
Fig. 5.
Effects of carbidopa on L-DOPA and NGF-induced
process outgrowth. PC12 cultures at 2 days after plating were treated
for 24 hr with vehicle (A), 50 µM L-DOPA (B), 25 µM carbidopa (CBD) (C), L-DOPA with
carbidopa (D), vehicle and NGF (50 ng/ml) (E), NGF with L-DOPA (F), NGF
with carbidopa (G), and NGF with L-DOPA and carbidopa (H). Carbidopa
did not block process outgrowth evoked by either L-DOPA or NGF.
Scale bar, 10 µm.
|
|
View larger version (65K):
[in this window]
[in a new window]
|
Fig. 6.
Effects of L-DOPA and carbidopa on neurite
induction. The percentage of cells that exhibited neurites after
exposure to L-DOPA, carbidopa (CBD), and NGF in
combination as in Fig. 5 are reported as mean ± standard error
(n = 10 cultures). Carbidopa did not block process
outgrowth evoked by either L-DOPA or NGF (p < 0.001, ANOVA); carbidopa and control groups were not different.
|
|
It has been noted that the tyrosine kinase inhibitor genistein blocks
NGF-mediated attenuation of peroxide accumulation and excitotoxicity in
hippocampal cultures (Mattson et al., 1995). As G
protein-linked receptors, DA receptors might be expected to either
stimulate tyrosine kinase or potentiate the tyrosine kinase activity
induced by neurotrophin binding. However, genistein (10 µM for 24 hr) had no effect on process outgrowth elicited by either NGF alone (50 ng/ml) or L-DOPA/NGF (12-200 µM;
data not shown).
To further exclude the possibility that the synergistic responses to
L-DOPA/NGF were mediated by DA, we examined the effects of DA receptor
ligands on NGF-induced neurite outgrowth. Consistent with the
interpretation that DA itself is not required for L-DOPA induction of
process outgrowth, we observed that 24 hr exposure of PC12 cultures to
the D2 agonist quinpirole did not induce
neurites, whereas the DA agonist SKF 38393-A produced a small but
significant potentiation of outgrowth (Fig.
7); similar results were observed at 48 hr exposures (not shown). Moreover, the D2
antagonist sulpiride (20-200 µM) did not block L-DOPA
effects (data not shown), confirming that activation of at least one of
these receptors was not required.
View larger version (42K):
[in this window]
[in a new window]
|
Fig. 7.
Effect of DA receptor agonists on process
outgrowth. The percentage of cells that exhibited neurites after
exposure to 50 µM L-DOPA, 10 µM
apomorphine, 80 µM quinpirole, 100 µM
SKF38393-A, or 50 ng/ml NGF. Data are reported as mean ± standard
error (n = 6 cultures). Of the DA receptor agonists
examined, only apomorphine induced large-scale process outgrowth
comparable with L-DOPA (p < 0.001 versus both
control and NGF-only treatments, ANOVA followed by post
hoc test), although there is a small but significant effect
with SKF38393-A (p < 0.01 versus both control
and NGF-only treatments).
|
|
Surprisingly, only the
D1/D2 agonist apomorphine
(10 µM) had a potent synergistic effect on neurite
outgrowth similar to that of L-DOPA (Fig. 7). However, there was no
synergistic effect on NGF-elicited outgrowth using a combination of
quinpirole and SKF 38393-A (data not shown), which indicates that the
effect of apomorphine was not caused by binding to both classes of DA
receptors. We noted that of the DA receptor ligands tested, apomorphine
was the only one to induce a markedly visible extracellular pigment, in
this case a greenish pigment reminiscent of the brownish pigment of
L-DOPA quinone oxidation products. To examine this product, we measured
relative extracellular pigments levels using spectrophotometric readings of the medium at 490 nm (a blue wavelength), which has been
used previously to measure the brown pigment (Mena et al., 1992; Mena et al., 1993) and was seen here to clearly
measure the greenish pigment (Table 3).
Although indirect, the correlation of quinone formation and
potentiation of NGF-induced outgrowth suggests that the synergistic
effect could be attributable to an oxyradical-mediated mechanism.
Because apomorphine and other compounds that act as mild pro-oxidants
are known to stimulate GSH synthesis in several cell types (Han
et al., 1996), we examined the effect of apomorphine on GSH
levels (Table 4). We found that apomorphine, as well as NGF, indeed elevated the presence of GSH in the
cultures.
View this table:
[in this window]
[in a new window]
|
TABLE 2
Amperometric spike characteristics
Analysis of quantal release from neurite varicosities. Exposure of PC12
cells treated with NGF alone to 80 mM K+ evoked 214 release events from nine varicosities of a total of 16 recorded (see
Fig. 3, inset), whereas stimulation of cells treated with
NGF and 50 µM L-DOPA evoked 439 release events from eight
varicosities of a total of 15 recorded. Data in the table are expressed
as mean per cell ± standard error. Analysis of the entire
population of events, rather than cell means, showed similar trends
(quantal size and interspike intervals of the population are shown in
Fig. 4).
|
|
View this table:
[in this window]
[in a new window]
|
TABLE 3
Induction of extracellular quinones
Five days after plating, cultures were exposed to 50 µM
L-DOPA, 500 µM LNAC, 50 ng/ml NGF, or 10 µM
apomorphine as indicated for 24 hr. The extracellular media were then
assayed for quinone levels. Data are mean ± standard error
(n = 6). These measurements were run with two batches
of sister cultures; in each case, n = 6 cultures per
condition.
|
|
Effects of elevated GSH synthesis and synthesis inhibition.
From the above experiments, it seemed reasonable that an alternate
pathway that could underlie neurite initiation by L-DOPA might occur
via stimulation of reactive oxygen species derived from L-DOPA
metabolism. To examine if antioxidants might promote neurite outgrowth,
we used LNAC, a hydrophilic antioxidant (Yan et al., 1995)
that also elevates intracellular GSH in PC12 cells (Ferrari et
al., 1995; Yan et al., 1995; Kranich et al.,
1996). We found that LNAC and L-DOPA elevated GSH levels in a manner similar to that of NGF and apomorphine (Table
5). Consistent with a general
neurotrophic effect by antioxidant actions, LNAC also potentiated
L-DOPA-induced temporary neurite formation at levels identical to those
of NGF (Fig. 8).
View this table:
[in this window]
[in a new window]
|
TABLE 4
Effects of apomorphine and NGF on GSH synthesis
Five days after plating, cultures were exposed to 50 µM
L-DOPA, 10 µM apomorphine, or 50 ng/ml NGF for 24 hr. The
cultures were then assayed for total GSH and protein levels. Data are
mean ± standard error (n = 6).
|
|
View larger version (77K):
[in this window]
[in a new window]
|
Fig. 8.
Effect of LNAC on process outgrowth. The percentage
of cells that exhibited neurites following exposure to LNAC, L-DOPA,
and NGF in combination. Data are reported as mean ± standard
error (n = 10 cultures). PC12 cultures at 2 days
after plating were treated for 3 days with vehicle, 50 µM
L-DOPA, or 500 µM LNAC with or without NGF (50 ng/ml).
LNAC induced process outgrowth similarly to L-DOPA (all categories
p < 0.001 versus control) and potentiated the
effects of NGF and LNAC (p < 0.001, ANOVA
followed by post hoc test).
|
|
To test whether GSH synthesis per se was required by
L-DOPA/NGF synergistic effects, we used BSO, which inhibits the GSH
synthetic enzyme gamma-glutamylcysteine synthase. We found that 10 µM BSO treatment had no effect on total protein levels,
which indicates that this level of exposure was relatively nontoxic but
decreased GSH to 45% of control values (Table
6). At 24 hr, BSO had no effect on the
temporary induction of neurites by L-DOPA, NGF, or combined NGF/L-DOPA
(Fig. 9A). (In some cases, we noted that BSO elicited short, temporary neurites and we noted a small (< 10%)
and temporary potentiation of the effects on NGF and NGF/DOPA at 48 hr;
data not shown). However, at 5 days, a time point when NGF alone has
nearly caught up with the synergistic effects of NGF/L-DOPA, BSO
decreased neurite outgrowth elicited by both NGF and NGF/L-DOPA (Fig.
9B).
View this table:
[in this window]
[in a new window]
|
TABLE 5
Effects of L-DOPA, LNAC, and NGF on GSH synthesis
Five days after plating, cultures were exposed to 50 µM
L-DOPA, 500 µM LNAC, or 50 ng/ml NGF for 24 hr. The
cultures were then assayed for total GSH and protein levels. Data are
mean ± standard error (n = 6).
|
|
View this table:
[in this window]
[in a new window]
|
TABLE 6
Effects of L-DOPA and BSO on GSH synthesis
Five days after plating, cultures were exposed to 50 µM
L-DOPA or 10 µM BSO for 24 hr. The cultures were then
assayed for total GSH and protein levels. Data are mean ± standard error (n = 6).
|
|
View larger version (29K):
[in this window]
[in a new window]
|
Fig. 9.
Effect of BSO on L-DOPA and NGF-dependent process
outgrowth. PC12 cultures at 2 days after plating were treated for 24 hr
or 5 days with vehicle, 10 µM BSO, or 50 µM
L-DOPA, with or without NGF (50 ng/ml). The percentage of cells that
exhibited neurites are reported as mean ± standard error (A,
n = 4 cultures; B, n = 6 cultures). A, At 24 hr, each intervention presented increased neurites
over control (p < 0.01, ANOVA followed by
post hoc test). BSO slightly potentiated the effects of
NGF (p < 0.05) and NGF with DOPA (P < 0.05). B, At 5 days, L-DOPA, NGF, and L-DOPA with NGF each presented
increased neurites over control (p < 0.01, ANOVA followed by post hoc test). BSO alone was not
different from controls but reduced the presence of neurites when
included with L-DOPA (p < 0.001), NGF
(p < 0.001), and L-DOPA with NGF
(p = 0.026).
|
|
| |
Discussion |
L-DOPA, a widely prescribed intervention used to treat
Parkinson's disease and related disorders, has both trophic and toxic effects. Recently, factors that are produced by glia have been implicated in the compound's neurotrophic effects (Mena et
al., 1997b), which suggests that release of an L-DOPA- or
L-DOPA metabolite-elicited neurotrophic factor may be involved. To
examine that possibility, we have used the PC12 cell line, which
expresses neurites on exposure to NGF. Strikingly, L-DOPA provided a
synergistic effect on the process outgrowth induced by this
neurotrophic factor. A physiologically relevant result of this
neurotrophic effect was that L-DOPA enhanced the effect of NGF on
quantal neurotransmitter release from the neurite varicosities. The
findings suggest that long-lasting effects of L-DOPA may play an
important role in the induction of neurite outgrowth and enhanced
presynaptic function because of effects on neurotrophic factor-mediated
pathways.
Oxidative metabolism and L-DOPA-induced neurite outgrowth.
L-DOPA alone induced unstable neurites in PC12 cells that usually
retracted within 3 days unless NGF was present. Interestingly, we
observed similar small, temporary processes after exposure to LNAC,
BSO, and apomorphine, all compounds that can promote relatively mild
oxidant effects. This suggests that a molecular pathway initiated by
mild pro-oxidant treatment can induce a form of early process
outgrowth, but that subsequent steps, likely triggered by tyrosine
receptor kinase receptor binding, are required for neurite
maintenance. Mild pro-oxidant treatments have been suggested to protect
a range of cells from death by up-regulating GSH (Han et
al., 1996). The effect we have observed on neurite outgrowth may
stem from the action of L-DOPA or an L-DOPA metabolite directly on NGF
signaling, analogous to the observation that
H2O2 generation is required
for signal transduction by platelet-derived growth factor (Sundaresan
et al., 1995). Alternatively, antioxidant mechanisms that
result from NGF binding [elevation of catalase expression, for
instance (Sampath et al., 1994)], may be potentiated by
L-DOPA. Indeed, our data indicate that GSH was up-regulated not only by
NGF but also by all of the treatments that potentiated neurite
outgrowth, including L-DOPA, LNAC, and apomorphine, as well these
compounds in combination. However, the full elucidation of the
biochemical pathways and roles played by oxyradical pathways in the
synergistic response between L-DOPA and NGF will require further study.
As in midbrain neurons (Mena et al., 1997a) L-DOPA-induced
outgrowth in PC12 cells did not depend on conversion to DA, because the
effects were independent of blockade of DA synthesis by carbidopa, were
not blocked by the D2-like antagonist sulpiride, and similar responses
were not observed after exposure to DA receptor agonists, with the
exception of apomorphine. However, apomorphine is itself a pro-oxidant,
as seen by its metabolism to a quinone. We suggest that apomorphine may
act in a manner similar to L-DOPA, as a mild pro-oxidant that
stimulates a cascade of downstream effects that may be synergistic with
neurotrophin activation. However, although GSH seemed to be required
for maintenance of L-DOPA-induced neurite outgrowth in both neurons and
PC12 cells, this effect was delayed in PC12 cells. Another difference
between findings for these cell types is that midbrain DA neurons
required unidentified glial factors for the L-DOPA-induced neurotrophic
action, as indicated by the requirement for glial-conditioned medium,
while L-DOPA-induced PC12 neurite outgrowth was strongly potentiated by
NGF alone. The differences may be consistent with the different
biological milieus of the cell types. In the ventral midbrain, DA
neurons are surrounded by glia (Damier et al., 1993) that
provide a range of compounds and actions that mediate oxyradical
pathways, including catalase activity, GSH, monoamine oxidase B
activity, and other mechanisms to sequester and degrade toxic compounds
(Tsacopoulos and Magistretti, 1996), including control of potentially
neurotoxic ambient glutamate levels via glial conversion to glutamine
(Rosenberg and Aizenman, 1989). However, immortalized cell lines such
as the adrenal chromaffin tumor PC12 cell line have acquired
characteristics that allow them to thrive independently of the factors
contributed by other cell types.
L-DOPA promotes the quantal release of DA from neurites.
To
examine neurotransmitter release from neurite varicosities, as opposed
to total release from the culture, we used new electrochemical approaches that measure quantal catecholamine secretion in real time.
We found that in PC12 cells, long term L-DOPA increased the quantal
release rate, the number of quanta per stimulation, and the quantal
size of neurotransmitter release. This contrasts with short term (30 min) L-DOPA exposure, where quantal size is potentiated but the
frequency of release is unchanged (Pothos et al., 1996). The
elevated quantal size is presumably caused by augmented vesicular DA
levels that persist even after more than 1 week of L-DOPA withdrawal.
The elevated release rate could be caused by a range of mechanisms,
including the presence of more vesicles in the varicosities, a larger
percentage of vesicles docked to the plasma membrane, or effects on
calcium channels or other proteins involved in exocytosis. Indeed, an
alteration in ion channel expression may be consistent with the finding
that long term L-DOPA administration in vivo increases the
firing rate of midbrain DA neuron action potentials (Harden and Grace,
1995). It is possible that a trophic mechanism similar to that which promotes neurite outgrowth may also be trophic for the presynaptic sites or synthesis of secretory granules and associated proteins. However, the salient point is that the elevated neurite outgrowth after
the synergistic response to L-DOPA and NGF was physiologically significant, providing an impressive elevation of transmitter release
at individual varicosities.
The long term effects of L-DOPA explored here may be relevant for
treatment of Parkinson's disease because of the important issue of
whether L-DOPA improves prognosis. L-DOPA has not been observed to
promote toxicity using in vivo models, and normal primates
treated with high doses of L-DOPA for 3 months exhibit an intact
nigrostriatal tract with no evidence of oxidative damage to lipids,
proteins, or DNA (Jenner, 1996). Nevertheless, L-DOPA may be toxic for
diseased neurons already prone to degeneration. If L-DOPA normally
contributes to DA release in the striatum in part by promoting neurite
outgrowth and potentiating presynaptic mechanisms of neurotransmitter
release, it may be worth considering additional therapy with LNAC,
which is effective in vivo (Zimmerman et al.,
1989) and has been used safely in humans for more than 30 years (Aruoma
et al., 1989).
In summary, long term exposure to L-DOPA seems to increase release of
DA by at least two mechanisms: potentiating the process outgrowth
induced by trophic factors and increasing quantal release from neurite
varicosities. The ability of other compounds that elicit subneurotoxic
oxidative stress to provide a similar trophic effect indicates that
oxidative metabolism may play an important role in the potentiation of
response to neurotrophic factors. A consequence of this may include
promoting mechanisms that protect sites that contain monoaminergic
synaptic vesicles from local oxidative stress.
We thank Dr. Lloyd A. Greene for PC12 cells, collagen, NGF and
helpful discussion.
This work was supported by the Parkinson's Disease Foundation,
Spanish MEC (PR-95-098 and SAF 96-1099; MAM), and National Instititue on Drug Abuse Grants DA10154 and DA07418 (D.S.).
![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Top
Summary
Introduction
