Department of Pharmacology, Nanjing Medical University, Nanjing,
Peoples Republic of China (N.C.); and Department of Chemistry, Emory
University, Atlanta, Georgia (N.C., C.G.T., J.B.J.)
Effects of cations on dopamine (DA) uptake into cells expressing the
human dopamine transporter and on inhibition of DA uptake by various
substrates and inhibitors were investigated by using rotating disk
electrode voltammetry. The Na+ dependence of DA uptake
varied with Na+ substitutes, hyperbolic with
Li+, almost linear at 1 µM DA but hyperbolic at 8 µM DA
with choline, and sigmoidal with K+. With Na+
substituted by Li+, K[DA]
decreased and Vapp remained constant with
increasing [Na+], whereas
K[Na+] decreased and
Vapp increased with increasing
[DA], suggesting an ordered sequence with Na+
binding before DA. Similar trends for the Na+-DA
interactions were observed in the presence of cocaine. Cocaine inhibited DA uptake solely by increasing K[DA], with its
Ki not significantly different at 55 and 155 mM [Na+], whereas it inhibited Na+
stimulation by reducing Vapp more than
K[Na+] at 1 µM DA, and
Vapp only and less potently at 8 µM DA. Thus, cocaine may compete with DA, not with Na+,
for the transporter, and might not follow a strictly ordered reaction
with Na+. With Na+ substituted by
K+, K[DA] or
K[Na+] became insensitive to
Na+ or DA. K+ impaired the DA uptake mainly by
reducing Vapp, but affected cocaine
inhibition by elevating Ki. Despite their
different patterns for inhibiting DA uptake, nontransportable
inhibitors cocaine, methylphenidate, mazindol, and
1-[2-[bis(4-fluorophenyl)methoxy]ethyl]-4-(3-phenyl-2-propyl)piperazine (GBR12909) showed similarly modest Na+ dependence in their
Ki values. In contrast, substrates DA,
m-tyramine, and amphetamine displayed a similarly
stronger Na+ requirement for their apparent affinities.
 |
Introduction |
The
dopamine (DA) transporter (DAT) mediates uptake of synaptic DA into
neurons (Amara and Kuhar, 1993
; Uhl and Johnson, 1994). It also takes
up structurally similar substrates, including tyramine and amphetamine
derivatives (Sitte et al., 1998). This uptake process is dependent on
external Na+ and Cl
, with
K+ being primarily inhibitory (Krueger, 1990;
Amejdki-Chab et al., 1992; McElvain and Schenk, 1992; Gu et al., 1994).
The DAT is also a target for nontransportable compounds such as
cocaine, mazindol, methylphenidate, and
1-[2-[bis(4-fluorophenyl) methoxy]ethyl]-4-(3-phenyl-2-propyl)piperazine (GBR12909; Andersen, 1989; Tatsumi et al., 1997). The reinforcing effects of cocaine are best correlated with its DAT-blocking property (Ritz et al., 1987; Giros et al., 1996), whereas the other three compounds are currently under investigation for potential
pharmacotherapeutic application to treat cocaine abuse (for reviews,
see Grabowski et al., 1997; Wu et al., 1997; Zhang, 1998).
Binding studies on the DAT have led to the conclusion that the DA and
cocaine sites are different but overlapped; Na+
stimulates the binding of DA and cocaine analogs allosterically, whereas K+ blocks their binding by recognizing a
cation site included in both DA and cocaine sites (Reith et al., 1992;
Chen et al., 1997a,b; Li and Reith, 1999). In those
studies, the binding affinity of DA was assessed indirectly
through inhibition of binding of radiolabeled cocaine analogs,
2
-carbomethoxy-3-(4-fluorophenyl)tropane
([3H] WIN 35,428) and
[125I]3-(4-iodophenyl)tropane-2-carboxylic
acid isopropyl ester (RTI-121), to the DAT. Similar suggestions for
cations to modulate the binding of a GBR analog,
[3H]1-[2-(diphenylmethoxy)ethyl]-4-(3-phenyl-2-propenyl)piperazine (GBR12783; Héron et al., 1996), and
[3H]mazindol (Wu et al., 1997) have also been
made. Furthermore, there is preliminary binding evidence for a
differential sensitivity of substrates and nontransportable inhibitors
to ambient Na+ and K+ (Chen
et al., 1997a,b; Li and Reith, 1999). An important question remains to
be answered: do these cationic effects apply to the binding step only
or to the entire transport process as well? DA transport is an
electrogenic process, composed of multiple steps such as external
ligand binding, translocation, internal ligand release, and carrier
reorientation (Rudnick, 1997). It is possible that an effect observed
at the binding level is not necessarily rate-limiting and may fail to
modify the process at the levels of substrate translocation and
transporter reorientation.
To explore the impact of cationic effects found at the binding level
(Chen et al., 1997a,b; Li and Reith, 1999) on the active DA transport
process, we performed the present functional study investigating the
relationship between monovalent cations, substrates, and
nontransportable inhibitors. First, in transport assays, the effect of
Na+ is often confounded by using substituted ions
to maintain the isosmolarity and membrane integrity. Thus, the effect
on Na+-coupled DA transport of the two most
commonly used Na+ substitutes,
Li+ and choline, was evaluated. Second, binding
of DA to the DAT cannot be measured directly, presumably due to its
rapid dissociation rate. Thus, the binding sequence of DA and
Na+ was characterized by complementary kinetic
analyses and mathematical modeling. This provides a basis for comparing
Na+ modulation between substrates and inhibitors.
Third, interactions of cocaine with DA, Na+, and
K+ during the entire transport process were
tested and compared with those at the binding step (Chen et al., 1997;
Li and Reith, 1999). Finally, the Na+ requirement
was addressed for the apparent affinity and the action patterns of
other DAT ligands. A difference between substrates and inhibitors or
between cocaine and the potential agents to ameliorate cocaine abuse
would suggest the involvement of mechanisms differently linked to the
Na+ site. We used rotating disk electrode (RDE)
voltammetry to monitor the rapid rate of DA clearance by the
recombinant human DAT (hDAT) stably expressed in heterogeneous cells.
This approach facilitates the measurement of the DA uptake by
circumventing problems encountered in previous DA uptake assays, such
as the release of endogenous/accumulated DA and metabolism of
accumulated DA. Our findings suggest that the Na+
modulation of the hDAT occurs mainly at the binding level, which impacts the entire transport process. In contrast, the
K+ modulation may occur at both binding and
translocation levels, and the modulation after binding steps seems
rate-limiting.
| |
Experimental Procedures |
Materials.
Dopamine hydrochloride, cocaine hydrochloride
(COC), and the buffer salts were from Sigma Chemical Co. (St. Louis,
MO). m-Tyramine hydrochloride, d-amphetamine
sulfate, mazindol, methylphenidate hydrochloride, and GBR12909
dihydrochloride were from Research Biochemicals International (Natick,
MA). The pcDNA3-hDAT was a generous gift from Dr. Amy J. Eshleman
(Department of Physiology and Pharmacology, Oregon Health Science
University, Portland, OR).
Stable Expression of hDAT in Human Embryonic Kidney Cells.
Human embryonic kidney cells (HEK-293; CRL 1573; American Type Culture
Collection, Manassas, VA) were transfected by Lipofectin (Life
Technologies, Grand Island, NY) with the pcDNA3-hDAT construct at 60% confluence. After 3 days, the cells were dissociated with versene (Sigma Chemical Co.) and split at a 1:80-100 ratio into the
growth medium containing 600 µg/ml geneticin. The resistant colonies
were isolated 2 weeks later with sterile clone rings. Lines stably
expressing hDAT (HEK-hDAT cells) were identified by
Na+-dependent and cocaine-sensitive DA uptake
with RDE voltammetry.
Cell Culture.
The parental HEK-293 cells were maintained in
Dulbecco's modified Eagle medium supplemented with 10%
heat-inactivated fetal bovine serum and 2 mM L-glutamine at
37°C and 5% CO2. The transfected cell lines
were maintained in the same medium except that geneticin was added at a
concentration of 200 µg/ml.
Transport Assays.
DA transport assays were performed as
described previously (Burnette et al., 1996; Chen and Justice, 1998;
Chen et al., 1998). Briefly, when cells had grown to confluence in
150-mm cell culture dishes, they were harvested by scraping and
centrifugation. The harvested cells were then resuspended in 300 µl
of 37°C assay buffer and used for RDE measurement. After the cell
suspension was placed into the electrochemical cell, the working
electrode was introduced just below the surface of the solution and
rotated with an AFMSRX Analytical Rotator System (Pine Instrument
Company, Grove City, PA) at 4000 rpm. A potential of 400 mV relative to a Ag/AgCl reference electrode was applied to the RDE with a
potentiostat (EI-400; Einsman Instrumentation, Bloomington, IN), and
the output of the signal was amplified and recorded. All experiments
were performed at 37°C. Origin software (version 4.0; MicroCal
Software, Northampton, MA) was used for data acquisition.
For uptake assays, cells were preincubated for 3 min, and then a 1-min
baseline was collected. Subsequently, DA was added to the cell
suspension and the decrease in the DA signal was recorded. In studies
involving inhibition of DA uptake by various compounds, m-tyramine and d-amphetamine, whose inhibition
decreases with time, were added simultaneously with DA; cocaine,
methylphenidate, mazindol, and GBR12909, whose inhibition increases
within a certain time period, were added 3 min earlier than DA.
Increasing the pretreatment time from 0 to 3 min had no effect on the
Na+ dependence of the inhibitors. A total of 100 µM cocaine was used to define nonspecific uptake. Initial transport
rates were obtained from linear regression analysis of the decrease in
medium DA concentration versus time over the first 10-15 s after an
addition of DA (Burnette et al., 1996; Chen and Justice, 1998; Chen et
al., 1998).
The assay buffer contained 150 mM NaCl, 4.7 mM KCl, 2.2 mM
CaCl2, 1.2 mM MgSO4, 1.2 mM
KH2PO4, 10 mM HEPES, and 10 mM glucose. The pH of the assay buffer was adjusted to 7.4 with either
NaOH (resulting in approximately 5 mM additional
Na+ in the assay buffer) or LiOH. For
Na+ dependence of DA uptake,
Na+ was substituted with isosmolar quantities of
Li+, choline, or K+ in
chloride salt form. The cell protein was determined with a Bio-Rad DC
Protein Assay Kit (Bio-Rad, Hercules, CA).
Data Analysis and Statistics.
The Hill coefficients of the
uptake curves were estimated from curve-fitting of the initial rate
data with the expression v = (Vapp
[S]n)/(K[s]n + [S]n), in which v is
the initial uptake rate, Vapp is the
apparent maximal initial rate of DA uptake, S is substrate
(Na+ or DA),
K[s] is the concentration for a
substrate to produce half-maximal initial rate of DA uptake
(K[Na+] or
K[DA]), and n is the Hill
coefficient. The resulting Hill values were then rounded up to the next
integer values as an estimated value of the true Hill number (1 or 2)
and used for further kinetic analysis. Multiple models for the binding
sequence of DA and Na+ were examined by both
linear (Stein, 1986; Segel, 1993) and nonlinear (see Results
for details) regressions of the experimental rate data.
K[Na+],
K[DA], and
Vapp were calculated with
Lineweaver-Burk transformation of the Michaelis-Menten equation. The
maximal initial rate of DA uptake at saturating concentration of
Na+ (Vmax),
dissociation constant for DA at saturating concentration of
Na+ (KDA), and
dissociation constant for Na+ at 0 concentration
of DA (KNa+) were calculated with
replots of the results obtained from Lineweaver-Burk analysis (Stein, 1986; Segel, 1993). The half-saturation inhibition constant
(Ki) of various compounds was
estimated with the Michaelis-Menten expressions modified for
competitive, noncompetitive, and mixed inhibitions (Segel, 1993).
Results are expressed as mean ± S.E. of n experiments,
unless indicated otherwise. The F-test was used to determine
whether the goodness of fits obtained with different models was
statistically different (Munson and Rodbard, 1980). In brief, the
"extra sum of squares" principle is applied: F = [(SS1 
SS2)/(df1 df2)]/[SS2/df2], in which SS is the residual sums of squares of the
deviations of the experimental points to the fitted curve and
df is the degrees of freedom, going from the original model
1 to the more complex model 2 with added parameters. Other statistics
included one-way ANOVA followed by the Dunnett's test, and paired or
unpaired Student's t test, as appropriate. The accepted
level of significance was .05.
| |
Results |
Effects of Substituted Cations on Na+-Dependent DA
Uptake.
When the uptake was determined as a function of
Na+ concentration (5-155 mM) in the presence of
different cation substitutes, the shape of the curve varied with the
substitute used (Fig. 1). At 1 µM DA,
the Na+ curve was hyperbolic for
Li+ substitution (Fig. 1A), almost linear for
choline substitution (Fig. 1B), and sigmoid for
K+ substitution (Fig. 1C). The fitted Hill value
for the Na+ curve was close to 1 with
Li+ substitution and close to 2 with
K+ substitution (Table
1). At this DA concentration, we failed to get a reliable Hill value for the choline-substituted
Na+ curve because of its linear trend. At 8 µM
DA, the Na+ curve was hyperbolic for both
Li+ and choline substitutions (Fig. 1, A and B),
and sigmoid for K+ substitution (Fig. 1C). This
concentration of DA did not significantly change the Hill value for the
Li+- or K+-substituted
Na+ curve, but it improved the fitting of the
Hill coefficient for the choline-substituted Na+
curve, giving a value close to 1 (Table 1).
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 1.
Influence of substituted cations and DA
concentrations on Na+ dependence of DA uptake. The initial
rate of DA uptake at the indicated concentrations was measured as a
function of the Na+ concentration (5-155 mM). The
Na+ concentration was altered by replacing NaCl with
isomolar quantities of LiCl (A), choline chloride (B), or KCl (C). The
solid curves show the fit of data to the expression
v = (Vapp
[Na+]n)/((K[Na+])n + [Na+]n) with n
fixed at 1 for Li+ and Cl+ substitutions and 2 for K+ substitution. Shown are means ± S.E. of three
to five experiments.
|
|
View this table:
[in this window]
[in a new window]
|
TABLE 1
DA uptake as a function of Na+ concentration with various
substituted cations
The Na+ concentration (5-155 mM) was altered by replacing NaCl
with isosmolar quantities of LiCl, choline chloride, or KCl. Shown are
means ± SE of three to five experiments.
|
|
With Li+ or choline substitution, the uptake rate
of 8 µM DA at the lowest [Na+] was generally
close to the uptake rate of 1 µM DA at the highest [Na+] (Fig. 2).
This result was not attributable to a nonspecific uptake at higher
[DA], because 100 µM cocaine reduced this uptake to a rate similar
to that measured in parent HEK-293 cells (Fig. 2).
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 2.
Specific uptake of DA by hDAT-HEK cells at various
cationic conditions. Indicated cations (150 mM) were added to a basal
medium containing approximately 5 mM Na+ delivered from
HEPES/NaOH buffer. Cocaine (100 µM) was added 3 min earlier than DA.
Shown are means ± S.E. of three to eight experiments.
|
|
The K[Na+] and
Vapp for the Na+
curves at each fixed [DA] were obtained by fitting the rate data to
the Lineweaver-Burk transformation of Michaelis-Menten equation with
the Hill coefficient fixed at 1 for Li+ and
choline substitutions, and at 2 for K+
substitution. Although we failed to assess the Hill coefficient for the
choline-substituted Na+ curve at 1 µM DA, we
chose a value of unity because this number allowed a reasonable
estimation of Vapp and was close to
the Hill value obtained from the choline-substituted
Na+ curve at 8 µM DA. At 1 µM DA, the rank
order of K[Na+] was
K+ > choline > Li+
substitution (Table 1). At 8 µM DA, the
K[Na+] with
Li+ and choline substitution were greatly reduced
to a similar level, whereas that with K+
substitution was not significantly changed (Table 1).
When the DA uptake was determined as a function of DA concentration
(1-8 µM) with Na+ concentration fixed at
either 155 mM or 55 mM, the Hill values for all curves were close to 1 (Table 2). At lower
[Na+] (55 mM), the change in parameters varies
with substituted cations. The K[DA]
was markedly raised, whereas the
Vapp was unchanged under the
Li+- or choline-substituted condition. In
contrast, the Vapp was prominently
reduced, whereas the K[DA] was not
significantly changed under the K+-substituted
condition (Table 2).
View this table:
[in this window]
[in a new window]
|
TABLE 2
DA uptake as a function of DA concentration at various cationic
conditions
The DA concentration varied from 1 to 8 µM. The indicated Na+
concentration was altered by replacing NaCl with isosmolar quantities
of LiCl, choline chloride, or KCl. Shown are means ± SE of
n experiments.
|
|
Interaction between DA and Na+ with Li+ as
a Substitute.
Initially, we measured the kinetic parameters of DA
uptake over an extended [DA] range (0.5-16 µM) either at a high
[Na+] (155 mM) or at low
[Na+] (approximately 5 mM). In the latter case,
LiCl rather than NaCl was added to the medium, and the
[Na+] was mainly contributed by the NaOH used
to adjust the pH of the assay buffer. Under this low
[Na+] condition, DA uptake remained
appreciable, with the K[DA] increasing and the Vapp constant (Fig.
3). This was not due to a stimulatory
effect of Li+ used to replace
Na+, because the uptake rate was further reduced
by a Na+-free, Li+-fully
substituted buffer (LiOH was used to adjust the pH of the buffer). Even
with a Na+-free buffer, the uptake remained
measurable. However, it is difficult to maintain a zero external
Na+ condition, because under the influence of the
reversed Na+ gradient, intracellular
Na+ unavoidably leaves the cells, which makes the
Na+ concentration in the external medium nonzero.
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 3.
Concentration dependence of DA uptake at a low or
high concentration of Na+. The initial rate of DA uptake
was measured as a function of DA concentration (0.5-16 µM). In low
[Na+] medium, 150 mM LiCl was added to a basal medium
containing approximately 5 mM Na+ delivered from HEPES/NaOH
buffer. In high [Na+] medium, 150 mM NaCl was added to
the basal medium. Eadie-Hofstee plot was used to determine the values
for Vapp and
K[DA]. The straight solid line represents
the result of the least-squares linear regression. The calculated
parameters for this experiment were as follows: at 5 mM
[Na+], Vapp = 39.2 ± 4.8 pmol/s/mg, and K[DA] = 18.7 ± 2.4 µM; at 155 mM [Na+],
Vapp = 41.5 ± 1.6 pmol/sec/mg,
and K[DA] = 1.98 ± 0.1 µM.
Inset shows the fit of data to the Michaelis-Menten expression
v = (Vapp
[DA])/(K[DA] + [DA]), which gave
similar values for Vapp and
K[DA]. Shown are means ± S.E. of
three experiments.
|
|
By keeping the concentration of Cl constant and
saturating (Gu et al., 1994; Povlock and Schenk, 1997), we can consider
the hDAT as a bireactant system, in which all binding and dissociation steps are very rapid compared with the translocation of the ternary complex. The assumption of rapid equilibrium is often used in modeling
transporters (Stein, 1986). Two binding models are proposed in such a
system: random binding and ordered binding (Stein, 1986; Segel, 1993).
To test the binding sequence of DA and Na+, we
performed complementary experiments in which initial rates of DA uptake
were determined at various concentrations of Na+
and DA. The data were then analyzed by both linear regression (Stein,
1986; Segel, 1993) and nonlinear regression.
In the linear regression approach, the Lineweaver-Burk plots of
1/v versus 1/[DA] constructed for several fixed
concentrations of Na+ revealed straight lines,
which intersected the ordinate at a single point, i.e., the reciprocal
of the Vapp (Fig.
4). Such a plot is an indication of an
ordered reaction with Na+ binding to the hDAT
before DA (Stein, 1986; Segel, 1993). In accordance, as the
[Na+] increased from 15 mM to 155 mM, the
K[DA] decreased, whereas the
Vapp remained the same (Table
3). A replot of
K[DA] or the slope of the primary
plot versus 1/[Na+] was linear (insets of Fig.
4). The Lineweaver-Burk plots of 1/v versus
1/[Na+] constructed for several fixed
concentrations of DA also revealed straight lines, which, however,
intersected at a common point to the left of the ordinate and above the
abscissa (Fig. 5). In accordance, as the
[DA] increased from 0.5 µM to 8 µM, the
K[Na+] decreased, whereas the
Vapp increased (Table
4). A replot of the intercept (the
reciprocal of the Vapp) or the slope
of the primary plot versus 1/[DA] was again linear (insets of Fig.
5). Notably, the replot of the slope versus 1/[DA] gave an intercept at the origin (0.02 ± 0.02; n = 5; also see
bottom inset of Fig. 5), i.e., when [DA] is infinitely high, the
K[DA] approaches to infinitely
small, or the slope of the primary reciprocal plot is 0. This is
another kinetic indication that the reaction is ordered, with
Na+ adding to the hDAT before DA (Segel, 1993).
For a random binding reaction, this replot should have intercepted the
ordinate at a value higher than 0 (Stein, 1986; Segel, 1993).
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 4.
Initial rate of DA uptake as a function of DA
concentrations at several fixed Na+ concentrations. The
Na+ concentration was altered by replacing NaCl with
isomolar quantities of LiCl. Lineweaver-Burk plot of the data was used
to analyze the reaction sequence. The straight solid line represents
the result of the least-squares linear regression. Shown is a
representative experiment with five concentrations (0.5-8 µM) of DA
and indicated concentrations of Na+ on the same generation
of cells. Each experiment was performed five times with similar
results. Top inset shows a replot of the slope of the Lineweaver-Burk
plot as a function of 1/[Na+]. Bottom inset shows a
replot of the K[DA] obtained from the
Lineweaver-Burk plot as a function of 1/[Na+]. The true
Vmax, KDA, and
KNa+ were estimated from these replots (see
Table 3).
|
|
View this table:
[in this window]
[in a new window]
|
TABLE 3
Kinetic parameters for DA uptake at various concentrations of
Na+
The DA uptake was determined as a function of the DA concentration
(0.5-8 µM) at the indicated fixed concentrations of Na+. The
Na+ concentration was altered by replacing NaCl with isosmolar
quantities of LiCl. Shown are means ± SE of five experiments.
Parameters calculated from the replots of Lineweaver-Burk analysis in
Fig. 4: KDA = 1.6 ± 0.4 µM,
 = 99 ± 17 mM,
Vmax = 44.8 ± 2.2 pmol/s/mg.
|
|
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 5.
Initial rate of DA uptake as a function of
Na+ concentrations at several fixed DA concentrations. The
Na+ concentration was altered by replacing NaCl with
isomolar quantities of LiCl. Lineweaver-Burk plot of the data was used
to analyze the reaction sequence. The straight solid line represents
the result of the least-squares linear regression. Shown is a
representative experiment with seven concentrations (15-155 mM) of
Na+ and indicated concentrations of DA on the same
generation of cells. Each experiment was performed five times with
similar results. Top inset shows a replot of the intercept of the
Lineweaver-Burk plot as a function of 1/[DA]. Bottom inset shows a
replot of the slope of Lineweaver-Burk plot as a function of 1/[DA].
The true Vmax,
KDA, and KNa+
were estimated from these replots (see Table 4).
|
|
View this table:
[in this window]
[in a new window]
|
TABLE 4
Kinetic parameters for DA uptake at various concentrations of DA
The DA uptake was determined as a function of the Na+
concentration (15-155 mM) at the indicated fixed concentrations of DA.
The Na+ concentration was altered by replacing NaCl with
isosmolar quantities of LiCl. Shown are means ± SE of five
experiments. Parameters calculated from the replots of Lineweaver-Burk
analysis in Fig. 5: KDA = 1.6 ± 0.4 µM,
= 100 ± 17 mM,
Vmax = 44.2 ± 4.0 pmol/s/mg.
|
|
The replots of Figs. 4 and 5 were used to calculate
KDA,
KNa+, and the
Vmax (see insets of Figs. 4 and 5).
The values estimated from the two sets of analyses were extremely
similar (Tables 3 and 4).
In the nonlinear regression approach, the same data were fit by
mathematical expressions describing various possible models for the
binding sequence of DA and Na+. Different
translocation complexes were also considered. The scheme and the
equation are as follows: 
|
(1)
|
Here, KDA1 and
KDA2 are the dissociation constants
for the binding of DA to the free hDAT (T) and
Na+-bound hDAT (T-Na+); and
KNa+1 and KNa+2 are the
dissociation constants for the binding of Na+ to
the free hDAT and DA-bound hDAT (T-DA);
V1 and
V2 are the maximal translocation rate
for the T-DA and T-Na+-DA complex, respectively.
Three models were tested: 1) ordered binding, Na+
binds before DA and only the T-Na+-DA form
translocates (fixing 1/KDA1 and
V1 at 0); 2) random binding, both
Na+ and DA bind the free transporter but only the
form of T-Na+-DA translocates (fixing
V1 at 0); and 3) random binding, both Na+ and DA bind the free transporter and both the
T-Na+-DA form and the T-DA form translocate
(allowing all parameters to be free). The fitted parameter values
corresponding to each model and the quality of the fits are shown in
Table 5. Ordered binding with
Na+ binding first gave the best fit, with the
derived values of KDA2, KNa+1, and
V2 similar to those of
KDA,
KNa+, and
Vmax obtained from the linear
regression approach mentioned earlier. Random binding with the
translocation of only T-Na+-DA gave a poor fit.
Random binding with the translocation of both T-DA and
T-Na+-DA did not improve the fit. In the later
two models, the relative S.D. values of the added parameter,
KDA1 or
V1, was greater than 80%. Therefore,
there was no strong evidence either for the binding of DA before
Na+ or for the translocation of the hDAT-DA
complex.
View this table:
[in this window]
[in a new window]
|
TABLE 5
Nonlinear curve fitting of the uptake rate data with various models for
interactions between DA and Na+
DA uptake was determined as a function of the concentrations of DA
(0.5-8 µM) and Na+ (15-155 mM). The Na+
concentration was altered by replacing NaCl with isosmolar quantities
of LiCl. The parameters were estimated from nonlinear curve fitting of
Equation 1 (see Results) to the average uptake rate of five
experiments. Precision in the fitted parameters is expressed as the
standard deviation of the regression.
|
|
Interaction between Cocaine and Cations at the hDAT.
First,
the DA uptake was determined as a function of Na+
in the presence of 0.5 µM cocaine. To facilitate comparison, we
performed these experiments in parallel with the experiments in Table
1. At 1 µM [DA], cocaine reduced both
K[Na+] and
Vapp. The relative reduction in
magnitude in K[Na+] was similar
regardless of the substituted cation used (Table
6), with the ratio of
K[Na+] at the three substitutions (1:1.3:1.8 for Li+, choline, and
K+, respectively) similar to that in the absence
of cocaine (1:1.3:1.9). With DA concentration elevated to 8 µM, the
effect of cocaine on the K[Na+] was
abolished, and its effect on the Vapp
was significantly reduced (Table 6). The data also were evaluated as a
percentage of inhibition as a function of
[Na+]. Cocaine-induced inhibition of DA uptake
(1 µM) was significantly less at 30 mM [Na+]
(for Li+ and choline substitution) or 55 mM
[Na+] (for K+
substitution) than at 155 mM [Na+] (49 ± 3%, 46 ± 2%, and 30 ±3% versus 59 ± 0.6%, 56 ± 0.7%, and 57 ± 1.5%; p < .05). At 8 µM DA,
cocaine-induced inhibition of DA uptake did not show a significant
difference over a [Na+] range from 15 to 155 mM.
View this table:
[in this window]
[in a new window]
|
TABLE 6
Na+ dependence of DA uptake in the presence of cocaine
The DA uptake at indicated concentrations was determined as a function
of the Na+ concentration (15-155 mM) in the presence of 0.5 µM cocaine. The Na+ concentration was altered by replacing
NaCl with isosmolar quantities of LiCl, choline chloride, or KCl.
Cocaine was added 3 min earlier than DA. The experiments were performed
in parallel with the experiments in Table 1. Shown are means ± SE
of three experiments.
|
|
Second, the DA uptake was determined as a function of DA in the
presence of 1 µM cocaine. These experiments were performed in
parallel with the experiments in Table 2. At 155 mM
[Na+], cocaine raised the
K[DA] value and did not change the
Vapp value (compare Table
7 with Table 2). When the
Na+ concentration was reduced to 55 mM, the
inhibition pattern of cocaine remained the same, irrespective of the
presence of different substituted cations. However, the change in
Ki of cocaine varied with the
substitute used. Thus, at 55 mM [Na+], the
slight increment in the Ki at
Li+ substitution did not reach statistical
significance. In contrast, the Ki at
choline or K+ substitution was significantly
raised (Table 7). As Li+ is relatively inert and
the Na+ concentration remains the same, a
difference observed between Li+ and
K+ or choline substitution could be considered as
an effect of K+ or choline.
View this table:
[in this window]
[in a new window]
|
TABLE 7
Inhibition of DA uptake by cocaine at various cationic conditions
The DA uptake was determined as a function of the DA concentration
(1-8 µM) in the presence of 1 µM cocaine. The Na+
concentration was altered by replacing NaCl with isosmolar quantities
of LiCl, choline chloride, or KCl. Cocaine was added 3 min earlier than
DA. The experiments were performed in parallel with the experiments in
Table 2. Shown are means ± SE of n experiments.
|
|
Additionally, the above experiments allowed the binding sequence of DA
and Na+ in the presence of cocaine to be
assessed. As shown in Tables 6 and 7 under Li+
substitution condition, with cocaine, the changes in
K[DA], Vapp, and
K[Na+] were in the same direction as those without cocaine, indicating that cocaine may not modify the
binding sequence.
Inhibition of DA Uptake by Other Substrates and Inhibitors and
Their Na+ Requirement.
The inhibition pattern and
potency of other substrates and nontransportable inhibitors were tested
at two levels of the Na+, 155 mM and 55 mM (with
Li+ as a substitute). Two substrates,
m-tyramine and d-amphetamine, inhibited the DA
uptake solely by elevating K[DA]
regardless of the Na+ concentration (Table
8). Among the nontransportable
inhibitors, methylphenidate only increased the
K[DA], regardless of the
Na+ concentration; mazindol affected both
K[DA] and
Vapp at 155 mM
[Na+], but mainly affected
K[DA] at 55 mM
[Na+]; and GBR12909 mainly reduced
Vapp, regardless of the
Na+ concentration (Table 8).
View this table:
[in this window]
[in a new window]
|
TABLE 8
Inhibition of DA uptake by various compounds and effect of Na+
The DA uptake was determined as a function of the dopamine
concentration (1-8 µM) in the absence and presence of each compound
at the indicated concentration. The Na+ concentration was
altered by replacing NaCl with isosmolar quantities of LiCl. Shown are
means ± SE of three to four experiments.
|
|
The apparent noncompetitive inhibition by GBR12909 of DA uptake in the
present study, although agreeing with a recent study on another GBR
analog, GBR12783 (Do-Régo et al., 1998), differs from previous
studies showing that GBR12909 exerted a competitive inhibition on
[3H]DA uptake in synaptosomal preparations
(Andersen, 1989). As the binding of a [3H]GBR
analog to membranes prepared from the transfected cells exhibits a
pharmacological profile indicative of the DAT (Eshleman et al., 1995),
it seems impossible that GBR12909 inhibits DA uptake by acting at a
protein distinct from the hDAT. One possibility is that the slow,
tight-binding feature of GBR12909 may not allow DA to fully displace it
during a short time (10 s in the present study). Other mechanisms may
also be involved (Pristupa et al., 1994; Do-Régo et al., 1998).
Compared with those at 155 mM [Na+], the
Ki values of m-tyramine and
d-amphetamine were more than doubled, whereas the
Ki values of the nontransportable
inhibitors were raised by less than 50% at 55 mM
[Na+] (Table 8). The ratios of the
Ki
(Km) value at 155 mM
[Na+] to the
Ki
(Km) value at 55 mM
[Na+]
(K155/55) for all tested substrates
and nontransportable inhibitors were calculated with data measured in
pairs (Fig. 6). The results showed that
the sensitivity of the two categories of compounds to
Na+ was different, with the
K155/55 for the substrates
significantly higher than for inhibitors (p < .01 on
pooled data between substrates and inhibitors).
View larger version (44K):
[in this window]
[in a new window]
|
Fig. 6.
Different Na+ sensitivity between
substrates and inhibitors. The ratios of the
Ki (Km) value at
155 mM [Na+] to the Ki
(Km) value at 55 mM [Na+]
(K155/55) were calculated with data measured
in pairs. The Na+ concentration was altered by replacing
NaCl with isomolar quantities of LiCl. Shown are means ± S.E. of
three to five experiments. One-way ANOVA on pooled data shows a
significant difference between substrates and inhibitors
(p < .01).
|
|
| |
Discussion |
Profile of Na+ Dependence.
With
Li+ as a Na+ substitute, we
observed a hyperbolic Na+ dependence of the DA
uptake, which is different from the sigmoid Na+
dependence reported by previous studies. In those studies, some measured [3H]DA accumulation in synaptosomes
(Krueger, 1990) or nonisotopic DA clearance by brain tissues or
transfected cells (McElvain and Schenk, 1992; Povlock and Schenk, 1997;
Earles and Schenk, 1999) with choline as a Na+
substitute, and others measured [3H]DA
accumulation in transfected cells with Li+ as a
Na+ substitute (Gu et al., 1994; Pifl et al.,
1997). Choline has been found inhibitory in both tissues and
transfected cells (Shank et al., 1987; Amejdki-Chab et al., 1992, Gu et
al., 1994; the present study). With an inhibitory
Na+ substitute, a sigmoid shape of the
Na+ curve is expected to be readily observed
because dual effects occur along the
Na+-substituted cation curve: reduced stimulation
by Na+ and enhanced inhibition by substituted
cation of DA uptake. Indeed, we failed to observe a hyperbolic
Na+ dependence with choline substitution at 1 µM [DA], a concentration close to that used by the previous
studies. Additionally, low Na+-induced release of
endogenous DA or accumulated [3H]DA (Levi and
Raiteri, 1993; Pifl et al., 1997) and metabolism by
catechol-O-methyltransferase of accumulated
[3H]DA (Eshleman et al., 1997) may confound the
measurement of DA clearance from tissues or
[3H]DA accumulation into transfected cells.
Although Li+ also appears to be a potent
inhibitor of DA uptake in synaptosomes (Shank et al., 1987;
Amejdki-Chab et al., 1992), it acts like a relatively inert
Na+ substitute in recent uptake and binding
studies on hDAT-expressing intact cells (Gu et al., 1994; Wu et al.,
1997). This is further confirmed by the present study. Thus,
Li+ was used as a Na+
substitute for more detailed kinetic studies.
The Hill value is commonly used to reflect cotransport stoichiometry.
However, this value technically varies with the
Na+ substitutes used, and, theoretically, is only
valid if the affinities for the different Na+
binding sites are similar (Rudnick, 1998). Under our experimental condition, it is possible that more than one Na+
ion is involved in DA binding/translocation, but that the initial transport rate is dependent on only one low-affinity
Na+-binding site. It is also possible that one
Na+ ion is required for DA binding/translocation,
but that more than one Na+ ion is actually cotransported.
Kinetic Binding Sequence of DA and Na+.
The
current study suggests a fixed binding order of DA and
Na+ at the hDAT, with Na+
binding before DA. Furthermore, translocation occurs only after the
binding of both Na+ and DA. The reaction scheme
could be shown as follows:
The presence of DA drives the reaction to the right, thereby the
apparent affinity of the hDAT for Na+ increases
with DA. Because a concentration of Na+ far below
the KNa+ may still be much higher than
the concentration of the hDAT, it would be expected that a saturating
concentration of DA would shift all the hDAT to the
hDAT-Na+-DA form. This finding agrees with our
observation that appreciable DA uptake remained at extremely low
[Na+] but saturating [DA]. In
electrophysiological studies with cloned hDAT, DA readily blocks leak
conductance carried by Li+ or
K+ in the absence of external
Na+, questioning the requirement of
Na+ for the binding of external DA (Sonders et
al., 1997). The difference in cationic selectivity of the DA transport
and leak conductance indicates that DA might bind to different sites or
states of the hDAT to block leak conductance. However, such an effect
is not measurable in the present study monitoring the specific
Na+-coupled DA transport.
Although the binding sequence proposed by the present study differs
from previous studies in which a random binding of one DA and two
Na+ ions was proposed (McElvain and Schenk, 1992;
Wheeler et al., 1993; Povlock and Schenk, 1997), a similar trend for
the ordered binding with choline substitution has recently been
observed (Earles and Schenk, 1999). This trend is also reflected in our
experiments using choline to replace Na+, even
though choline is not a satisfactory Na+
substitute at low [DA]. With K+ as a
substitute, K[DA] or
K[Na+] became insensitive to
Na+ or DA as if the hDAT system were a random
system with one ligand having no effect on the binding of the other.
Such a pattern possibly is due to the fact that
K+ is not the right substitute for analysis of
the binding order.
Distinct Modulation by Na+ and K+ of
Cocaine Inhibition and DA Uptake.
The results in Tables 6 and 7
indicate that the presence of cocaine may not modify the binding
sequence of DA and Na+. In this scenario, an
apparently competitive inhibition of DA uptake by cocaine could arise
from interactions of cocaine with either DA or
Na+ sites. Our data are in favor of interactions
of cocaine with DA sites because of the failure of cocaine to raise
K[Na+] and vice versa.
Consistent with its stimulatory effect on the binding of cocaine
analogs (Chen et al., 1997; Li and Reith, 1999),
Na+ enhances the inhibition by cocaine of DA
uptake (1 µM). This is revealed by the reduced
K[Na+] in the presence of cocaine.
Noticeably, the reduction in magnitude is independent of the
substituted cations, excluding the possibility that the reduced
K[Na+] is caused by interactions
between cocaine and substituted cations. However, cocaine affects the Na+ stimulation by reducing
Vapp more than
K[Na+], suggesting that cocaine
inhibition is not entirely dependent on the presence of
Na+. This raises the question whether
Na+ and cocaine bind in a strictly ordered
sequence. Indeed, both stimulatory and inhibitory effects of
Na+ have been observed on the binding affinity of
cocaine analogs (Reith and Coffey, 1993; Chen et al., 1997a,b; Wu et
al., 1997). Furthermore, Na+ accelerates the
dissociation of a cocaine analog (Chen et al., 1997a) and can therefore
bind to the cocaine analog-DAT complex. This complication remains to be
investigated further.
At the binding level, K+ reduces the binding
affinity of DA and cocaine analogs in a similar fashion (Li and Reith,
1999). In the present transport study, K+ affects
the effect of cocaine again by reducing the affinity of cocaine, as
deduced from the increased Ki, whereas
it affects the DA uptake mainly by reducing the
Vapp. However, our results do not rule
out a possible effect of K+ on the binding
affinity of external DA. According to the simplest kinetic scheme
(Bönisch, 1998), K[DA] is
proportional to both the binding affinity of DA and
k2, a rate constant reflecting multiple steps associated with DA translocation, whereas
Vapp is proportional to both the total
concentration of hDAT and k2. The
reduced Vapp observed with
K+ most probably results from a reduced
k2 caused by
K+-induced membrane depolarization (Krueger,
1990; Sonders et al., 1997). Accordingly, a decrease in
k2 should have reduced the
K[DA] if the binding affinity did
not change. In contrast, we observed an almost unchanged
K[DA] in the presence of
K+, which may actually indicate that both binding
and translocation of DA are impaired. However, it remains possible that
the rate-limiting step in DA transport occurs subsequent to substrate
binding and depends on the membrane potential. This may explain why
cocaine, despite competing with K+ for the DAT at
the binding level (Chen et al., 1997; Li and Reith, 1999), failed to
antagonize the inhibitory effect of K+ on
K[Na+].
Different Na+ Dependence between Substrates and
Nontransportable Inhibitors.
The present study reveals that one
property distinguishing between substrates and inhibitors may be the
different Na+ dependence. Thus, raising
[Na+] independent of
[Cl] enhances the apparent affinity of
substrates for the hDAT more than that of inhibitors. Interestingly,
despite their diverse chemical structures and different inhibition
patterns, the inhibitory potency of all nontransportable inhibitors
tested show a similarly modest response to Na+.
This finding suggests that there is a similarity among the interactions of cocaine, mazindol, and GBR12909 with Na+,
albeit, some differences between their interactions with DA.
In transport assays, the Ki of a
substrate depends on both its binding affinity and its
k2, whereas the
Ki of an inhibitor is identical with
its binding affinity (Bönisch, 1998). Could the higher
Na+ sensitivity of the substrates represent a
Na+-induced change in their translocation rate?
This is unlikely because we have demonstrated that
Na+ changes only the
K[DA], not the
Vapp, of DA uptake. It is more likely
that Na+ differently influences the binding
affinity of substrates and inhibitors. In consonance, a binding study
shows that lowering [NaCl] reduces the binding affinity of DA and
amphetamine more than that of cocaine, mazindol, and GBR12935 for the
[125I]RTI-121-labeled DAT sites. (Chen et al.,
1997b). The present transport assays and the previous binding assays
were done with hDAT on intact cells and rat DAT on membranes,
respectively. The same temperature and a similar range of
Na+ were employed in both investigations, but
Li+ substitution was not used in the membrane
work. Nevertheless, the two different experimental environments both
gave 2.3 for the ratio of Ki at low
[Na+] to Ki at
high [Na+] for substrates; the corresponding
ratios found for inhibitors were also close for cell-hDAT and
membrane-rDAT, 1.3-1.5 and 1.2, respectively. The consistency between
the two studies argues that the difference in Na+
sensitivities between substrates and inhibitors may be at least partially due to reactions occurring at the binding level and impacting
the entire transport process.
DA, dopamine;
DAT, dopamine transporter;
hDAT, human DAT;
RDE, rotating disk electrode;
K[DA], concentration for DA to produce
half-maximal initial rate of DA uptake at a given concentration of
Na+;
KDA, dissociation constant
of DA at saturating concentrations of Na+;
K[Na+], concentration for Na+
to produce half-maximal initial rate of DA uptake at a given
concentration of DA, KNa+, dissociation
constant of Na+ at 0 concentration of DA;
Vapp, apparent maximal initial rate of DA
uptake at a given concentration of DA and saturating concentrations of
Na+ or at saturating concentrations of DA and a given
concentration of Na+;
Vmax, maximal initial rate of DA uptake at saturating concentrations of DA
and Na+;
GBR12909, 1-[2-[bis(4-fluorophenyl)methoxy]ethyl]-4-(3-phenyl-2-propyl)piperazine;
GBR12783, 1-[2-(diphenylmethoxy)ethyl]-4-(3-phenyl-2-propenyl)piperazine;
RTI-121, 3-(4-iodophenyl)tropane-2-carboxylic acid isopropyl
ester;
WIN 35,428, 2-carbomethoxy-3-(4-fluorophenyl)tropane.
![[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
Abstract
Introduction
