Professional Documents
Culture Documents
Amini, B., J. W. Clark, Jr., and C. C. Canavier. Calcium dynamics observed, presumably due to the loss of synaptic afferents. In
sodium rather than calcium dependent (Johnson et al. 1992) ion pumps, a sodium-calcium exchanger, and other ionic cur-
and is also not addressed in this study. rents. The lumped fluid compartment model (Fig. 1B) consists
The goal of this study was to test hypothesized mechanisms of an intracellular compartment containing constant concentra-
for the SOP and the square wave. On the basis of data that tions of Na1 and K1, and a calcium buffer (presumably cal-
suggest that dihydropyridines abolish the SOP (see DISCUSSION), modulin). A material balance for Ca21 describes the time rate
we hypothesize that the SOP is generated by interplay between of change in cytosolic Ca21 concentration. The extracellular
the L-type calcium current ICa,L and the calcium-activated space is assumed to have a relatively large volume, so that the
IK,Ca,SK. We further hypothesize that in the presence of apamin, ionic concentrations of Ca21, Na1, and K1 there are assumed
the calcium-mediated inactivation of ICa,L is responsible for to be constant.
repolarization of the square wave. We also hypothesize that the
mechanisms driving these phenomena are located in or near the MEMBRANE CURRENTS
soma. Hence we constructed a quantitative somatic model
based on voltage-clamp, morphological, and calcium-imaging Under space-clamp conditions, the differential equation de-
data to confirm that the currents described under voltage clamp scribing the time-dependent changes in the membrane potential
are capable of generating the electrical activity observed under (V) is
current clamp. Minimum expectations for model performance
V̇ 5 2~I Ca,T 1 I Ca,L 1 I Ca,N 1 I Ca,HVA 1 I K 1 I A 1 I h 1 I Na,K
included production of SOPs similar to those of Ping and
Shepard (1996) in the presence of TTX and TEA, SOPs under 1 I Ca,pump 1 I Na,Ca !/C m (1)
conditions where either the calcium current ICa,T or ICa,N were
blocked but not when ICa,L is blocked, and square waves where Cm is the whole cell membrane capacitance. Model
similar to those observed (Ping and Shepard 1996) in the currents include a T-type calcium current (ICa,T), an L-type
presence of apamin. calcium current (ICa,L), an N-type calcium current (ICa,N), a
residual high-voltage activated calcium current (ICa,HVA), a
MODEL DEVELOPMENT
delayed rectifier current (IK), a transient outward current (IA),
a small conductance calcium-dependent potassium current
Our model of the DA neuron consists of a single compart- (IK,Ca,SK), a hyperpolarization activated current (Ih), a sodium-
ment Hodgkin-Huxley (HH)-type parallel conductance mem- potassium pump current (INa,K), a calcium pump current
brane model (Fig. 1A) and a lumped fluid compartmental (ICa,pump), and a sodium-calcium exchanger (INa,Ca).
model (Fig. 1B). The HH equivalent circuit is composed of a In the current descriptions that follow, the HH-type activa-
somatic membrane capacitance of 15.8 pF (Kang and Kitai tion and inactivation gating variables are solutions of the
1993b), shunted by resistive ion-selective channels, as well as familiar first-order differential equations described as
CALCIUM DYNAMICS IN PACEMAKER AND BURST FIRING 2251
z# ~V! 2 z~V, t! variation in the waveshape of the ionic current response from
ż~V, t! 5 (2)
t z ~V! cell to cell.
where z#(V) is the steady-state value of the general gating Calcium currents—ICa
variable z at membrane voltage V. We have characterized the
steady-state gating variable z#(V) by a sigmoidal, or Boltzman- The mathematical equations used in the model for the dif-
type relationship, and the time constant tz(V) by a Gaussian ferent calcium currents are given in Table 1, and the voltage
relationship. dependence of the steady-state activation and inactivation char-
Individual ionic membrane currents were characterized by acteristics of these currents are shown graphically in Fig. 3.
fits to published voltage clamp experiments where available. The reversal potential for the calcium currents has been set to
The temperature at which the experiments were performed a constant 50 mV by extrapolating current-voltage curves from
ranged between 30 and 35°C, thus temperature adjustments Kang and Kitai (1993b).
(e.g., Q10) were not needed during final integration of currents T-TYPE LVA CALCIUM CURRENT—ICA,T. The T-type calcium cur-
into the model. To obtain fits to the currents, membrane po- rent is based on data from Kang and Kitai (1993b). The
tential was held constant and the differential equations char- membrane potential was held at 288 mV and clamped at
acterizing the gating variables for each current were integrated various voltages from 268 mV through 254 mV in increments
numerically and substituted into the appropriate equation for of 2 mV. Their experiments show a two-stage inactivation: a
ionic current. Examples of fits to the component currents of the quick inactivation to a small current followed by a slow inac-
model are shown in Fig. 2. tivation that maintains the small current for the duration of the
Fits to voltage-clamp data are not the only criteria for voltage clamp (300 ms). Characterization for the long second-
formulating ionic current descriptions, which may have to be ary component is based on data from neostriatal neurons
modified to fit whole cell transmembrane potential data. These (Hoehn et al. 1993), whereas the activation and fast inactiva-
additional adjustments are justified considering that the volt- tion component was characterized by fitting data from Kang
age-clamp and the free-running SOP and square-wave re- and Kitai (1993b). The voltage-clamp fits to data are shown in
cordings were obtained from different cells and in different Fig. 2A. The current description fits the quick activation and
laboratories and that the voltage-clamp experiments were biphasic inactivation well at the expense of matching the peak
performed on a particular cell, and there is a considerable current values at some clamp potentials.
2252 B. AMINI, J. W. CLARK, AND C. C. CANAVIER
FIG. 3. Steady-state characteristics of calcium currents. d and f represent FIG. 4. Steady-state calcium-dependent characteristics associated with var-
the activation and inactivation gating variables, respectively, for calcium ious ionic currents. A: summary of calcium-dependent ionic currents. fCa,L,
currents. Steady-state activation (d# T) and inactivation (f#T) characteristics of fCa,N, and pK,Ca are indicated. B: activation of IK,Ca,SK by calcium during slow
ICa,T were obtained from voltage-clamp studies of Kang and Kitai (1993b). oscillation potential (SOP). IK,Ca activation variable (pK,Ca) is plotted against
Steady-state activation of ICa,N (d# N) was obtained by fitting voltage-clamp data the internal calcium concentration, [Ca21]i. Extent of [Ca21]i excursion during
from the same source. Steady-state activation of ICa,L (d# L) is more hyperpo- SOP is indicated by the shaded region. C: inactivation of ICa,L and ICa,N by the
larized than that of ICa,N. Steady-state activation (d# HVA) and inactivation (f#HVA) calcium-dependent fCa,L and fCa,N, respectively, during square-wave oscilla-
of ICa,HVA, are adapted from Cardozo and Bean (1995) and Kang and Kitai tions. Extent of [Ca21]i excursion during the square-wave oscillations is
(1993b), respectively. indicated by the shaded region.
CALCIUM DYNAMICS IN PACEMAKER AND BURST FIRING 2253
IK 5 g# Kn3(V 2 EK)
perpolarization-activated cation current (Ih) has been shown to
be important for pace-making activity in thalamic and cortical
1.0 neurons (McCormick and Pape 1990). In DA neurons, how-
RESULTS
neither Ni21 nor v-conotoxin can block the SOP in guinea pig
SNc DA neurons, whereas nifedipine application alone is ef- FIG. 9. Underlying currents of square-wave oscillations. A: membrane po-
fective. Figure 8 shows that blockade of ICa,T, ICa,N, and ICa,L tential during square-wave oscillations. B: [Ca21]i variation during square-
in the model for the purposes of simulating Ni21, v-conotoxin, wave oscillations is greater than that seen during SOPs. C: ICa,L is the main
ionic current. The residual IK,Ca,SK current and ICa,N are minor contributers in
and nifedipine application, respectively, conforms to these this mode of oscillation. D: calcium and sodium background currents, IB,Ca and
experimental results. The increase in amplitude after blockade IB,Na. E: remaining calcium currents ICa,T and ICa,HVA. F: sodium-dependent
of ICa,T is due to the removal of a depolarizing drive and pump currents, INa,K and INa,Ca. G: transient outward current (IA), potassium
reduction in calcium entry, which allows ICa,L to remain active background current (IB,K), and the calcium pump (ICa,P).
longer, raising the peak potential.
the square-wave oscillations need not depend on the presence
of a distinct hyperpolarization current in this mode of oscilla-
Square-wave oscillations tion. Without the activation of IK,Ca,SK by calcium, internal
calcium concentration continues to rise and achieves a higher
The component currents underlying the square-wave oscil- peak level than in sinusoidal oscillations (Fig. 9B), and the
lations are shown in Fig. 9 in the same order as those for the frequency of oscillation decreases. ICa,L is inactivated more
sinusoidal case (Fig. 6). The major depolarizing currents are strongly at the elevated levels of [Ca21]i. As ICa,L declines due
again ICa,L and ICa,N; however, the outward K1 current IK,Ca,SK to inactivation, the remaining “residual” currents (IRes, Fig. 10)
has been blocked. Our model predicts that the mechanism for sum to repolarize the membrane.
DISCUSSION
FIG.11. Effect of TEA on the model. A: control SOP in the presence of
TTX and TEA. B: SOP with bath application of TTX alone results in lower We have developed a semiquantitative model that can mimic
amplitude oscillations. the behavior of DA neurons under a wide range of experimen-
2258 B. AMINI, J. W. CLARK, AND C. C. CANAVIER
tion that the somatic currents and calcium dynamics drive the
SOP is justified by several experimental observations. First,
during the SOP the calcium transients in the soma are a
substantial fraction of those observed in the dendrites (Calla-
way and Wilson 1997) despite the large disparity in surface
area to volume ratio. Hence there is a significant amount of
calcium entry into the soma. Second, the PLD can be evoked
by brief depolarizations of the soma, and its rate of activation
is strongly dependent on the membrane potential in the soma
(Grace and Onn 1989). Finally, one study found that sectioning
the distal dendrites of DA neurons had no significant effect on
the pacemaker firing frequency (Nedergaard and Greenfield
1992). Assuming the firing frequency is strongly influenced by
the SOP, the distal dendrites do not contribute significantly to
the SOP as recorded at the soma.
30 mM nifedipine blocked spontaneous pacemaker activity in nism that they proposed for NMDA-induced bursting is that
rats, and by inference, the SOP as well. This is consistent with same as that proposed by a later, more physiologically based
findings in rat that the calcium conductance mediating the SOP model by Canavier (1999), namely the interaction of the
is dihydropyridine sensitive (Ping and Shepard 1997) and with NMDA-induced current (INMDA) and the sodium pump. Re-
those of Kang and Kitai (1993a,b), who found that the calcium generative voltage activation of INMDA is postulated to be
channel blocker Cd abolished the SOP in rat. These findings responsible for the depolarization during a burst, and sodium
are also consistent with the results of Nedergaard et al. (1993), accumulation due to entry via INMDA is hypothesized to acti-
who reported that 0.5–20 mM nifedipine abolished the SOP in vate the electrogenic sodium pump, which is a net outward
guinea pig. On the other hand, Fujimura and Matsuda (1989) current, until a regenerative hyperpolarization due to the volt-
reported that although the calcium channel blockers Cd21 and age-dependent closing of INMDA channels. Sodium is removed
Co21 as well as Ca-free saline blocked the SOP in guinea pig, during the hyperpolarizing phase, allowing the cycle to begin
100 mM nifedipine did not. Furthermore Yung et al. (1991) again. On the other hand, the mechanism proposed by Li et al.
reported that 500 mM Ni21 abolished the SOP in guinea pig, (1996) for apamin-induced, calcium-dependent bursting as-
whereas Nedergaard et al. (1993) reported that 500 mM Ni21 signs a crucial role to ICa,T rather than ICa,L, which is unlikely
only slightly attenuated it. in view of the more physiologically based simulations in the
Thus, in guinea pig at least, there are conflicting data. current study. In addition to the models described above,
However, the evidence argues for a dominant role for ICa,L. For Kotter and Feizelmeier (1998) also have modeled DA neurons,
This work was funded by a Biomedical Engineering Research grant from the
Whitaker Foundation and National Institute of Neurological Disorders and
Stroke Grant NS-37963.
Address for reprint requests: C. Canavier, Dept. of Psychology, University
of New Orleans, New Orleans, LA 70148.
REFERENCES
CALLAWAY, J. AND WILSON, C. Imaging voltage-dependent calcium oscillations
in the dopaminergic neurons of the substantia nigra, pars compacta. Soc.
Neurosci. Abstr. 23: 1278, 1997.
CANAVIER, C. C. Sodium dynamics underlying burst firing and putative mech-
FIG. A1. DA neuron nullclines, associated trajectories, and kinetics of the
anisms for the regulation of the firing pattern in midbrain dopamine neurons:
reduced model. Oscillation trajectory in the V-[Ca21]i plane for SOPs and a computational approach. J. Comput. Neurosci. 6: 49 – 69, 1999.
square-wave oscillations (A). Time course of the potential and calcium oscil- CANAVIER, C. C., CLARK, J. W., AND BYRNE, J. H. Simulation of the bursting
lations is shown in B. Reduced model (- - -) has a smaller calcium excursion activity of neuron RIS in Aplysia: role of ionic currents, calcium balance,
and a higher frequency compared with the complete model (—). and modulatory transmitters. J. Neurophysiol. 66: 2107–2124, 1991.
CARDOZO, D. L. Midbrain dopaminergic neurons from postnatal rat in long
term primary culture. Neuroscience 56: 409 – 421, 1993.
dV CARDOZO, D. L. AND BEAN, B. P. Voltage-dependent calcium channels in rat
50
dt midbrain dopamine neurons: modulation by dopamine and GABAB recep-
tors. J. Neurophysiol. 743: 1137–1148, 1995.
d@Ca 21 # i CARLSON, N. R. Fundamentals of Physiological Psychology (2nd ed.). Need-
50 (A1) ham Heights, MA: Allyn and Bacon, 1992.
dt CARTER, C. J. Topographical distribution of possible glutaminergic pathways
from the frontal cortex to the stratum and substantia nigra in rats. Neuro-
Additionally, we assumed that a rapid equilibrium is reached between pharmacology 21: 379 –383, 1982.
the free calcium ([Ca21]i) and the buffered calcium (n[B]OC). We CHAY, T. R. AND KEIZER, J. Minimal model for membrane oscillations in the
then used the ratio of the free calcium to the total calcium to scale the pancreatic beta cell. Biophys. J. 42: 181–190, 1983.
derivative of [Ca21]i (Chay and Keizer 1983). If we assume that the DEFAZIO, T. AND WALSH, J. P. Differential ontogeny of calcium currents in the
[Ca21]i kinetics are slow compared with those of V, we can use the rat nigrostriatal system. Soc. Neurosci. Abstr. 21: 1577, 1995.
DEMIR, S. S., CLARK, J. W., MURPHY, C. R., AND GILES, W. R. A mathematical
slope of the potential nullcline at the equilibrium point as an indica- model of a rabbit sinoatrial node cell. Am. J. Physiol. 35: 832– 852, 1994.
tion of the stability of the system. If the slope is negative, we have a DUNIA, R., BUCKWALTER, G., DEFAZIO, T., VILLAR, F. D., MCNEIL, T., AND
stable equilibrium as any perturbations return the trajectory to the WALSH, J. Decreased duration of Ca21-mediated plateau potentials in striatal
equilibrium point. A positive slope indicates an unstable equilibrium neuron from aged rats. J. Neurophysiol. 764: 2353–2363, 1996.
point as it repels all points in its vicinity to a limit cycle. The DA ELLINOR, P. T., ZHANG, J. F., RANDALL, M., ZHOU, A. D., SCHWARZ, T. L.,
neuron nullclines for the SOP and square-wave oscillations along with TSIEN, R. W., AND HORNE, W. A. Functional expression of a rapidly
inactivating neuronal calcium channel. Nature 363: 455– 458, 1993.
their trajectories are shown in Fig. A1A. The waveforms generated by FEIGENSPAN, A., GUSTINCICH, S., BEAN, B. P., AND RAVIOLA, E. Spontaneous
the reduced (two dimensional) model (Fig. A1B, - - -) in are compared activity of solitary dopaminergic neurons of the retina. J. Neurosci. 18:
with those of the complete model (—). In both SOP and square-wave 6776 – 6789, 1998.
oscillations, the reduced model has a smaller calcium excursion, and FISHER, T. E. AND BOURQUE, C. W. Calcium-channel subtypes in the somata
as a result, a higher frequency. and axon terminals of magnocellular neurosecretory cells. Trends Neurosci.
Rough tunings of the model were performed through analysis of the 19: 440 – 444, 1996.
FOX, A. P., NOWYCKY, M. C., AND TSIEN, R. W. Kinetic and pharmacological
nullclines of the reduced system. First, the slope of the potential properties distinguishing three types of calcium currents in chick sensory
nullcline at the equilibrium point of the reduced system was a good neurones. J. Physiol. (Lond.) 394: 149 –172, 1987a.
predictor of the stability of the complete system. Using this rule, we FOX, A. P., NOWYCKY, M. C., AND TSIEN, R. W. Single-channel recordings of
were able to make large changes in the system equations while three types of calcium channels in chick sensory neurones. J. Physiol.
maintaining the model in the correct range for oscillation. Also, (Lond.) 394: 173–200, 1987b.
because [Ca21]i varies slowly relative to the potential, the trajectory FREEMAN, A., MELTZER, L., AND BUNNEY, B. Firing properties of substantia
nigra dopaminergic neurons in freely moving rats. Life Sci. 3620: 1983–
in phase plane remains more or less on the potential nullcline, jumping 1994, 1985.
between stable branches when an unstable region is encountered. This FUJIMURA, K. AND MATSUDA, Y. Autogenous oscillatory potentials in the
was very helpful in estimating from the nullclines, the extent of neurons of the guinea pig substantia nigra pars compacta in vitro. Neurosci.
[Ca21]i and V excursion without running the full simulations. Lett. 104: 53–57, 1989.
CALCIUM DYNAMICS IN PACEMAKER AND BURST FIRING 2261
GRACE, A. AND BUNNEY, B. The control of firing pattern in nigral dopamine PING, H. AND SHEPARD, P. Apamin-sensitive Ca21 activated K1 channels
neurons: burst firing. J. Neurosci. 4: 2877–2890, 1984. regulate pacemaker activity in nigral dopamine neurons. Neuroreport. 73:
GRACE, A. A. AND ONN, S.-P. Morphology and electrophysiological properties 809 – 814, 1996.
of immunochemically identified rat dopamine neurons recorded in vitro. PING, H. AND SHEPARD, P. Intracellular Ca21 chelation evokes bursting plateau
J. Neurosci. 910: 3463–3481, 1989. potential in nigral dopamine neurons in vitro. Soc. Neurosci. Abstr. 23:
GU, X., BLATZ, A., AND GERMAN, D. Subtypes of substantia nigra dopaminer- 1211, 1997.
gic neurons revealed by apamin: autoradiographic and electrophysiological REGAN, L. J. Voltage-dependent calcium currents in purkinje cells from rat
studies. Brain Res. Bull. 28: 435– 440, 1992. cerebellar vermis. J. Neurosci. 11: 2259 –2269, 1991.
HAIRER, E. AND WARNER, G. Solving Ordinary Differential Equations. II. Stiff RINZEL, J. AND ERMENTROUT, G. B. Analysis of neuronal excitability and
and Differential-Algebraic Problems. Springer Series in Computational
oscillations. In: Methods in Neuronal Modeling: From Ions to Networks,
Mathematics. New York: Springer-Verlag, 1990.
edited by C. Koch and I. Segev. Cambridge, MA: MIT Press, 1998, p.
HILLE, B. Ionic Channels of Excitable Membranes (2nd ed.). Sunderland: MA:
Sinauer Associates Sunderland, 1992. 251–292.
HOEHN, K., WATSON, T. W., AND MACVICAR, B. Multiple types of calcium ROBERTSON, S., JOHNSON, J., AND POTTER, J. The time course of Ca21 exchange
channels in acutely isolated rat neostriatal neurons. J. Neurosci. 133: 1244 – with calmodulin, troponin, and myosin in response to transient increases in
1257, 1993. Ca. 21. Biophys. J.34: 559 –569, 1981.
HSIAO, C. F., DEL NEGRO, C. A., TRUEBLOOD, P. R., AND CHANDLER, S. H. Ionic ROBLEDO, P. AND FEGER, J. Excitatory influence of rat subthalamic nucleus to
basis for serotonin-induced bistable membrane properties in guinea pig substantia nigra pars reticulata and the pallidal complex: electrophysiolog-
trigeminal motoneurons. J. Neurophysiol. 796: 2847–2856, 1998. ical data. Brain Res. 518: 47–54, 1990.
JOHNSON, S. W. AND NORTH, R. A. Two types of neuron in the rat ventral ROTH, R. H. AND ELSWORTH, J. D. Psychopharmacology: The Fourth Gener-
tegmental area and their synaptics inputs. J. Physiol. (Lond.) 450: 455– 468, ation of Progress (4th ed.). New York: Raven Press 1995, vol. 1, chapt. 21,