ABSTRACT

Autorhythmic bursting pacemaker cells have been hypothesized
to function by a calcium-coupled system. Bursting pacemaker cells
in Aplysia californica, the RI5's and LOB's were flushed with saline
solutions containing .01, 10, and 60 mM concentrations of Cazt
I found that the burst duration is affected by the extracellular
concentration of Ca2+, shortening in high Ca2+ and lengthening in low
Ca2+ solutions. This is consistent with an increase and a decrease in C
current(calcium dependent potassium current) activation. The
mechanism behind the interburst interval, however, is still unclear.
The changes in its duration point to a homeostatic mechanism.
INTRODUCTION
Bursting pacemaker cells are present in a variety of living

organisms from human beings to mollusks. These neurons fire in a
repeating cycle of a train of action potentials, followed by a quiescent
period of hyperpolarization. This firing pattern is involved in the
regulation of constant, rhythmic activities such as respiration,

peristalsis, sensory perception, and circadian rhythms. In the
Aplysia californica, the R15 bursting pacemaker neuron may signal
peptide release for homeostatic osmotic regulation.
Several models from RI5 bursting pacemakers have been
developed in order describe how this mechanism of firing may be
controlled, yet a quantitative characterization of the membrane
conductances has not yet been produced.
In the theory presented
by Thompson and Smith on bursting pacemaker cells in the Tritonia
nudibranch, the recurrent, firing pattern is depicted as being under
7
the regulation of a calcium-coupled system. In their studies, two
ionic conductances have been pinpointed as the major determinants
of the burst pacemaking activity. Other studies have also proposed
that one conductance is a slow, inward Ca2+ current found solely in
bursting pacemaker cells(Gormann, Hermann, and Thomas, 1981);
this conductance is voltage dependent and insensitive to
TTX(Thompson and Smith, 1987). The opposing slow, outward
current could be a Ca2+ dependent potassium conductance which
might cause the hyperpolarization during the quiescent
period. (Johnston, 1978: Gormann and Hermann, 1979). The
concentration of the intracellular Ca2+ could also be a regulatory
determinant, for it has been found that injecting Ca2+ into cells had
produced an outward current which had significant magnitude to
create the postburst hyperpolarization.(Gormann and Thomas, 1977).
The outward current could be directly activated by the free calcium
ions which are found in close proximity to the inner membranes.
Although information has been found on these currents, the specific
details on which conductances are the most crucial at different points
in the cycle has not been completely elucidated.
Smith and Thompson have created a Ca2+ - coupled model which
describes when these ionic currents are highly activated or in a state
of decay. The three currents of central importance are the:
1) B current, slow Ca2+ and Nat current, activates rapidly during
depolarization
2) D current, fast Ca2+ current, responsible for the greatest influx of
Ca2t
3) C current, slow Ca2+ dependent K current, decays within 20-30
seconds back to repolarization.
The bursting pattern appears to be caused by an interaction of these
three currents. At the start of the cycle with the first action
potential, the depolarizing pulse activates the fast D current and slow
B current. Simultaneously, C current is activated, yet at a slower rate
than B, so the depolarizing action potentials proceed. B current is
soon saturated, and so much Ca2+ has accumulated near the cell
membrane that C current reaches larger values of activation. Thus,
the firing pattern terminates, and the silent pattern of
hyperpolarization begins. While B current decays, the C current also
is inactivating since it is Ca2+ dependent. B current increasingly
decays as Ca2+ is extruded from the cell by a Ca2+ pump. As C
current decays, the membrane can once again undergo depolarization
so that B current can be activated. The the series of action potentials
çan be fired once again.
Furthermore, several ambiguities still exist on the regulation of
intracellular calcium concentration. The exact concentration of free
Ca2+ is hard to detect since it is involved in several mechanisms. In
the model, the factors affecting calcium are separated into three

processes:
1) influx across the cell surface membrane
2) the efflux across the surface membrane
3) the intracellular diffusion and binding of Ca
Firstly, Ca2+ may primarily influx by channels of the B and D
currents, while the efflux is activated by a Ca2+ extrusion pump. It is
activated in approximately direct proportion to the concentration of
calcium at the inner surface membrane. The free intracellular Ca2+ is
hypothesized to diffuse according to Fick's Law of diffusion. In

äddition, it is assumed that Ca2+ is subject to several sequestration
reactions and buffering once it enters the cell. In this study, the
extracellular Ca2+ concentration was varied from a low to high
concentration. These extreme conditions could contribute in
exhibiting calcium's primary roles in the burst pacemaking activity¬

to discover if the cell would adapt to the altered conditions in a
homeostatic fashion or if it would change its firing pattern in a
predictable pattern. This paper attempts to elucidate more on
calcium's specific roles in this rhythmic pattern of firing.
7
Materials and Methods

Specimens of Aplysia californica, were attained from Sea Life
supply (Sand City, CA) and Marinus Animals(southern California).
They weighed from 100-500 grams and were placed in tanks with
continuously running sea water from Monterey Bay. The abdominal
ganglia (Figure 1) was isolated, transferred to a plastic dish lined with
silgard, and immersed in artificial sea water (ASW).
After 1 hour treatment with dispase enzyme and 1 hour of ASW
wash at 10 degrees celsius, the sheath of the abdominal ganglia was
spliced away, using special forceps and a razor blade chip. Select
parts of the ganglia were removed, exposing the R15 cell and the left
quadrant bursting cells.
Kwik-fil glass capillaries were utilized. The microelectrodes,
which measured the membrane potential, were filled with 3 M KCL,
and a silver wire which was plated with silver chloride represented
the electrical contact between the KCL solution and the amplifier. The
microelectrodes had resistances of 1-5 megaohms.
Membrane potential was monitered by utilizing a microelectrode
amplifier and was recorded on a chart recorder. Data were
analyzed by hand and plotted using computer program Lotus-123.
Äfter desheathing, an exposed bursting pacemaker cell such as R-

15 or a left upper quadrant cell was impaled by a microelectrode
while a constant perfusion system was running with ASW(Figure 2)
The perfusion system was constituted by a drip system which flowed
at a constant rate and a suction system which removed excess saline.
The experiment consisted of a series of physiological saline changes
beginning with artificial sea water and then alternating this control
with sea water containing varied amounts of calcium. Each saline
washed over the animal for 15-60 minutes, and measurements were
obtained through out the entire experiment. The solutions were
altered once the cell appeared to reach a steady state in that saline.
The last 10-15 periods of bursting were selectively chosen for
analyzation to ensure that the changed solution had surrounded the
cell.
Physiological salines of ASW containing varying concentrations of
calcium were created at pH of 7.8, combining different amounts of
IM stock solutions and dry salts. Table I displays the varying ionic
solutions for the saline solutions in mM.
Table I. Compostition of Physiological Salines (mM)
Nacl KC MgC Cacl
Hepe
470
50
10
ASW
10
470
O Ca ASW
470
30 Ca Asw
30
470
10
60 Ca ASW
Computer Modelling
I used a mathematical model of bursting pacemaker mechanisms
created by Smith and Thompson. The model is based on the ionic
conductances and other parameters which affect the Tritonia
bursting pacemaker pattern. Parameters were altered to simulate a
higher concentration and a lower concentration of extracellular
calcium. The Ca-+ extrusion pump rate constant was increased in
order to mimic the results of lowering external calcium
concentrations. This would allow the same amount of Ca2+ to enter
the cell, yet it would be pumped out of the cell more rapidly. Thus,
the C current would not be as highly activated. To simulate high
external calcium, the pump rate was decreased so that the external
Ca2+ concentration was maintained, yet the calcium remained in the
cell for a longer period. In this manner, the cell's Ca2+ dependent
potassium current would be highly activated for a much longer time.
RESULTS
Five bursting pacemaker cells, two RI5 and three LOB cells were
flushed with different physiological salines of 0 Ca2+, artificial sea
water with 10 mM calcium, and 60 mM calcium; their membrane
oscillations were recorded.
The RI5's and LOB's were separated into
two groups because of their varying characteristics. One of the LOB
cells which was recorded was silent upon microelectrode
impalement, so the measurements were taken after the cell had been
stimulated with seretonin. The LOB's were recorded for
approximately 2.7 hours, while the RIS's were monitored for an
average of 5.3 hours.
In the 10 mM Ca2+ ASW solution, before any solution changes,
there were some variations between the LOB and RI5 cell firing
patterns. The R15 cells fired 3-5 action potentials per burst, and the
burst durations were approximately 3.22 seconds long. The
interburst intervals were 83.99 seconds on average for the two R15
cells. The LOB's, on the other hand, had much shorter burst
durations with two cells which had a single spike per burst in ASW.
The other cell, however, had bursts lasting 8.73 seconds with 7-8
spikes per burst. The interburst intervals for the LOB's were much
shorter on average than the R15 cells, lasting 34.7 seconds.
Several parameters were measured in order to compare Smith
and Thompson's model to the experimental cells- the average length
of the burst durations for the RI5's and LOB's, the average number of
action potentials for all the cells, and the average length of the
interburst interval for the two types of cells. In the unaltered
model which was juxtaposed to the control ASW, the burst duration
lasted 19 seconds with 17 spikes per burst. The interburst interval,
on the other hand, had a duration of 63 seconds. For all the
experimental cells, the measurements appear to have lower values
Both the RI5's and the LOB's have average burst durations of 18 and
3.55 seconds respectively, while there are five action potentials per
burst in the control solution. Finally the interburst intervals were
also shorter, 52 seconds for the R15's, 24 seconds for the LOB's, as
compared to the 63 second interburst interval in the model.

In Figure 3, the computer model in the control(lOmM Ca2*) is
depicted in the center while the computer runs which parallel high
and low extracellular Ca2+ are also shown. Those with altered
parameters had distinct changes of shortened burst durations and

lengthened interburst intervals in high Ca2+. This occurred because
Ca2+ remained in the cell for a longer period with a decreased pump
rate, which caused the activation of C current. With low calcium, the
model predicted increased burst durations and decreased interburst
intervals because calcium was extruded more quickly. Thus C
current was not activated for very long, and the hyperpolarizing
force was decreased.
As seen in Figure 4, the different cells were subjected to 0, 10,
and 60 mM Ca2+ artificial sea water. In the model, the average burst
durations decreased with increased calcium concentrations. The
burst durations were 3.28 times longer in the 0 Ca2+ as compared to
the control saline, while they were 1.9 times longer in the 60 mM
saline solution as compared to the control solution. The burst
durations of the R15 cells decrease in length with increased Caz
concentrations, starting from 0 to 10 to 60 mM Ca2+ This correlates
with the predictions of the model, yet the LOB cells display a deviant
behavior since the burst durations increase from 0 to 10 mM Cazt
while a drop in the length of the burst durations is visualized as the
salines are changed from 10 mM to 60 mM Ca. This deviation could
have resulted from the exceptionally long burst durations of one out
of the three LOB's which were tested. For this reason, the burst
duration average for the LOB's is unusually lengthy. While
examining the individual cells, it was found that 75% of the cells had
shorter burst durations in high Ca2+ as compared to ASW, while 80%
of the cells had longer burst durations in the 0 Ca2+ solution. This
manifests that burst duration, for the most part, increases with less
Ca concentration.
In Figure 5, the number of action potentials per burst was
averaged for the whole data set in the different physiological salines.
An increasing number of spikes with lowered Ca2+ concentration
levels corresponds with the lengthened burst duration with lowered
Ca2+ concentration. The lengths of the burst duration and the
number of action potentials per burst both display that burst
durations are dependent on the activation of the Ca2+ dependent K-
channels, since they undergo the greatest activation with higher
amounts of intracellular Ca2+ The model also exhibits a decrease
number of action potentials with increased calcium concentrations
from 39 to 19 to 9 in 0 mM, lOmM, and 60 mM Ca concentrations
respectively, but the slope of the decrease is much steeper with the
model as compared with the actual cells.
While examining Figure 6, it is clear that the interburst intervals
lengthened in duration with increasing concentrations of calcium in
the calculated model. In 60 mM Ca2+, the interburst intervals are
1.59 times longer than in the control solution, and the control saline
has interburst intervals which are 1.19 times the size of 0 Caz
However, the interburst intervals of both the R15 and the LOB cells
did not increase as markedly. The interburst intervals exhibit a
small increase from the 0 mM Ca2+ to the control of 10 mM Ca2+. The
R15 cells have average interburst intervals which are 1.32 times

longer in the control than in the 0 Ca solution and the LOB cells have
interburst intervlas which are 1.23 times longer in the control as
compared to the 0 Ca solution. However, the interburst intervals for
both types of cells do not increase in length when the concentration

is altered from 10 to 60 mM Ca? This implies that the prediction of
increased interburst interval duration because of the highly
activated C current in higher concentrations of calcium is not always
valid. While looking at the detailed accounts for individual cells, it
was found that only 40% of the total cells mimiced the model and

had longer intervals while changing solutions from control to 60 mM
Ca2+ and returning to control. However, when examining the
interburst intervals which were altered from control to 0 Ca2+ and
back to control, 75% of the time the cells had shorter intervals during
the 0 Ca2+ wash as compared to the control.
DISCUSSION
By altering the extracellular amount of Ca2+ several factors
could be affected such as the influx of Ca2+ across the cell surface, the
efflux across the surface membrane, and the intracellular diffusion
and binding of Ca2+ to different sites. I developed a hypothesis on
how the bursting pacemaker pattern would be transformed in higher
and lower extracellular Ca2+ based on the model.
1) In the high calcium saline solution, it was predicted that a
greater calcium influx would occur because of the new availability
and greater concentration gradient. Both the slow, inward Ca2+ or B
current and the fast, inward Ca2+ or D current would be activated to a
greater extent. As a result, the B current would saturate more
rapidly in comparison to the control solution. In addition, more
calcium channels would inactivate as a result of greater Ca2+ binding
to the channel. While the B and D current are active, Ca2+ rushes into
the cell and accumulates near intracellular membrane. This leads to
a greater activation of the Ca2+-dependent B current because Caz
binds and activates these channels which cause hyperpolarization of
the cell. This hyperpolarization eventually terminates the burst of
action potentials more rapidly than in the control solution. In the
model, the Ca2+ extrusion pump is activated in direct proportion to
the intracellular concentration of submembranous Ca2+ If this is
valid, the Ca2+ pump should have a larger activation value with the
higher Ca2+ concentration. The decrease of Ca2+ concentration still
occurs more slowly than the control, though, because of the increased
Ca2+ load from the inward B and D currents. Thus, the interburst
intervals should be longer. Since the extracellular Ca2+ concentration
is much greater than the intracellular concentration, if the Caz
pump which extrudes Ca2+ from the cell is concentration dependent,
then Ca2+ will remain in the cell for a longer period than in the
control solution. C current will maintain activation, and the
interburst interval will be lengthened by the two effects of greater
calcium load and the concentration gradient.
2) In 0 Ca2+ solution, the influx will be lessened so that minute
amounts or no calcium will enter the cell through B and D currents.
Therefore, no Ca2+ binding will occur, and calcium channels will not
be inactivated by Ca2+. Additionally, C current will not be activated.
In this way, the typical periods of hyperpolarization will also be
absent. Either continuous, repetitive firing will occur, or the burst
durations will be lengthened. The absence of C current will either
completely abolish the periods of hyperpolarization or shorten their
durations.
The most striking finding of this study is that extracellular Caz
affects the length of the burst durations by decreasing their duration
in increased Ca2+ salines and increasing the duration in 0 Ca2+ salines.
This is consistent with the idea that the coupled interaction of the
Ca2+ inward currents with the Ca2+ dependent K+ current determines
the length of the burst duration. The experimental results match the
prediction of the model. In 0 Ca2+ solutions, no Ca2+ would influx, and
the burst durations would be altered because the B, D, and C current
are either present with extremely decreased conductance or absent.
The normal, continuous firing still occurs, however, because other
currents such as the fast sodium and slow potassium currents are
still active. In low calcium, the burst durations were lengthened, yet
with high external calcium, the burst durations were shortened, often
reduced to single spikes per burst from the normal, multiple spikes
per burst. Increased [Calo causes a greater influx into the cell, and
this increased influx leads to a higher concentration of intracellular
Ca2+. This causes a greater accumulation of intracellular Ca2+ which
then leads to inactivation of Ca2+ channels by intracellular Caz
binding. Simultaneously, the higher intracellular Ca2+ concentration
causes an increased activation of the Ca2+ dependent K current. The
outward C current then activates at a larger value than the inward B
current. Thus, the burst duration is terminated more quickly.
Although the results from the actual cells match the results from the
model, the model depicts a greater change in the different salines.
This could be a result of the fact that this bursting pacemaker was
modelled after a Tritonia nudibranch, not an Aplysia.
The results from the actual cells also matched the predictions for
the model in the case of the number of action potentials per burst.
Similar reasoning for the burst durations help explain the lowered
number of action potentials in higher concentrations of Ca2+ and the
increased number in low Ca2+. A more highly activated C current in
the high concentration of Ca2+ results in a stronger depolarizing force
which makes it difficult for the cell to fire.
By comparing the lengths of the interburst intervals, I expected
to discover more about the method of Ca2+ extrusion from the cell.
While calcium accumulates in close proximity to the membrane, C
current would be activated, causing hyperpolarization. If the Caz
pump functioned independently of Ca2+ concentration, then the
interburst intervals should have remained constant. This hypothesis
can be rejected because the length of the interburst intervals does
vary with the different saline solutions. If the Ca2+ pump was
dependent on the concentration, the interburst intervals should have
been prolonged in 60 mM Ca and shortened in 0 mM Ca. In the

model, the interburst intervals do display this increase in length with
increasing Ca2+ concentration. However, while looking at the
averages from the R15's and the LOB's, no definite lengthening
appears as the calcium concentration is increased. The interburst
intervals increase slightly from the nominal 0 Ca2+ level to the
control of 10 mM Ca2*, yet they appear to remain stablilized at a
certain average when the concentrations are examined from the
control to 60 mM Ca2+.
This suggests that the Ca2+ pump does not function in simply - in
7
direct proportion to the intracellular Ca2+ as predicted by the model.
Some homeostatic mechanism could be functioning, so that the Ca¬
pump is regulated to maintain some steady level of Ca2+ if the cell
was stressed to extreme conditions. A positive feedback system
could also be active where with increased levels of intracellular Ca-
the Ca2+ pump would be increasingly activated. The mechanism of
this postulated homeostasis is quite unclear, so further
experimentation must be done.
The next logical set of experiments is to alter the amounts of

intracellular calcium buffering by injecting calcium chelators such as
EGTA. The extracellular concentration of Ca2+ could also be reduced
by replacing the calcium molarity with magnesium, and
subsequently injecting EGTA for the lum Ca2+ found in de-ionized
water. The voltage trace could then be recorded for cells with high
and low intracellular calcium. Their recrurrent, bursting patterns
could be compared to the bursting characteristics in high and low
The time courses of the intracellular Ca
extracellular calcium.
concentration could then be monitored by using fluorescent dyes
such as fura-2. More information could be found out about the
mechanism of the Ca pump. If it was Ca2+ dependent, then Ca-
should be extruded at a faster rate with a higher concentration of
intracellular Ca2+
Literature Cited
Gorman, A. L. F. and A. Hermann. (1979) "Internal Effects of Divalent
Cations on Potassium Permability in Molluscan Neurones.
Journalof Physiology 296, 393-410.

Gorman, A. L. F., A. Hermann, and M. V. Thomas. (1981) "Intracellular
Calcium and the Neuronal Pacemaker Activity. Federation
Proceedings. 40, 8, 2233-2239.
Gorman, A. L. F. and M. V Thomas. (1978) "Changes in the
Intracellular Concentration of Free Calcium lons in a Pace¬

Maker Neurone, Measured with the Metallochromic Indicatoi
dye Arsenazo III." Journal of Physiology 275, 357-376.
Johnston, Daniel. (1978) "Voltage, Temperature, and lonic
Dependence. Journal of Physiology 289, 145-157.
Jones, Bradley R. Slow lonic Currents Underlying Post Inhibitory

Rebound In Aplysia Neurons, Stanford University, 1986.
Kandel, Eric K. Behavioral Biology of Aplysia. W. H. Freeman and
Company, U. S. A., 1979.
Smith, Stephen and Stuart Thompson (1987) "Slow Membrane
Currents in Bursting Pace-Maker Neurones of Tritonia. ournal

of Physiology 382, 425-428.
Thompson, Stuart H. Membrane Currents Underlying Bursting in
Mollusçan Pacemaker Neurons. University of Washington, 1976.
Figure Legend.
figure 1. Diagram of a dorsal view of the abdominal ganglia from
Aplysia californica, R15 and LOB cells were used in the experiments.
firgure 2. Diagram of the experimental set-up with the constant
perfusion system(drip and suction system) and a single
microelectrode in cell.
figure 3. Graphs of Various Computer Runs on Smith and Thompson's
Model (A) Computer Run of Increased Pump Rate at 5.0 E* as
compared to 0 [Ca2+0 experimentally (B)Computer Run with no
parameters altered, pump rate at 4.0 E* as compared to control (C)
Computer Run with Decreased Pump Rate at 3.0 E* as compared to
60 mM Ca2+ experimentally
figure 4. Average Burst Durations in .01, 10, and 60 mM Ca2+
solutions. Burst durations were measured for single periods.
Measurements were taken from the last 10-15 periods in each saline
to ensure saturation of the solution.
figure 5. Average Number of Spikes per Burst. The number of spikes
per individual burst was measured in .01, 10, and 60 mM Ca2+
Measurements were obtained from last 10-15 periods in each saline.
figure 6. Average Length of Interburst Interval. The last 10-15
periods of silent hyperpolarization were measured in .01, 10, and 60
mM Ca2+
L. connective
Left rostral
quarter-ganglion.
LD
LD,
LD,,
LDE¬
Left caudal
quarter-ganglion
Siphon n

Genital-pericardial n.
DORSAL
VR
R6/ R7188
11

De

AD
rR
YRI5

1c 10 R0.
Her
LBycLBLD
LBy
Figure 1
R. connective
Right rostral
quarter-ganglion
RD,
Right caudal
-quarter-ganglion
Branchial n.
DRIP System
S
Hingle Electrode. Impaled
in Burster
—
6UCTION
770

APLYSIA
A. GANGLIA
Figure 2
Ca pump rate = 5.0E4

1 7—
8e0 900
o
Seconds
Ca pump rate = 3.OE4
240
102
1000
100
seconds
Figure3
(A)
(B)
(C)
kvatakaa-
s 0
ES
D

4
NN
(spuos o suo) suogoing ising
&
HuS