Abstract:
Zebrafish, Danio rerio, are a useful model system for vertebrate cardiovascular
research. Their transparent bodies during the larval stage of development allow for in
vivo experimentation. Pharmacological experimentation reveals important similarities
between zebrafish hearts and those of mammals. Muscarinic cholinergic receptor activity
was demonstrated through the dose dependent decrease in heart rate caused by the
acetylcholine analog carbachol. Adrenergic receptor responses were elicited by
norepinephrine and isoproterenol. The inhibition of intracellular calcium release
channels by ryanodine also had a decelerating effect on heart rate, as did the disruption of
sodium-potassium active pumps by ouabain. The stimulation of an undefined nitric
oxide pathway by two nitric oxide donors also slowed heart rate. Clearly, the
pharmacologic responses of the zebrafish heart should be further explored as a model for
vertebrate cardiovascular physiology and as a foundation for research in the genetics and
cardiovascular development of zebrafish.
Introduction
Danio rerio, commonly known as zebrafish, are fresh water tropical fish that
naturally occur in India. Zebrafish are used for biological research due to their short
generation time, the relative ease of their maintenance, their sufficiently small body size
permitting the storage of large populations. The defining characteristic that sets them
apart from other potential vertebrate model systems is that their bodies in the larval stage
are transparent. Their transparency allows for detailed observations of the cardiovascular
system in vivo without the complications of surgical procedures.
I began studying the zebrafish cardiovascular system with the goal of studying
vascular development, growth and repair. Technical problems with this goal included the
difficulty in obtaining cross-reacting mammalian antibodies to receptors and to endothelial
cell growth factors and the lack of a visible response to certain pharmacological agents
that have been shown to play a role in the vascular actions of other model systems. The
size of zebrafish, while allowing for easy storage, made controlled wounding extremely
difficult. These problems highlighted the need for more knowledge describing the
cardiovascular physiology of the normal zebrafish.
The short generation time and transparent bodies of the zebrafish caused the fields
of genetics and developmental biology to be interested in zebrafish as a model system to
study cardiovascular mutations among others. However, before abnormal zebrafish
hearts can be characterized, better understanding of the normal zebrafish heart is needed.
The zebrafish heart has developed two distinct chambers before the larva hatches (Kimmel
255). Heart beat developes even sooner, within 24 hours of fertilization. Vasodilation,
vasoconstriction, changes in stroke volume, changes in heart rate are essential actions of
vertebrate cardiovascular systems allowing response to stimuli. These processes can be
witnessed in vivo in zebrafish hearts through stimulation of muscarinic cholinergic
receptors, adrenergic receptors, intracellular calcium pathways, sodium-potassium active
pumps, and nitric oxide pathways. The hypothesis of this project was that it would be
possible to examine the muscarinic cholinergic and adrenergic receptor activity and the
intracellular pathways of the larval zebrafish heart through manipulation with known
pharmacological agents.
Cholinergic receptors are responsible for actions of the parasympathetic system.
They include both nicotinic and muscarinic receptors. To activate these receptors an
acetylcholine analog, carbachol, was used. To control for the fact that muscarinic rather
than nicotinic receptors were being activated, Curare was used as an immobilizer to keep
the fish in place and also as an inhibitor of nicotinic receptors.
Adrenergic receptors are part of the sympathetic system. Fish have been shown to
have such receptors and respond to adrenaline and norepinephrine but with less increase in
heart rate than expected by mammals (Satchell 162). Norepinephrine activates both alpha
and Bi adrenergic receptors, while isoproterenol only activates beta receptors (Katzung
127-128).
The pharamacological agent Ryanodine inhibits ryanodine receptors necessary in
the intracellular release of Calcium into the cytosol. It is clinically used to treat the
excessive release of calcium from the sarcoplasmic reticulum that may result in the fatal
condition of malignant hyperthermia (Taylor 188). Calcium is needed to although cardiac
cells to contract properly.
In addition to inhibition of ryanodine receptors by ryanodine, intracellular sodium
potassium active pumps were inhibited by ouabain. As a cardiac glycoside, it inhibits the
natural exchange of sodium and potassium and causes the cells to lose potassium and
accumulate calcium (Kelly and Smith 814). Thus the cells become depolarized.
The role of nitric oxide in relation to heart rate was explored using to nitric oxide
donors, Spermine NONOate and S-nitroso-N-acetylpenicillamine (SNAP). Spermine
NONOate releases higher levels of nitric oxide but its release period is shorter than that of
SNAP (Shih 12). SNAP releases low levels of nitric oxide for a considerable time period.
Materials and Methods
Adult zebrafish were maintained in fresh water at 28° Celsius, under a controlled
light source with a regular light cycle of 9 dark hours in a 24 hour period. In order to
obtain fertilized eggs, marbles were placed in the tank the night before. I then syphoned
the bottom of the tank in the morning and manually separated the fertilized eggs. Next, I
placed the eggs in 10% normal hanks and 90% de-ionized water solution in an incubator
at 28° Celsius.
Larvae between 6 and 8 days post fertilization were used for each experiment. I
immobilized each larva with 2 mg d-Tubocurarine in 1 ml 10% Hanks solution. The
larvae were placed in a temperature controlled stage of the light microscope at 28
Celsius. A robotic arm was used to inject all the pharmacological agents except
NONOate through a glass capillary tube directly into the caudal end of the pericardium.
luM Carbachol, 100 uM Isoproterenol,100 uM norepinephrine, 200 uM SNAP, 1 uM
Quabain, and 100 uM Ryanodine were the concentrations injected. NONOate was
added to the diluted Hank solution surrounding the larva to a final concentration of 100
UM, instead of injection. Vacuum grease helped secure the larva in the same place on the
slide. All injections were made at a pressure of 60 psi.
Each injection and reaction was observed under a magnification of 10 with
Nomarski optics and videotaped. We digitized the heart rate from the visual data of the
video tape by using a light sensor and the computer program Dempster. The interpeak
interval, or period, was measured using this program. The heart beat is the frequency, or
inverse of the period, converted to beats per minute.
The control for each experiment was a baseline reading of that particular larva’s
heart beat and circulation before the administration of the pharmaceutical agent. In
addition to these individual controls, I recorded larval heart beat under anesthesia by
luM MS 222, to control for the effects of d-Tubocurarine. A control was needed for d-
Tubocurarine because it is a vasodilator and it releases histamine (Taylor 187).
Results
There were no clear differences between the average heart rate of six larvae
immobilized by Curare and one anesthetized by MS 222 (212 bpm and 189 bpm
respectively as illustrated in figure 1). Figure 2A comprises the raw data obtained from
the light sensor for carbachol. Carbachol demonstrated a dose dependent deceleration of
heart rate.
The heart rate of larval zebrafish increased in response to 100 uM isoproterenol
and 100 uM norepinephrine, with a greater increase in response to isoproterenol (Figure
3). The acceleration in response to norepinephrine reached a plateau at approximately
220 bpm even with further doses of norepinephrine (Figure 4).
Both nitric oxide donors decreased the heart rate. Within 15 minutes of injection
of 200 uM SNAP, the heart rate had decreased by half its original rate (Figure 5). Within
the same time period of 15 minutes, NONOate had less of an effect on heart rate
(decreasing it by 40 beats per minute) than SNAP did (decreasing it by 140 beats per
minute) (Figure 6). Overall, SNAP decreased heart rate much more dramatically then
NONOate (Figure 7).
Within 12 minutes, 100 uM ryanodine decreased the heart rate from 212 beats per
minute to 89 beats per minute (Figure 8). Increasing doses of 1 uM ouabain slowed
the heart rate from 216 beats per minute to 82 beats per minute.
In addition to the change in heart rate, the standard deviation of heart rate is a
measurement of cardiac arrhythmia. Ryanodine, both nitric oxide donors, and oubain
caused profound cardiac arrhythmia. However, ryanodine caused the most dramatic
arrhythmia (Figure 10). The maximum standard deviation of ryanodine was 472. The
standard deviation is a slightly misleading measurement of arrhythmia however because
the data are skewed heavily to the side o
slower heart rates; the major arrhythmias were
all long pauses in heart rate rather than irregular accelerations. Figure 11 shows the raw
data from a portion of the ryanodine experiment, including two long pauses in heart rate.
Discussion
The difference in the degree to which both nitric oxide donors decreased the heart
rate could be due to the procedure of bathing the larva with NONOate rather than directly
injecting it into the pericardium. Another possible reason for this difference could be the
differences in concentration. NONOate had been expected to release more nitric oxide
than SNAP.
Both carbachol and the nitric oxide donors dramatically slowed the heart rate.
Further experimentation is needed to show the relation, if any, between the activation of
muscarinic cholinergic receptors and the nitric oxide pathway. In a 1996 study by
Feldman et. al., the muscarinic cholinergic receptors in rats that were involved in
increases in heart rate were linked to a nitric oxide production pathway by
experimentation that inhibited this cardiac acceleration by the application of nitric oxide
inhibitors L-NAME and Methylene blue (21 1). Now that both nitric oxide and carbachol
show acceleration in zebrafish heart rate, important further experimentation includes
manipulation with nitric oxide inhibitors. Another important test of the stimulatory
action of muscarinic cholinergic receptors by carbachol would be to block such
stimulation with atropine. This manipulation could be combined with nitric oxide donors
to see if artificial activation of the nitric oxide pathway down stream was enough to
overcome the inhibition of the proposed pathway upstream. An important component of
future experiments would be to compare light sensor information from different regions
of the heart because Davies et. al. states that cholinergic innervation only occurs in the
fish atrium not the ventricle (261).
The actual reaction of the heart to the catecholamines was more than the expected
one. Fish heart rates are expected to increase “15-20% compared with the 100%
increase reported in some mammals" (Satchell 162). Although the direct effect of
norepinephrine in humans is an accelerated heart rate, the culmination of effects is often a
slowed heart rate because of the compensatory mechanisms. In the case of larval
zebrafish, these compensatory mechanisms were not detected. The effect may be masked
by part of the experiment's design: increasing doses of norepinephrine were added
(Figure 4). We believed the multiple doses would not have an additive effect because of
the short time period of the drug’s action. However, the increase in dose caused by the
increase in duration may have overcome the body’s compensatory mechanisms. The
extremely rapid heart rate of more than 220 beats per minute may be possible only in
these larval fish. Fetal mammals have a faster heart rate than that of adult mammals of
the same species. The response to catecholamines may be mediated through increases in
calcium flow within the contractile mechanisms of the cardiac cell (Satchell 162). This
idea is supported by the changes in heart rate and rhythm pattern caused by ryanodine.
Ryanodine produced not only a dramatically slower heart rate, but also large
standard deviations in this rate. These arrhythmias may be caused by the dramatically
slowed rate and the need (according to Starling’s Law) for the heart to be sufficiently full
and the cardiac muscle thus sufficiently stretched, before the heart can contract. In fact,
calcium is the molecule responsible for contraction of cardiac cells in relation to
Starling’s law because cardiac muscles are more sensitive to being stretched because the
troponin in cardiac muscle cells is more likely to bind to calcium than is other muscle
troponin (Satchell 38).
Calcium may also contribute to the change in rhythm not only because of this
difference of the affinity of cardiac cell fibers for calcium, but also in some other
mechanism. Calcium plays a role in the regular rhythm of the vasomotion of arteries, and
ryanodine can inhibit that role (Griffith and Edwards H1696). The data collected by
ryanodine is surprising for a teleost because previous studies argued that extracellular
rather than intracellular sources of calcium played more prominent roles (Vornanen
R1432). Bradykinin, verapamil, and diltiazem have been used to show differences in the
roles of extracellular and intracellular calcium sources in canine cells and should be used
in future experimentation to explore the role of calcium in the zebrafish heart (Yang et.
al. 59).
Quabain's mode of action was to disrupt the active exchange of sodium and
potassium. Theoretically as a cardiac glucoside, ouabain should have contributed to an
increase in intracellular calcium and thus had opposite effects of ryanodine’s inhibition of
intracellular release of calcium. A study by Gjini, et. al. show that a cardiac glucosides
reduce the calcium current (95). A reduction in calcium current could have had similar
effects as an inhibition of calcium release, thus explaining the similar responses to
ryanodine and ouabain.
A significant next step in this research will be to apply the tool of the light sensor
to examine perfusion rates and quantify vasoconstriction and vasodilation. Nitric oxide is
a vasodilator (Eckenhoff and Longnecker 356). Histamine which may be stimulated by
Curare is also a vasodilator (although the control showed it did not exhibit clear
contributions) (Babe and Serafin 585). The light sensor technique will be especially
important with pharmacological agents like norepinephrine that act as vasoconstrictors in
many areas of the body while acting as vasodilators in the heart muscle. It will help
compare norepinephrine’s effects to those of the vasodilator isoproterenol.
Discuss problems with model?? Such as the fact that an important hypoxia response is
the coupling of cardiac and respiratory rate. Since larval fish 'respirate" through
absorption through the skin rather than gill tissue or lungs, the mechanisms are so
different that the coupling response is not seen Satchell 172.
Suggest another pathway to explore? There are many interesting articles on the role of
adenosine and the response of the AV node including Dennis, DM et. al. 1995. And in
comparison with carbachol (Koglin, J 1994)
Conclusions
Through manipulation by pharmacological agents, these experiments support the
hypothesis that the larval zebrafish heart has muscarinic cholinergic receptor activity,
adrenergic receptor activity, an intracellular calcium release pathway necessary for
proper contraction, and a nitric oxide pathway necessary for normal rhythm similar to
that of mammals. Therefore zebrafish are an effective vertebrate model system for
cardiovascular research. Further experimentation with the vascular responses to these
pharmacological agents is needed, and the transparency of the larval zebrafish bodies will
allow for a new technique of data collection on cardiovascular physiology by a light
sensor.
Literature Cited
Babe, Kenneth S. and Serafin, William E. "Histamine, Bradykinin, and Their
Antagonists." The Pharmacological Basis of Therapeutics. New York: McGraw
Hill Inc.; 1996; 581-600.
Davies, Philip J., Donald, John A., & Campbell, Graeme. "The distribution and
colocalization of neuropeptides in fish cardiac neurons.” Journal of the Autonomic
Nervous System 46. 1994; 261-272.
Eckenhoff, Roderic G. and Longnecker, David, E. "The therapeutic gases." Goodman
and Gilman's The Pharmacological Basis of Therapeutics. New York: McGraw
Hill Inc.; 1996; 349-359.
Feldman, D. S.; Terry, A.V.; and Buccafusco, J. J.. "Spinal muscarinic and nitric oxide
systems in cardiovascular regulation." European Journal of Pharmacology 313(3);
October 1996; 211-220.
Gjini, V. et. al. Frequency dependence in the action of the class III antiarrhythmic drug
dofetilide is modulated by altering L-type calcium current and digitalis glucoside.
Journal of Cardiovascular Pharmacology 3 1(1); January 1997; 95-100.
Griffith, T. M. and Edwards, D. H. "Calcium sequestration as a determinant of chaos
and mixed-mode dynamics in agonist-induced vasomotion." American Journal of
Physiology (272(4(2)) April 1997; H1696-H1709.
Katzung, Bertram, ed. Basic and Clinical Pharmacology. Stamford, Connecticut:
Appleton and Lange; 1998.
Kelly, Ralph A. and Smith, Thomas W. "Pharmacological Treatment of Heart Failure.
Goodman and Gilman's The Pharmacological Basis of Therapeutics
New York: McGraw Hill Inc.; 1996; 809-839.
Kimmel, Charles, et. al. "Stages of embryonic development of the Zebrafish.
Developmental Dynamics 202; 1995; 253-310.
Satchell, George. Physiology and Form of Fish Circulation. Cambridge: Cambridge
University Press; 1991.
Shih, Karen. "Nitric Oxide Promotes while Cholinergic Agonist and Cyclic GMP Inhibit
Neurite Outgrowth of NIE-115 Mouse Neuroblastoma Cells." Pacific Grove,
California: Hopkins Spring Class 1997 Papers.
Taylor, Palmer. "Agents acting at the nueromuscular junction and autonomic ganglia.
Goodman and Gilman's The Pharmacological Basis of Therapeutics.
New York: McGraw Hill Inc.; 1996; 177-197.
Vornanen, Matti. "Sarcolemmal Ca influx through L-type Ca channels in ventricular
myocytes of a teleost fish." The American Physiological Society. 1997; R1432-
R1440.
Yang, CM. et. al. "Bradykinin-stimulated calcium mobilization in cultured canine
tracheal smooth muscle cells." Cell Calcium 2; August 1994; 59-70.
Acknowledgements:
I would like to thank Dr. Stuart Thompson and Dr. Adawia Alousi for their
incredible dedication, instruction, and guidance. It has been an honor to work with them.
I would also like to thank Christian Riley and Matt McFarlane for their continual support
and patient answers to my questions.
Figure Legend
For all figures, the error bars represent standard deviations in heart rate
Figure 1. Although the average larval heart rate under Curare is slightly higher (212 beats
per minute) than that under MS 222 (190 beats per minute), there is no clear difference
between them.
Figure 2 A is a collection of raw data collected from the light sensor on the response to
carbachol: The amplitude merely reflects the change in optical density of the sample site
and is thus unimportant. As the dose is increased through a longer duration of the pulse
of injection, the interval between contractions increased. B: The increase of this interval
was then normalized with respect to the actual dose in picoliters. The actual dose was
calculated by injecting different durations of pulses into oil. There is a clear dose
dependent negative response of heart rate to Carbachol.
Figure 3. The heart rate increased in response to norepinephrine by 20 beats per minute.
It increased by approximately 30 beats per minute.
Figure 4. Increased injections of 100 uM Norepinephrine did not dramatically increase the
response of the heart.
Figure 5: Within 15 minutes of injection of 200 uM SNAP, heart rate decreased by 50%.
Figure 6. Within 15 minutes of administration of 100 uM NONOate, heart rate decreased
By 40 beats per minute.
Figure 7. Administration of 200 uM SNAP decreased the heart rate more than that of
100(M NONOate.
Figure 8. Following the injection of 100 uM ryanodine, the heart rate decreased from 212
beats per minute to 89 beats per minute.
Figure 9. Increasing injections of ouabain decelerated the heart rate even more.
Figure 10. The decrease in heart rate after injection of 100 uM ryanodine was dramatic
and continual. The huge standard deviations (as illustrated by large error bars) at different
sampling periods show cardiac arrhythmia.
Figure 11. Four beats into this sequence, there is a larger interval between peaks. It
represents a pause between heart beats. The heart recovers to normal and then there is an
even more dramatic pause between contractions of a couple of seconds.
300
250
200
150
100
Figure 1. Control


Curari
MS 222
Immobilized & Anesthetized
Figure 2: Carbachol: raw data and dose response curve
A
50ms
150
300
400
veeeete
500
B
2.0 -
1.8
1.6
1.4
1.2
1.0 %
0.0 0.5
10sec
1.5 2.0
CCh injection (pl)
delta
OD
250
200
150 -
Figure 3: Adrenergic Receptor Agonists
.


NE Control
Control Isoproterenol
NE
300
200
100
40 ms
40 ms
Figure 4: Norepinephrine
80 ms
120 ms
Time (min.)
250 ms
20
300
200
100
200 microM SNAP
Figure 5: SNAP
Time (min.)
20
25
300
200
100 -
100micron
NONOate
0
Figure 6: NONOate


2 4 6 8 10 12 14 16 18
time (min.)
200
100
S Control
Figure 7: Nitric Oxide Donors
N Control
SNAP
NONOate
350
300
250
200
150
100
Figure 8: Inhibitor of Intracellular Ca Release Channels

...
Ryanodine
Control
300
200
10
Figure 9: Quabain

50 ms
150 ms
350 ms
Time (min.)
450 ms
15
20
500
400
3 300
200
100
Figure 10: Ryanodine

I
50ms pulses of
100microM Ryanodine
2
8
Time (min.)
28
10
12 14
Figure 11:
A section of raw data from a ryanodine experiment

5 sec