Mechanism of excitation and
contraction in dissociated muscle cells
from the mantle of Loligo opalescens
Adam Stein
Advisor: Prof. William Gilly
Biology 175H
June 3, 1994
Abstract
Muscle cells from the mantle of the squid Loligo opalescens were
dissociated and allowed to settle on plates coated with collagen. They were
bathed in either high calcium medium (HCM, 10.8 mM Ca2*) or low calcium
medium (LCM, 1.3 mM Ca2*) and stimulated with a series of electrical shocks.
The initial medium was then replaced by one with a differing calcium
concentration, and the same series of shocks was again administered. Cells
exhibited a variety of responses to shock, but the most typical was a rapid
contraction that grew stronger with increasing voltage.
Switching the cells from LCM to HCM slightly increased contraction
strength and slightly lowered the minimum voltage necessary to produce a
twitch. Switching from high calcium to low had no discernable effect on
contraction strength or threshold voltage. Likewise, the addition of
tetrodotoxin until it reached a concentration of 200 nM also had no effect,
regardless of the external calcium concentration.
External calcium concentration in the culture medium strongly
affected the cell mortality rate. Almost all cells (»90%) kept in HCM for 20 hrs
were completely unresponsive to electrical stimulation. By contrast, about
50% of cells kept in LCM continued to respond to stimulation even after a
period of several days.
Introduction
Little is known about the mechanism of excitation and contraction in
squid mantle muscle. Most invertebrates studied so far excite via local
membrane depolarizations or graded calcium action potentials (Junge, 1992).
These depolarizations often directly supply the calcium influx necessary to
produce contraction. Several pieces of evidence, however, point to a different
scenario in squid mantle. Voltage clamp recordings have detected large
sodium currents in some muscle cells from the mantle of 3-month-old Loligo
opalescens (W. F. Gilly, personal communication). This finding is
particularly interesting in light of the fact that there are few examples of
vertebrate muscle that fire sodium action potentials (Schwartz and Stühmer,
1984). Furthermore, ultrastructural studies of another species of squid,
Symplectoteuthis oualaniensis, indicate the presence of a sarcoplamic
reticulum in mantle muscle cells (Moon and Hulbert, 1974), suggesting that
these fibers possess an internal calcium store and may release Ca2t from this
organelle to mediate excitation-contraction coupling. To further explore the
mechanism of excitation and contraction, the responses of dissociated mantle
muscle cells to electrical stimulation were examined under a variety of
conditions.
Materials and Methods
Experimental animal
The squid, Loligo opalescens, all between 33 and 45 days old, were
hatched and raised in the Monterey Bay Aquarium on a diet of natural
plankton and algae-enriched Artemia nauplii.
Muscle dissociation
Muscle from an entire squid mantle was first bathed in a 5 mg/ml
Sigma Type IX protease solution for 1 hour at 17°C. For LCM cells, the
protease solution was made with calcium-free artificial sea water, consisting
of 470 mM NaCl, 50 mM MgCl2, and 10 mM HEPES, at pH 7.8. For HCM cells,
the protease solution was made with filtered natural sea water. A glass
micropipette was used to tease muscle cell fragments into a 100 ul drop of the
appropriate culture medium (see below) in a 35 mm plastic culture dish.
Dishes had previously been coated with Vitrogen 100 purified collagen in .012
N HCl, which was allowed to dry completely before dishes were washed 6
times with sterile distilled water. Cells settled for 1 hour, after which 2 ml of
the appropriate medium was added to the dish.
Cells were cultured at 4°C in either a low calcium medium (LCM)
3— 3 mM NaCl, 4.6 mM KCl, 49.5 mM MgCl2, added to Gibco LIE
made up of 2/7
medium at pH 7.6, or high calcium medium (HCM) containing 263 mM
Nacl, 4.6 mM KCl, 49.5 mM MgCl2, 9.5 mM CaClz, added to Gibco L15
medium at pH 7.6. Final calcium concentrations were 1.3 mM in the LCM,
and 10.8 mM in the HCM.
Electric shocks, delivered with an electrolytically sharpened tungsten
wire, were spaced 45 s apart to permit a standard recovery time that was
deemed ample. A Grass Model SD9 Stimulator was used to deliver a standard
series of shocks, which varied in voltage and duration. The stimulating
electrode was always at a positive voltage with respect to a Ag:AgCl wire in
the culture dish. During testing, cells were kept at 16-17°C.
A pasteur pipette was used to change media without disturbing the cell
being viewed under the microscope. Estimates of the efficacy of solution
transfer using a dye showed that approximately 5% of the original solution
was left behind after a change.
Cells were viewed at 400 x magnification with an inverted microscope
(Olympus IMT2) using Köhler illumination. All trials were recorded on 8
mm videotape with a Sony model SSC-C374 camera operating with a shutter
speed of 1/250 sec.
Results
Basic responses of dissociated muscle cells
Dissociation of mantle muscle yielded muscle cell fragments mostly 30-
50 microns in length and 5-8 microns in diameter. A great deal of smaller
breakdown products, as well as some much longer cells, were also produced,
but these rarely contracted repeatedly. The proportion of these fragments
containing nuclei is unknown. Most fibers gave graded responses to electrical
stimulation, shortening slightly when shocked just above threshold, and
contracting most vigorously when shocked at 3-4 times threshold (See also
below). When the electrode was placed near the center of the long axis of the
cells, they usually contracted asymmetrically, curling toward the electrode tip
(Fig. 1). Contraction occured rapidly, usually within one video frame (33 ms),
whereas relaxation was slow, usually taking at least several seconds. Äfter
stimulation, each cell underwent a refractory period, during which an
additional shock brought only a diminished response, or none at all. The
length of this refractory period depended on the size of the shock, as well as
the speed of relaxation. 45 s was judged to be adequate to permit full recovery
Muscle fibers exhibited a variety of responses to stimulation. Some
cells fibrillated for as long as a minute after receiving a shock. These spasms
of activity consisted of fast, repetitive contractions and relaxations in localized
parts of the fiber. The fibrillations were almost always asymmetrical, causing
the cell to curl and uncurl. Some cells contracted fully into balls upon
stimulation and failed to relax. At other times stimulation would induce
swelling and the appearance of a band of central structures, presumably
mitochondria, running the length of the cell. However, no obvious
morphological differences divided the muscle fibers into distinct groups, and
the various response patterns appeared only sporadically. A graded
contraction/relaxation cycle was the most common response.
Responses to calcium
Switching from LCM to HCM caused an immediate, though small,
increase in contractile strength in response to shocks of 0.4 ms duration. In
individual cells (Fig. 2) and on the population level (Table 1), an 8-fold
increase in external calcium caused a slight boost of percent contraction, a
measure of the degree of cell shortening. Percent contraction was calculated
by determining the difference in length between the contracted and relaxed
states, and then dividing this value by the relaxed length. Results from the
two cells in figure 2 are characteristic, rising most rapidly at low voltages and
reaching a plateau at 50-60 V. Population results were analyzed by computing
the mean percent contractions at several shocks strengths (all of 0.4 ms
duration) and comparing mean values at each voltage in LCM and HCM by T-
test. HCM produces a significantly stronger response to 40 V and 60 V shocks.
The difference isn’t significant at 20 V, because this stimulus is very near
threshold and many cells failed to respond.
Behavior near threshold was more carefully studied by using shocks of
various durations and determining the minimum voltage that could produce
a twitch at each duration. On the basis of results from individual cells (Fig. 3)
and from a population analysis (Table 1), switching from LCM to HCM
lowered threshold voltage at every duration. Differences in the population
means are statistically significant between LCM and HCM at every duration.
Although the increase in contractile activation described above is
small, it was reliably seen. Switching from HCM to LCM, on the other hand,
did not produce the opposite effect of impairing contraction. For individual
cells, response curves from before and after the switch closely agree (Fig. 4).
Similarly, analysis of the population data shows no significant effect of
switching from HCM to LCM on contractile strength at any voltage (Table 1)
Analysis of threshold voltages also indicate that the switch from HCM to
LCM produced no measurable effect (Fig. 5, Table 1).
Responses to tetrodotoxin
Adding tetrodotoxin (TTX) until it reached a concentration of 200 nm
had no effect on the pattern or strength of contraction or threshold voltage in
either LCM or HCM (Fig. 6, Fig. 7, Table 1). Additionally, TTX had no
apparent effect on cell survival, nor did it modify the effect of calcium on cell
survival. (Fig. 8)
Calcium and cell survival
Chronic exposure to a high calcium environment led to premature cell
death. Immediately after fibers were dissociated, typically half of the cells
would respond to electric shock, regardless of the medium. Äfter 20 hours in
HCM, few if any cells would contract. Intact cells were sparse; most of the
organic matter consisted of small, shriveled blobs. After 20 hours in LCM, by
contrast, response rate remained steady (Fig. 8), and a significant number of
cells appeared visually to be viable. Whole cells did become more scarce over
time, but a large percentage of those present still contracted. Cells have been
kept viable in LCM for up to 4 days. Attempts to culture cells at 15-17°C also
resulted in premature cell death even in LCM.
Discussion
Excitation
Several lines of evidence suggest that mantle muscle cells tested under
the conditions used in this study are unlikely to depolarize via a sodium
action potential. First, TTX, a potent and highly selective antagonist of the
voltage-gated sodium channels responsible for the all-or-none response in
neurons and vertebrate skeletal muscle (Costantin, 1975), had no discernible
effect on contraction. Second, twitches were usually asymmetrical, occuring
more strongly on the side closer to the electrode. This behavior is
inconsistent with the occurence of a propagated action potential, which in
such a small cell would simultaneously depolarize the entire membrane.
Third, the data in figure 1 indicate that contractile strength is smoothly graded
over a large range of stimulus strengths. The graded nature of the
contractions was most clearly seen during the low-voltage trials to determine
threshold, during which cells would barely twitch. Such gradation would not
be expected in response to an all-or-none action potential.
Unlike Nat action potentials, calcium action potentials are often
graded, such as those in the giant muscle fiber of the barnacle (Hagiwara et al.,
1968). However, the drop from 10.8 mM calcium to 1.3 mM barely affected
contraction strength or threshold voltage. Though this evidence is not
conclusive, it does indicate that contraction is not highly dependent on
external calcium concentration, as might be expected of a cell firing a calcium
action potential. The possibility exists that calcium influx through channels
is virtually saturated even at 1.3 mM Ca2+, and that calcium levels would
have to be even lower to produce an inhibiting effect. Several trials were
attempted in which the free calcium concentration was brought to very low
levels (sub-micromolar) with EGTA, a chelating agent. However, these
experiments failed for technical reasons, and offer an obvious avenue for
future study.
Rather than the firing of action potentials, a more likely mechanism of
membrane excitation is graded local depolarization at the site of stimulation.
Such a sheme accounts for the asymmetry of the twitches. When a shock is
delivered from the point source electrode, the voltage drops rapidly across the
resistance of the medium. Therefore, the side of the muscle fiber closer to the
stimulating electrode receives the stronger shock. A model of graded, local
excitation predicts that this side would contract more strongly than the other,
causing the cell to curl toward the electrode, as was observed. The gradation
of responses seen in figures 2 and 4 also support this model.
The cells in this study were tested under nonphysiological conditions,
making speculation about their behavior in vivo difficult. As previously
mentioned, voltage clamp experiments have detected the presence of large
Nat currents in mantle muscle fibers. A complete model of cell behavior
must account for why the dissociated cells failed to show Nat spikes. It is
possible that the dissociation process interfered with the firing of action
potentials, and that the large shocks used to stimulate the cells were bypassing
the normal excitation phase of muscle contraction and instead directly
activating some internal voltage-triggered calcium release. Alternatively,
squid mantle may consist of muscle types that have various excitation
mechanisms. This experimental technique might favor the survival of those
that are excited in vivo by only the local membrane depolarizations due to
neuromuscular transmission. A small number of cells did have atypical
responses to electrical shock: contracting into blobs, twitching wildly, etc.
Often these cells were quite long (»150 microns), and the entire cell would
respond to stimulation of just one end, possibly indicating active or
propogated spread of depolarization. Because this type of cell was rare and
this response could not be quantified, these cells weren’t included in the data
analysis, but their behavior might reflect a different excitation mechanism,
involving Nat channels.
Contraction
All muscle cells must couple excitation with mechanical activity.
Calcium acts as a direct link between these stages, rising in response to
depolarization, and directly binding to the contractile machinery to allow
contraction (Hoyle, 1983). A priori, the surge in cytoplasmic calcium
concentration can be attributed to one of two sources: influx from an external
pool or release from intracellular stores. Influx works through a
straightforward mechanism. Depolarization opens voltage-sensitive
membrane channels which allow calcium to flow down an electrochemical
gradient into the cell. Release of calcium from internal stores requires a
specialized structure, the sarcoplasmic reticulum (SR). This organelle actively
sequesters calcium, helping to keep internal resting concentrations low.
Depolarization triggers release of calcium from the SR back into the
cytoplasm (Ebashi, et al., 1980). When the stimulus is removed, the SR
actively takes up calcium, thereby causing relaxation.
The relative independence of contraction strength from external
calcium concentration suggests that squid mantle muscle has internal
calcium stores. This model is lent plausibility by the fact that SR is evident in
electron micrographs of the mantle muscle of Symplectoteuthis oualaniensis
another species of squid (Moon and Hulbert, 1974). Furthermore, this model
suggests a possible explanation of the discrepancy between cells switched from
LCM to HCM and those switched from HCM to LCM. Perhaps the fibers are
unable to maintain a high calcium gradient between the SR and the medium,
so the calcium reserves of cells in LCM are lowered, resulting in weaker
contractions. When placed in HCM, these cells actively take up calcium, and
contraction strength increases. Cells moved from HCM to LCM, on the other
hand, only lose their reserves slowly, so no effect is immediately seen. The
dependence of SR stores on external calcium concentration has also been
proposed to explain loss of contractile strength in arthropod muscle (Caputo,
1978) and in mammalian cardiac muscle (Costantin, 1975). A full analysis of
the problem would require knowledge of the rates of both calcium uptake
into the SR and eflux into the medium. However, simply tracking the
contraction strength over time of cells switched from HCM to LCM might
shed some light on the problem. And, of course, stimulating the cells in a
near-zero calcium medium buffered with EGTA would also yield useful data.
Calcium and cell death
Because a higher percentage of cells in LCM responded after 20 hrs than
after 5, figure 5 seems to indicate that LCM actually increases survivorship
over time. This effect is almost certainly an artifact caused by experimental
method. Following cell dissociation, plates were completely coated with
organic material, most of it consisting of small cellular breakdown products.
For all experiments, only cells that looked intact received shocks.
Immediately after dissociation, roughly half of the healthy-looking cells failed
to contract. After 20 hours, many of these unresponsive cells probably broke
down, leaving a higher percentage of intact cells that would actually respond.
In accordance with this theory, intact cells were observed to become more
10
sparse in all dishes over time, although no quantitative density studies were
performed.
A high external calcium concentration had a strong effect on the
percentage of cells that would respond to stimulation. Qualitative
observations revealed that HCM did quickly lead to cellular breakdown,
although the mechanism of cell death is unknown.
Final considerations
Several studies have demonstrated the existence of two distinct types of
mantle muscle cell. One type, which contains high levels of mitochondria
and oxidative enzymes, is roughly analogous to vertebrate slow-twitch fibers,
while the other, which contains fewer mitochondria and a greater amount of
glycolytic enzymes, is roughly analogous to vertebrate fast-twitch fibers
(Mommsen et al., 1981). Because these cell types were indistinguishable
under the microscope, a complete analysis of the data in this paper is
impossible. If cell dissociation is to be a reliable method for investigating
electrophysiology, there must be a way to unambiguously determine cell type.
Histochemistry may provide a method; for example, researchers have
differentiated the cells by staining for succinic dehydrogenase, an oxidative
enzyme that indicates mitochondrial abundance. However, this procedure
kills the cells, so it could only be used after cells were tested.
A more plausible method of differentiation may rely on the distinct
metabolic properties of the cells. The refractory period following shock could
reflect the depletion of the cell’s energy store, which is replenished only
slowly. Because fast- and slow-twitch cells have drastically different metabolic
profiles, perhaps they also exhibit different recovery responses. A search for
signature refractory patterns might prove productive. Although no such
analysis was performed on the data in this paper, refractory times were
observed to vary widely, ranging from several seconds to almost a minute.
Finally, morphology might also provide a means of distinguishing cell
types. In certain cells, a band of midline structures appeared in response to
stimulation, and the cells failed to respond to any subsequent shocks.
Mitochondria form a central core in some cephalopod muscle (Hochachka et
al., 1978), and this pattern has been observed in squid slow-twitch fibers
(Mommsen et al., 1981). If histochemical analysis can correlate the
appearance of this central structure to cell type, morphology alone will
provide a basis for discrimination.
Another way to improve the experimental technique is to simulate
physiological conditions more closely. A simple step in this direction would
be to determine the neurotransmitter used to produce contraction in vivo
and to use this as a stimulus in place of electric shock. Although not as easily
controllable or as minutely adjustable as electrical shock, stimulation with
neurotransmitter would yield interesting supplemental data, as well as
provide evidence that dissociated cells have retained some normal function.
Bibliography
Caputo, C.; Dipolo, R.; Contractile Activation Phenomena in Voltage-clamped
Barnacle Muscle Fiber. Jour. Gen. Phys., Vol. 71: 467-488, 1978
Costantin, L. L.; Activation in striated muscle. Chapter 7, Handbook of
Physiology — The Nervous System I, 1975
Ebashi, S.; Maruyama, K.; Endo, M.; Muscle Contraction, Its Regulatory
Mechanisms. Springer-Verlag, New York, 1980
Hagiwara, S.; Takahashi, K.; Junge, D.; Excitation-contraction coupling in a
barnacle muscle fiber as examined with voltage clamp technique. Jour. Gen.
Phys., Vol. 51: 157-175, 1968
Hochachka, P. W.; French, C. J.; Meredith, J.; Metabolic and Ultrastructural
Organization in Nautilus Muscles. Jour. Exp. Zool., Vol. 205: 51-62, 1978
Hoyle, G.; Muscles and Their Neural Control. John Wiley & Sons, New York,
1983
Junge, D.; Nerve and Muscle Excitation, Third Edition. Sinauer Associates,
Sunderland, Mass., 1992
Mommsen, T. P.; Ballantyne, J.; MacDonald, D.; Gosline, J.; Hochachka, P. W.;
Analogues of red and white muscle in squid mantle. Proc. Natl. Acad. Sci.
USA, Vol. 78: 3274-3278, 1981
Moon, T. W.; Hulbert, W. C.; The ultrastructure of the mantle musculature
of the squid Symplectoteuthis oualaniensis. Comp. Bioch. Physiol., Vol. 52B:
145-149, 1975
Schwartz, L. M.; Stuhmer, W.; Voltage dependent sodium channels in an
invertebrate striated muscle. Science, Vol. 225: 532-537, 1984
13
Figure legends
Table 1 MEAN CONTRACTILE STRENGTHS AND THRESHOLD VOLTAGES. Each
value is the mean response of a population of cells. The lefthand three
columns represent the degree of shortening in response to a 0.4 ms shock of
20 V, 40 V, or 60 V. (A 25% contraction means the contracted cell is 3/4 the
length of the relaxed cell.) The righthand three columns indicate the
minimum voltage necessary to induce contraction in response to shocks of
4 ms, 5 ms, or 10 ms. The rows are paired together, each pair containing
data taken from the same population of cells. Between members of a pair, a
change was made in the external medium, as indicated by the arrows in the
row labels. For the trials with tetrodotoxin (TTX), the toxin was added to the
medium already in the dish until the concentration reached 200 nM.
Underlined values indicate a statistically significant difference between data
points in paired rows (t-test, p«.05). Only the switch from LCM to HCM
caused a significant change in response pattern. The number of cells in a
population are indicated by n.
Figure 1 VIDEO PRINT OF AN ASYMMETRICAL CONTRACTION The two pictures
are consecutive video frames, representing a timespan of about 30 ms. The
30.4 micron cell is bathed in LCM, and receiving a +60 V, 0.4 ms shock. The
arrow indicates the muscle fiber, and the asterisk indicates the approximate
location of the electrode tip, which cannot be seen because the electrode is
viewed from above.
Figure 2 CONTRACTILE STRENGTH IN CELLS SWITCHED FROM LCM TO HCM
Each graph represents data taken from a single cell. While in LCM, a cell
received a series of increasingly strong 0.4 ms shocks spaced 45 s apart. The
change to HCM was made carefully so that the cell remained in the
microscope's field of view. The cell then received an identical series of
shocks. At every voltage, the cells contracted more strongly in HCM than in
LCM (exept when the percent contraction was 0, indicating that the voltage
was subthreshold in both media).
Figure 3 THRESHOLD VOLTAGE IN CELLS SWITCHED FROM LCM TO HCM The
two cells in figure 1 were also tested for threshold voltage at a series of shock
durations. The thresholds were always lower in HCM than in LCM.
Figure 4 CONTRACTILE STRENGTH IN CELLS SWITCHED FROM HCM TO LCM
Each graph represents data from a single cell tested first in HCM and then in
LCM. The values at each voltage closely agree.
Figure 5 THRESHOLD VOLTAGE IN CELLS SWITCHED FROM HCM TO LCM The
two cells in figure 3 were also tested for threshold. The switch from HCM to
LCM had no consistent effect on threshold level.
Figure 6 TTX'S EFFECT ON CONTRACTILE STRENGTH Each graph represents data
from a single cell. The plot on the left is data taken from a cell bathed in
LCM to which TTX was added in a concentration of 200 nm. The plot on
right is from an identical experiment, except that HCM was used. The
values at each voltage closely agree.
Figure 7 TTX'S EFFECT ON THRESHOLD VOLTAGE Each graph represents data
from a single cell. The plot on the left is data taken from a cell bathed in
77
X was added in a concentration of 200 nm. The plot on
LCM to which 11.
right is from an identical experiment, except that HCM was used. The
values at each duration closely agree.
Figure 8 CALCIUM'S EFFECT ON CELL VIABILITY Cells which reacted in any way
to an 80 V shock were considered responsive, and therefore alive. Even
immediately after dissociation, only about half of the cells contracted.
15

O





1
1

4





—
o0

8

S
8
8
o%

8
o%
0
8
5
.
o
—
o


4
SIOA

STIOA

95
55
s
5
uo %
8
8
uo %
D
S
62

e

.
SIOA
S1OA
vv-
en
uo %
o %
E
S10A
8
SIOA
60
Calcium's effect on cell viability
HCM
LCM +TIX HCM + TIX
Medium
CM
Figure 8
Adam Stein
hours
20 hours