Proceedings of the 20th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, Vol. 20, No 6,1998
THEORETICAL ANALYSIS OF MODE OF HEPTANOL ACTION ON
SMOOTH MUSCLE SYNAPTIC POTENTIALS
Rohit Manchanda*, K. Venkateswarlu, S. Sourav, and K. Moudgalya**
School of Biomedical Engineering, **Department of Chemical Engineering,
Indian Institute of Technology-Bombay, Powai, Mumbai - 400 076, INDIA.
*E-mail: rmanch@cc.iitb.ernet.in
ABSTRACT
A theoretical analysis has been undertaken of the synaptic
or "junction" potentials of syncytial smooth muscle in order
to examine the most likely mode of action of a chemical, 1heptanol, that has profound effects on neurot"ission in
this tissue. The smooth muscle syncytium of the mammalian
vas deferens has been modelled as a bidomain electrical grid
which allows simulation of the spontaneous and evoked
excitatory junction potentiak (SEWSand eEJPs) observed
experimentally in this organ. Two hypotheses of heptanol
action raised by its effects on the electrical responses of the
vas have been examined, (i) inhibition of cell-to-cell electrical
coupling in smooth muscle; (ii) inhibition of the stimulationevoked release of neurot"itter from the autonomic nerve
supply. We find that the range of experimental findings can
be explained consistently on the basis of the second, but not
the first, hypothesis. Our results thus cast doubt on the current
belief that heptanol, at the concentrations used empirically,
specifically uncouples smooth muscle cells from one another,
and indicatea novel, previously unsuspected biophysical made
of action which merits close scrutiny.
KEYWORDS: Electrical syncytium, quantal depolarizations,
smooth muscle, electricaluncoupling, physiological modelling.
1. INTRODUCTION
In syncytial muscles of the body (e.g. smooth and cardiac
muscle) electrical continuity between cells is believed to be
served by "gap junctions", which are low-impedance proteinic
channels spanning the apposed membranes of neighbouring
cells [l]. Important clues to the role of syncytial coupling in
normal organ function can therefore a c m e from studying the
effects of disruption of gap junctional function. Recently, the
effects of a chemical, I-heptanol, which is suggested to block
gap junctions specifically [11, were explored experimentally
on synaptic potentials of the smooth muscle of the guinea-pig
vas deferens [2], i.e. the spontaneous and evoked excitatory
junction potentials (&IPS and e m s respectively). The salient
observations were as follows. (i) In most cells ("passive" cells
which do not receive innervation by close-contact varicosities
or CCVs), eEJPs were reversibly abolished by 2.0 mM
heptanol without change in their time course [2]. (ii) In
"active" cells which receive CCVs, a novel kind of stimulus-
0-7803-5164-9/98/$10.000 1998 IEEE
locked, evoked elecbical signal was induced by heptanol.
These signals, which replaced the eEJPs, were strikingly
similar to sEJPs,and were termed quantaI EJPs (qEJPs) [3].
(ii) Heptanol left unaffected the randomly occurring
spontaneous Elps (sEJPs) in all their properties, including
time courses and amplitude distribution.
From a qualitative viewpoint, this set of observations
could not be consistently explained on the basis of cell-to-cell
uncoupling by heptanol [3]. Therefore, it was considered
whether alternative modes of action of heptanol might offer
a more consistent framework to explain the results. In
particular, it was conjectured that heptanol may interfere
specifically with stimulation-evoked release of
neurotransmitter.
In this paper, we present the results of a theoretical
exploration of syncytial electrophysiology, which helps
discriminate between the competing hypotheses on hand
Smooth muscle junction potentials have been simulated
recently in a discrete model of the syncytium [41. In the
present study we undertook to use this model so as to
incorporate the conditions appropriate to each of the
hypotheses under examination. We report here the predictions
made by the model under conditions of (i) gap junctional
inhibition between individual cells throughout the syncytium
(Hypothesis 1); and (ii) inhibition of stimulation-evoked
neurotransmitter release (Hypothesis 2).
On setting the predictions arising from this study against
empirical observations, the most plausible hypothesis to
emerge is that heptanol inhibits stimulation-evoked neurotransmitter release from the innervation of the vas deferens.
Such a mode of action has not hitherto been proposed for
heptanol, and we discuss its biophysical implications.
II. THEORY AND METHODS
Following Bennett & Gibson [4], the smooth muscle
syncytium is modelled as a discrete 3-D grid, each node of
which represents the internal medium of a muscle cell. Cellto-cell connections are modelled as resistive linksbetween the
nodes. A second grid represents the adjacent interstitial
region. The two grids are connected at every pair of
corresponding nodes by an RC circuit representing the
membrane of a muscle cell, For the activation of cells by
neurotsansmitter released from varicosities, membrane
3100
Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY BOMBAY. Downloaded on February 11, 2009 at 23:39 from IEEE Xplore. Restrictions apply.
representation incorporates current injection, modelled in
terms of a driving potential E, and a series conductance
change g(T)=g,aTexp(l-aT). Here go is the maximum
membrane conductance and T is time normalized with respect
to membrane time constant (T=~/T,); 7, is given by the
product of membrane resistance and capacitance C,; a is
a “forcing function”dictating conductance time course, given
by a=r&,, where $,=time to peak of g(T) [SI.
The change in membrane potential at any node ijk (V,&
following injection of current into the syncytium is given [4]
by the solution of :
s + v i k =Ax(vi+ljk
2
+
[a,
[7]) as well as physiological probability of activation of
CCVs (-0.001 to 0.05, see [8]) in the guinea-pig vas deferens.
SIMULATION OF sEJPs: The sEJP was generated by
making one CCV at the centre of the syncytium [designated
node 01 active. sEJP amplitude histograms were obtained by
collecting peak sEJP amplitudesthat developed at all nodes in
a 7 x 7 ~ 7lattice centered on the active node. This size of
sampling lattice takes into account all significant sEJP
amplitudessince nodes 23 nodes from node 0 in any direction
show negligible sEJP amplitudes.
Cells that received direct CCV input (node a),and those
s 2 nodes away, showed eEJPs characteristicof “active“cells
whereas cells 24 nodes from node 0 showed “passive“eEJPs.
-
dT
III. RESULTS
where A = J[%J(R,+Rk)],
being the resistance between
adjacent intracellularnodes while R, the interstitial resistance
between the adjacent extracellularnodes. Here 6 =1if ijkqqr,
and denotes an active node, else 6-0 for passive nodes.
The set of differentialequationsresulting from application
of (1) to all nodes in the syncytium was solved numerically
to obtain solution for V, with respect to time. The boundary
conditions used were those appropriate for an insulated
surface of the syncytium [4]. Simulations were performed
using Fortran 77 code on a DEC ALPHA server running
under OSF/l (UNIX). A doubleprecision ordinary differential
equation solver implementing the implicit Adams method
(ODEPACK, Alan C. Hindmarch, 1987,Lawrence Livermore
National Laboratory, California, USA) was used.
GENERATION OF JUNCTION POTENTIALS: To
generatejunction potentials we have used a 29x29~29system
as the S i t e syncytium, since this size was found in pilot
studies to satisfy the requirement of retention of syncytial
behaviour. To generate eEJPs, two kinds of inputs,
corresponding to known morphological and fimctional
properties of smoothmuscle innervation by autonomic nerves,
were provided: (i) close-contact varicosity (CCV) input, using
g,=30 nS, a=14.3, (ii) loosscontact varicosity (LCV)input,
using g =O. 1nS, a=4.5. These parame er values and hose for
other parameters in equation (1) are taken from 141, and their
use provides eEJPs and sEJPsthat closely resemble those
experimentally recorded (see Results). We have used a CCV
density of 1 in 343 cells along with one LCV for every cell
%at does not receive CCV input. The CCV density chosen
-dls in the range expected from known properties of
wrphological occurrence of CCVs (on -1@25% of cells, see
Hypothesis 1: Cell-to-cell uncoupling in smooth muscle by
heptanol.
The conditionof cell-to-cell uncoupling in smooth muscle
was simulated by increasing intercellular resistance Ri
uniformly in the syncytium by factors of 10,100 and 1OOO.
This resulted in the following general effects.
e U P s in passive cells declined partially in peak
amplitude (by about 20%), accompanied by a noticeable
increase in rise time (tR) (Fig. 1A). In active cells, eEJPs
actually increased in amplitude, with marked changes in
and time constant of decay (“decay> (Fig. 1B). No qEJP-like
events were obtained in active cells.
S U P peak amplitudes progressively increased as Ri
increased, accompanied by a flattening of the peak and
increase of T~~ (Fig. 1C). In Fig. 1D we show amplitude
histograms of sEJPs derived from the normal syncytium (Fig.
1Da) and at 1000Ri (Fig. 1Db). Note, at 1000Ri (i) the
increased clustering of observations in the lowest-amplitude
bins; (ii) the emergence of very large SEWS(peakamplitude
30-50 mV, which would OCCUT at node 0). In contrast, the
amplitude histogram of experimentally recorded sEJPs in the
presence of heptanol does not change (not shown).
Hypothesis 2: Electrical@ unaltered smooth muscle
syncytium, evoked neurotransmitter release inhibited
This effect was simulated by reducing the spatial density
of active CCVs and LCVs. The density of activation was
reduced (without changing go) to 1 in 2197 cells receiving
CCV input and 1 in 8 receiving LCV input. In passive cells,
the eEJP was strongly suppressed, by “7040% without
change of time c o m e (Fig. 2B). Simultaneously, in active
cells, rapid evoked depolarizations,resembling sEJPs in time
course, occurred at a short latency, thus generating qEJP-like
signals (Fig. 2B). This pattern of changes is strikingly similar
to that which emergesfollowing heptanol action in active cells
which exhibit qEJPs in experimental recordings [3].
qEJPs ride on a background depolarization. By
subtracting the background depolarization (generated in a
31 01
Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY BOMBAY. Downloaded on February 11, 2009 at 23:39 from IEEE Xplore. Restrictions apply.
A: eEJPs
a
4
b
1
100 m
B: sEJPs
a
335
IOOOR,
330
4
10
5
-0
2
-/+
50 55
0
0
2
h---Ja-
+
4 45
mV
50
55
mV
Fig. 1. Effects of increasing R, (Hypothesis 1) on eEJPs (A) and sEJPs (B). Nomd cell-tocell coupling resistance is indicated by 1 (=RJ.
Extent of increase of R, over normal levels indicated as follows:2=10R, 3=100Ri,4=1000R1.Au: eEJPs in a passive cell (6 nodes away
fiom node 0), showing moderate inhibition and increase of rise times at larger 9's. Ab:eEJPs in an active cell (node 0), showing increase
of amplitude at larger R,'s. Ba:Change of sEJP configuration and amplitude at node 0 due to increase of R. Bb,c: Showing change of
sEJP amplitude distribution due to increase of R,. No observations over interval of break in x-axis.
passive cell) from the net depolarization recorded in an active
cell, the "true" time course of the qEJP was obtained (Fig.
2C). qEJPs thus obtained agreed very closely with the time
course and shape of sHPs recorded at the same nodes (Fig.
2D). This indicates strongly that the quantal depolarizations
underlying the broader eEJP can be revealed by inhibiting
evoked transmitter release. sEJPs did not change in any of
their properties under this hypothesis, as expected, since
muscle electrical properties are left intact.
V. DISCUSSION
Three salient features of the electrophysiology of the vas
deferens in the presence of heptanol require to be explained
consistently for any proposed mechanism of action of
heptanol to be accepted, namely: i) decline of eEJPs in
passive cells without cbange in time course; ii) emergence of
qEJPs after the suppression of background depolarization in
active cells; iii) persistence of sEJPswithout alteration.
The commonly accepted hypothesis of heptanol action is
that it uncouples cells from one another [11. Our fiidings on
simulated junction potential behaviour under conditions of
increased R, (mimicking various degrees of probabilistic
"block" of gap junction channels) suggest strongly that this
mechanism of action cannot account for the experimental
observations with 2.0 mM heptanol. Superficially the decline
of eEJP amplitudes at passive nodes on increasing Ri
resembles the suppression of eEJPs by heptanol. However at
a level of increase of R, that produces only '20% suppression
of e m s at passive nodes, the following effects are also
observed: (i) eEJPs change in time course, especially in their
rising phases (Fig. 1A); (ii) sEJPs are profoundly altered, in
configuration as well as amplitude distribution (Fig. 1B). A
combination of effects such as this was not observed
experimentally. Furthermore the appearance of qEJPs could
not be simulated by increasing R,. On account of these
discrepancies, it is difficult to accept that heptanol uncouples
individual smooth muscle cells from one another to produce
its experimentally observed effects.
In mtrast, the al&mative hypothesis, Le. that of
inbibition of evoked neurotransmitter release by heptanol,
succeeded in predicting experimental observations adequately
including (i) suppr&ion of eEJPs in passive cells without.
3102
Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY BOMBAY. Downloaded on February 11, 2009 at 23:39 from IEEE Xplore. Restrictions apply.
B
A .
1
6
D
C
Fig. 2. Effect of inhibiting stimulation evoked transmitter release (Hypothesis2) on eEJPs. Numbers indicate node number. A: eEJPs in
active cells (0,l) and a passive cell (6) at normal levels of release. B: eEJPs at the same nodes as A after inhibition of release (see text)
C : Signal obtained (qEJP) after subtraction of eEJP at node 6 from eEJP at node 0 in B. D: qEJP of C superimposedon an sEJP on an
expanded time scale (see bar), showing near-identicalityof the signals.
change of time course, and (ii) emergence of distinct sEJPlike qEJPs in active cells (Fig. 2). If heptanol is found to have
such an effect experimentally, this would have a number of
significant implications. First, the effects of heptanol observed
in some previous studies might require reevaluation, since
their explanation has rested on the inhibition of cell-to-cell
coupling [l]. Secund, it will be interesting to identify the
precise target of heptanol action. In the process of evoked
transmitter release, or "stimulus-secretion coupliig" in the
autonomic varicosity, several steps are intepsed, including
(i) Ca++influx linked to action potential invasion; (ii)
interaction of Ca" with secretory vesicles; (iii) migration of
vesicles to the varicosity surface membrane and their docking;
(iv) fusion and poration of vesicle and surface membranes,
leading to exocytosis of transmitter [9]. Very few agents are
known that interfere specifically with any of these processes,
and the action of heptanol would offer valuable insight into
the mechanisms of evoked transmitter release. However, a
necessary proviso is that the locus of heptanol action should
be consistent with its lipophilic nature, and its proposed partitioning into the membrane lipid compartments of cells [l].
ACKNOWLEDGMENTS
REFERENCES
[11 G.J.Christ, ''Modulation of a , adrenergiccontractility in isolated
vascular tissues by heptanol," Life Sci., vol. 56, No. 10, pp. 709721,1995.
121 R. Mancbanda, and K. Venkateswarlu, "Effectsof heptanol on
electrical activity in the guinea-pig vas deferens," Br. J.
PharmacoL, vol. 120, pp. 367-370, 1997.
[3] R. Manchanda, and K. Venkateswarlu, "Intracellular detection of
evoked quantal depolarizations in smooth muscle," Pmc. IntL
Con$ Biomed. Engg., Coimbatore, India, pp. III.12-15, 19%.
[4] M.R. Bennett, and W.G. Gibson, "Onthe contribution of quantal
secretion from close-contact and loosecontact varicosities to the
synaptic potentials in the vas deferens," Phil. Trans. Roy. SOC.
(Lon)B., vol. 347, pp. 187-204, 1995.
[5] R.D. Purves, "Current flow and potential in a three-dimensional
syncytium,"J. Theor. BioL, vol. 60, pp. 147-162, 1976.
[6]M.R. Bennett, Autonomic Neuromuscular Transmission, London:
Cambridge University Press, 1972.
[7] N.C.R. Merriilees, "The nervous environment of individual
smooth muscle cells of the guinea pig vas deferens," J. Cell
L k L , vol. 37, pp. 794-817, 1968.
[8] N.A. Lavidis, and M.R. Bennett, "Probabilistic secretion of
quanta from visualized sympathetic nerve varicosities in mouse
vas deferens," J. PhysioL, vol. 454, 9-26, 1992.
3103
Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY BOMBAY. Downloaded on February 11, 2009 at 23:39 from IEEE Xplore. Restrictions apply.