• The effects of the putative gap junction uncoupler, 1-heptanol, on the neurogenic and myogenic contractile responses of guinea-pig vas deferens were studied in vitro.
• Superfusion of 2.0mM heptanol for 20–30min produced the following reversible changes in the biphasic neurogenic contractile response (8 trials): (i) suppression of both phases; (ii) delayed development of both the first as well as the second phase, accompanied by complete temporal separation of the two phases; (iii) prominent oscillations of force during the second (noradrenergic) phase only.
• To eliminate prejunctional effects of heptanol, myogenic contractions were evoked by field stimulation of the vas in the presence of suramin (200μM) and prazosin (1μM). Heptanol (2.0mM) abolished these contractions reversibly.
• These results show that (i) heptanol inhibits both excitatory junction potential (EJP)-dependent and non EJP-dependent contractions of the vas; (ii) a postjunctional site of action of heptanol, probably intercellular uncoupling of smooth muscle cells, contributes to the inhibition of contraction.

Smooth muscle cells form ‘functional syncytia' (Tomita, 1975), in which biochemical or electrical changes originating in one cell are communicated to neighbouring cells through pathways of intercellular communication (Tomita, 1967; Christ et al., 1996). These pathways, or ‘gap junctions', are believed to be located at regions of close apposition of the surface membranes of neighbouring cells, in the form of intercellular ion channels connecting their interiors (Unwin & Zampighi, 1980; Loewenstein, 1981).

The contractile response of mammalian smooth muscle may depend upon the spread of excitation through gap junctions. In neurogenic contractions such as those evoked in the guinea-pig vas deferens by stimulation of the innervating hypogastric nerve, only a small proportion of cells are directly influenced by transmitter released from the sympathetic motor innervation, because of two reasons. (1) Only about a fifth of the cells receive direct innervation by close-contact axonal varicosities (Merrillees, 1968; Bennett, 1973); (2) Varicosities do not release transmitter in response to every invasion by the axonal action potential because of the low probability (≈amp;0.01) of evoked transmitter release (Cunnane & Stjarne, 1984; Brock & Cunnane, 1988). Therefore, spread and co-ordination of excitation from the few directly activated cells to other cells probably requires the involvement of gap junctions.

Contraction of smooth muscle cells is triggered finally by an increase in concentration of cytosolic Ca²⁺ ([Ca²⁺]_i, Himpens & Somlyo, 1988). Activation leading up to elevation of [Ca²⁺]_i and the ensuing contraction in the guinea-pig vas deferens is of two kinds. The first is primarily electrical activation, mediated by adenosine 5′-triphosphate (ATP) which is released as a co-transmitter from the sympathetic innervation along with the ‘classical' sympathetic neurotransmitter noradrenaline (NA) (Burnstock, 1995; von Kugelgen & Starke, 1994). NA itself acts primarily via biochemical activation. Corresponding to its activation by the two transmitters, the contraction of the guinea-pig vas in response to tetanic (8–20Hz) motor nerve stimulation is biphasic, consisting of a rapid twitch, mediated mainly by ATP and a slower tonic phase, mediated mainly by NA (Sneddon & Westfall, 1984; Kennedy et al., 1996). The purinergic phase is triggered via P_2x purinoceptor-mediated excitatory junction potentials (EJPs) leading to regenerative muscle action potentials associated with Ca²⁺ influx (Blakeley et al., 1981; Sneddon & Westfall, 1984; Sneddon, 1992; Driessen et al., 1994), whereas the noradrenergic phase is mediated by elevation of [Ca²⁺]_i through the inositol phosphate (IP₃) second messenger pathway following α₁ adrenoceptor activation (Berridge, 1993).

The proposal that gap junctions modulate contractile activity has been explored, and strengthened, in vascular smooth muscle by the use of substances that appear selectively to block gap junctional communication, in particular the alkanol 1-heptanol (Christ, 1995; Christ et al., 1996). These studies have concentrated on the effects of heptanol on contractions evoked by externally applied agonists. Neurogenic contractions, however, which may involve considerably different spatial patterns of directly excited and indirectly recruited cells, have not been similarly investigated.

The aim of the present investigation was to assess the contribution of cell-to-cell communication to the neurogenic contractile response of the guinea-pig vas deferens. It was found that both phases of the neurogenic contraction were reversibly suppressed by heptanol. To clarify the site of heptanol action, we also explored its effects on myogenic contractions, and present evidence supporting a postjunctional mechanism of action of heptanol.

The force transducer was connected to an amplifier (Polyrite, Biodevices, Chandigarh, India) whose output was directly displayed on a digital storage oscilloscope (1425, L&T Gould, Mumbai, India) and simultaneously stored on a digital tape recorder (DTR-1204, Biologic, Claix, France). The tissue was mounted under a resting tension of 0.1–0.5g. The hypogastric nerve was stimulated via a pair of Ag/AgCl ring electrodes (separation about 3mm), through which its proximal end was passed, using rectangular voltage pulses delivered by an electronic stimulator. A second pair of Ag/AgCl ring electrodes was placed surrounding the muscle, one electrode at the epididymal end and the other at the prostatic end, for direct stimulation of the muscle. Supramaximal tetanic nerve stimulation was carried out at 8Hz, using a pulse amplitude of 10V and pulse width 5ms. A gap of 10min was provided between consecutive tetani (~20s duration each) to allow for recovery of the tissue. The muscle was stimulated selectively using pulses of amplitude 3V and width 200ms at 1Hz, these parameters of stimulation being known to result in direct muscle stimulation (Holman et al., 1977).

All data are expressed as mean±s.e.mean number of observations). Student's unpaired t-test was used to assess the statistical significance of the difference between means, a P value of <0.05 being taken to indicate a statistically significant difference.

Figure 1 Effect of 2.0mM heptanol on neurogenic contractions of the guinea-pig vas deferens in two tissues. (A) Tissue showing complete suppression of two phases and (B) showing only partial suppression of purinergic phase a: Control; b: after 10min (more ...)

Figure 4 Generation of myogenic contractions of the guinea-pig vas deferens. (A) Suppression of neurogenic force, generated by hypogastric nerve stimulation at 8Hz in the presence of suramin and prazosin (S+P). a: Control; b,c: 40 and 70min (more ...)

In general, superfusion of 2.0mM heptanol solution for about 20min gradually suppressed both phases of the neurogenic contraction (Figure 1). In the majority of trials (seven out of nine), both phases were inhibited similarly. The peak force of the first phase was suppressed without change of shape by 83±7% (n=7) on 20min exposure to heptanol. Inhibition of the second phase was more complex, as it developed irregular fluctuations during its suppression, as described in greater detail below.

Figure 1A shows results of a trial in which the inhibition was complete for both phases at 20min (Figure 1A, c). The effect was completely reversed on washout with normal Krebs for about 20min (Figure 1A, e). In the other two trials the purinergic phase was partly resistant to the action of heptanol, its force being suppressed by only 50% following 20min of heptanol exposure, whereas the adrenergic phase was completely abolished during this interval (see Figure 1B, c).

Figure 2 (A,B) Comparison of control neurogenic response with that in the presence of heptanol. Note the increase in latency, separation of the two phases and oscillations in the second phase induced by heptanol. Stimulation carried out at 8Hz (10V, (more ...)

The onset of the noradrenergic phase was more strikingly delayed by heptanol (Figure 2A). The latency of the second phase in control contractions is difficult to determine with certainty because its onset is obscured by the preceding purinergic phase. However, the latency of the second phase can reasonably be assumed to be similar to that of the first since both ATP and NA, the two neurotransmitters mediating the phases, are known to be released simultaneously from the innervation following stimulation, though the noradrenergic phase is slower to develop (von Kugelgen & Starke, 1994; von Kugelgen et al., 1994; Stjarne & Stjarne, 1995). On the basis of this assumption the latency of the second phase was increased from 0.83±0.11s in controls to 14.93±1.23s (n=7) following 20min exposure to heptanol (Figure 2A).

Since the increase of latency of the second phase was much greater than that of the first phase, the induced suppression of contractile force was accompanied by a progressive temporal separation of the two phases. The extent of separation at 20min exposure to heptanol was such that there was no longer any overlap between the phases (Figure 2A).

Since the first phase of contraction is EJP-dependent, the increase in latency of this phase suggests that EJPs inhibited by heptanol (Manchanda & Venkateswarlu, 1997) take longer to reach threshold for action potential generation. To test this, we recorded EJPs in the presence of heptanol at 8Hz, the same frequency of stimulation as that employed in the contractile studies. In control conditions, EJPs at 5Hz readily summate to reach action potential threshold and produce contractions (Blakeley et al., 1981). In the presence of heptanol, the development of threshold depolarization was considerably delayed, and in one trial EJPs at 8Hz did not reach threshold in 8s of stimulation (Figure 2C). When the frequency of stimulation was increased gradually beyond 8Hz to about 70Hz, an action potential and contraction were finally generated (Figure 2D). Threshold for regenerative excitation is achieved at higher frequencies of stimulation since a larger degree of summation results in a larger mean depolarization. Correspondingly in contractile studies, the heptanol inhibited neurogenic contraction, too, was partially restored by increasing the stimulus frequency to 70Hz (not shown).

A possible interpretation of the oscillations observed in the presence of heptanol is that they are the result of muscle action potentials triggered intermittently by residual purinergic EJPs which may not have been inhibited completely by heptanol. To confirm the nature of the transmitter responsible for the delayed response in the presence of heptanol, the vas was superfused with suramin (200μM) before the application of heptanol. Suramin inhibited powerfully the first phase of contraction and a monophasic contraction remained, corresponding to the second, noradrenergic phase of the normal biphasic neurogenic contraction (Figure 3). These suramin-resistant contractions were affected by heptanol in the same way as the second (noradrenergic) phase of the normal neurogenic contraction, i.e. in suppression of force, induction of oscillations in force during suppression, and marked increase of latency of the same order (Figure 3). Thus these oscillations are likely to be due to effects exerted specifically on noradrenergic mechanisms by heptanol.

Figure 3 Persistence of heptanol-induced contractile oscillations in the second phase in the presence of suramin (200μM). Exposure to suramin for 40min inhibits profoundly the first (purinergic) phase of the neurogenic contraction, while (more ...)

To elicit myogenic contractions, stimulation was switched to the muscle stimulation electrodes using appropriate pulse parameters (see Methods), which elicited contractions of the type illustrated in Figure 4B, a. These contractions, since they arise despite the presence of suramin and prazosin, are therefore likely to be of purely myogenic origin. To confirm this, the same pair of electrodes was used to deliver stimulation selective for nerve stimulation (10V, 5-ms pulses at 8Hz). Figure 4B, b shows that in the presence of the antagonists, this stimulation failed to generate any detectable force, indicating that neurogenic contributions were entirely eliminated. The latency of the myogenic response (0.32±0.02s, n=6) was about 60% less than that of the neurogenic response (0.83±0.11s, n=9) (see Figure 4B, a). This is expected from the elimination of neuronal conduction, synaptic delay and time taken for the summation of EJPs to reach threshold for generation of muscle action potential.

Following the application of 2.0mM heptanol for 20–30min, peak myogenic force was almost entirely suppressed in all trials, by 98±1% (n=5) (Figure 5A,B,C). The reduction of force observed when using the long duration pulses required for myogenic contraction could possibly occur as a result of electrode polarization. However, the suppression of force persisted when stimulation was delivered after reversing electrode polarity (Figure 5D). This indicates that the suppression of the myogenic force was unrelated to electrode polarization and was indeed a heptanol mediated effect. The suppression was reversible on washout of heptanol (Figure 5E,F).

Figure 5 Reversible suppression of myogenic force by 2.0mM heptanol. (A) Control; (B,C) 20 and 30min respectively after application of heptanol. (D) absence of force generation following reversal of stimulation polarity using the same parameters (more ...)

The role of gap junctional communication in the generation and maintenance of smooth muscle contraction has received strong support in previous studies from the effects of heptanol on the contractions (Christ et al., 1993; Christ, 1995). Heptanol at concentrations between 1.0 and 5.6mM has been shown to selectively block gap junctions in various tissues including smooth muscle (Christ et al., 1991; Bastide et al., 1995; Blennerhassett & Garfield, 1991; Bukauskas et al., 1992; Peracchia, 1991; Christ, 1995), and to inhibit agonist evoked contractions in smooth muscle such as rat aorta and corpus cavernosum (Christ et al., 1991; 1993). However, neurogenic contractions have not yet been similarly investigated. In the present inquiry, heptanol has been used at a concentration of 2.0mM. This concentration falls within the range mentioned above for specific action on gap junctional communication. Furthermore, in electrical recordings, heptanol at concentrations of 2.0mM completely and reversibly abolished the EJP of the vas deferens (Manchanda & Venkateswarlu, 1997). Reversible suppression of EJPs was observed at concentrations greater than 2.0mM (up to 10mM) as well, while at concentrations below 2.0mM there was a dose-dependent suppression, but not abolition, of the EJP (Manchanda & Venkateswarlu, 1997; unpublished observations). Since 2.0mM was the lowest concentration required for complete, reversible abolition of the EJP in the absence of non-specific actions (Manchanda & Venkateswarlu, 1997), we used this concentration in these studies.

The results presented here show that 1-heptanol inhibits, profoundly and reversibly, the neurogenic contractions of the guinea-pig vas deferens. These findings may, therefore, throw additional light on the significance of intercellular coupling in the modulation of smooth muscle tone in vivo, to which neuronal input makes significant contributions. However the interpretation of the effects of heptanol on neurogenic contractions is complicated by the contribution to these contractions by prejunctional processes. This raises the possibility that the observed effects of heptanol arise from interference with neuronal mechanisms rather than with postjunctional cell-to-cell coupling.

Perhaps the most direct confirmation that heptanol affects selectively postjunctional mechanisms in the vas deferens stems from our observation that it abolishes the myogenic contractions, which do not include neuronal contribution. To place confidence in these results it is essential to be confident that the method used here for direct muscle stimulation excites only postjunctional smooth muscle, without involving prejunctional elements. This has been ensured in the present work by the choice of stimulation parameters for muscle excitation (Geddes & Baker, 1975; Holman et al., 1977), and the concomitant use of suramin and prazosin at concentrations that completely abolished the neurogenic response (Figure 4).

The suppression of myogenic contractions by heptanol indicates strongly that this agent has a postjunctional, post-receptor site of action in the guinea-pig vas deferens. Such a site of action could be located in any of the processes leading from receptor activation to contraction, including receptor-mediated second messenger generation, activation of the contractile machinery by cytosolic Ca²⁺ (Somlyo & Somlyo, 1960; Somlyo et al., 1985), and spread and co-ordination of contractile activity in smooth muscle through intercellular pathways. A large body of evidence gathered in studies on several smooth muscle organs indicates, however, that the mechanism lies probably in the specific inhibition of cell-to-cell coupling by this agent (Christ et al., 1991; 1993; 1996).

Such an action would result in inhibition of contractions as follows. Electrically evoked smooth muscle contraction results from influx of Ca²⁺ into the cells through voltage gated calcium channels (Somlyo & Somlyo, 1960; Mekata, 1981). In the present experiments, action potentials during muscle stimulation would be directly evoked in only a narrow band of cells directly encircled by the cathode. A much larger number of cells is presumably recruited for contraction by the spread of the action potentials, or of elevated cytosolic [Ca²⁺], through gap junctions. If heptanol blocks gap junctional communication, the recruitment of cells for contraction would be reduced, resulting in inhibition of the myogenic response.

In view of the arguments presented above, we restrict attention below to blockade of gap junctional communication by heptanol as the mechanism underlying its effects on neurogenic contractions.

The suppression of the noradrenergic phase by heptanol can be explained along lines similar to those mentioned for the purinergic phase above, and for its inhibition of agonist-induced contraction of vascular smooth muscle (Christ et al., 1991, 1993; Christ, 1995). In many smooth muscle organs, including the vas deferens, noradrenaline evokes contraction not via direct electrical activation involving EJPs but via biochemical pathways involving second messengers such as IP₃ and Ca²⁺ (Somlyo et al., 1985; Han et al., 1988). Because of the features of neuronal activation mentioned above, NA, like ATP also activates directly a small fraction of smooth muscle cells. Contractile activation of the noninnervated cells would then depend upon the cell-to-cell diffusion of Ca²⁺ presumably through gap junctions (Christ et al., 1992). Hence heptanol, if it blocks gap junctional communication, would profoundly reduce the number of cells recruited for contraction, inhibiting the noradrenergic phase.

The increase in latency of the contractions in the presence of heptanol may be explained as follows. Since EJPs are inhibited by heptanol, therefore at a given frequency of stimulation a larger number of low-amplitude EJPs would be required to achieve threshold through summation for action potential generation (see Figure 2C and D). Furthermore, once action potentials have been generated, block of gap junctions would delay the diffusion of activator Ca²⁺ ions into indirectly recruited cells, delaying the onset of the contractile response.

In conclusion, the present study shows that a presumptive uncoupler of cell-to-cell communication in smooth muscle, 1-heptanol, suppresses neurogenic as well as myogenic contractile force. In many smooth muscle organs, neurogenic transmitter evoked contractions contribute significantly to the maintenance of tone or to bursts of propulsive contractions. Our results, therefore, suggest an important role for modulation of intercellular communication via gap junctions in regulating the contractility of these organs.

Financial support from the Department of Science and Technology, India, under project no. SP/SO/NO6/93 is gratefully acknowledged. We thank Bayer plc, U.K. and Pfizer Ltd, Mumbai, India for supplying suramin sodium salt and prazosin hydrochloride respectively. We are thankful to Dr G.J. Christ for his valuable commnets on an earlier version of the paper.