- Email: [email protected]
Neuromodulation in molluscan smooth muscle: the action of 5-HT, FMRFamide and purine compounds
~
Gen. Pharmac.Vol. 25, No. 3, pp. 539-552, 1994 Copyright © 1994 ElsevierScienceLtd Printed in Great Britain. All rights reserved 0306-3623/94 $7.00 + 0.00
Pergamon
Neuromodulation in Molluscan Smooth Muscle: the Action of 5-HT, FMRFamide and Purine Compounds I. D. N E L S O N A N D H. H U D D A R T Division of Biological Sciences, Institute of Environmental and Biological Sciences, Lancaster University, Lancaster LA 1 4YQ, England (Tel. (0524) 65201; Fax (0524) 843854)
(Received 2 August 1993)
Abstract--l. The RR, OR, RS and RP muscles of Buccinum did not respond directly to 5-HT, but this monoamine converted their normally tonic ACh responses to fast twitch contractions with lowered tonic force. This action was not accompanied by significant membrane potential changes. 2. Pre-treatment with dibutyryl cAMP potentiated ACh responses and enhanced 5-HT modification of the responses. 3. All muscles responded strongly to FMRFamide with twitch contractions but this was not accompanied by significant membrane potential changes. 4. FMRFamide enhanced ACh contracture force and converted the responses into fast twitch activity. FMRFamide responses were dramatically inhibited by 5-HT with loss of all tonic force and fast twitch activity. 5. While dibutyryl cAMP did not affect FMRFamide responses, the IP 3 inhibitor lithium, at very high concentrations, caused a significant diminution of FMRFamide responses. 6. All four muscles were unresponsive to adenosine and ATP but all except the RP responded in a dose-dependent manner to GTP and GTP-?-S over the 10-7-10 -4 mol 1-i range. The responses showed moderate fast twitch activity which was unaccompanied by action potential discharges. Guanosine was without effect, except at very high concentrations where it inhibited FMRFamide responses. 7. ACh and GTP acted additively to increase muscle force and to enhance ACh-induced depolarization. Similarly both GTP and GTP-?-S acted additively, considerably enhancing FMRFamide responses. 8. It is proposed that 5-HT, FMRFamide and GTP may, via their separate receptors or by possible interaction with ion channels, activate secondary messenger systems to modify the calcium released by ACh-induced depolarization to modulate excitation-contraction coupling and force generation in these muscles.
KeyWords:Buccinum undatum, dibutyryl cAMP, INTRODUCTION A few previous studies have reported that molluscan smooth muscles have a complex innervation which may involve more than one type of nerve ending (Hill and Sanger, 1974; Heyer et aL, 1973; Huddart et aL, 1977; Brooks, 1982). A recent fine structural study of Buccinum radular retractor muscle (Nelson and Huddart, 1994) has revealed several types of nerve ending with at least four distinct types of transmitter vesicle, several vesicle types coexisting in the same endings, implying the possibility of pre- and postsynaptic neuromodulation by co-transmission. F r o m their appearance, and similarities with other described nerve endings in Busycon (Hill and Sanger, 1974) and Neptunea (Brooks, 1982) a tentative
FMRFamide, GTP, 5-hydroxytryptamine, lithium
classification into physiological types was attempted. These included cholinergic, serotonergic, peptidergic and purinergic types. Although acetylcholine (ACh) is undoubtedly the primary neurotransmitter in molluscan smooth muscle, Welsh and Moorhead (1959) identified serotonin (5-HT) in Busycon radular muscles. This monoamine or tryptamine itself have been shown either to induce small muscle twitches (Nystrom, 1967; Dorsett et al., 1989) or to interact with ACh-induced activity by lowering force and inducing rhythmic contractions (Hill et al., 1970; Heyer et al., 1973; Hill and Licis, 1981, 1985). Early work by Kerkut and Laverack (1960) and Kerkut and Cottrell (1963) demonstrated the existence of "cardioexcitatory" agents other than 5-HT in molluscs, the most important of which is the 539
540
I.D. NELSONand H. HU19DART
tetrapeptide FMRFamide, isolated and identified by Price and Greenberg (1977). This agent is widespread in molluscs (see Cottrell, 1989) and many molluscan non-cardiac smooth muscles respond strongly to very low concentrations of FMRFamide (Price and Greenberg, 1979; Greenberg and Price, 1979; Hill and Langton, 1988; Yanagawa et al., 1988; Huddart et al., 1992) so much so that the radular protractor muscle of Busycon was developed as a microbioassay for this peptide (Nagle and Greenberg, 1982). How this peptide interacts with other putative transmitters to modulate muscle activity has been little studied except for the briefly reported observation by Huddart et al. (1992) that FMRFamide treatment enhanced subsequent ACh responses in the odontophore protractor muscle of Busycon. In dissociated muscle cells of Helisoma, FMRFamide was shown to have some inexplicable effects (Zoran et al., 1989) and these authors commented that these may be due to the elimination of interactive effects between multiple transmitters which may be expected to occur in the intact in vivo preparation. Even more problematic is the potential role of purines in neuromuscular transmission in the invertebrates. Although adenosine and ATP seem to have clear pre- and postsynaptic modulatory roles in mammalian smooth muscle (Burnstock, 1972, 1980, 1981; Williams and Cusack, 1990), they have now been shown to be widespread in several invertebrate phyla including the molluscs with clear evidence for invertebrate purinoreceptors (Hoyle and Greenberg, 1988). More recently, Knight et al. (1992) have shown that adenosine, ADP and ATP have a variety of actions on Helix heart, but at uncomfortably high concentrations. In his fine structural study of Neptunea oesophageal muscle Brooks (1982) showed the clear presence of nerve endings with dense black vesicles quite unlike the normal small lucent cholinergic, dark granular aminergic or large grey peptidergic types. These he suggested may be purinergic in nature but with no in vitro pharmacological evidence to back this up. Such vesicles are present in Buccinum radular retractor muscle (Nelson and Huddart, 1993), so the proposition must be taken seriously that some type of purinergic transmission may exist in molluscan smooth muscle. In their study of the pharmacology of Octopus digestive tract, Andrews and Tansey (1983) concluded that ACh and monoamines may interact in vivo to regulate activity of the gut. The seeming complexity of the innervation of molluscan proboscis muscles prompted us not just to examine individual responses to ACh, monoamines, FMRFamide and purines, but to see how they interacted with each other to cast light on how interactive neuromodu-
lation may regulate proboscis muscle activity and hence the operation of the feeding mechanism. MATERIALS AND METHODS Mature specimens of Buccinum undatum were obtained from the University Marine Station at Millport, Isle of Cumbrae and kept in large aerated aquaria at 12°C. The muscles were dissected from the proboscis as described in Huddart et al. (1992) and Nelson and Huddart (1994), ligated at each end with monofilament nylon and placed in aerated jacketed organ baths cooled to 13°C. Preparations were suspended between the bath hooks and Grass FT.03 force displacement transducers with a passive load of 0.5 g. The transducer outputs were connected to four Grass 7PI22D low level d.c. amplifiers and from there to the analogue inputs of an Intracell $200 A D C A D interface coupled to an Elonex PC 386SXM/16 microcomputer for digital capture and display. A single sucrose gap system, described elsewhere by Huddart et al. (1992), was used to examine the electrical activity of the muscle fibres. The set up of the gap system and the protocols for its operation have also been described elsewhere (Huddart and Hill 1988). Each experiment began by depolarizing the cells of the right-hand compartment to a reference zero with isotonic 0.56moll -t KCI. From this the compound resting potential of the cells of the left (test) compartment could be measured via flat tip IVM Ag/AgCI electrodes in both compartments. These were connected to the high impedance probe of an HSE Type 309a/310 Microelectrode/Voltage Clamp amplifier, then via Grass low level d.c. amplifiers to the microcomputer for digital capture and display. Drugs and depolarizing salines were passed through the left (test) compartment and the induced changes in membrane potential and tension were simultaneously recorded via the A D C A D system. All drugs were obtained from Sigma and were made up freshly as stock concentrates in sea water. RESULTS Interaction o f 5 - H T and ACh
The four muscles used routinely in this study were the radular retractor (RR), the odontophore retractor (OR), the radular sac (RS) and radular protractor (RP). None of these muscles responded directly to 5-HT application but, as is the case in Busycon RP muscle (Hill and Licis, 1985) and Mytilus ABRM (Murakami et al., 1992), application of 5-HT during the course of acetylcholine-induced responses
Neuromodulation in molluscan muscle resulted in a marked change in the response, with a decline in tonic force but a stimulation of furious fast twitch contractions [Fig. l(a)]. When simultaneous membrane potential and mechanical responses were recorded by single sucrose-gap analysis, 5-HT was found to induce no change in membrane potential after the initial ACh-induced depolarization. A typical example of such a response of the RR muscle is shown in Fig. l(b). It was difficult to wash 5-HT out of the muscles, even after 20 min continuous sea water wash the ACh response was still characterized by abnormal fast twitch activity [Fig. l(c)l. It has been suggested that 5-HT effects in molluscan muscle may be mediated by the secondary messenger cAMP (Kohler and Lindl, 1980; Ishikawa et al., 1981; Ishii et al., 1989) as appears also to be the case in molluscan neurones (Siegelbaum et al., 1986), so the effect of the membrane permeant analogue dibutyryl cAMP was examined in these muscles. After a 30 rain incubation with 100 # M 1-j dibutyryl cAMP control ACh responses were potentiated while subsequent doses of 5-HT led to an even larger relaxation of tonic force and an increase in the level of twitch activity over that in controls untreated with dibutyryl cAMP (Fig. 2).
The effect of FMRFamide F M R F a m i d e was found to be extremely excitatory in all four muscles, the active concentration range being [10-8]-[5 × 10 -6] mol 1-1, above this responses declined dramatically indicative of receptor desensitization (Fig. 3). Responses were characterized by a large initial contraction followed by a long continuous burst of fast oscillatory twitch activity. Similar responses were seen with the related tetrapeptide F M R F a m i d e (not shown). When examined in the sucrose gap, F M R F a m i d e surprisingly induced only a tiny membrane depolarization of c. 1 mW at a concentration of 5 × 10 6 m o l l - I which induces maximal contractile activity.
Interaction of ACh, FMRFamide and 5-HT F M R F a m i d e and ACh were found to be additive in their effects in all the muscles examined. F M R F amide at 10 6tool 1-t not only enhanced ACh contractures but converted smooth ACh responses into fast twitch activity [Fig. 4(a)]. However, when the muscles were activated to contract by FMRFamide, application of 5-HT caused dramatic inhibition with loss of all tonic and fast twitch activity [Fig. 4(b)]. These results suggested that ACh, F M R F a m i d e and 5-HT may act via different mutually exclusive transduction routes to activate or inhibit cellular
541
calcium release. In an attempt to clarify which transduction routes might be involved we resorted to modulators of the two common secondary messenger systems. Whereas 5-HT modulation of ACh responses was influenced by dibutyryl cAMP, this latter agent had only a slight effect on control F M R F a m i d e responses [Fig. 5(a)]. A second system involving inositol trisphosphate (Ins{1,4,5}P 3 or IP3) was examined. The action of IP 3 has been reviewed by Berridge (1989) and it has been shown that lithium at high concentrations can block the IP 3 pathway. The effect of lithium was thus examined to see if IP 3 might be involved in F M R F a m i d e signal transduction. At the very high concentration of 100 mmol 1- J lithium did cause a significant diminution of F M R F amide responses [Fig. 5(b)].
The action of purines Neither adenosine nor ATP exerted any effect on the four proboscis muscles examined even at concentrations as high as 1 m m o l l -j. However all muscles except the RP responded strongly to G T P and to its stable analogue GTP- 7-S over the concentration range [10-7]-[10 4] mol 1-l. Figure 6(a) shows a typical cumulative dose addition response of the muscles to G T P while Fig. 6(b) and (c) show typical single shot responses of the muscles. The muscles were found to be very prone to receptor desensitization with both compounds above the upper limit of 1 0 - 4 m o l l -1. The responses were characterized by moderate levels of fast twitch activity. When the R R muscle was examined in the sucrose gap [Fig. 6(d)] while 5 × l0 5moll -1 G T P induced a depolarization of c. 2.5 mV the twitch activity was not seen to be accompanied by action potential discharges. In complete contrast the muscles were rather unresponsive to guanosine over the normal physiological range but at the very high and quite unphysiological concentration of 1 0 m m o l l -~, partial inhibition of F M R F a m i d e responses was seen and two examples of this are shown in Fig. 7.
Interaction of ACh, FMRFamide and GTP ACh, when administered after G T P application, enhanced the responses by superimposition of considerably higher force levels than that generated by G T P alone [Fig. 8(a)] and when the RR muscle was examined in the sucrose gap this was mirrored in the electrical responses [Fig. 8(b)] where 10 - 5moll 1 ACh alone induced typically c. 1 mV depolarization and the subsequent addition of 10-Smoll ~ G T P generated a further 4 mV depolarization. Both G T P and GTP-~,-S were found to also considerably enhance responses to F M R F a m i d e (Fig. 9).
542
I . D . NELSON and H. HUDDART
Y
Y
_________~
tt'~~
v-,
m
¸¸
<
J
g.h
O
-!
l
®~ ©
e',l
Neuromodulation in molluscan muscle
©
543
ACh
W
l
Ilog
!
RR
OR
1
RS
RP •
, . . . . . . .
~ .~. . . .
~
.
.
~
200 s Fig. 1. (a) The effect of 10 -s mol 1-i 5-HT on the response to a 10-5 mol 1-i dose of ACh in the proboscis muscles with decline in tonic force and stimulation of fast twitch activity. Upper scale applies to the two upper panels, lower scale to the two lower panels. (b) Typical single sucrose gap recording showing membrane depolarization induced by 10-s mol I -I ACh and the lack of effect of 10 -5 mol l -~ 5-HT on membrane potential. (c) Responses to l0 -~ mol 1-~ ACh after a 20 min washout following exposure to 5-HT with persistence of fast twitch activity. Upper scale applies to the two upper panels, lower scale to the two lower panels. In this and all subsequent figures W indicates washout.
An odd but consistent observation was that pre-treatment with F M R F a m i d e induced the RP muscle, which is normally insensitive to G T P and GTP-?-S, to respond to a subsequent dose of these purines. DISCUSSION The observation that 5-HT interacts with acetylcholine-induced responses to produce relaxation is not new, having been observed in the A B R M of Mytilus and the RP muscle of Busycon. Here, however, we examined the R R muscle in the sucrose gap and found that although 5-HT lowered tonic force and induced fast twitch activity this was not accompanied by any significant membrane potential
changes, suggesting that 5-HT action here was by a secondary messenger system. Cole and Twarog (1972) showed that 5-HT relaxation of catch in the Mytilus A B R M could be mimicked by the membrane permeant dibutyryl cAMP, suggesting that the 5-HT action was mediated via the adenylate cyclase/cAMP pathway. This view is strengthened by the observations of Kohler and Lindl (1980) in that same preparation that 5-HT between [10-6]-[10 -4] mol 1-~ greatly increased c A M P content in vitro. Our own experiments here showed that dibutyryl c A M P did indeed enhance the 5-HT-induced relaxation and fast twitch activity in all four proboscis muscles of Buccinum adding weight to the view that 5-HT operates via the adenylate cyclase/cAMP route. However there are a number of conflicting reports on the
1. D. NELSON and H. HUDDART
544
5-HT
ACh
W 4g
RR OR RS RP
30 mins 100pmol1-1 200s diB cAMP RR
_
~
OR RS
~
....
4g
RP 200s Fig. 2. Upper panel, responses to a control dose of 10-5 mol I-t ACh followed by 10-5 moll -~ 5-HT in the four proboscis muscles. Lower panel, same protocol as in upper panel following a 30 min exposure to 100 ttmol 1-~ dibutyryl cAMP. Note enhancement of ACh responses and enhancement of the 5-HT effect. Upper calibration applies to all traces in upper panel, lower calibration applies to all traces in lower panel.
relationship between 5-HT action a n d c A M P in o t h e r molluscan tissues. Whereas lshikawa et al. (1981) showed t h a t 5-HT dose-dependently elevated c A M P and reduced c G M P levels in Mytilus A B R M , M u r a k a m i et al. (1992) showed in t h a t same preparation that the relaxing responses mediated by the 5-HT2 receptor subtype were not linked to adenylate or guanylate cyclase systems. In contrast in Helix
neurones, Paupardin-Tritsch et al. (1986) showed that 5-HT increased Ca 2+ m e m b r a n e c o n d u c t a n c e but this was not mimicked by intracellular injection o f c A M P but was mimicked by intracellular injection of c G M P and zaprinast, a c G M P phosphodiesterase inhibitor. T o add further confusion, by using patch clamp analysis of Aplysia neurones with single channel recording, Siegelbaum et al. (1986) showed that
(Fig. 3 Opposite) Fig. 3. (a) Typical effects of 10-6 mol 1-~ FMRFamide on proboscis muscles. Upper scale applies to RR muscle trace, lower scale to the other three traces. (b) Mean FMRFamide dose responses curves for all four proboscis muscles. Note the sharp fall off of responses above 10-5 mol 1-~. (c) Typical sucrose gap electrical recording from the RR muscle in response to 10-6 mol 1-t FMRFamide, showing a depolarization of only c. 1 mV.
RP
RS
OR
RR
®
...-,
I
FMRFamide
120s
2g
4g
©
E..
4o
o
"4
M
®
Fig. 3--caption opposite.
!
W
0 10 - 9
20
40
60
80
100
4"
/
/
/
//
10-7
/,/
///
ff
FMRFamide
',,
4
"
'q
"+,
| W
'mY / 1
'
( t o o l 1-1 )
10-B
n=4,4,4,4,4,4,4
10-a
t
'" ," /~ ..."
....:,
F~lRlWamide C o n c e n t r a t i o n
10-0
T,','/
.:'
RP
*
OR RS
• •
RR
•
10-4
O ,--e
o
=_
o
c-
z
ACh
200s
FMRFamide
4g
!2g
,I
W
O
RP
RS
OR
RR
5-HT
1
FMRFamide
|
200s
g
W
2g
,I
4g
Fig. 4. (a) Interaction of ACh and FMRFamide responses. The response to a control dose o f 10 -5 tool 1-~ A C h is considerably enhanced by a subsequent dose of 10 -6 tool 1-~ FMRFamidc. Upper scale applies to the RR muscle, lower scale to the others. (b) Strong inhibition by 10 -5 mol 1-~ 5-HT of responses to 10 -6 mol 1-~ F M R F a m i d e . Upper scale applies to the R R muscle, lower scale to the others.
RP
RS
OR
RR
®
e~
Z
Z
ox
1
~
FMRFamide
~
200s
.
.
.
.
..
W
3g
4g
RP
200s
~
. . . .
..-~2.
= ' = ~ ~7. 2.-Y_31101_~='_, Z_:;_.~Z"_Z_~S77_.;-":2~Y.Y;Y'L'L~-
~
10 m i n 1 0 0 r a M L i 3÷
1
FMRFamide
RR
RP
RS
OR
RR
®
L
1
W
2g
4g
200S 200s Fig. 5. (a) Upper panel, control responses to 10 6 mol 1-~ FMRFamide, lower panel, FMRFamide responses after 30 min incubation with dibutyryl cAMP, showing no significant effect. In both panels the tension scale applies to all traces of the panel. (b) Upper panel, control responses to 10-6 tool 1-~ FMRFamide, lower panel, FMRFamide responses after 10 rain incubation with 100 mmol 1-~ lithium. Note the significant inhibition, but only at very high lithium concentration. Upper scale applies to the RR muscle of both panels, lower scale applies to the other muscles of both panels.
RP
RS
o.
RR
30 m i n s 1 0 0 u m o l 1-1 diB cAMP
aP
RS
OR
RR
®
o
=_.. ga
o
c-
z
RP
RS
OR
RR
0b
t2
t3
110-Smol 1-l GTP
+1
Fig.
&caption
4
WI
opposite.
0C
RP
Neuromodulation in molluscan muscle
10 mmoi I-I Guanosine
1
549
2g
RR
120s
lg
RS
Fig. 7. Typical inhibitory effects of guanosine on FMRFamide-treated RR and RS muscles, but only seen at the very high concentration of 10 mmol 1-~. 5-HT induced prolonged closure of the S potassium channel mediated by cAMP-dependent phosphorylation of a membrane protein which may be the channel itself. The relationship between 5-HT, ACh and contractile activation has been studied in suspensions of isolated cells from Mytilus ABRM which were fura-2 loaded to follow changes in [Ca]i (Ishii et al., 1989). These studies clearly showed that 5-HT suppressed the elevation of [Ca]i induced by carbachol, giving a direct explanation for the 5-HT relaxatory effect seen in the whole muscle. The view that 5-HT modulatory effects on ACh responses are mediated by a rise in cAMP and a fall in [Ca]i thus modifying ACh-induced electromechanical coupling are consistent with our results in Buccinum proboscis muscles. Only single channel recording of muscle by patch clamp will yield more detailed information about 5-HT and cAMP actions. The experiments here showed that Buccinum proboscis muscles were extremely sensitive to FMRFamide but without significant membrane potential
changes, and that FMRFamide responses were unaffected by dibutyryl cAMP but inhibited by lithium, albeit at high concentration. This suggests that FMRFamide action may be routed via an IP3 transduction pathway. Interaction between ACh and FMRFamide responses was additive and this suggests that FMRFamide, via intracellular IP3 release was able to access and release a pool of cellular Ca independently of that released by AChinduced depolarization, thus significantly enhancing fibre free Ca and promoting excition--contraction coupling, the reverse of the actions seen with 5-HT on ACh responses. In complete contrast, 5-HT inhibited FMRFamide responses in all proboscis muscles examined. If indeed FMRFamide activates IP 3 to release a distinct cellular Ca pool to induce classic pharmacomechanical coupling when 5-HT, via the cAMP pathway, must induce re-sequestration of [Ca]i to induce pharmacomechanical uncoupling as effectively as Ca-free conditions which uncouple ACh-induced electro-
(Fig. 6 Opposite) Fig. 6. (a) Responses of the RR, OR and RS muscles to GTP. Concentrations were I, 10-6moll-~; 2, 5 × 10-6 mol 1-~; 3, 10-5 mol I-~. Upper tension calibration applies to the RR muscle, lower calibration applies to the OR and RS muscle. (b) Muscle responses to GTP, note lack of response in the RP muscle. Upper tension scale applies to the RR, OR and RP, lower scale to the RS muscle. (c) Responses of proboscis muscles to GTP-y-S, again note lack of response of the RP muscle. (d) Sucrose gap recording of the RR muscle response to 5 × 10-5 mol 1-I GTP showing lack of action potential discharges during fast twitch activity.
550
I.D. NELSONand H. HUDDART
®
10 .5 m o l I -I G T P
W 1
10 -5 m o l I -I A C h
1
1
4g OR
RS
~
.....
....
2g
~ 120s
10 -5 m o l 1-I A C h
®
RR
1
10 -5 m o l 1-1 G T P
1
W 1 Eleclrical~ 1mV Mechanical
lg
120s Fig. 8. (a) Interaction of GTP and ACh in RR, OR and RS muscles, with enhancement of GTP responses caused by a subsequent exposure to ACh. Upper tension scale applies to the RR muscle, lower scale to the OR and RS muscle. (b) Sucrose gap recording showing that the membrane depolarization to ACh increases considerably by a subsequent exposure to 10-5 moll -~ GTP.
mechanical coupling (Huddart and Hill, 1988; Nelson and Huddart, 1992). None of the Buccinum proboscis muscles were responsive to either adenosine or ATP at concentrations up to 1 0 - 3 m o l l -~, as is also the case in Busycon proboscis muscles (Huddart, unpublished observations) and in Helix and Arion heart in what can be considered the normal (i.e. < 1 0 3 m o l l - l ) physiological range (Knight et al., 1992). However the RR, O R and RS muscles of Buccinum all showed strong dose-dependent responses to G T P and its
stable analogue GTP-7-S at low concentrations, as is also the case in Busycon RP muscle and atrium (Huddart, unpublished). Our Buecinum muscles were inhibited by guanosine, but only at the high non-physiological level of 10 mmol I -~. The cumulative dose-addition responses [Fig. 6(a)] were clearly suggestive of agonist/receptor interactions but with what type of recognition site is unknown, whether A 1, A 2 or the various P2 receptor subtypes (Williams and Cusack, 1990) or even the ACh receptor or another ion channel. The G T P responses were
Neuromodulation in molluscan muscle
10 -6
RR
tool 1-1 FMRFamide
10 -5 tool 1-1 GTP-y-S
...... j I"
551
W
~ 2g
OR
..... r~.
.
.
.
0.5g
RS
120s Fig. 9. Interaction of FMRFamide and GTP-y-S responses. FMRFamide-induced tension was considerably enhanced by subsequent treatment with GTP-y-S. Upper tension scale applied to the RR and OR muscles, the lower scale bar to the and RP muscles. Note that with combined FMRFamide and GTP-y-S treatment the normally unresponsive RP muscle develops modest force.
additive to ACh responses, and sucrose gap recordings showed that GTP at 10-Smoli - ' induced a modest membrane depolarization which was additive to that induced by ACh. This is suggestive that GTP may not have acted via a secondary messenger but may have interacted with a receptor or an ion channel to promote Na + influx and additionally raise [Ca]~ promoting electromechanical coupling. This point could only be resolved using voltage clamp analysis of the muscles with measurement of ACh-induced inward current in normal, in GTP-containing salines and when challenged with GTP in Na-free conditions. Unfortunately the unusual ACh receptor possessed by these muscles is unresponsive to the so called "usual" agonists/antagonists of the mammalian nicotinic and muscarinic AChR (Nelson and
Huddart, 1992), only mytoion appears to be partially effective in any predictable manner (Kohler and Lindl, 1980; Brooks et al., 1990). These muscles possess not just a complex innervation but also a novel interactive pharmacology. There appear to be multiple pathways for modulating the calcium released by classic ACh-promoted electromechanical coupling and hence several ways to modulate both force development and its pattern of expression during radular activity.
Acknowledgements--This work was carried out while IDN was in receipt of an SERC postgraduate research studentship. The authors are grateful to Professor R. B. Hill and Dr D. D. Brooks for much helpful discussion.
I. D. NELSONand H. HUDDART
552
REFERENCES Andrews P. L. R. and Tansey E. M. (1983) The digestive tract of Ocropus vulgaris: the anatomy, physiology and pharmacology of the upper tract. J. mar. biol. Ass. U.K. 63, 1099134. Berridge M. J. (1989) lnositol trisphosphate, calcium, lithium and cell signaling. J. Am. med. Ass. 262, 1834-1841. Brooks D. D. (1982) Membrane currents associated with contraction in molluscan visceral muscle. Ph.D. thesis, University of Lancaster. Brooks D. D.. Huddart H.. Lennard R. and Hill R. B. (1990) Calcium utilization in contractures induced by acetylcholine or high-potassium saline in molluscan proboscis muscles. Jy exp. Biol. 149, 379-394. Burnstock G. (1972) Purinergic nerves. Pharmac. Rev. 24. 5099581. _ Burnstock G. (1980) Purinergic nerves and receptors. Prog. Biochem. Pharmac. 16, 141-154. Burnstock G. (1981) Neurotransmitters and trophic factors in the autonomic nervous system. J. Physiol. 313, I-35. Cottrell G. A. (1989) The biology of the FMRFamide series of peptides in molluscs with special reference to Helix. Comp. Biochem. Physiol. 93A, 4145. Dorsett D. A., Skelton M. E. and Evans C. G. (1989) The innervation and physiology of the extrinsic buccal retractor muscles of Philine aperta (Linnaeus). J. molluscan Srud. 55, 193-208. Greenberg M. J. and Price D. A. (1979) FMRFamide, a cardioexcitatory peptide of molluscs; an agent in search of a mission. Am. Zool. 19, 1633174. Heyer C. B., Kater S. B. and Karlsson U. L. (1973) Neuromuscular systems in molluscs. Am. Zool. 13, 247-270. Hill R. B. and Langton P. D. (1988) Use of a sucrose gap apparatus to record electrical responses of gastropod radular protractor muscle to FMRFamide. Comp. Biothem. Pi?ysiol. 9OC, 2077214. Hill R. B. and Licis P. (1981), Effect of neurohumors during shutdown of electrogenic sodium pumping in a molluscan muscle. Adv. Physiol. Sci. 22, 339-360. Hill R. B. and Licis P. (1985) Lanthanum and caffeine affect both membrane response and force in a molluscan muscle, the radular protractor of Busycon canaliculatum. Comp. Biochem. Physiol. 82C, 363-376. Hill R. B. and Sanger J. W. (1974) Anatomy of the innervation and neuromuscular junctions of the radular protractor muscle of the whelk, Bucycon canaliculatum (L.) Biol. Bull. 147, 369-385. Hill R. B., Greenberg M. J., Irisawa H. and Nomura H. (1970) Electromechanical coupling in a molluscan muscle, the radula protractor of Busycon canaliculatum. J. exp. Zool. 174, 331-348. Hoyle C. V. and Greenberg M. J. (1988) Actions of adenyl compounds in invertebrates from several phyla; evidence for internal purinoreceptors. Comp. Biochem. Physiol. 9OC, 113-122. Huddart H. and Hill R. B. (1988) Electromechanical uncoupling in a molluscan smooth muscle examined by the sucrose gap technique. The effect of calcium antagonist and agonist agents. J. camp. Physiol. B 158, 501-512. Huddart H., Hunt S. and Oates K. (1977) Calcium movements during contraction in molluscan smooth muscle, and the loci of calcium binding and release. J. exp. Biol. 68, 45-56. Huddart H., Nelson I. D., Brooks D. D. and Hill R. B. (1992) Electrical and mechanical responses of the odontophore protractor muscle of Busycon canaliculatum: a sucrose gap voltage clamp study. Comp. Biochem. Physiol. 102C, 149-158.
Ishii N.,. Simpson A. W. M. and Ashley C. C. (1989) Effects of 5-hydroxytryptamine (serotonin) and forskolin on intracellular free calcium in isolated and furaloaded smooth muscle cells from the anterior byssus retractor (catch) Mytilus edulis. Pfugers Arch. 414, 162-170. Ishikawa T., Murakami H. and Iwayama Y. (1981) Changes in CAMP and cGMP levels induced by relaxing drugs in ACh and K treated molluscan smooth muscle fibres. Comp. Biochem. Physiol. 7OC, 171Ll76. Kerkut G. A. and Cottrell G. A. (1963) Acetylcholine and 5-hydroxytryptamine in the snail brain. Comp. Biochem. Physiol. 8, 5363. Kerkut G. A. and Laverack M. S. (1960) A cardio-accelerator present in tissue extracts of the snail Helix aspersa. Comp. Biochem. Physiol. 1, 62-71. Knight G. E., Hoyle C. H. V. and Burnstock G. E. (1992) Effects of nucleosides and nucleotides on the isolated hearts of the snail Helix aspersa and the slug Arion ater. Comp. Biochem. Physiol. lOlC, 175-181. Kohler G. and Lindl T. (1980) Effects of 5-HT. dopamine and acetylcholine on accumulation of CAMP and cGMP in the anterior byssus retractor muscle of Mytilus edulis L. (Mollusca). Pflugers Arch. 383, 257-262. Murakami H., Kizawa Y., Sano M., Edamura N., Maruyama C. and Yamazaki A. (1992) Possible site of action of 2-methylserotonin in inducing relaxation of acetylchohne-induced contraction in the molluscan (Mytilus edulis) smooth muscle. Comp. Biochem. Physiol. lOlC, 3433347. Nagle G. T. and Greenberg M. J. (1982) A highly sensitive microbioassay for the molluscan neuropeptide FMRFamide. Comp. Biochem. Physiol. 71C, 101-105. Nelson I. D. and Huddart H. (1992) The nature of the acetylcholine receptor in a Buccinum proboscis muscle examined by the sucrose-gap voltage clamp technique. Gen. Pharmac. 23, 317-323. Nelson I. D. and Huddart H. (1994) The fine structure of a molluscan muscle, the radular retractor of Buccinum undatum in relation to excitation-contraction coupling. Submitted. Nystrom R. A. (1967) Spontaneous activity of clam intestinal muscle. Comp. Biochem. Physiol. 21, 601-610. Paupardin-Tritsch D., Hammond C. and Gerschenfeld H. M. (1986) Serotonin and cyclic GMP both induce an increase of the calcium current in the same identified molluscan neurones. J. Neurosci. 6, 2715-2723. Price D. A. and Greenberg M. J. (1977) Structure of a molluscan cardioexcitatory neuropeptide. Science 197, 67067 I. Price D. A. and Greenberg M. J. (1979) Pharmacology of the molluscan cardioexcitatory neuropeptide FMRFamide. Gen. Pharmac. 11, 237-241. Siegelbaum S. A., Belardetti F., Camardo J. S. and Shuster M. J. (1986) Modulation of the serotonin-sensitive potassium channel in Aplysia sensory neurone cell body and growth cone. J. exp. Biol. 124, 2877306. Welsh J. H. and Moorhead M. (1959) Identification and assay of 5-hydroxytryptamine in molluscan tissues by fluorescence method. Science 129, 1491-1492. Williams M. and Cusack N. J. (1990) Neuromodulatory roles of purine nucleosides and nucleotides: their receptors and ligands. Neurotransmissions 6, l-6. Yanagawa M., Fujiwara M., Takabatake I.. Muneoka Y. and Kobayashi M. (1988) Potentiating effects of some invertebrate neuropeptides on twitch contraction of the radula muscles of a mollusc, Rapana thomasiana. Comp. Biochem. Physiol. WC, 73-77. Zoran M. J., Haydon P. G. and Matthews P. J. (1989) Aminergic and peptidergic modulation of motor function at an identified neuromuscular junction in Helisoma. J. exp. Biol. 142, 2255243.
Gen. Pharmac.Vol. 25, No. 3, pp. 539-552, 1994 Copyright © 1994 ElsevierScienceLtd Printed in Great Britain. All rights reserved 0306-3623/94 $7.00 + 0.00
Pergamon
Neuromodulation in Molluscan Smooth Muscle: the Action of 5-HT, FMRFamide and Purine Compounds I. D. N E L S O N A N D H. H U D D A R T Division of Biological Sciences, Institute of Environmental and Biological Sciences, Lancaster University, Lancaster LA 1 4YQ, England (Tel. (0524) 65201; Fax (0524) 843854)
(Received 2 August 1993)
Abstract--l. The RR, OR, RS and RP muscles of Buccinum did not respond directly to 5-HT, but this monoamine converted their normally tonic ACh responses to fast twitch contractions with lowered tonic force. This action was not accompanied by significant membrane potential changes. 2. Pre-treatment with dibutyryl cAMP potentiated ACh responses and enhanced 5-HT modification of the responses. 3. All muscles responded strongly to FMRFamide with twitch contractions but this was not accompanied by significant membrane potential changes. 4. FMRFamide enhanced ACh contracture force and converted the responses into fast twitch activity. FMRFamide responses were dramatically inhibited by 5-HT with loss of all tonic force and fast twitch activity. 5. While dibutyryl cAMP did not affect FMRFamide responses, the IP 3 inhibitor lithium, at very high concentrations, caused a significant diminution of FMRFamide responses. 6. All four muscles were unresponsive to adenosine and ATP but all except the RP responded in a dose-dependent manner to GTP and GTP-?-S over the 10-7-10 -4 mol 1-i range. The responses showed moderate fast twitch activity which was unaccompanied by action potential discharges. Guanosine was without effect, except at very high concentrations where it inhibited FMRFamide responses. 7. ACh and GTP acted additively to increase muscle force and to enhance ACh-induced depolarization. Similarly both GTP and GTP-?-S acted additively, considerably enhancing FMRFamide responses. 8. It is proposed that 5-HT, FMRFamide and GTP may, via their separate receptors or by possible interaction with ion channels, activate secondary messenger systems to modify the calcium released by ACh-induced depolarization to modulate excitation-contraction coupling and force generation in these muscles.
KeyWords:Buccinum undatum, dibutyryl cAMP, INTRODUCTION A few previous studies have reported that molluscan smooth muscles have a complex innervation which may involve more than one type of nerve ending (Hill and Sanger, 1974; Heyer et aL, 1973; Huddart et aL, 1977; Brooks, 1982). A recent fine structural study of Buccinum radular retractor muscle (Nelson and Huddart, 1994) has revealed several types of nerve ending with at least four distinct types of transmitter vesicle, several vesicle types coexisting in the same endings, implying the possibility of pre- and postsynaptic neuromodulation by co-transmission. F r o m their appearance, and similarities with other described nerve endings in Busycon (Hill and Sanger, 1974) and Neptunea (Brooks, 1982) a tentative
FMRFamide, GTP, 5-hydroxytryptamine, lithium
classification into physiological types was attempted. These included cholinergic, serotonergic, peptidergic and purinergic types. Although acetylcholine (ACh) is undoubtedly the primary neurotransmitter in molluscan smooth muscle, Welsh and Moorhead (1959) identified serotonin (5-HT) in Busycon radular muscles. This monoamine or tryptamine itself have been shown either to induce small muscle twitches (Nystrom, 1967; Dorsett et al., 1989) or to interact with ACh-induced activity by lowering force and inducing rhythmic contractions (Hill et al., 1970; Heyer et al., 1973; Hill and Licis, 1981, 1985). Early work by Kerkut and Laverack (1960) and Kerkut and Cottrell (1963) demonstrated the existence of "cardioexcitatory" agents other than 5-HT in molluscs, the most important of which is the 539
540
I.D. NELSONand H. HU19DART
tetrapeptide FMRFamide, isolated and identified by Price and Greenberg (1977). This agent is widespread in molluscs (see Cottrell, 1989) and many molluscan non-cardiac smooth muscles respond strongly to very low concentrations of FMRFamide (Price and Greenberg, 1979; Greenberg and Price, 1979; Hill and Langton, 1988; Yanagawa et al., 1988; Huddart et al., 1992) so much so that the radular protractor muscle of Busycon was developed as a microbioassay for this peptide (Nagle and Greenberg, 1982). How this peptide interacts with other putative transmitters to modulate muscle activity has been little studied except for the briefly reported observation by Huddart et al. (1992) that FMRFamide treatment enhanced subsequent ACh responses in the odontophore protractor muscle of Busycon. In dissociated muscle cells of Helisoma, FMRFamide was shown to have some inexplicable effects (Zoran et al., 1989) and these authors commented that these may be due to the elimination of interactive effects between multiple transmitters which may be expected to occur in the intact in vivo preparation. Even more problematic is the potential role of purines in neuromuscular transmission in the invertebrates. Although adenosine and ATP seem to have clear pre- and postsynaptic modulatory roles in mammalian smooth muscle (Burnstock, 1972, 1980, 1981; Williams and Cusack, 1990), they have now been shown to be widespread in several invertebrate phyla including the molluscs with clear evidence for invertebrate purinoreceptors (Hoyle and Greenberg, 1988). More recently, Knight et al. (1992) have shown that adenosine, ADP and ATP have a variety of actions on Helix heart, but at uncomfortably high concentrations. In his fine structural study of Neptunea oesophageal muscle Brooks (1982) showed the clear presence of nerve endings with dense black vesicles quite unlike the normal small lucent cholinergic, dark granular aminergic or large grey peptidergic types. These he suggested may be purinergic in nature but with no in vitro pharmacological evidence to back this up. Such vesicles are present in Buccinum radular retractor muscle (Nelson and Huddart, 1993), so the proposition must be taken seriously that some type of purinergic transmission may exist in molluscan smooth muscle. In their study of the pharmacology of Octopus digestive tract, Andrews and Tansey (1983) concluded that ACh and monoamines may interact in vivo to regulate activity of the gut. The seeming complexity of the innervation of molluscan proboscis muscles prompted us not just to examine individual responses to ACh, monoamines, FMRFamide and purines, but to see how they interacted with each other to cast light on how interactive neuromodu-
lation may regulate proboscis muscle activity and hence the operation of the feeding mechanism. MATERIALS AND METHODS Mature specimens of Buccinum undatum were obtained from the University Marine Station at Millport, Isle of Cumbrae and kept in large aerated aquaria at 12°C. The muscles were dissected from the proboscis as described in Huddart et al. (1992) and Nelson and Huddart (1994), ligated at each end with monofilament nylon and placed in aerated jacketed organ baths cooled to 13°C. Preparations were suspended between the bath hooks and Grass FT.03 force displacement transducers with a passive load of 0.5 g. The transducer outputs were connected to four Grass 7PI22D low level d.c. amplifiers and from there to the analogue inputs of an Intracell $200 A D C A D interface coupled to an Elonex PC 386SXM/16 microcomputer for digital capture and display. A single sucrose gap system, described elsewhere by Huddart et al. (1992), was used to examine the electrical activity of the muscle fibres. The set up of the gap system and the protocols for its operation have also been described elsewhere (Huddart and Hill 1988). Each experiment began by depolarizing the cells of the right-hand compartment to a reference zero with isotonic 0.56moll -t KCI. From this the compound resting potential of the cells of the left (test) compartment could be measured via flat tip IVM Ag/AgCI electrodes in both compartments. These were connected to the high impedance probe of an HSE Type 309a/310 Microelectrode/Voltage Clamp amplifier, then via Grass low level d.c. amplifiers to the microcomputer for digital capture and display. Drugs and depolarizing salines were passed through the left (test) compartment and the induced changes in membrane potential and tension were simultaneously recorded via the A D C A D system. All drugs were obtained from Sigma and were made up freshly as stock concentrates in sea water. RESULTS Interaction o f 5 - H T and ACh
The four muscles used routinely in this study were the radular retractor (RR), the odontophore retractor (OR), the radular sac (RS) and radular protractor (RP). None of these muscles responded directly to 5-HT application but, as is the case in Busycon RP muscle (Hill and Licis, 1985) and Mytilus ABRM (Murakami et al., 1992), application of 5-HT during the course of acetylcholine-induced responses
Neuromodulation in molluscan muscle resulted in a marked change in the response, with a decline in tonic force but a stimulation of furious fast twitch contractions [Fig. l(a)]. When simultaneous membrane potential and mechanical responses were recorded by single sucrose-gap analysis, 5-HT was found to induce no change in membrane potential after the initial ACh-induced depolarization. A typical example of such a response of the RR muscle is shown in Fig. l(b). It was difficult to wash 5-HT out of the muscles, even after 20 min continuous sea water wash the ACh response was still characterized by abnormal fast twitch activity [Fig. l(c)l. It has been suggested that 5-HT effects in molluscan muscle may be mediated by the secondary messenger cAMP (Kohler and Lindl, 1980; Ishikawa et al., 1981; Ishii et al., 1989) as appears also to be the case in molluscan neurones (Siegelbaum et al., 1986), so the effect of the membrane permeant analogue dibutyryl cAMP was examined in these muscles. After a 30 rain incubation with 100 # M 1-j dibutyryl cAMP control ACh responses were potentiated while subsequent doses of 5-HT led to an even larger relaxation of tonic force and an increase in the level of twitch activity over that in controls untreated with dibutyryl cAMP (Fig. 2).
The effect of FMRFamide F M R F a m i d e was found to be extremely excitatory in all four muscles, the active concentration range being [10-8]-[5 × 10 -6] mol 1-1, above this responses declined dramatically indicative of receptor desensitization (Fig. 3). Responses were characterized by a large initial contraction followed by a long continuous burst of fast oscillatory twitch activity. Similar responses were seen with the related tetrapeptide F M R F a m i d e (not shown). When examined in the sucrose gap, F M R F a m i d e surprisingly induced only a tiny membrane depolarization of c. 1 mW at a concentration of 5 × 10 6 m o l l - I which induces maximal contractile activity.
Interaction of ACh, FMRFamide and 5-HT F M R F a m i d e and ACh were found to be additive in their effects in all the muscles examined. F M R F amide at 10 6tool 1-t not only enhanced ACh contractures but converted smooth ACh responses into fast twitch activity [Fig. 4(a)]. However, when the muscles were activated to contract by FMRFamide, application of 5-HT caused dramatic inhibition with loss of all tonic and fast twitch activity [Fig. 4(b)]. These results suggested that ACh, F M R F a m i d e and 5-HT may act via different mutually exclusive transduction routes to activate or inhibit cellular
541
calcium release. In an attempt to clarify which transduction routes might be involved we resorted to modulators of the two common secondary messenger systems. Whereas 5-HT modulation of ACh responses was influenced by dibutyryl cAMP, this latter agent had only a slight effect on control F M R F a m i d e responses [Fig. 5(a)]. A second system involving inositol trisphosphate (Ins{1,4,5}P 3 or IP3) was examined. The action of IP 3 has been reviewed by Berridge (1989) and it has been shown that lithium at high concentrations can block the IP 3 pathway. The effect of lithium was thus examined to see if IP 3 might be involved in F M R F a m i d e signal transduction. At the very high concentration of 100 mmol 1- J lithium did cause a significant diminution of F M R F amide responses [Fig. 5(b)].
The action of purines Neither adenosine nor ATP exerted any effect on the four proboscis muscles examined even at concentrations as high as 1 m m o l l -j. However all muscles except the RP responded strongly to G T P and to its stable analogue GTP- 7-S over the concentration range [10-7]-[10 4] mol 1-l. Figure 6(a) shows a typical cumulative dose addition response of the muscles to G T P while Fig. 6(b) and (c) show typical single shot responses of the muscles. The muscles were found to be very prone to receptor desensitization with both compounds above the upper limit of 1 0 - 4 m o l l -1. The responses were characterized by moderate levels of fast twitch activity. When the R R muscle was examined in the sucrose gap [Fig. 6(d)] while 5 × l0 5moll -1 G T P induced a depolarization of c. 2.5 mV the twitch activity was not seen to be accompanied by action potential discharges. In complete contrast the muscles were rather unresponsive to guanosine over the normal physiological range but at the very high and quite unphysiological concentration of 1 0 m m o l l -~, partial inhibition of F M R F a m i d e responses was seen and two examples of this are shown in Fig. 7.
Interaction of ACh, FMRFamide and GTP ACh, when administered after G T P application, enhanced the responses by superimposition of considerably higher force levels than that generated by G T P alone [Fig. 8(a)] and when the RR muscle was examined in the sucrose gap this was mirrored in the electrical responses [Fig. 8(b)] where 10 - 5moll 1 ACh alone induced typically c. 1 mV depolarization and the subsequent addition of 10-Smoll ~ G T P generated a further 4 mV depolarization. Both G T P and GTP-~,-S were found to also considerably enhance responses to F M R F a m i d e (Fig. 9).
542
I . D . NELSON and H. HUDDART
Y
Y
_________~
tt'~~
v-,
m
¸¸
<
J
g.h
O
-!
l
®~ ©
e',l
Neuromodulation in molluscan muscle
©
543
ACh
W
l
Ilog
!
RR
OR
1
RS
RP •
, . . . . . . .
~ .~. . . .
~
.
.
~
200 s Fig. 1. (a) The effect of 10 -s mol 1-i 5-HT on the response to a 10-5 mol 1-i dose of ACh in the proboscis muscles with decline in tonic force and stimulation of fast twitch activity. Upper scale applies to the two upper panels, lower scale to the two lower panels. (b) Typical single sucrose gap recording showing membrane depolarization induced by 10-s mol I -I ACh and the lack of effect of 10 -5 mol l -~ 5-HT on membrane potential. (c) Responses to l0 -~ mol 1-~ ACh after a 20 min washout following exposure to 5-HT with persistence of fast twitch activity. Upper scale applies to the two upper panels, lower scale to the two lower panels. In this and all subsequent figures W indicates washout.
An odd but consistent observation was that pre-treatment with F M R F a m i d e induced the RP muscle, which is normally insensitive to G T P and GTP-?-S, to respond to a subsequent dose of these purines. DISCUSSION The observation that 5-HT interacts with acetylcholine-induced responses to produce relaxation is not new, having been observed in the A B R M of Mytilus and the RP muscle of Busycon. Here, however, we examined the R R muscle in the sucrose gap and found that although 5-HT lowered tonic force and induced fast twitch activity this was not accompanied by any significant membrane potential
changes, suggesting that 5-HT action here was by a secondary messenger system. Cole and Twarog (1972) showed that 5-HT relaxation of catch in the Mytilus A B R M could be mimicked by the membrane permeant dibutyryl cAMP, suggesting that the 5-HT action was mediated via the adenylate cyclase/cAMP pathway. This view is strengthened by the observations of Kohler and Lindl (1980) in that same preparation that 5-HT between [10-6]-[10 -4] mol 1-~ greatly increased c A M P content in vitro. Our own experiments here showed that dibutyryl c A M P did indeed enhance the 5-HT-induced relaxation and fast twitch activity in all four proboscis muscles of Buccinum adding weight to the view that 5-HT operates via the adenylate cyclase/cAMP route. However there are a number of conflicting reports on the
1. D. NELSON and H. HUDDART
544
5-HT
ACh
W 4g
RR OR RS RP
30 mins 100pmol1-1 200s diB cAMP RR
_
~
OR RS
~
....
4g
RP 200s Fig. 2. Upper panel, responses to a control dose of 10-5 mol I-t ACh followed by 10-5 moll -~ 5-HT in the four proboscis muscles. Lower panel, same protocol as in upper panel following a 30 min exposure to 100 ttmol 1-~ dibutyryl cAMP. Note enhancement of ACh responses and enhancement of the 5-HT effect. Upper calibration applies to all traces in upper panel, lower calibration applies to all traces in lower panel.
relationship between 5-HT action a n d c A M P in o t h e r molluscan tissues. Whereas lshikawa et al. (1981) showed t h a t 5-HT dose-dependently elevated c A M P and reduced c G M P levels in Mytilus A B R M , M u r a k a m i et al. (1992) showed in t h a t same preparation that the relaxing responses mediated by the 5-HT2 receptor subtype were not linked to adenylate or guanylate cyclase systems. In contrast in Helix
neurones, Paupardin-Tritsch et al. (1986) showed that 5-HT increased Ca 2+ m e m b r a n e c o n d u c t a n c e but this was not mimicked by intracellular injection o f c A M P but was mimicked by intracellular injection of c G M P and zaprinast, a c G M P phosphodiesterase inhibitor. T o add further confusion, by using patch clamp analysis of Aplysia neurones with single channel recording, Siegelbaum et al. (1986) showed that
(Fig. 3 Opposite) Fig. 3. (a) Typical effects of 10-6 mol 1-~ FMRFamide on proboscis muscles. Upper scale applies to RR muscle trace, lower scale to the other three traces. (b) Mean FMRFamide dose responses curves for all four proboscis muscles. Note the sharp fall off of responses above 10-5 mol 1-~. (c) Typical sucrose gap electrical recording from the RR muscle in response to 10-6 mol 1-t FMRFamide, showing a depolarization of only c. 1 mV.
RP
RS
OR
RR
®
...-,
I
FMRFamide
120s
2g
4g
©
E..
4o
o
"4
M
®
Fig. 3--caption opposite.
!
W
0 10 - 9
20
40
60
80
100
4"
/
/
/
//
10-7
/,/
///
ff
FMRFamide
',,
4
"
'q
"+,
| W
'mY / 1
'
( t o o l 1-1 )
10-B
n=4,4,4,4,4,4,4
10-a
t
'" ," /~ ..."
....:,
F~lRlWamide C o n c e n t r a t i o n
10-0
T,','/
.:'
RP
*
OR RS
• •
RR
•
10-4
O ,--e
o
=_
o
c-
z
ACh
200s
FMRFamide
4g
!2g
,I
W
O
RP
RS
OR
RR
5-HT
1
FMRFamide
|
200s
g
W
2g
,I
4g
Fig. 4. (a) Interaction of ACh and FMRFamide responses. The response to a control dose o f 10 -5 tool 1-~ A C h is considerably enhanced by a subsequent dose of 10 -6 tool 1-~ FMRFamidc. Upper scale applies to the RR muscle, lower scale to the others. (b) Strong inhibition by 10 -5 mol 1-~ 5-HT of responses to 10 -6 mol 1-~ F M R F a m i d e . Upper scale applies to the R R muscle, lower scale to the others.
RP
RS
OR
RR
®
e~
Z
Z
ox
1
~
FMRFamide
~
200s
.
.
.
.
..
W
3g
4g
RP
200s
~
. . . .
..-~2.
= ' = ~ ~7. 2.-Y_31101_~='_, Z_:;_.~Z"_Z_~S77_.;-":2~Y.Y;Y'L'L~-
~
10 m i n 1 0 0 r a M L i 3÷
1
FMRFamide
RR
RP
RS
OR
RR
®
L
1
W
2g
4g
200S 200s Fig. 5. (a) Upper panel, control responses to 10 6 mol 1-~ FMRFamide, lower panel, FMRFamide responses after 30 min incubation with dibutyryl cAMP, showing no significant effect. In both panels the tension scale applies to all traces of the panel. (b) Upper panel, control responses to 10-6 tool 1-~ FMRFamide, lower panel, FMRFamide responses after 10 rain incubation with 100 mmol 1-~ lithium. Note the significant inhibition, but only at very high lithium concentration. Upper scale applies to the RR muscle of both panels, lower scale applies to the other muscles of both panels.
RP
RS
o.
RR
30 m i n s 1 0 0 u m o l 1-1 diB cAMP
aP
RS
OR
RR
®
o
=_.. ga
o
c-
z
RP
RS
OR
RR
0b
t2
t3
110-Smol 1-l GTP
+1
Fig.
&caption
4
WI
opposite.
0C
RP
Neuromodulation in molluscan muscle
10 mmoi I-I Guanosine
1
549
2g
RR
120s
lg
RS
Fig. 7. Typical inhibitory effects of guanosine on FMRFamide-treated RR and RS muscles, but only seen at the very high concentration of 10 mmol 1-~. 5-HT induced prolonged closure of the S potassium channel mediated by cAMP-dependent phosphorylation of a membrane protein which may be the channel itself. The relationship between 5-HT, ACh and contractile activation has been studied in suspensions of isolated cells from Mytilus ABRM which were fura-2 loaded to follow changes in [Ca]i (Ishii et al., 1989). These studies clearly showed that 5-HT suppressed the elevation of [Ca]i induced by carbachol, giving a direct explanation for the 5-HT relaxatory effect seen in the whole muscle. The view that 5-HT modulatory effects on ACh responses are mediated by a rise in cAMP and a fall in [Ca]i thus modifying ACh-induced electromechanical coupling are consistent with our results in Buccinum proboscis muscles. Only single channel recording of muscle by patch clamp will yield more detailed information about 5-HT and cAMP actions. The experiments here showed that Buccinum proboscis muscles were extremely sensitive to FMRFamide but without significant membrane potential
changes, and that FMRFamide responses were unaffected by dibutyryl cAMP but inhibited by lithium, albeit at high concentration. This suggests that FMRFamide action may be routed via an IP3 transduction pathway. Interaction between ACh and FMRFamide responses was additive and this suggests that FMRFamide, via intracellular IP3 release was able to access and release a pool of cellular Ca independently of that released by AChinduced depolarization, thus significantly enhancing fibre free Ca and promoting excition--contraction coupling, the reverse of the actions seen with 5-HT on ACh responses. In complete contrast, 5-HT inhibited FMRFamide responses in all proboscis muscles examined. If indeed FMRFamide activates IP 3 to release a distinct cellular Ca pool to induce classic pharmacomechanical coupling when 5-HT, via the cAMP pathway, must induce re-sequestration of [Ca]i to induce pharmacomechanical uncoupling as effectively as Ca-free conditions which uncouple ACh-induced electro-
(Fig. 6 Opposite) Fig. 6. (a) Responses of the RR, OR and RS muscles to GTP. Concentrations were I, 10-6moll-~; 2, 5 × 10-6 mol 1-~; 3, 10-5 mol I-~. Upper tension calibration applies to the RR muscle, lower calibration applies to the OR and RS muscle. (b) Muscle responses to GTP, note lack of response in the RP muscle. Upper tension scale applies to the RR, OR and RP, lower scale to the RS muscle. (c) Responses of proboscis muscles to GTP-y-S, again note lack of response of the RP muscle. (d) Sucrose gap recording of the RR muscle response to 5 × 10-5 mol 1-I GTP showing lack of action potential discharges during fast twitch activity.
550
I.D. NELSONand H. HUDDART
®
10 .5 m o l I -I G T P
W 1
10 -5 m o l I -I A C h
1
1
4g OR
RS
~
.....
....
2g
~ 120s
10 -5 m o l 1-I A C h
®
RR
1
10 -5 m o l 1-1 G T P
1
W 1 Eleclrical~ 1mV Mechanical
lg
120s Fig. 8. (a) Interaction of GTP and ACh in RR, OR and RS muscles, with enhancement of GTP responses caused by a subsequent exposure to ACh. Upper tension scale applies to the RR muscle, lower scale to the OR and RS muscle. (b) Sucrose gap recording showing that the membrane depolarization to ACh increases considerably by a subsequent exposure to 10-5 moll -~ GTP.
mechanical coupling (Huddart and Hill, 1988; Nelson and Huddart, 1992). None of the Buccinum proboscis muscles were responsive to either adenosine or ATP at concentrations up to 1 0 - 3 m o l l -~, as is also the case in Busycon proboscis muscles (Huddart, unpublished observations) and in Helix and Arion heart in what can be considered the normal (i.e. < 1 0 3 m o l l - l ) physiological range (Knight et al., 1992). However the RR, O R and RS muscles of Buccinum all showed strong dose-dependent responses to G T P and its
stable analogue GTP-7-S at low concentrations, as is also the case in Busycon RP muscle and atrium (Huddart, unpublished). Our Buecinum muscles were inhibited by guanosine, but only at the high non-physiological level of 10 mmol I -~. The cumulative dose-addition responses [Fig. 6(a)] were clearly suggestive of agonist/receptor interactions but with what type of recognition site is unknown, whether A 1, A 2 or the various P2 receptor subtypes (Williams and Cusack, 1990) or even the ACh receptor or another ion channel. The G T P responses were
Neuromodulation in molluscan muscle
10 -6
RR
tool 1-1 FMRFamide
10 -5 tool 1-1 GTP-y-S
...... j I"
551
W
~ 2g
OR
..... r~.
.
.
.
0.5g
RS
120s Fig. 9. Interaction of FMRFamide and GTP-y-S responses. FMRFamide-induced tension was considerably enhanced by subsequent treatment with GTP-y-S. Upper tension scale applied to the RR and OR muscles, the lower scale bar to the and RP muscles. Note that with combined FMRFamide and GTP-y-S treatment the normally unresponsive RP muscle develops modest force.
additive to ACh responses, and sucrose gap recordings showed that GTP at 10-Smoli - ' induced a modest membrane depolarization which was additive to that induced by ACh. This is suggestive that GTP may not have acted via a secondary messenger but may have interacted with a receptor or an ion channel to promote Na + influx and additionally raise [Ca]~ promoting electromechanical coupling. This point could only be resolved using voltage clamp analysis of the muscles with measurement of ACh-induced inward current in normal, in GTP-containing salines and when challenged with GTP in Na-free conditions. Unfortunately the unusual ACh receptor possessed by these muscles is unresponsive to the so called "usual" agonists/antagonists of the mammalian nicotinic and muscarinic AChR (Nelson and
Huddart, 1992), only mytoion appears to be partially effective in any predictable manner (Kohler and Lindl, 1980; Brooks et al., 1990). These muscles possess not just a complex innervation but also a novel interactive pharmacology. There appear to be multiple pathways for modulating the calcium released by classic ACh-promoted electromechanical coupling and hence several ways to modulate both force development and its pattern of expression during radular activity.
Acknowledgements--This work was carried out while IDN was in receipt of an SERC postgraduate research studentship. The authors are grateful to Professor R. B. Hill and Dr D. D. Brooks for much helpful discussion.
I. D. NELSONand H. HUDDART
552
REFERENCES Andrews P. L. R. and Tansey E. M. (1983) The digestive tract of Ocropus vulgaris: the anatomy, physiology and pharmacology of the upper tract. J. mar. biol. Ass. U.K. 63, 1099134. Berridge M. J. (1989) lnositol trisphosphate, calcium, lithium and cell signaling. J. Am. med. Ass. 262, 1834-1841. Brooks D. D. (1982) Membrane currents associated with contraction in molluscan visceral muscle. Ph.D. thesis, University of Lancaster. Brooks D. D.. Huddart H.. Lennard R. and Hill R. B. (1990) Calcium utilization in contractures induced by acetylcholine or high-potassium saline in molluscan proboscis muscles. Jy exp. Biol. 149, 379-394. Burnstock G. (1972) Purinergic nerves. Pharmac. Rev. 24. 5099581. _ Burnstock G. (1980) Purinergic nerves and receptors. Prog. Biochem. Pharmac. 16, 141-154. Burnstock G. (1981) Neurotransmitters and trophic factors in the autonomic nervous system. J. Physiol. 313, I-35. Cottrell G. A. (1989) The biology of the FMRFamide series of peptides in molluscs with special reference to Helix. Comp. Biochem. Physiol. 93A, 4145. Dorsett D. A., Skelton M. E. and Evans C. G. (1989) The innervation and physiology of the extrinsic buccal retractor muscles of Philine aperta (Linnaeus). J. molluscan Srud. 55, 193-208. Greenberg M. J. and Price D. A. (1979) FMRFamide, a cardioexcitatory peptide of molluscs; an agent in search of a mission. Am. Zool. 19, 1633174. Heyer C. B., Kater S. B. and Karlsson U. L. (1973) Neuromuscular systems in molluscs. Am. Zool. 13, 247-270. Hill R. B. and Langton P. D. (1988) Use of a sucrose gap apparatus to record electrical responses of gastropod radular protractor muscle to FMRFamide. Comp. Biothem. Pi?ysiol. 9OC, 2077214. Hill R. B. and Licis P. (1981), Effect of neurohumors during shutdown of electrogenic sodium pumping in a molluscan muscle. Adv. Physiol. Sci. 22, 339-360. Hill R. B. and Licis P. (1985) Lanthanum and caffeine affect both membrane response and force in a molluscan muscle, the radular protractor of Busycon canaliculatum. Comp. Biochem. Physiol. 82C, 363-376. Hill R. B. and Sanger J. W. (1974) Anatomy of the innervation and neuromuscular junctions of the radular protractor muscle of the whelk, Bucycon canaliculatum (L.) Biol. Bull. 147, 369-385. Hill R. B., Greenberg M. J., Irisawa H. and Nomura H. (1970) Electromechanical coupling in a molluscan muscle, the radula protractor of Busycon canaliculatum. J. exp. Zool. 174, 331-348. Hoyle C. V. and Greenberg M. J. (1988) Actions of adenyl compounds in invertebrates from several phyla; evidence for internal purinoreceptors. Comp. Biochem. Physiol. 9OC, 113-122. Huddart H. and Hill R. B. (1988) Electromechanical uncoupling in a molluscan smooth muscle examined by the sucrose gap technique. The effect of calcium antagonist and agonist agents. J. camp. Physiol. B 158, 501-512. Huddart H., Hunt S. and Oates K. (1977) Calcium movements during contraction in molluscan smooth muscle, and the loci of calcium binding and release. J. exp. Biol. 68, 45-56. Huddart H., Nelson I. D., Brooks D. D. and Hill R. B. (1992) Electrical and mechanical responses of the odontophore protractor muscle of Busycon canaliculatum: a sucrose gap voltage clamp study. Comp. Biochem. Physiol. 102C, 149-158.
Ishii N.,. Simpson A. W. M. and Ashley C. C. (1989) Effects of 5-hydroxytryptamine (serotonin) and forskolin on intracellular free calcium in isolated and furaloaded smooth muscle cells from the anterior byssus retractor (catch) Mytilus edulis. Pfugers Arch. 414, 162-170. Ishikawa T., Murakami H. and Iwayama Y. (1981) Changes in CAMP and cGMP levels induced by relaxing drugs in ACh and K treated molluscan smooth muscle fibres. Comp. Biochem. Physiol. 7OC, 171Ll76. Kerkut G. A. and Cottrell G. A. (1963) Acetylcholine and 5-hydroxytryptamine in the snail brain. Comp. Biochem. Physiol. 8, 5363. Kerkut G. A. and Laverack M. S. (1960) A cardio-accelerator present in tissue extracts of the snail Helix aspersa. Comp. Biochem. Physiol. 1, 62-71. Knight G. E., Hoyle C. H. V. and Burnstock G. E. (1992) Effects of nucleosides and nucleotides on the isolated hearts of the snail Helix aspersa and the slug Arion ater. Comp. Biochem. Physiol. lOlC, 175-181. Kohler G. and Lindl T. (1980) Effects of 5-HT. dopamine and acetylcholine on accumulation of CAMP and cGMP in the anterior byssus retractor muscle of Mytilus edulis L. (Mollusca). Pflugers Arch. 383, 257-262. Murakami H., Kizawa Y., Sano M., Edamura N., Maruyama C. and Yamazaki A. (1992) Possible site of action of 2-methylserotonin in inducing relaxation of acetylchohne-induced contraction in the molluscan (Mytilus edulis) smooth muscle. Comp. Biochem. Physiol. lOlC, 3433347. Nagle G. T. and Greenberg M. J. (1982) A highly sensitive microbioassay for the molluscan neuropeptide FMRFamide. Comp. Biochem. Physiol. 71C, 101-105. Nelson I. D. and Huddart H. (1992) The nature of the acetylcholine receptor in a Buccinum proboscis muscle examined by the sucrose-gap voltage clamp technique. Gen. Pharmac. 23, 317-323. Nelson I. D. and Huddart H. (1994) The fine structure of a molluscan muscle, the radular retractor of Buccinum undatum in relation to excitation-contraction coupling. Submitted. Nystrom R. A. (1967) Spontaneous activity of clam intestinal muscle. Comp. Biochem. Physiol. 21, 601-610. Paupardin-Tritsch D., Hammond C. and Gerschenfeld H. M. (1986) Serotonin and cyclic GMP both induce an increase of the calcium current in the same identified molluscan neurones. J. Neurosci. 6, 2715-2723. Price D. A. and Greenberg M. J. (1977) Structure of a molluscan cardioexcitatory neuropeptide. Science 197, 67067 I. Price D. A. and Greenberg M. J. (1979) Pharmacology of the molluscan cardioexcitatory neuropeptide FMRFamide. Gen. Pharmac. 11, 237-241. Siegelbaum S. A., Belardetti F., Camardo J. S. and Shuster M. J. (1986) Modulation of the serotonin-sensitive potassium channel in Aplysia sensory neurone cell body and growth cone. J. exp. Biol. 124, 2877306. Welsh J. H. and Moorhead M. (1959) Identification and assay of 5-hydroxytryptamine in molluscan tissues by fluorescence method. Science 129, 1491-1492. Williams M. and Cusack N. J. (1990) Neuromodulatory roles of purine nucleosides and nucleotides: their receptors and ligands. Neurotransmissions 6, l-6. Yanagawa M., Fujiwara M., Takabatake I.. Muneoka Y. and Kobayashi M. (1988) Potentiating effects of some invertebrate neuropeptides on twitch contraction of the radula muscles of a mollusc, Rapana thomasiana. Comp. Biochem. Physiol. WC, 73-77. Zoran M. J., Haydon P. G. and Matthews P. J. (1989) Aminergic and peptidergic modulation of motor function at an identified neuromuscular junction in Helisoma. J. exp. Biol. 142, 2255243.