324

Brain Research, 556 (1991) 324-328 © 1991 Elsevier Science Publishers B.V. All rights reserved. 0006-8993/91/$03.50 ADONIS 000689939124781G

BRES 24781

The role of pacemaker properties and synaptic input in generation and modulation of spiking activity in a pair of electrically coupled peptidergic neurons W.C. Wildering, C. Janse and T.A. de Vlieger Department of Biology, Vrije UniversiteitAmsterdam, Amsterdam (The Netherlands) (Accepted 30 April 1991)

Key words: Electrical synapse; Spike synchronization; Mollusc; Pacemaker; Neuronal oscillator; Identified neuron; Lymnaea stagnalis The origin of patterned electrical activity in two electrotonically coupled peptidergic neurons, VD~ and RPD2, in the CNS of Lymnaea stagnalis was investigated. VD 1proved to have intrinsic beating pacemaker properties. Hybrid current/voltage clamp experiments demonstrated that in the intact CNS generation of spike activity in the coupled cell system is dominated by VD1. Modulation of spiking activity of VD~/RPD2 appears to originate mainly from chemical synaptic input. The electrical coupling of VD 1 and RPD2 proved essential for spike synchronization between the cells. Many different types of neurons and neuronal circuits are able to generate a rhythmic output of action potentials in the absence of synaptic input. Endogenous rhythmic spiking activity may arise either from intrinsic membrane properties of individual neurons or from special patterns of connectivity within a neuronal network 1'4'7-9'11A8. A feature frequently encountered in oscillatory neuronal networks is the electrical synapse 5' 8,20,22 In addition to effects on temporal and spatial integrative properties 3'4"7, electrical synapses are thought to contribute to modulation of the oscillatory behavior of some electrically coupled neuronal systems4'7'14"16. The present paper is concerned with a neuron system in the central nervous system (CNS) of Lymnaea stagnalis, consisting of 2 peptidergic cells. These cells, called VD 1 and RPD2, are electrically coupled and fire in dose synchrony 2'13"23. The VD1/RPD 2 system receives excitatory and inhibitory synaptic input ~2'21. Histological and immunocytochemical studies 15 as well as electrophysiological studies ~2'2~ indicate that the V D J R P D 2 system is involved in modulation of cardiorespiratory functions. Peptides produced by VD 1 and RPD 2 share considerable homology with the Aplysia R15 peptides 6. We focussed on the origins of spike generation and modulation of spiking activity in the VD1/RPD 2 system. Therefore, the individual electrophysiological properties of VD~ and RPD2 and their contribution to the system's activity pattern in the intact isolated CNS were investigated. In addition, possible involvement of the electrical coupling and chemical synaptic input in modulation of

the system's spiking pattern was considered. Experiments were performed both using isolated CNS preparations and isolated somata of VD 1 and RPD 2. Experimental animals were raised under standard conditions 19. Unless mentioned otherwise a HEPES-buffered saline was used (HBS; 30 mM NaCI, 1.7 mM KCI, 10.0 mM NaCHaSO 4, 1.5 mM MgCI2, 4.0 mM CaCI 2, 5.0 mM N a H C O 3, 10 mM H E P E S adjusted to p H 7.8 with 3 M NaOH). In some cases chemical synaptic activity was blocked with low Ca2+]high Mg 2+ saline (0 mM Ca 2+, 4 mM Mg2÷-HBS). Measurements were made with 2 discontinuous single electrode current/voltage clamp amplifiers (SEVC, sampling frequency 1.5-3.5 kHz, for details concerning preparations and operation of the SEVC see reference 23). Both cells were impaled with glass microelectrodes (Clark GF 150-15) filled with 0.5 M KCI (impedance 15-30 MI~). Isolated V D 1 and RPD 2 were obtained as follows (modified from Moed et alj7): the isolated CNS was incubated for 30 min in HBS containing 0.5 mg/ml Trypsin (Sigma bovine pancreas type IH) at a temperature of 37 °C. Subsequently, the CNS was rinsed several times in HBS. Connective tissue and perineurium were removed, VD 1 and RPD 2 were isolated from the CNS by gently sucking them into glass microtubes. The cells were transferred to a culture dish (Costar 3035) containing HBS. The normal activity pattern of the VD1/RPD 2 system in the isolated CNS is shown in Fig. 1A. Both excitatory as well as inhibitory modulatory influences affect the spike interval (median spike interval 0.7-1.4 s). VD 1 and

Correspondence: C. Janse, Department of Biology, Vrije Universiteit Amsterdam, P.O. Box 7161, 1007 MC Amsterdam, The Netherlands.

325 R P D 2 fire their spikes in a close 1:1 synchrony (Fig. 1B). In a large majority of the preparations VD 1 leads the spike sequence. Spikes in R P D 2 follow within a very short delay (range 4-15 ms; see Fig. 2C). V D 1 and R P D 2 are electrotonicaUy coupled to each other. Figure 1B illustrates the hyperpolarizing current spread in each direction of the VD1/RPD 2 network upon current injec-

tion in one of the cells. Previous studies showed that the electrical synapse of VD1/RPD 2 is electrically linear in both directions 2'23. Reciprocal electrical interactions between V D 1 and R P D 2 may thus occur. In order to investigate the properties of V D t and R P D 2 separately, the electrical activity of single isolated neurons was recorded. In nearly all cases (18 out of 20),

A VD-I

RPD2

__l

40 mV

10 sec

B

VD1

RPD2

20 mV 500 msec

C ....

RPD2

J 50mV 10 sec

Fig. 1. A: simultaneously recorded spontaneous spiking activity of VD 1 and R P D 2 in the isolated central nervous system. B: voltage recordings of simultaneous spiking activity of VD 1 and RPD2 drawn on larger timescale. Note the synchronized spiking activity and electrical coupling between VD1 and RPD 2 (arrow indicates current injection in either VD 1 or RPD2). The DC-coupling between VD~ and RPD2 is nearly equal in both directions. C: spiking activity of isolated VD t and RPD2. VD~ spontaneously develops a regular beating type of spiking activity with a frequency of 0.7-1.2 Hz. RPD 2 is mostly silent, but sometimes generates irregular low frequency spike patterns.

326 isolated V D 1 showed a very regular spike pattern (Fig. 1C; u p p e r trace). M e a n spike frequency of isolated V D 1 ranged between 0.7 and 1.2 Hz. Interspike m e m b r a n e potential of VD~ ranged between - 4 5 and - 5 3 mV. In contrast, regular spiking activity was never o b s e r v e d in isolated R P D 2 (n = 12). In most cases, this cell was silent and h a d a m e m b r a n e potential lower than - 6 0 mV. Only in a few instances (3 out of 12), one of t h e m being shown in Fig. 1B, isolated R P D 2 p r o d u c e d irregular spiking activity. A l t h o u g h differential susceptibility of the 2 neurons to d a m a g e inflicted by the isolation p r o c e d u r e

Control

cannot be totally excluded as a cause of the differences in intrinsic activity, these results suggest that unlike RPD2, V D 1 can be r e g a r d e d as an intrinsic p a c e m a k e r of the beating type. A s will be shown below, hybrid current/voltage clamp studies p e r f o r m e d on V D 1 and RPD2 in the intact CNS affirmed this view (see below). Considering the differences in s p o n t a n e o u s spiking b e t w e e n isolated V D 1 and R P D 2, it is conceivable that in the intact CNS, VDx might act as a driver and R P D 2 as a follower neuron. In addition, however, in the isolated CNS the activity p a t t e r n of V D 1 and R P D 2 could also be

VD1 = master

RPD2

A

i

VD1

RPD2

= master

illiiNiIIIi

'

i

40 mV 10 sec

B

FI '11 r,, VD1

RPD2

tll111tt1111111t d IHHHIttH

40 mV 10 sec

C VD1

RPD2

40 mV 20 msec

Fig. 2. Simultaneous voltage recordings in current clamp conditions (left panels) and in hybrid current/voltage clamp conditions (middle and fight panels) of VD 1 and R P D 2 in the intact isolated CNS. A: experiments performed in standard saline. B: experiments performed in low Ca2+/high Mg2÷ saline in order to block synaptic activity. C: sections of the voltage traces from part B, drawn on a larger timescale. Note the absence of phase differences between spikes in VD 1 and RPD 2 in the VDl-enforced situation as opposed to the control situation. In contrast to forced transfer of voltage control in the VD1/RPD 2 system to VD 1 only (VD~ = master), the transfer of voltage control to RPD 2 (RPD 2 = master) completely abolished the spike pattern normally observed in VD~/RPD 2 (control). Minor differences in transient amplitudes of both cells under hybrid current/voltage clamp conditions are caused by differences in bandwidth of the SEVCs voltage and current clamp circuits. As is shown in Fig. 2C, however, the phase lock between spikes under hybrid current/voltage clamp was not discredited by these differences.

327 influenced by factors such as reciprocal electrotonic interactions between the cells, or synaptic input on either one of the cells. To investigate these aspects, a hybrid current/voltage clamp technique was used in which the membrane potential of one of the cells (master) was recorded under current clamp conditions and forced upon the other one (slave) by means of voltage clamp. In this way, the transjunctional potential difference is permanently held at zero and consequently no current flows across the junction. The technique enabled us to shortcut the electrical interactions between the cells and to assign control of the electrogenesis of the VD1/RPD 2 system in the intact CNS to either one of the cells alone. Under these conditions any change in the electrical activity of VD1/RPD 2 can be attributed to the cell put in charge of the network (-- master). Results of these experiments (n = 10) are represented by an example shown in Fig. 2. With both cells recorded under current clamp conditions, VD1/RPD 2 produced the familiar spiking activity presented previously in Fig. 1A. Transfer of the voltage command to VD 1 under hybrid current/voltage clamp conditions did not introduce large changes in the firing pattern of VDI/RPD 2 except for a slight increase in median spike frequency (compare Fig. 2A, control and VD 1 = master). Note that under these conditions the delay between VD 1 and RPD 2 is absent in contrast to the control situation (compare Fig. 2C left and middle panels). Transfer of the voltage command to RPD2, however, dramatically changed the spiking activity (compare Fig. 2A, control and RPD 2 = master). In standard HBS, the membrane potential dropped to values o f - 6 0 mV and lower (mean + S.E.M. = -68 + 0.9 mV). Several rapid electrotonic potentials were seen, which probably should be interpreted as excitatory postsynaptic potentials (EPSP) since they remained absent under low Ca2+/high Mg 2+ conditions (Fig. 2B). Spiking activity of the system became highly irregular. RPD 2 initiated spikes every now and then, apparently as a result of summating EPSPs. Moreover, it is important to note that despite several EPSPs were seen under these conditions, we never observed repetitive excitatory inputs on RPD 2 which could explain the approximately 1 Hz firing pattern of the intact preparations. It seems unlikely therefore that repetitive excitatory synaptic inputs operating synchronously on both VD~ and RPD 2 contribute to the spike synchrony of both cells. Thus we conclude that the electrical synapse between VD x and RPD 2 is very likely the only mediator of spike synchrony in the network. Chemical synaptic inputs on VD~/RPD 2 apparently only modulate the basic firing pattern. In order to investigate the effects of synaptic input on the firing activity of VD1/RPD 2 we minimized synaptic activity by replacement of HBS with low Ca2+/high Mg 2÷

HBS. Under these conditions excitatory and inhibitory modulations disappeared from the spiking pattern in both the control situation as well as with VD 1 in forced control of the system (Fig. 2B, left and central panels). Spike generation was apparently unaffected under these conditions. When under the same conditions the systems voltage control was transferred to RPD2, spiking activity was completely abolished (compare Fig. 2B, control and RPD 2 = master), neither was there any sign of rapid synaptic potentials previously observed in normal saline (see Fig. 2A). Together, these results are consistent with the idea that in the CNS, the spiking activity of the VD1/RPD 2 system is driven by VD 1. Deprived of both its chemical synaptic inputs and its electrical communication with VD~, RPD 2 proved incapable of driving the V D 1 / R P D 2 system in an oscillatory firing pattern. Therefore, we conclude that, in the absence of synaptic inputs, the firing pattern of RPD 2 is dictated by the beating spiking activity generated by VD 1. Even in the presence of synaptic activity, however, the firing pattern controlled by RPD 2 is completely different from the normal rhythmic firing pattern of VD1/RPD 2. Thus the pacemaker activity developed by VD~ is an essential element of the normal spiking activity of the VD1/RPD 2 network. The fact that blockade of chemical synaptic transmission prevented modulation of the VD~/RPD2 activity pattern, indicates that synaptic inputs rather than intrinsic mechanisms are involved in modulation of the spiking activity of VDI/RPD 2. However, many of the membrane currents involved in oscillatory and bursting electrical activity of molluscan neurons are Ca2+-dependent L~°. The role of intrinsic membrane mechanisms in modulation of the VD~/RPD 2 spiking activity can therefore not be excluded entirely. Since isolated VD~ somata under normal saline conditions invariably developed an unmodulated steady beating activity (Fig. 1A) the involvement of intrinsic mechanisms in short-term modulation of the spiking activity seems, however, not very likely. With respect to the role of the electrical coupling between VD 1 and RPD 2 in generation and modulation of the VD1/RPD 2 activity pattern there are some additional observations worth noting. From comparison of left and middle panels of Fig. 2 it can be seen that shortcutting the electrical connection between the cells by activation of the hybrid current/voltage clamp, increases the average spike frequency. Apparently, the pacemaker cycle in VD 1 is slowed down when RPD 2 is allowed electrical interaction with V D 1. This phenomenon could be related to loading effects of RPD 2 on the spike generating currents in V D 1. It is conceivable that the precise effects of transjunctional loading on spiking activity of the V D 1 / R P D 2 network as a whole, depends on the phase

328 differences between V D 1 and R P D 2. Getting and Willows7 provided evidence that shifts in the phase difference between neurons in an electrically coupled network can be instrumental in the burst behavior of the network. In a preliminary study (unpublished), we o b t a i n e d indications that variations occurring in the delay between spike pairs in V D 1 and R P D 2 might indeed correlate with changes in the network's spike frequency. In addition, the differences in electrophysiological p r o p e r t i e s of V D 1 and R P D 2 established so far, gain special relevance when the study of K e p l e r et al. ~4 is considered. These authors investigated the interactions of intrinsic p a c e m a k e r properties, electrical coupling and spike dynamics in a numerical m o d e l consisting of one p a c e m a k e r and one follower neuron. T h e y pointed out that in such an oscillatory neuronal network, complex interactions between these components rather than simple loading effects might affect the networks oscillatory p e r i o d and burst behavior. Their model predicts that synaptic inputs m a y have complex effects on the firing

pattern of such a network. Considering the evidence given above, the VD1/RPD 2 system m a y be considered as a physiological c o u n t e r p a r t of the numerical m o d e l p r o p o s e d by K e p l e r et al. T h e V D 1 / R P D 2 system thus seems to provide an attractive physiological m o d e l to study the interaction of intrinsic p a c e m a k e r properties, electrical coupling, dynamic aspects of action potentials and chemical synaptic input. We have i n d e e d indications that ontogenetical changes in the impulse transmission properties of the electrical synapse contribute to a changed susceptibility of the network to burst-inducing chemical transmitters.

1 Adams, W.B. and Benson, J.A., The generation and modulation of endogenous rhythmicity in the Aplysia bursting pacemaker neurone R15, Prog. Biophys. Molec. Biol., 46 (1985) 1-49. 2 Benjamin, P.R. and Pilkington, J.B., The electrotonic location of low-resistance intercellular junctions between a pair of giant neurones in the snail Lymnaea stagnalis, J. Physiol., 370 (1986) 11-126. 3 Bennet, M.V.L., Electrical transmission: a functional analysis and comparison to chemical transmission. In E.R. Kandel (Ed.), Handbook of Physiology, 1Iol. 1, American Physiological Society, Bethesda, MD, 1977, pp. 357-416. 4 Berry, M.S. and Pentreath, V.W., The integrative properties of electrotonic synapses, Comp. Biochem. Physiol., A 57 (1977) 289-295. 5 Blankenship, J.E. and Haskins, J.T., Electrotonic coupling among neuroendocrine cells in Aplysia, J. Neurophysiol., 42 (1979) 347-355. 6 Bogerd, J., Geraerts, W.P.M., Van Heerikhuizen, H., Kerkhoven, R.M. and Joosse, J., Characterization and evolutionary aspects of a prohormone encoding a transcript from Lymnaea neurons, VD 1 and RPD2, Mol. Brain Res., in press. 7 Getting, P.A. and Willows, A.O.D., Modification of neuron properties by electrotonic synapses. II. Burst formation by electrotonic synapses, J. Neurophysiol., 37 (1974) 858-868. 8 Getting, P.A., Comparative analysis of invertebrate central pattern generators. In A.H. Cohen, S. Rossignol and S. Grillner (Eds.), Neural Control of Rhythmic Movements in Vertebrates, Wiley, New York, 1988, pp. 101-127. 9 Getting, P.A., Emerging principles governing the operation of neural networks, Annu. Rev. Neurosci., 12 (1989) 185-204. 10 Gorman, A.L.E and Hermann, A., Quantitative differences in the currents of bursting and beating molluscan pacemaker neurones, J. Physiol., 333 (1982) 681-699. 11 Harris-Warrick, R.M., Chemical modulation of central pattern generators. In A.H. Cohen, S. Rossignol and S. Griliner (Eds.), Neural Control of Rhythmic Movements in Vertebrates, Wiley, New York, 1988, pp. 285-331. 12 Janse, C., van der Wilt, G.J., van der Plas, J. and van der Roest, M., Central and peripheral neurones in oxygen perception in the pulmonate snail Lymnaea stagnalis (Mollusca, Gastropoda), Comp. Biochem. Physiol., A 82 (1985) 459-469. 13 Janse, C., van der Roest, M. and Slob, W., Age-related decrease

in electrical coupling of two identified neurons in the mollusc Lymnaea stagnalis, Brain Research, 376 (1986) 208-212. 14 Kepler, T.B., Marder, E. and Abbot, L.E, The effect of electrical coupling on the frequency of neuronal oscillators, Science, 248 (1990) 83-85. 15 Kerkhoven, R.M., Croll, R.P., van Miunen, J., Bogerd, J., Ramkema, M.D., Lodder, H. and Boer, H.H., Axonal mapping of the giant peptidergic neurons VD 1 and RPD 2 located in the CNS of the Pond Snail Lymnaea stagnalis, with particular reference to the innervation of the Auricle of the Heart, Brain Research, (1991) in press. 16 Ter Maat, A., Roubos, E.W., Lodder, J.C. and Buma, P., Integration of biphasic synaptic input by electrotonicallycoupled neuroendocrine caudodorsal cells in the Pond Snail, J. Neurophys., 49 (1983) 1392-1409. 17 Moed, P.J., Bos, N.P.A. and ter Maat, A., Morphology and electrophysiologicalcharacteristics of candodorsal cells of Lymnaea stagnalis in dissociated cell culture, Comp. Biochem. Physiol., A 92 (1989) 445-453. 18 Selverston, A.I. and Moulins, M., Oscillatory neuronal networks, Ann. Rev. Physiol., 47 (1985) 29-48. 19 Van der Steen, W.J., van den Hoven, N.P. and Jager, J.C., A method for breeding and studying freshwater snails under continuous water change, with some remarks on growth and reproduction in Lymnaea stagnalis (L.), Neth. J. Zool., 19 (1969) 131-139. 20 Van Swigchem, H., On the endogenous bursting properties of 'light yellow' neurosecretory cells in the freshwater snail Lymnaea stagnalis (L.), ./. Exp. Biol., 80 (1979) 55-67. 21 Van tier Wilt, G.J., van der Roest, M. and Jause, C., Neuronal substrates of respiratory behavior and related functions in Lymnaea stagnalis. In H.H. Boer, W.P.M. Geraerts and J. Joosse (Eds.), Neurobiology, Molluscan Models, North-Holland, Amsterdam, 1987. 22 De Vlieger, T.A., Kits, K.S., ter Maat, A. and Lodder, J.C., Morphology and electrophysiology of the ovulation hormone producing neuro-endoerine cells of the freshwater snail Lymnaea stagnalis (L.), J. Exp. Biol., 84 (1980) 259-271. 23 Wildering, W.C., van der Roest, M., de Vlieger, T.A. and Janse, C., Age-related changes in junctional and non-junctional conductance in two electrically coupled peptidergic neurons of the mollusc Lymnaea stagnalis, Brain Research, 547 (1991) 89-98.

W.C.W.'s work is supported by the Foundation for Fundamental Biological Research (BION), which is subsidized by the Netherlands Organization for the Advancement of Research (NWO). The authors thank Drs. R.F. Jansen, K.S. Kits, A. ter Maat and other members of the neurophysiological unit of the Faculty of Biology, Vrije Universiteit Amsterdam, for critically reading the manuscript.