- Email: [email protected]
Synaptic Gating: The Potential to Open Closed Doors
Current Biology, Vol. 13, R554–R556, July 15, 2003, ©2003 Elsevier Science Ltd. All rights reserved. DOI 10.1016/S0960-9822(03)00471-8
Synaptic Gating: The Potential to Open Closed Doors Paul S. Katz
Synaptic gating is normally thought to be a mechanism for excluding synaptic input, but three recent studies show how the resting membrane potential interacts with integrative properties to act as a permissive synaptic gate.
“When one door closes another door opens; but we often look so long and so regretfully upon the closed door, that we do not see the ones which open for us.” Alexander Graham Bell
The ability of neural circuits to ‘gate’ inputs by suppressing inappropriate synapses is well known, but neural circuits also can open gates that are closed and allow signals to enter. Primary afferent depolarization is a classic example of synaptic gating, whereby central locomotor circuits inhibit the release of neurotransmitter from sensory afferents and thereby prevent sensory input from arriving during inappropriate phases of the locomotor cycle [1,2]. Sensory gating also has been proposed as a way of focussing attention during cognitive processing by suppressing inputs [3]. Recent papers [4–6] suggest that, at some synapses, the figurative ‘gate’ is normally closed but can be opened by depolarization of the membrane potential. These studies show that, besides excluding inputs, synaptic gating also can act in a permissive fashion, and that the gating can result from the integrative properties of the neurons rather than simply through the actions of presynaptic receptors. Permissive synaptic gating differs from other types of synaptic change that potentiate synapses. For example, long-term potentiation is thought to play a role in Hebbian-type plasticity, increasing the efficacy of connections by unmasking silent synapses or enhancing active synapses [7,8]. These learningrelated changes help set up networks for future use, but do not gate the input on a moment-to-moment basis. Similarly, forms of homosynaptic short-term plasticity such as synaptic facilitation and augmentation can increase synaptic strength [9], but are not thought of as gating the synaptic input. These types of change do not serve a gating function, because they are entirely dependent upon activity of the presynaptic neuron. Implicit in the notion of synaptic gating is the concept of a ‘gatekeeper’, whereby gating involves the momentary control of synaptic efficacy by a neuron other than the neuron being gated.
Department of Biology, Georgia State University, MSC 8L0389, 33 Gilmer St. SE Unit 8, Atlanta, Georgia 30303-3088, USA. E-mail: [email protected]
Dispatch
Neuromodulatory systems can act as gatekeepers, selectively enhancing synaptic inputs [10]. There are many known examples of presynaptic receptors that enhance neurotransmitter release. For example, studies on the sea slug Aplysia californica have shown that serotonin enhances neurotransmitter release from sensory neuron terminals [11]. Recent work with a related mollusc, Tritonia diomedea, has demonstrated that serotonergic neurons intrinsic to a central pattern generator (CPG) enhance release of neurotransmitter from other members of the same circuit by acting at presynaptic G-protein-coupled receptors [12,13]. This so-called ‘intrinsic neuromodulation’ can be thought of as an internal gate whereby synapses within the circuit are selectively enhanced when the circuit is in use [14]. Neurotransmitter release can also be enhanced by presynaptic ionotropic receptors, which momentarily depolarize the presynaptic terminal or allow calcium entry that facilitates release [15,16]. Thus, synapses can be permissively gated through the actions of presynaptic receptors. Recent papers by Ivanov and Calabrese [4], Evans et al. [5] and Herberholz et al. [6] propose mechanisms for permissive gating that are based upon the integrative electrophysiological properties of the neurons involved, rather than simply the actions of presynaptic receptors. In all three cases, the resting membrane potential has permissive gating effects on synapses that result in functional consequences for the neural networks. Although a depolarization of the membrane potential is required for these gating mechanisms, they are more complicated than simple synaptic summation. Ivanov and Calabrese [4] found that spike-mediated synaptic release in leech heart interneurons can be permissively gated by membrane potential oscillations acting on low threshold Ca2+ channels. The low threshold Ca2+ channels in the presynaptic neuron cause a background level of intracellular Ca2+ that determines whether an action potential will evoke an inhibitory postsynaptic current (IPSC) in the postsynaptic neuron. Calcium imaging showed that the rise in intracellular Ca2+ was correlated with the rise in synaptic strength (Figure 1A). Furthermore, uncaging a Ca2+ chelator or Ca2+ itself showed that the rise in intracellular Ca2+ was both necessary and sufficient for the membrane potential depolarization to increase synaptic strength. Thus, when the membrane potential of the neuron is sitting at rest, action potentials are ineffective at causing transmitter release; the synapses must be permissively gated by voltage to be effective. The function of the permissive gating examined by Ivanov and Calabrese [4] appears to be related to the pattern-generating role of the neurons, which form a half-center oscillator as part of the leech heartbeat CPG. The mechanism assures that spikes occurring
Current Biology R555
A
B ∆F/F
C a
a
a SN2
b
b
SN1
i
i
b
B21
c
R-HN
c d L-HN
B8
c
i
LG
Current Biology
Figure 1. Mechanisms by which membrane potential permissively gates synapses. (A) In the leech, the right heart interneuron (R-HN) makes an inhibitory synapse on the left heart interneuron (L-HN). No synaptic current is recorded in a voltage-clamped L-HN when a spike is elicited in R-HN from its resting membrane potential (light gray box). But when the membrane potential of R-HN is depolarized by current injection (i), then inhibitory post-synaptic currents (IPSCs) can be recorded in L-HN (dark gray box). The size of the IPSC corresponds to the level of intracellular calcium, as determined by the increase in fluorescence (∆F/F) of a Ca2+-sensitive dye (trace a). (B) In the Aplysia feeding system, the mechanosensory neuron, B21 makes excitatory connections on neuron B8. A spike initiated by mechanical stimulation (trace a) fails to invade the central branches of B21 and thus fails to evoke an excitatory postsynaptic potential (EPSP) in B8 (light gray box). Depolarization of the B21 soma by current injection (i) allows the spike to invade the central branches and evoke an EPSP in B8 (dark gray box). (C) In the crayfish terminal abdominal ganglion, sensory neurons (here labeled SN1, SN2) make rectifying electrical synapses with the lateral giant (LG) neuron and non-rectifying electrical synapses with each other. Stimulation of SN1 evokes a small EPSP in LG (light gray box). But when LG is depolarized by current injection (i), an action potential in SN1 recruits action potentials in nearby sensory neurons, such as SN2. This increases the apparent synaptic efficacy of SN1 and causes a larger EPSP in LG (dark gray box).
outside of the correct heartbeat phase will be ineffective. But more importantly, the efficacy of the synapse will be controlled by the very low threshold Ca2+ channels that play a critical role in setting the strength and duration of the depolarizing phase of the oscillation [17]. Thus, the gating mechanism elegantly matches the synaptic strength to the phase of the cycle at which it would be most advantageous to inhibit the postsynaptic partner in the half-center oscillator. A different role for voltage regulation of spikemediated synaptic transmission was uncovered by Evans et al. [5] examining synapses made by a mechanoafferent in Aplysia. The neuron, known as B21, fires action potentials in response to mechanical stimulation of a mouthpart tissue [18]. It was previously observed that, if the membrane potential of B21 is at rest when the neuron is activated by a sensory stimulus, then it fails to evoke a postsynaptic potential in a follower neuron, but if the membrane potential of B21 is depolarized from rest by current injection into the soma or by synaptic input, then the synapse is unsilenced [18,19] (Figure 1B). The mechanism underlying this permissive gating of B21 synaptic output is based on the failure of action potentials to invade central axon branches; depolarization of the cell soma relieves the conduction block and allows the action potential to invade the synaptic output region of the neuron [5] (Figure 1B). As in the case of the leech heartbeat interneurons, the membrane potential of B21 in Aplysia is rhythmically depolarized at a particular phase of a rhythmic motor behavior. In this case, B29 is depolarized as a result of synaptic excitation by interneurons in the feeding CPG. Thus, the feeding motor pattern permissively gates sensory input through B21; mechanosensory input would be effective at evoking a postsynaptic potential only during a particular phase of the motor pattern.
A third example of synaptic input being gated by voltage was demonstrated by Herberholz et al. [6] in the crayfish ‘tailflip escape’ system [20]. Mechanosensory neurons that respond to touch of the crayfish tailfan make rectifying electrical synapses with the lateral giant (LG) neuron. It was found that the afferents are themselves electrically coupled (Figure 1C), allowing stimulated sensory neurons to recruit other sensory neurons. This ‘lateral excitation’ is gated by the membrane potential of LG: when LG is depolarized, action potentials in some afferent axons will trigger spikes in neighboring sensory axons (Figure 1C). A computer simulation study suggests that the mechanism underlying the recruitment of unstimulated sensory axons is related to the voltage-sensitivity and kinetics of rectifying electrical connections to LG [6]; current flows antidromically from the depolarized LG to the afferents, and at the same time the rectifying electrical synapses are back-biased, effectively increasing the input resistance of the afferents and making it easier for them to be recruited by neighboring axons. The net effect of this mechanism is that the membrane potential of LG determines whether mechanosensory input will effectively excite LG and elicit a tailflip response. Thus, LG gates its own synaptic input by causing stimulated sensory neurons to spread excitation to other afferent axons. Although synaptic gating is commonly understood to mean closing a gate, gates can also be opened. These three examples show that membrane potential can act through very different mechanisms to permissively gate synaptic transmission. They illustrate how important it is to understand the integrative properties of neurons rather than assume that an action potential reliably leads to neurotransmitter release. Thus, in thinking about synaptic gating, it is important to remember that while one door may close, another may open.
Dispatch R556
References 1. Rudomin P. (1999). Presynaptic selection of afferent inflow in the spinal cord. J. Physiol. (Paris) 93, 329-347. 2. Büschges A., El Manira A. (1998). Sensory pathways and their modulation in the control of locomotion. Curr. Opin. Neurobiol. 8, 733739. 3. Castro-Alamancos M.A., Oldford E. (2002). Cortical sensory suppression during arousal is due to the activity-dependent depression of thalamocortical synapses. J. Physiol. 541, 319-331. 4. Ivanov A.I., Calabrese R.L. (2003). Modulation of spike-mediated synaptic transmission by presynaptic background Ca2+ in leech heart interneurons. J. Neurosci. 23, 1206-1218. 5. Evans C.G., Jing J., Rosen S.C., Cropper E.C. (2003). Regulation of spike initiation and propagation in an Aplysia sensory neuron: gating-in via central depolarization. J. Neurosci. 23, 2920-2931. 6. Herberholz J., Antonsen B.L., Edwards D.H. (2002). A lateral excitatory network in the escape circuit of crayfish. J. Neurosci. 22, 90789085. 7. Montgomery J.M., Pavlidis P., Madison D.V. (2001). Pair recordings reveal all-silent synaptic connections and the postsynaptic expression of long-term potentiation. Neuron 29, 691-701. 8. Poncer J.C., Malinow R. (2001). Postsynaptic conversion of silent synapses during LTP affects synaptic gain and transmission dynamics. Nat. Neurosci. 4, 989-996. 9. Zucker R.S., Regehr W.G. (2002). Short-term synaptic plasticity. Annu. Rev. Physiol. 64, 355-405. 10. Katz P.S. (1999). Beyond Neurotransmission: Neuromodulation and its Importance for Information Processing. (Oxford: Oxford University Press). 11. Byrne J.H., Kandel E.R. (1996). Presynaptic facilitation revisited: State and time dependence. J. Neurosci. 16, 425-435. 12. Katz P.S., Frost W.N. (1995). Intrinsic neuromodulation in the Tritonia swim CPG: The serotonergic dorsal swim interneurons act presynaptically to enhance transmitter release from interneuron C2. J. Neurosci. 15, 6035-6045. 13. Clemens S., Katz P.S. (2003). G Protein signaling in a neuronal network is necessary for rhythmic motor pattern production. J. Neurophysiol. 89, 762-772. 14. Katz P.S., Frost W.N. (1996). Intrinsic neuromodulation: Altering neuronal circuits from within. Trends Neurosci. 19, 54-61. 15. MacDermott A.B., Role L.W., Siegelbaum S.A. (1999) Presynaptic ionotropic receptors and the control of transmitter release. Annu. Rev. Neurosci. 22, 443-485. 16. Vitten H., Isaacson J.S. (2001). Synaptic transmission: exciting times for presynaptic receptors. Curr. Biol. 11, R695-R697. 17. Hill A.A., Lu J., Masino M.A., Olsen O.H., Calabrese R.L. (2001). A model of a segmental oscillator in the leech heartbeat neuronal network. J. Comput. Neurosci. 10, 281-302. 18. Cropper E.C., Evans C.G., Rosen S.C. (1996). Multiple mechanisms for peripheral activation of the peptide-containing radula mechanoafferent neurons B21 and B22 of Aplysia. J. Neurophysiol. 76, 1344-1351. 19. Rosen S.C., Miller M.W., Cropper E.C., Kupfermann I. (2000). Outputs of radula mechanoafferent neurons in Aplysia are modulated by motor neurons, interneurons, and sensory neurons. J. Neurophysiol. 83, 1621-1636. 20. Edwards D.H., Heitler W.J., Krasne F.B. (1999). Fifty years of a command neuron: the neurobiology of escape behavior in the crayfish. Trends Neurosci. 22, 153-161.
Synaptic Gating: The Potential to Open Closed Doors Paul S. Katz
Synaptic gating is normally thought to be a mechanism for excluding synaptic input, but three recent studies show how the resting membrane potential interacts with integrative properties to act as a permissive synaptic gate.
“When one door closes another door opens; but we often look so long and so regretfully upon the closed door, that we do not see the ones which open for us.” Alexander Graham Bell
The ability of neural circuits to ‘gate’ inputs by suppressing inappropriate synapses is well known, but neural circuits also can open gates that are closed and allow signals to enter. Primary afferent depolarization is a classic example of synaptic gating, whereby central locomotor circuits inhibit the release of neurotransmitter from sensory afferents and thereby prevent sensory input from arriving during inappropriate phases of the locomotor cycle [1,2]. Sensory gating also has been proposed as a way of focussing attention during cognitive processing by suppressing inputs [3]. Recent papers [4–6] suggest that, at some synapses, the figurative ‘gate’ is normally closed but can be opened by depolarization of the membrane potential. These studies show that, besides excluding inputs, synaptic gating also can act in a permissive fashion, and that the gating can result from the integrative properties of the neurons rather than simply through the actions of presynaptic receptors. Permissive synaptic gating differs from other types of synaptic change that potentiate synapses. For example, long-term potentiation is thought to play a role in Hebbian-type plasticity, increasing the efficacy of connections by unmasking silent synapses or enhancing active synapses [7,8]. These learningrelated changes help set up networks for future use, but do not gate the input on a moment-to-moment basis. Similarly, forms of homosynaptic short-term plasticity such as synaptic facilitation and augmentation can increase synaptic strength [9], but are not thought of as gating the synaptic input. These types of change do not serve a gating function, because they are entirely dependent upon activity of the presynaptic neuron. Implicit in the notion of synaptic gating is the concept of a ‘gatekeeper’, whereby gating involves the momentary control of synaptic efficacy by a neuron other than the neuron being gated.
Department of Biology, Georgia State University, MSC 8L0389, 33 Gilmer St. SE Unit 8, Atlanta, Georgia 30303-3088, USA. E-mail: [email protected]
Dispatch
Neuromodulatory systems can act as gatekeepers, selectively enhancing synaptic inputs [10]. There are many known examples of presynaptic receptors that enhance neurotransmitter release. For example, studies on the sea slug Aplysia californica have shown that serotonin enhances neurotransmitter release from sensory neuron terminals [11]. Recent work with a related mollusc, Tritonia diomedea, has demonstrated that serotonergic neurons intrinsic to a central pattern generator (CPG) enhance release of neurotransmitter from other members of the same circuit by acting at presynaptic G-protein-coupled receptors [12,13]. This so-called ‘intrinsic neuromodulation’ can be thought of as an internal gate whereby synapses within the circuit are selectively enhanced when the circuit is in use [14]. Neurotransmitter release can also be enhanced by presynaptic ionotropic receptors, which momentarily depolarize the presynaptic terminal or allow calcium entry that facilitates release [15,16]. Thus, synapses can be permissively gated through the actions of presynaptic receptors. Recent papers by Ivanov and Calabrese [4], Evans et al. [5] and Herberholz et al. [6] propose mechanisms for permissive gating that are based upon the integrative electrophysiological properties of the neurons involved, rather than simply the actions of presynaptic receptors. In all three cases, the resting membrane potential has permissive gating effects on synapses that result in functional consequences for the neural networks. Although a depolarization of the membrane potential is required for these gating mechanisms, they are more complicated than simple synaptic summation. Ivanov and Calabrese [4] found that spike-mediated synaptic release in leech heart interneurons can be permissively gated by membrane potential oscillations acting on low threshold Ca2+ channels. The low threshold Ca2+ channels in the presynaptic neuron cause a background level of intracellular Ca2+ that determines whether an action potential will evoke an inhibitory postsynaptic current (IPSC) in the postsynaptic neuron. Calcium imaging showed that the rise in intracellular Ca2+ was correlated with the rise in synaptic strength (Figure 1A). Furthermore, uncaging a Ca2+ chelator or Ca2+ itself showed that the rise in intracellular Ca2+ was both necessary and sufficient for the membrane potential depolarization to increase synaptic strength. Thus, when the membrane potential of the neuron is sitting at rest, action potentials are ineffective at causing transmitter release; the synapses must be permissively gated by voltage to be effective. The function of the permissive gating examined by Ivanov and Calabrese [4] appears to be related to the pattern-generating role of the neurons, which form a half-center oscillator as part of the leech heartbeat CPG. The mechanism assures that spikes occurring
Current Biology R555
A
B ∆F/F
C a
a
a SN2
b
b
SN1
i
i
b
B21
c
R-HN
c d L-HN
B8
c
i
LG
Current Biology
Figure 1. Mechanisms by which membrane potential permissively gates synapses. (A) In the leech, the right heart interneuron (R-HN) makes an inhibitory synapse on the left heart interneuron (L-HN). No synaptic current is recorded in a voltage-clamped L-HN when a spike is elicited in R-HN from its resting membrane potential (light gray box). But when the membrane potential of R-HN is depolarized by current injection (i), then inhibitory post-synaptic currents (IPSCs) can be recorded in L-HN (dark gray box). The size of the IPSC corresponds to the level of intracellular calcium, as determined by the increase in fluorescence (∆F/F) of a Ca2+-sensitive dye (trace a). (B) In the Aplysia feeding system, the mechanosensory neuron, B21 makes excitatory connections on neuron B8. A spike initiated by mechanical stimulation (trace a) fails to invade the central branches of B21 and thus fails to evoke an excitatory postsynaptic potential (EPSP) in B8 (light gray box). Depolarization of the B21 soma by current injection (i) allows the spike to invade the central branches and evoke an EPSP in B8 (dark gray box). (C) In the crayfish terminal abdominal ganglion, sensory neurons (here labeled SN1, SN2) make rectifying electrical synapses with the lateral giant (LG) neuron and non-rectifying electrical synapses with each other. Stimulation of SN1 evokes a small EPSP in LG (light gray box). But when LG is depolarized by current injection (i), an action potential in SN1 recruits action potentials in nearby sensory neurons, such as SN2. This increases the apparent synaptic efficacy of SN1 and causes a larger EPSP in LG (dark gray box).
outside of the correct heartbeat phase will be ineffective. But more importantly, the efficacy of the synapse will be controlled by the very low threshold Ca2+ channels that play a critical role in setting the strength and duration of the depolarizing phase of the oscillation [17]. Thus, the gating mechanism elegantly matches the synaptic strength to the phase of the cycle at which it would be most advantageous to inhibit the postsynaptic partner in the half-center oscillator. A different role for voltage regulation of spikemediated synaptic transmission was uncovered by Evans et al. [5] examining synapses made by a mechanoafferent in Aplysia. The neuron, known as B21, fires action potentials in response to mechanical stimulation of a mouthpart tissue [18]. It was previously observed that, if the membrane potential of B21 is at rest when the neuron is activated by a sensory stimulus, then it fails to evoke a postsynaptic potential in a follower neuron, but if the membrane potential of B21 is depolarized from rest by current injection into the soma or by synaptic input, then the synapse is unsilenced [18,19] (Figure 1B). The mechanism underlying this permissive gating of B21 synaptic output is based on the failure of action potentials to invade central axon branches; depolarization of the cell soma relieves the conduction block and allows the action potential to invade the synaptic output region of the neuron [5] (Figure 1B). As in the case of the leech heartbeat interneurons, the membrane potential of B21 in Aplysia is rhythmically depolarized at a particular phase of a rhythmic motor behavior. In this case, B29 is depolarized as a result of synaptic excitation by interneurons in the feeding CPG. Thus, the feeding motor pattern permissively gates sensory input through B21; mechanosensory input would be effective at evoking a postsynaptic potential only during a particular phase of the motor pattern.
A third example of synaptic input being gated by voltage was demonstrated by Herberholz et al. [6] in the crayfish ‘tailflip escape’ system [20]. Mechanosensory neurons that respond to touch of the crayfish tailfan make rectifying electrical synapses with the lateral giant (LG) neuron. It was found that the afferents are themselves electrically coupled (Figure 1C), allowing stimulated sensory neurons to recruit other sensory neurons. This ‘lateral excitation’ is gated by the membrane potential of LG: when LG is depolarized, action potentials in some afferent axons will trigger spikes in neighboring sensory axons (Figure 1C). A computer simulation study suggests that the mechanism underlying the recruitment of unstimulated sensory axons is related to the voltage-sensitivity and kinetics of rectifying electrical connections to LG [6]; current flows antidromically from the depolarized LG to the afferents, and at the same time the rectifying electrical synapses are back-biased, effectively increasing the input resistance of the afferents and making it easier for them to be recruited by neighboring axons. The net effect of this mechanism is that the membrane potential of LG determines whether mechanosensory input will effectively excite LG and elicit a tailflip response. Thus, LG gates its own synaptic input by causing stimulated sensory neurons to spread excitation to other afferent axons. Although synaptic gating is commonly understood to mean closing a gate, gates can also be opened. These three examples show that membrane potential can act through very different mechanisms to permissively gate synaptic transmission. They illustrate how important it is to understand the integrative properties of neurons rather than assume that an action potential reliably leads to neurotransmitter release. Thus, in thinking about synaptic gating, it is important to remember that while one door may close, another may open.
Dispatch R556
References 1. Rudomin P. (1999). Presynaptic selection of afferent inflow in the spinal cord. J. Physiol. (Paris) 93, 329-347. 2. Büschges A., El Manira A. (1998). Sensory pathways and their modulation in the control of locomotion. Curr. Opin. Neurobiol. 8, 733739. 3. Castro-Alamancos M.A., Oldford E. (2002). Cortical sensory suppression during arousal is due to the activity-dependent depression of thalamocortical synapses. J. Physiol. 541, 319-331. 4. Ivanov A.I., Calabrese R.L. (2003). Modulation of spike-mediated synaptic transmission by presynaptic background Ca2+ in leech heart interneurons. J. Neurosci. 23, 1206-1218. 5. Evans C.G., Jing J., Rosen S.C., Cropper E.C. (2003). Regulation of spike initiation and propagation in an Aplysia sensory neuron: gating-in via central depolarization. J. Neurosci. 23, 2920-2931. 6. Herberholz J., Antonsen B.L., Edwards D.H. (2002). A lateral excitatory network in the escape circuit of crayfish. J. Neurosci. 22, 90789085. 7. Montgomery J.M., Pavlidis P., Madison D.V. (2001). Pair recordings reveal all-silent synaptic connections and the postsynaptic expression of long-term potentiation. Neuron 29, 691-701. 8. Poncer J.C., Malinow R. (2001). Postsynaptic conversion of silent synapses during LTP affects synaptic gain and transmission dynamics. Nat. Neurosci. 4, 989-996. 9. Zucker R.S., Regehr W.G. (2002). Short-term synaptic plasticity. Annu. Rev. Physiol. 64, 355-405. 10. Katz P.S. (1999). Beyond Neurotransmission: Neuromodulation and its Importance for Information Processing. (Oxford: Oxford University Press). 11. Byrne J.H., Kandel E.R. (1996). Presynaptic facilitation revisited: State and time dependence. J. Neurosci. 16, 425-435. 12. Katz P.S., Frost W.N. (1995). Intrinsic neuromodulation in the Tritonia swim CPG: The serotonergic dorsal swim interneurons act presynaptically to enhance transmitter release from interneuron C2. J. Neurosci. 15, 6035-6045. 13. Clemens S., Katz P.S. (2003). G Protein signaling in a neuronal network is necessary for rhythmic motor pattern production. J. Neurophysiol. 89, 762-772. 14. Katz P.S., Frost W.N. (1996). Intrinsic neuromodulation: Altering neuronal circuits from within. Trends Neurosci. 19, 54-61. 15. MacDermott A.B., Role L.W., Siegelbaum S.A. (1999) Presynaptic ionotropic receptors and the control of transmitter release. Annu. Rev. Neurosci. 22, 443-485. 16. Vitten H., Isaacson J.S. (2001). Synaptic transmission: exciting times for presynaptic receptors. Curr. Biol. 11, R695-R697. 17. Hill A.A., Lu J., Masino M.A., Olsen O.H., Calabrese R.L. (2001). A model of a segmental oscillator in the leech heartbeat neuronal network. J. Comput. Neurosci. 10, 281-302. 18. Cropper E.C., Evans C.G., Rosen S.C. (1996). Multiple mechanisms for peripheral activation of the peptide-containing radula mechanoafferent neurons B21 and B22 of Aplysia. J. Neurophysiol. 76, 1344-1351. 19. Rosen S.C., Miller M.W., Cropper E.C., Kupfermann I. (2000). Outputs of radula mechanoafferent neurons in Aplysia are modulated by motor neurons, interneurons, and sensory neurons. J. Neurophysiol. 83, 1621-1636. 20. Edwards D.H., Heitler W.J., Krasne F.B. (1999). Fifty years of a command neuron: the neurobiology of escape behavior in the crayfish. Trends Neurosci. 22, 153-161.