{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

Selverston1980 - THE BEHAVIORAL AND BRAIN...

Info icon This preview shows pages 1–13. Sign up to view the full content.

View Full Document Right Arrow Icon
Image of page 1

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 2
Image of page 3

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 4
Image of page 5

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 6
Image of page 7

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 8
Image of page 9

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 10
Image of page 11

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 12
Image of page 13
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: THE BEHAVIORAL AND BRAIN SCIENCES (1980)3,535—571 Printed in the United States of America Are central pattern generators understandable? _ Allen I. Selverston , Department or Sioiogy. University or California. San Diego,- La Jolla'. Calif. 92093 Abstract: Most rhythmic behaviors are produced by a specialized ensemble of neurons found in the central nervous system. These central pattern generators {CPGs} have become a cornerstone of neuronal circuit analysis. Studying simple invertebrate nervous systems may reveal the interactions of the neurons involved in the production of rhythmic motor output. There has recently been progress in this area, but due to certain intrinsic features of UFOs it is unlikely that present techniques will ever yield a complete understanding of any but the simplest of them. The chief impediment seems to be our inability to identify and characterize the total interneuronal pool making up a CPG. In addition, our general analytic strategy relies on a descriptive, reductionist approach, with no analytical constructs beyond phenomenological modeling. Detailed descriptive data are usually not of sufficient depth for specific model testing, giving rise instead to ad hoc explanations of mechanisms which usually turn out to be incorrect. Because they make too many assumptions, modeling studies have not added much to our understanding of CPGs; this Is due not so much to inadequate simulations as to the poor quality and incomplete nature of the data provided by experimentalists. _ A basic strategy that would provide sufficient information for neural modeling would include: (1) identifying and characterizing each element in the CFO network; (2) specifying the synaptic connectivity between the elements; and (3) analyzing nonlinear synaptic properties and interactions by means of the connectivity matrix. Limitations based on our present technical capabilities are also discussed, Keywords: central pattern generator; motor patterns; neural circuits; neural modeling Animal behavior is the result of sequential muscular contrac— tions controlled by patterns of impulses generated in the central nervous system. To understand how such motor patterns are produced one must know what neurons are involved, how they are turned on and off, and how they are modified as the behavior progresses. Behavior which is being continuously changed by both internal and external ones is referred to as episodic. Capturing prey and playing the piano are examples of such behaviors. Since these are by nature nonrepetitive behaviors, they are extremely difficult to study. New pools of motor neurons are constantly being shifted around inside the CNS (central nervous system) so that the motor patterns are never exactly the same. Repetitive behaviors, on the other hand, are, by definition, relatively consistent in their motor output patterns. Locomotion, flight, ' - chewing, and swimming use essentially the same groupslof " motor neurons over and over again, and although the motor pattern can be modified to meet changes which may be encountered in the environment, the basic rhythm is reason- ably consistent. Most of the progress in understanding the neural control of behavior has come from studying rhythmic behaviors in both vertebrates and invertebrates. In fact, it is only rhythmic behavior which has the possibility of being explainable in terms of all the neurons involved. At the heart of the analysis is the concept of the central pattern generator (CPG), a group of neurons which can produce the rhythmic pattern. It is now well established that almost all such CPCs can operate without the need of peripheral sensory input (Wilson 1961; Grillner 1977). A fundamental question. then, is how are the neurons which comprise the CPG arranged and h0w are they able to generate rhythmic patterns — i.e., how do they work? In the light of our present knowledge the question could perhaps be . phrased: what constitutes our current understanding of a 0 i930 Cambridge University Press 0 140-525X/80/040535-37/Mflflf CPG? We can begin with a more detailed look at the idea of the CPG and the present limits on its neurophysiological analysis. The central pattern generator concept The concept of a CPG arose from experiments which demonstrated that a discrete group of neurons could produce a rhythmic motor pattern when completely isolated from sensory inputs (Wilson 1961); such an ensemble of neurons represents a CPG. In some cases one CPG can drive several structures, while in other cases there is a separate CPG for each. When there are multiple CPGs, special pathways are necessary to coordinate their activities (Stein 1976). Only in rare instances do CPGs work in complete isolation (Selverston et al. 1976); Usually some form of generalized excitation is required (Grillner 1977; Wilson 1961). The CPG has been found to be made rip of motor neurons (Maynard GI Selverston 1975), interneurons (Getting et al. 1980). or some combination of the two (Mulloney 81 Solver- ston 1974). When the UPC is made up of premotor inter- neurons, some of them may also be a part of a CPG for other rhythmic behaviors; that is, different CPGs may share intemeurons. For example, the CPGs for locomotion and for scratching in the cat probably use neurons from the same pool (Orlovsky and Shik 1976). The CPG may be turned on by special input pathways [command fibers — see Kupfermann Ex Weiss: “The Command Neuron Concept" BBS 1(1) 1978], increased activity of excitatory axons (Wilson 1961). or the action of various circulating substances such as peptides [Barker and Gainer 1974 — see also Dismukes: "New Concepts of Molec- ular Communication" BBS 2(3) 1978]. Once activity is under 535 Selverston: Central pattern generators way, it is often modulated by sensory feedback or specialized inputs from other areas of the CNS (Grillner 1977). Activity of the CPG can be terminated by removal of the input excitation {Wilson 1964) or by degradation of any neuro- chemical agents (Barker et al. 1975}. Present strategies Most of the recent progress in CPG analysis has come from work on invertebrate preparations in which large, identifiable cells can be easily penetrated. One can record from pairs of pre- and postsynaptic cells while the pattern is being generated. With the use of such “simple" systems, the goal of “understanding" the mechanisms involved in the production of rhythmic motor patterns at one time seemed theoretically possible. The general approach consisted of attempting to: (1) describe all of the components; (2) explain how these components interact to produce a rhythmic motor program; and (3) ultimately develop a model which not only replicates the dynamic range of the motor output but also can predict the results of altering any component of the system. Component description The CPG is a machine - a complex operating system made up of interrelated parts. To understand its operation we must know the number of parts it contains and as much as possible about the characteristics of each part. Although this would seem to be straightforward, there is enormous variation in what workers consider to be an adequate description. The most comprehensive list I have seen is that proposed by Bullock (1976), which includes forty-six parameters. for units and groups that should be known in order to understand a network's function. This list assumes that all the cells in the network can be identified. For most CPGs, however, just finding all the participating cells has proven to be an enormous problem. Locating the cell bodies of the motorneurons is not too difficult, since soma potentials can usually be correlated with action potentials in the motor roots, and all of the somata can be located by backfilling through the motor roots with marker dyes such as cobalt or Lucifer Yellow. If the CFO is made up solely of motor neurons, then a complete description can, in theory, be obtained. But for cases in which the UPC is made up exclusively or partly of interneurons, the problem is much greater. Such interneurons can often be quite small, which makes them hard to find by intracellular probing. In addition, the motor pattern must be operative during such probing for the interneurons to be recognizable as components of the CFO (this is not the case for motor neurons). Unless one knows the total number of interneurons present, or their individual properties, any circuit diagram will be incomplete. (One may argue that some of the interneurons are less important than others and so can be omitted from the circuit without much harm. This is probably true. They may serve only minor “tuning functions" or have redundant or overlapping activity with other cells. In some cases, their omission will not affect a first approximation of the circuit.) A CPG does not produce a motor Output pattern which is either present or absent. but instead shows various levels of activity. Ideally, the pattern produced by the preparation in its experimental condition is equivalent to the pattern observed in the freely behaving animal. When one searches for interneurons which are part of the CPS, the identification is based mostly on the fact that their activity correlates with motor output. Unless one can account for every interneuron present in a ganglion, it is never known with certainty that all 536 THE BEHAVIORAL AND BRAIN SCIENCES (193m, 3 of the CFO interneurons have been found. Unless they are all known, the formulation of any mechanism is difficult. But what is more disturbing is that omissions may lead to a faulty explanation of the mechanism, especially when some sort of model is used to verify the circuit. If the progress made in describing various CPG circuits is examined carefully, it is clear that most descriptions are incomplete, and, with our present tools, they are likely to remain that way. That is, for most rhythmic circuits, the total number of cells involved is not known. But suppose we could identify all of the neuronal compo— nents of a CPG. First, we would need to know each neuron 's cellular properties. If one did not know the elemental properties of each component, providing a mechanistic explanation of a network would be as difficult as assembling a radio with unlabeled resistors and capacitors. Knowing the properties does not lead directly to an explanation; however, a reductionistic experimental approach can be valuable in formulating a hypothesis. Again using the radio analogy, although properties of the components do not explain the overall function of the circuits, no mechanistic explanation would be possible without them. What properties of the component cells should be known? A minimal list would include: I. the excitability of their membranes, i.e., current-voltage curve; 2. their thresholds relative to their resting potentials; 3. their firing characteristics — steady, irregular, bursty — as well as the amount of accommodation they show; 4. any endogenous activity such as slow—wave generating mechanisms; 5. stability or instabilities, such as plateauing properties; 6. extent of postinhibitory rebound. Ideally, these values should be obtained from each cell in isolation, but since this is often impossible, the values must be referred to the particular activity state of the CFO at the time of measurement — i.e., whether the elemental properties were obtained during pattern—generating activity or during a silent period. Another area of purely descriptive information is the connectivity between components. Connectivity can most easily be determined by recording pairwise from the neurons that make up the CPG. Spontaneous or elicited activity in one cell can be correlated with some reSponse in a synaptically connected follower cell. Such paired recordings are the basis for determining the functional relations between cells but do not necessarily demonstrate that a monosynaptic pathway is responsible. While ascertaining the presence of monosynaptic connec- tions is not technically difficult (Berry & Pentreath 1976), most CPG circuits have been constructed without rigorous testing for monosynapticity. At a minimum, one should be able to demonstrate a clear, unitary postsynaptic potential following a spike in the presynaptic cell. Showing that the firing of one cell can excite or inhibit another cell without showing postsynaptic- potentials is poor evidence for a monosynaptic connection (Friesen et al. 1976). Several other tests can help to distinguish mono- and polysynaptic path- ways. Raising the Ca” concentration of the saline increases synaptic effectiveness but at the same time raises the threshold for firing (Berry 81 Cottrell 1974). With this technique, neurons in multisynaptic pathways may be blocked. Another method is to inject the presynaptic neuron with TEA (tetraethyl ammonia) to block K+ channels (Hille 1967). Since this prolongs the action potential and augments the release of transmitter from the presynaptic terminal, a larger than. normal postsynaptic response should occur if the connection is monosynaptic. An additional problem associated with determining the monosynapticity of connections is the possibility of measuring PSPs (postsynaptic potentials) from cell bodies which are electrically distant from the postsynaptic site. This may present difficulties, especially with some arthropod systems which not only have electrically passive soma membranes, but also have low resistance cables connecting the somata to the neuropile synaptic regions (Selverston & Remler 1972). This particular problem can be overcome by neuropile penetrations closer to the synaptic areas, so I would not regard it as an intractable diEiculty. The last type of information which must be obtained is a description of the synaptic properties. I would include among these: 1. Synaptic strength - determined functionally and not merely by the size of a postsynaptic response measured at a distant recording site. By functionally I mean the effectiveness of presynaptic cell firing in changing the ongoing activity in the postsynaptic cell. 2. An index of facilitation or antifacilitation. 3. Frequency-dependent conductance changes. 4. The strength of electrical coupling, if present. The strength of the synapse is probably the most important of these factors since an extremely weak Synapse may, for all practical purposes, be omitted from the wiring diagram. (There are no generally accepted criteria for ignoring neurons with weal: synaptic interactions. Presumably such synapses will play only minor roles; however, they may form part of a pool of similar neurons which may sum their synaptic inputs. On the other hand, as suggested earlier, they may perform fine—tuning functions that are important but not critical to the central pattern generating mechanism.) It is hard for me to see how any explanation of mechanism can be attempted without having all of this information, yet many models are put forward based on the connectivity alone. As in the determination of the basic cellular properties, the synaptic parameters must be indexed to the activity or lack of activity of the CFO network. Synaptic. properties measured in an inactive system are likely to be quite different from those derived during normal activity (Spira et al. 1976). Nonlinear interactions may play a crucial role for many synaptic relationships. This is especially true for neurons which have both electrical and chemical interactions. Bennett and his colleagues have shown that the strength of a chemical synapse may be altered by current flowing in nearby electrical synapses. Similarly, chemical synapses can alter the coupling coefficient of electrical synapses (Spira and Bennett 1972). These phenomena suggest that synaptic strength and electrical coupling coefficients may vary continuously in some networks so that any static description of these prop— erties would be an oversimplification. Mechanisms Identifying and characterizing the components of a CPG will not, except in the most trivial cases, explain how the CFO works any more than - to use our earlier analogy — a similar classification of its parts will tell us how a radio works. Such information must be fitted into an existing model or, if necessary. a new one must be formulated. Several models have been proposed to account for the production of rhythmic motor patterns. These fall into two principal categories: a. ensembles of neurons driven by cells with spontaneously oscillating membrane potentials (endogenous bursters), or b. ensembles in which the synaptic and/or electrical connectivity (circuitry) causes an oscillatory or “reso- nant" mode of activity. Selverston: Central pattern generators Endogenous bursters Endogenous bursters are cells whose membrane potential oscillates rhythmically in such a way that bursts of impulses are triggered on the depolarizing phase of the oscillation (see Figure 1). Such cells have been well studied in mollusks, in which the ionic mechanisms underlying the oscillations have been examined by means of voltage clamping (Gala 1976). All such cells show a negative resistance characteristic due to an increased inward conductance during the depolarizing phase. Repolarization appears to be caused by an increase in K + conductance. Cells can be identified as endogenous bursters by showing that their bursting continues when all synaptic input to them has been blocked or removed. This can be done experimen- tally by removing or tying off the soma (Alving 1968) or by using high Mgfi+ lov.r Ca++ saline (Selverston 1976). In the latter method, care must be taken to ensure that the ionic imbalance does not affect the endogenous burst-generating mechanism directly. Some cells can be converted into endogenous bursters by certain peptides, so this property may be under the control of humoral agents as well (Barker 8; Gainer 1974). Despite the widespread occurrence of endogenous bursters, there are very few examples showing how they are used in pattern generation. The cardiac ganglion of the lobster contains only nine cells, some of which appear to be electrically coupled bursters (Tameyasu 1976). The pyloric rhythm of the lobster stomatogastric ganglion also contains three endogenous bursters, which play an important role in the generation of the pyloric rhythm (Maynard 61 Selverston 1976). Endogenous bursters have been implicated in the control of the heartbeat in both leech (Thompson iii Stent 1976) and Aplysia (Koester et al. 1974). Except in the case of the lobster pyloric rhythm, endogenous bursters appear to be used mainly as driving elements in the simplest rhythmic behav- iors, such as the heartbeat. However, if the burster cells are connected to other neurons with inhibitory synapses, then tonically firing neurons will be interrupted during the endogenously active cell burst and an antagonistic burst pattern will be produced. An example of this type is the pyloric rhythm .generator of the lobster stomatogastric ganglion (Maynard 8: Selverston 1976). In this system, a group of three electrically coupled endogenous bursters periodically inhibits other tonically firing neurons in the circuit. Although most cells of this type generate action potentials, there are several examples in arthropods which may indicate that no spikes are involved in their presynaptic functioning. Mendelson has described cells in the lobster which drive Hypothetical circuits for the production of rhythmic motor output driven by an endogenously bursting cell. The burst- er eculd (l) periodically excite two silent motor neurons; (2) peri— odically inhibit two tonically active motor neurons; or (3') simul- taneously excite and inhibit two follower neurons. Figure 1. THE BEHAVIORAL AND BEAN SCiENCES (1980}. 3 537 Selverston: Central pattern generators rhythmic activity of the gill balers by graded release of transmitter {Mendelson 1971). Depolarization or hyperpolar— ization of this cell can speed up ...
View Full Document

{[ snackBarMessage ]}

What students are saying

  • Left Quote Icon

    As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

    Student Picture

    Kiran Temple University Fox School of Business ‘17, Course Hero Intern

  • Left Quote Icon

    I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

    Student Picture

    Dana University of Pennsylvania ‘17, Course Hero Intern

  • Left Quote Icon

    The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

    Student Picture

    Jill Tulane University ‘16, Course Hero Intern