Central Pattern Generators I: From Behavior to a Circuit

Introduction

• In the previous units, we have started to explore very simple forms of behavior (a neuromuscular system, a reflex). We now turn to the study of more complex behaviors, and the neural circuitry that underlies them.

• How do patterns of neural connections and of neural activity generate complex behavior?

• We will first discuss some important general issues raised by the goal of connecting neural circuitry to behavior:
  • How does one relate behavior to underlying neural mechanisms?
  • What behaviors are best to study to understand underlying neural mechanisms?
  • What is a pattern generator, and what advantages does it provide for studying neural circuitry?
  • How does one choose an animal system in which to study pattern generation?

• We will begin our exploration of neural circuitry by focusing on an example in which the relationship between behavior and neural circuitry has been worked out in detail, many of the key elements of the neural circuit controlling the behavior are known, and the role of modulation in generating the behavior has been studied at the level of single nerve cells. You will then have the opportunity to work with a simulation of this neural circuit, and analyze how it generates patterned activity.

Relating Behavior to Neural Mechanisms: Neuroethology

• In the first half of the twentieth century, Konrad Lorenz and Niko Tinbergen pioneered the field of ethology, the study of animal behavior in its natural environment.

• Investigators found that natural animal behavior was flexible and complex. Some behaviors appeared to be innate and relatively fixed; others were strongly shaped by experience.

• Tinbergen emphasized that understanding a behavior required answering a variety of questions:
  • What was the immediate stimulus that evoked a behavior, and then what are the underlying mechanisms within the animal that generated the behavior?
  • What is the adaptive significance of a behavior, i.e., how does it help an animal survive and reproduce?
  • How does the behavior change over the early life of the animal, i.e., how does it develop? Are there critical environmental experiences that must occur for the behavior to be expressed, or does it unfold independently of external influences?
  • What is the evolutionary history of the behavior? In similar species, how is the behavior expressed, and how does it differ? How could the behavior have arisen over the course of phylogeny?

• Ethologists also emphasized the importance of understanding an animal's sensory world (Umwelt), which is determined by an animal's sense organs and how it deploys them.

• J. J. Gibson also emphasized that features of the environment shape behavior. For example, optic flow as an organism moves through an environment provides important cues about the structure of the environment, and how close an animal may be to an object or surface. He described this approach as ecological psychology.

• These lines of investigation, which focused on animal behavior within that animal's natural environment, were often at odds with classical physiological studies, in which animals were brought into a laboratory, and were either strongly restrained, or were reduced to specific organs and tissues for detailed analysis and study. Investigators who focused on natural behavior in the field raised the objection that laboratory studies might reflect little or nothing of the mechanisms that operate when an animal is intact and behaving freely in its natural setting.

• This tension was particularly acute in studies of the neural basis of behavior, since the very process of attempting to analyze neural activity was usually incompatible with allowing an animal to exhibit its natural behavior. For example, if one attempts to impale a neuron intracellularly in part of the nervous system, or to precisely position an extracellular electrode near a specific neuron, any movement can disrupt the recording. But most behaviors require that the animal be able to move, and are thus incompatible with such recordings!

• The same tension has affected research in sensory physiology. In the visual system, studying single spots of light, bars, and other highly artificial stimuli in anesthetized animals provided some of the first insights into the organization of the visual system. However, new insights have begun to emerge as investigators study neuronal responses to natural scenes in awake animals. It is likely that still more will be discovered if neural activity is monitored as animals are able to move through an environment while encountering natural stimuli.

• Attempts to combine the study of natural behavior with relatively non-invasive measures of neural activity have flourished, despite these difficulties, and have given rise to the field of neuroethology, the study of the neural mechanisms underlying natural animal behaviors.

• Promising technical advances in the last few decades have made it possible to get closer to the ideal of allowing animals to freely behave while monitoring many aspects of their neural activity. Techniques to monitor electrical fields in the brain (electroenchephalography, EEG) can be accomplished using surface electrodes; chronically implanted extracellular electrodes can also be used to monitor the activity of large numbers of neurons simultaneously. Techniques to monitor magnetic fields in the brain (magnetoencephalograhpy, MEG), or changes in blood flow (e.g., positron emission tomography (PET), or functional magnetic resonance imaging (fMRI)) also provide measures of activity of large populations of nerve cells.

• Ingenious new techniques are further improving the ability to combine measures of brain activity with normal behavior. For example, it is possible to chronically image neural activity using calcium-sensitive or voltage-sensitive dyes, to control neural activity by inducing nerve cells to express special light-activated channels that can selectively turn neurons on or off (optogenetic techniques), to induce animals to run on a ball suspended on air so that they move their limbs without moving their whole bodies, and to show them virtual reality environments that are controlled by their movements. These are still not a full solution to the problem, but they are far closer than studying an isolated component of the brain.

Choice of Behaviors to Study

• Which behaviors are best studied to understand their underlying neural mechanisms?

• On the one hand, what may be of most interest are higher order behaviors, such as language, thought, and creativity, and these are certainly very important to understand.

• On the other hand, many of these behaviors are fleeting, and may not generate overt, easily observable changes in behavior for long periods of time. It may not be a good idea to focus one's entire research career waiting for a particular musician to compose his greatest symphony, or for a particular mathematician to discover the theorem for which she is known to posterity. Furthermore, it is not clear that any of these studies could be performed in animals, and thus access to underlying brain mechanisms will be very limited. Finally, it is worth noting that many complex behaviors depend upon or are even elaborated from simpler behaviors, and thus understanding the simpler behaviors may provide important insights into the most complex behaviors.

• Thus, investigators have focused much study on behaviors that are easily evoked, readily repeated by the subject, and that can be studied in animals as well as in humans.

• As a consequence, a great deal of work has focused on escape responses, which often will override any other behavior since their success determines whether an animal survives, and on rhythmic behaviors, whose repetition allows the investigator to accumulate a large amount of data.

• Escape responses are dominated by the need for speed. If an animal does not respond sufficiently quickly, it may become another animal's meal. Investigators have found several characteristics that are typical of escape behaviors:
  • There are relatively few synapses between sensory inputs signaling danger and motor outputs that generate the escape response; this allows the behavior to occur quickly.
  • The neurons that generate the behavior often have very large axons so that they can communicate signals to the periphery very rapidly (e.g., the squid giant axon, which is part of the squid's escape response; the Mauthner neuron, which generates the rapid startle responses in vertebrates). In some escape systems, electrical synapses are more common than chemical synapses, again because messages can be conducted through them very rapidly.
  • The complexity of escape circuitry is related to the complexity of the periphery that must be controlled. For example, the giant interneurons in the crayfish that mediate tail flip escape responses need to differentiate between touches from the front or from the rear of the animal, and generate a tail flip using different tail segments to ensure that the animal escapes away in the correct direction.
  • In the cockroach escape response, the motor circuitry needs to respond to rapid wind puffs (often generated by the tongue of a predator) around the entire circumference of the animal with a turn and run in the opposite direction. All six of the animal's legs need to be properly coordinated to generate the appropriate response, and the legs are often in variable positions at the time the wind puff is sensed; all of these factors require somewhat more complex neural circuitry. Despite this complexity, the initial turn in response to a wind puff is completed within 30 milliseconds, which is why cockroaches have done a very good job of surviving and reproducing even in urban environments (like your apartment).

• Rhythmic behaviors have also been extensively studied. These studies led to the concept of a pattern generator even before the neural basis of such behaviors was worked out.

Studying Pattern Generators

• Rhythmic behaviors are ubiquitous:
  • Breathing and heartbeat happen at fairly regular and constant intervals.
  • Swimming, flying, and walking are forms of locomotion in water, air or on land that occur through rhythmic alternations of body parts such as fins, wings or limbs.

• Many rhythmic behaviors are relatively easy to evoke in the laboratory.

• Because the behavior is rhythmic, both motor movements and neural activity should repeat regularly.

• For some time, there was significant controversy about how rhythmic behaviors were generated:
  • Some investigators were certain that the behavior was due to a chain reflex, a sequence of actions each of which triggered the next action.
    • For example, during walking, if a leg is down in the stance phase, and is pulled backwards as far as it can stretch and is no longer bearing load, the sensory inputs reporting its length and load will trigger that leg to lift and swing forward. Similarly, once the leg has swung as far forward as it can go, sensory inputs reporting its position will trigger it to move down and contact the ground. Thus the stance movement induces the sensory trigger to activate the swing phase, and sensory inputs from the swing movement in turn trigger the switch to the stance phase. Symbolically, if stance is motor response $R_{1},$ the sensory outputs generated by stance are $S_{1},$ swing is motor response $R_{2},$ and the sensory outputs generated by swing are $S_{2},$ then the chain reflex model of walking would be $R_{1} \to S_{1} \to R_{2} \to S_{2} \to R_{1}.$
  • Other investigators were equally certain that rhythmic movements were due to a central pattern generator, a neural network that generated rhythmic outputs in the absence of any sensory input.
    • For example, investigators showed that it was possible to dissect out the nervous system of the lamprey, and by bathing it in glutamate or glutamate agonists, to induce patterns of rhythmic output on motor nerves that were similar to those measured in intact, swimming animals. Such fictive locomotion was a strong argument that the rhythmic pattern was generated primarily or solely by the nervous system.
  • Given this data, are rhythmic behaviors generated due to chain reflexes or due to central pattern generators? As in most scientific controversies in biology, the answer to the question after further investigation was clearcut: Yes.
  • Most rhythmic behaviors have a mix of both characteristics. They are strongly influenced by cycle-to-cycle sensory inputs, especially for slow movements. They can also generate rhythmic movements even in the absence of sensory input, especially during fast movements, in which it may take too long for sensory inputs to propagate into the nervous system, and inertial forces may dominate movements.

• Understanding the neural basis of a pattern generator requires
  • identifying the motor neurons that control the muscles that mediate the rhythmic pattern;
  • identifying the sensory neurons that trigger the pattern, and that provide inputs that may modify it;
  • identifying the neurons responsible for generating the rhythmic pattern (which are often, but not always, pure interneurons).

• How can one determine whether a neuron is a crucial element of the pattern generator for a rhythmic behavior, rather than simply being driven by it?

• The critical way to answer this question is to perform a phase reset experiment.

• Since the behavior is rhythmic, it repeats at a regular, fixed interval. You could map the entire sequence onto a circle, similar to a clock. Just like a clock, you could mark each location along the circle with some number indicating how far along you are in the cycle. If you chose to use the marking on a clock, halfway around the cycle would correspond to the number 6, and three-quarters around the cycle would correspond to to the number 9. Usually, the markings correspond to those used for describing angles around a circle; they could be in degrees, and range from 0 degrees to 360 degrees; or they could be in radians, and range from 0 to $2 \pi$. The location on the circle is referred to as the phase angle.

• If you inhibited a neuron enough to turn it off, and then released it, and it was not part of the circuit generating the rhythmic cycle, it would go back to an activity pattern that was identical to what it had before you inhibited it. So if it had started to burst at about halfway through the cycle (at a phase angle of $\pi$), the next time around the cycle it would again start to burst at the same phase angle.

• In contrast, if the neuron was actually a part of the rhythm generator, then inhibiting it would stop the pattern for some time. Let's say you inhibited it for a quarter of the cycle. Now, the next time around the cycle, it would turn on a quarter of a cycle later than it had before you inhibited it. Thus, if it had turned on at a phase angle of $\pi$, it would now turn on at a phase angle of $\frac{\pi}{4} + \pi = \frac{5 \pi}{4}.$

• In other words, if you were successful in resetting the phase of the cycle, then the neuron you had inhibited would almost certainly be part of the pattern generator for the rhythm.

Rationale for Studying Invertebrate Nervous Systems

• Choosing the right preparation for answering a question in biology is every bit as important as asking the right questions, and using (or inventing) the right tools to answer the question.

• As we saw earlier, Hodgkin and Huxley chose the squid giant axon because it was so large, and with the tools at their disposal, they were able to analyze the mechanisms of excitability. Because it also had only two voltage-gated conductances, and they had the right tool to dissect these apart (the substitution of choline for sodium ions), they could fully analyze the problem. In the process, they created a model, a research methodology, and a conceptual framework that has guided biophysical studies of excitability to this day. Imagine what would have happened if they had decided that the only interesting neurons were those in the human or primate cerebral cortex! We might still be waiting to understand how excitability works in the nervous system.

• Similarly, understanding neural circuitry requires choosing the right animal, asking the right questions, and using (or inventing) the right tools to answer the question.

• For several decades, much of the best work on circuitry analysis and pattern generation was done through the analysis of invertebrate nervous systems.

• No animal is simple; each is the end product of complex evolutionary process, and has a complex suite of behaviors adapted to its environmental niche.

• But some animals are more experimentally tractable to analysis than others.

• For example, molluscan nervous systems have been a major focus of investigation because many of the neurons are large, and in some animals they are pigmented and thus easy to locate and impale with an intracellular electrode; they often have relatively stereotyped positions in the collections of nerve cells controlling parts of the body (ganglia), relatively similar inputs and outputs, similar neurotransmitters, and similar biophysical properties from animal to animal. This makes it possible to repeatedly identify individual nerve cells, which in turn makes it far easier to work out detailed circuitry. In some mollusks, the cell bodies of the neurons, which are the largest part of the nerve cell, are also electrically compact; that is, injecting inhibitory currents at the soma actually blocks all outputs of the neuron. This has been the basis for extensive studies of behavior and neural control in mollusks such as the sea slug Aplysia californica, which has been used extensively for the study of non-associative and associative forms of learning and memory, and the sea slug Tritonia diomedea, whose escape swimming behavior has been studied in detail.

• In crustaceans (e.g., the lobster or the crab), there is a ganglion containing 30 neurons, the stomatogastric nervous system, whose connections and biophysical properties were precisely mapped by Allen Selverston and his students. This ganglion controls teeth that grind food, a gastric mill, and the pyloric sphincter, all of which are crucial for assimilating the complex detritus on which these animals feed. Eve Marder and her students showed that a very large number of modulatory substances were present in the neurons of the ganglion, as well as in the nerves innervating the ganglion that originated from other parts of the nervous system. Modulators could not only modify the activity of the neurons within the ganglion; they could act to "sculpt out" entirely different patterns of neural activity from the same anatomical circuit. Thus, it became clear that simply having anatomical connectivity was not enough to understand a neural circuit; its functional, dynamic patterns are a crucial key to understanding its behavioral significance.

• Insect neurobiologists have also made remarkable strides in the analysis of neural circuitry. Great advances have been made by combining behavioral and neural studies in animals such as locusts and cockroaches, and entirely new approaches have become possible for studying the neural circuitry of the fruit fly Drosophila melanogaster, whose genetics have been extensively analyzed and can thus be manipulated in highly precise and sophisticated ways.

• The nematode Caenorhabditis elegans has 302 neurons, whose anatomical structures have been fully mapped. Because it is hard to penetrate the worm's cuticle, studies of the physiology of its neurons have been challenging, but have nevertheless progressed by using laser ablation, genetic manipulations, and calcium imaging.

• Optogenetic techniques and careful mapping of the mouse genome, which have made it possible to precisely knock out or knock in specific genes at specific location, during specific times in response to well-controlled stimuli have all made it possible to begin to apply many of the circuit analysis techniques that were once the sole province of invertebrates to the vertebrate nervous system. Combined with some of the noninvasive techniques for measuring the activity of large populations of neurons in humans, such as EEG, MEG, PET, or fMRI, great strides have been made within the last few years in starting to connect behavior and neural activity in higher vertebrates and in humans, though much remains to be done. As of yet, no mammalian system has been studied with the level of specificity that is possible in invertebrate nervous systems, though progress in analyzing the circuitry in the retina at the neural level is closing the gap.

Tritonia Escape Swimming

• When the marine mollusk Tritonia diomedea is contacted by a predator (for example, by the tube feet of a starfish), it generates an escape response:
  • It stiffens its entire body, turning its body into a paddle,
  • It then "flaps" this paddle around its midline by alternating ventral and dorsal flexions of its entire body.
  • The lift generated by its flapping rapidly moves it off the surface and allows it to tumble some distance away from the predator.

• A movie of the escape swim is here.

• The escape swim also has aspects of a rhythmic behavior, because of the alternating dorsal and ventral flexions of the body.

• Unlike most swimming, however, Tritonia's flapping behavior is not steered or guided by the animal in a particular direction.

• The rate of flapping also slows down and then stops, so the behavior is not a regular rhythm, but one that slows down from cycle to cycle.

• Because the behavior is an escape response, it can be evoked readily under laboratory conditions. At the same time, it is possible to immobilize the head region of the animal, and insert intracellular electrodes into the brain ganglia, so that recordings can be made from them during the escape swim.

• Using this technique, motor neurons generating dorsal flexions (the dorsal flexion neurons, DFN) or generating ventral flexions (the ventral flexion neurons, VFN) were identified. Not only were they rhythmically active during the pattern just before the appropriate motor movements, but when depolarized, they induced dorsal or ventral flexions (respectively). However, phase reset experiments demonstrated that they were not part of the pattern generator.

• How does one do phase reset experiments if the pattern is slowing down? It was found that one could reliably evoke a fictive swim pattern in isolated ganglia by shocking an appropriate sensory nerve. This made it possible to determine the average durations of each successive swim cycle, and in turn, to define a predicted phase for the behavior in each cycle. It was then possible to determine whether changing a neuron's activity altered the expected phase for that cycle.

• Focusing on the isolated ganglion preparation, evoking the rhythmic pattern, and doing phase reset experiments, it was possible to identify three classes of interneurons:
  • the dorsal swim interneurons (DSI) that activated the dorsal swim motor neurons (three were found),
  • the ventral swim interneurons (VSI) that activated the ventral swim motor neurons (two were found), and
  • cerebral neuron C2, whose burst was interposed between the DSI and the VSI.

• By recording from these neurons in high divalent cation solutions, which suppress polysynaptic activity, Peter Getting was able to characterize the direct synaptic connections between each of these neurons.

• He found that the synaptic connections had both slow and fast components, some of which were excitatory, and some of which were inhibitory.

• None of the neurons generated bursts in isolation, suggesting that the pattern was not due to intrinsic bursting of any neuron, but that it was a network oscillator.

• More recent work, by Paul Katz and Bill Frost, demonstrated that serotonin released by the DSIs acted both as a neurotransmitter and a neuromodulator within the circuit, making it possible for it to generate swimming patterns. They also identified a neuron, the DRI (dorsal ramp interneuron) that acted as a command neuron, i.e., a neuron that was both necessary and sufficient for inducing the swim pattern in the circuit.

• An excellent recent review of the Tritonia escape swim network is here.

Identifying Neuronal Functions

The first task in analyzing a neural circuit is to determine whether a neuron is a sensory neuron, a motor neuron, or an interneuron. An important tool for doing this analysis was introduced in the unit Synaptic Physiology II: Presynaptic Mechanisms and Quantal Analysis, the use of high divalent cations solution (HiDi) to determine whether connections are monosynaptic or polysynaptic.

To make your lives somewhat easier, we have set up the experimental preparation (in simulation, of course), and have helpfully impaled every one of the cells with its own intracellular electrode, which allows you to inject currents into each of the neurons. We have also helpfully bathed the entire preparation in a HiDi solution, so you can focus on determining the monosynaptic connections. Finally, we have helpfully labeled the neurons A through G (for a total of seven neurons). Your goals will be:

• To determine which neurons are sensory neurons, which neurons are motor neurons, and which neurons are interneurons;
• To create a first sketch of the neural circuit (i.e., a sketch of the different neurons and their connections, indicating which connections are excitatory and which are inhibitory).
• To determine, using phase reset experiments, which neurons are part of the pattern generator, and which are not.
• Based on these results, to match the neurons with those described in the Tritonia swim circuit.

Create a table in your lab notebook for organizing your data. Click this link, and copy the table template into your notebook. In each cell of the table, you can record the synaptic interaction between a presynaptic neuron and a postsynaptic neuron, or whether there is a motor effect.

• Thus, for example, if presynaptic neuron X creates an initial excitation, followed by an inhibition of neuron Y, you can enter this into the table as E/I (for excitation, followed by inhibition).
• If there is no effect, you can enter a dash ( - ).
• If the Body Flexion Angle (shown in the last panel) becomes a positive value (by more than just a few degrees), this means that the two halves of the body are closing together dorsally (i.e., undergoing a dorsal flexion), and you can record this using the letter D.
• If the Body Flexion Angle (shown in the last panel) becomes a negative value (by more than just a few degreees), this means that the two halves of the body are closing together ventrally (i.e, undergoing a ventral flexion), and you can record this using the letter V.

A schematic three neuron "ball and stick" diagram.

Once you have filled in the table, you will also create a schematic "ball and stick" diagram of the circuit, an example of which is shown in the figure for neurons X, Y and Z. The table that served as the basis for this diagram looked like this:

Synaptic Connection Table

Presynaptic neuron	Postsynaptic neuron
	X	Y	Z
X	-	E	I/E
Y	I	-	E/I
Z	-	I	-

• Note that the connection from neuron X to neuron Y had a simple excitatory connection. In the "ball and stick" diagram, this is represented as a line at the end of the connection from neuron X to neuron Y, which you can clearly see in the diagram.
• Note that the connection from neuron Z to neuron Y had a simple inhibitory connection. In the "ball and stick" diagram, this is represented as a small filled ball at the end of the connection from neuron Z to neuron Y, which you can clearly see in the diagram.
• Note that the connection from neuron X to neuron Z had a fast inhibition, followed by slow excitation, which is recorded as I/E in the table. Note that in the "ball and stick" diagram, the initial effect of the synapse from neuron X to neuron Z (the small black ball representing inhibition) is closer to neuron Z (i.e., the black ball is closer to neuron Z), followed by the line representing the slower excitation from neuron X to Z.
• Note that the connection from Y to Z has fast excitation followed by slow inhibition, and so the symbols are reversed in the diagram.

You now have enough information to begin identifying the functions and the connections of the neurons. You may want to open up the following simulation in its own page.

• Question 1. Determine the connections among the neurons, their motor effect, and use this to fill in the connection table and classify the neurons as motor neurons, sensory neurons or interneurons.
  • Click on the button labeled Hi Di treatment. Look through the panels. Each pair of panels shows a neuron membrane potential, and then the current that is being used to stimulate that neuron. The very last panel, Body Flexion Angle, shows the motor output, as described above.
  • Note carefully that not all the neurons are at their final resting potential until some time has elapsed; make a note of these neurons, so that you can distinguish these changes from those induced by synaptic connections.
  • To determine the connections from neuron A to all the other neurons, under Cell A Current Clamp, change the number of pulses from 0 to 1, and examine the effects on all the other neurons, and on the Body Flexion Angle. What do you observe? Now change the number of pulses to 2, and then to 3. What do you observe? Record your results in your Synaptic Connection Table.
  • Re-set the number of pulses under Cell A Current Clamp to 0. Now repeat the same protocol for neurons B, C, D, E, F, and G. What do you observe? Record your results in your Synaptic Connection Table.
  • Now sketch a "ball and stick" diagram of the circuit. You may choose to do this by hand, and then turn it into a digital format (e.g., by using your phone to take a picture), and upload it to your lab notebook; or you may use your favorite drawing program. Drawing this diagram and entering it into your lab notebook are a critical part of this assignment!
  • Based on your analysis, which neurons are likely to be motor neurons, which are likely to be sensory neurons, and which are likely to be interneurons? Write down your predictions in your lab notebook.

• Question 2. Determine which neurons are elements of the pattern generator, using a phase reset experiment.
  • Click on the button labeled Natural Swim with Current Clamp. Set the number of pulses under Touch Stimulus to 0. Run the simulation. What do you observe?
  • Now, under Cell A Current Clamp, set the stimulus current to +1 nA (from -1 nA), and the Number of pulses to 1. What do you observe? How do all the other cells respond to this change? Is this consistent with the results you obtained in Question 1?
  • Do a phase reset experiment on Cell B (do not reset the simulation itself):
    • First, measure the time at which the next burst begins after 25000 ms in each neuron that has multiple bursts.
    • Second, change the Cell B Current Clamp as follows: set the Stimulus delay to 25000 ms, the stimulus current to -10 nA, the Pulse Duration to 3000 ms, and the Number of pulses to 1. By doing this, you will have an inhibitory current during the time the cell would have been bursting, completely shutting off its firing during that period (Note that we are completely shutting off the 6th burst). Click this link and copy the provided template into your notebook. Record the parameters you used in the Cell B current clamp in the table.
    • Third, run the simulation again, and again measure the time that the next burst begins in all of the bursting cells you measured above, after 25000 ms. Take the difference between these times and the original times measured above, and record these values in the above table.
    • Fourth, what do you observe? What does this imply about the role of Cell B in the pattern generator? Remember that you are delaying Cell B when you inhibit it, so a delay in its own firing is expected regardless of its function. However, if it is part of the pattern generator, there should be large (greater than 500 ms) delays in the activity of the other bursting neurons.
  • Repeat this for each neuron that is bursting after about 25000 ms. Do not re-use a delay of 25000 ms if the neuron you need to inhibit ends its burst later. For example, if a neuron stops bursting at 27000 ms, use 27000 ms for your delay before applying inhibition to that neuron. Also, make sure the inhibitory pulse is long enough. If the cell originally burst for 4500ms, use this duration, rather than 3000 ms. When identifying the delay for each neuron, again make sure you are targeting the 6th burst. This will allow for uniform observations of how inhibiting the 6th burst of any individual neuron alters the timing of the 7th burst in the rest of the neurons.
    • Remember that we only want to test one cell at a time, so turn off the inhibitory pulse in the previous cell before starting to inhibit the next cell.
    • Make sure that the inhibitory pulse overlaps with the time that the burst was about to begin in the neuron into which the inhibitory pulse is injected, so that you delay the start of its burst. The pulse should be long enough to prevent most of the firing in the neuron you are studying during the burst that would have occurred in that cell after 25000 ms.
    • Record the parameters used in the current clamp of whatever cell you are currently studying in your table.
    • What happens to the time of onset of the other neurons after the injection of the inhibitory pulse? Use this information to classify neurons as either elements of the pattern generator, or not part of the pattern generator using the criteria laid out above. Make sure to write down your conclusions in your lab notebook.

• Question 3. Cell identification.
  • Click on the button labeled "Natural Swim Pattern". Look at the timing of the response of the entire system to a touch stimulus. Does this support your hypothesis about the identity of the sensory neuron? Explain.
  • The correct names of the cells you have studied, and their functions, are listed below:
    • TSN - touch sensory neuron;
    • DRI - dorsal ramp interneuron; receives input from the touch sensory neuron, drives all the other interneurons;
    • DFN - dorsal flexor motor neuron; induces the body to flex dorsally;
    • VFN - ventral flexor motor neuron; induces the body to flex ventrally;
    • DSI - dorsal swim interneuron; drives the DFN;
    • VSI - ventral swim interneuron; drives the VFN;
    • C2 - an interneuron interposed between DSI and VSI.
  • Click this link, and copy the table template into your notebook. Given the data you have obtained in these three questions, match the labels A through G with these cell names; write this down in your notebook. These are your predictions for the underlying circuit, which you will study in further detail in the next unit.

• • Please check to see if the matching you have done is consistent with your predictions in Questions 1 and 2. If not, please explain the discrepancies; this might mean that you will need to re-examine the conclusions you reached in those questions, and perhaps even do additional experiments. Taking this extra step can help to ensure full credit for the laboratory.