Central Pattern Generators II: Analysis of Tritonia Escape Swim Circuit Interneurons

Introduction

• In the previous unit, you classified neurons as sensory neurons, motor neurons, or interneurons.

• You also worked out the connections among the neurons, and, using phase reset experiments, were able to establish which neurons were part of the pattern generator.

• Just as investigators refer to the entire set of genes that an organism possess as the genome, and the entire set of proteins that an organism uses as the proteome, they have begun to refer to the entire set of connections that are used in an organism's nervous system as the connectome.

• In the previous unit, you worked out the connectome for a simple circuit.

• However, having the connections is not enough to understand circuit function. An excellent recent review of this problem is here.

• Understanding how a circuit works requires an understanding of the dynamics of the circuit, the pattern of activation of the different neurons over time.

• This may still not be enough, because ultimately the circuit function must be studied in the context of natural behavior, which may significantly alter the outputs of the circuit; but it is certainly a critical next step.

• In this unit, you will analyze the interneurons of the Tritonia swim circuit to obtain a fuller understanding of circuit dynamics.

Command Neurons

• One way in which behaviors could be organized is for specific neurons to switch on an entire behavior. Sensory input would be channeled to that neuron, and it, in turn, would be responsible for initiating behavior. Such an organization might be especially effective for escape behaviors, which override other behaviors once they are triggered.
• Although investigators informally called neurons "command neurons" if they could induce an entire complex motor pattern when fired, the review of Kupfermann and Weiss 1978 first suggested formal criteria for defining a command neuron:
  • A command neuron is defined as
    • A single interneuron interposed between sensory neurons and motor pattern generating neurons;
    • Activating it is sufficient to induce the entire pattern of activity;
    • If it is inhibited, the pattern of activity cannot be generated, so that it is necessary for that pattern.

• For a command neuron to be responsible for activating an entire pattern of activity, it should fire before the neurons that it controls.

• A command neuron should directly control the activity of the key neurons responsible for the neural pattern.
  • The command neuron should directly excite pattern generator elements.
  • Showing that there is an excitatory connection is not enough; it is also important to show that activating the system using the sensory input that ordinarily induces an escape swim motor pattern does so through activity of the command neuron.
  • If other neurons within the pattern generator can maintain excitation through their own activity, it is important to show that the command neuron is sufficient to excite the neuron that initiates the pattern, even if these neurons are inhibited.
  • If the command neuron is inhibited, this should block the normal response to excitation; thus, the command neuron is necessary for the normal pattern to be generated.

Analysis of Circuit Dynamics

• The simulation is based on a model of the network that was published by Bill Frost, Paul Katz and their students, which can be found here.

• Question 1. What are the different characteristics of the key interneurons that contribute to the swim central pattern generator? Please click the button at the bottom of the simulation (you may have to scroll down to see it) that is labeled "Isolated Cell Defaults". This disconnects the neurons from one another by disabling their synaptic connections.
  • Let us explore the responses of the individual neurons to identical inputs. Set each of the "Current Clamp" entries for C2, DSI, VSI-B and DRI to 500 ms Stimulus Delay, 2.5 nA Stimulus Current, 5000 ms Pulse duration, 500 ms Inter-stimulus interval, and set the Number of pulses to 1. Run the simulation.
  • Use the cursor to measure the initial and final firing interspike interval of the C2. You may need to "zoom in" on spikes at the beginning and end of the current pulse using your browser's ability to zoom in on an image. Use the peaks of the spikes for your computations.
  • To compute the firing frequency (in cycles/second, or Hertz, usually abbreviated Hz), you need to take the reciprocal of the number you measured, and convert it to seconds.
  • For example, if the interspike interval were 50 milliseconds, then the firing frequency between those two spikes would be $$\begin{align} \frac {50 \ \text{ms}}{\text{cycle}} &= \frac{1}{50 \times 10^{-3} \text{second/cycle}} \\ &= \frac{10^{3}}{50} \ \text{cycles/second} \\ &= 20 \ \frac{\text{cycles}}{\text{second}} \\ &= 20 \ \text{Hz}. \end{align}$$
  • To compute the percentage change in the firing frequencies, assume that the initial firing frequency is $f_\text{initial}$, and the final firing frequency is $f_\text{final}$. Then the percentage change in firing frequency is $$\frac{(f_\text{final} - f_\text{initial})}{f_\text{initial}}.$$
  • For example, if the C2 neuron initially fired at 20 Hz, and by the end of the stimulation pulse, it fired at 10 Hz, the percentage change would be $$\begin{align} \frac{(10 - 20)}{20} &= \frac{-10}{20} \\ &= -0.5 \\ &= -50\%, \end{align}$$ or a 50% decrease in its firing frequency over the course of the stimulation pulse.
  • Perform the same measurements and calculations for DSI, VSI-B, and DRI. Compare and contrast the changes in firing frequency in the neurons. How do they differ from one another? What ionic conductances could account for these differences? Explain.
  • Please measure the delay from the onset of stimulation to the first spike. Which neuron shows the longest delay? Describe a conductance that you learned about earlier in the course that could account for this delay.
  • Design an experiment to determine whether the conductance that you described could account for the delay in the VSI-B neuron, carry out this experiment, and show your results. Hint: Review Question 9 in Action Potential IV: Hodgkin-Huxley Equations and Other Conductances. Take pictures.

• Question 2. Develop a prediction for the spread of excitation and inhibition through the circuit.
  • Open the simulation and click on the button labeled "Modulated Swim Defaults". Scroll down to the Simulation Settings, and change the simulation duration to 10000 ms. Run the simulation.
  • Measure the time of the first spike of the C2, DSI, VSI-B and DRI neurons.
  • From these measurements, and from the synaptic connections that you determined in the previous unit, formulate and write down your hypothesis for the sequence in which these neurons are activated.
  • Modify your "ball and stick" diagram to use these names for the appropriate neurons in the circuit. Make sure to show this in your notebook.
  • How could you test your hypothesis? Propose an experiment, and please write it down in your notebook.

• Question 3. Trace the inhibition and excitation through the circuit.
  • Use the same settings as you did for Question 2.
  • Set the number of DRI pulses to zero.
  • Using the current clamp controls, examine the relationship among C2, DSI, and VSI-B.
  • To do this systematically, inject excitatory current into one of the cells (e.g., C2), while injecting inhibitory current into the other two cells (e.g., DSI and VSI-B) to prevent them from spiking. Then, remove the inhibition from each of the two cells, one at a time (i.e., first from DSI, then re-inhibit DSI, and remove inhibition from VSI-B).
  • Record your results in your notebook. Take pictures!
  • Carefully examine your data. How well does the data support your original hypothesis for how the circuit works? Refine and revise your hypothesis if necessary.
  • Based on these results, describe the circuit dynamics, i.e., how excitation and inhibition spread through the circuit. Use the connectivity data and the results you obtained from Question 2.

• Question 4. Which neurons can generate oscillations?
  • Open the simulation and click on the button labeled "Modulated Swim Defaults".
  • Set the number of DRI pulses to zero. Run the simulation. What do you observe?
  • Can a long depolarization of any one of the neurons C2, DSI or VSI-B generate a rhythmic pattern? To examine this question, under the current clamp for each neuron, do the following:
    • Make sure that the number of pulses in each of the three neurons is zero initially (otherwise, you will be exciting more than one neuron).
    • Set the pulse duration to 50000 ms.
    • For each neuron in turn, set the Number of pulses to 1.
    • Set the stimulus current to values that induce repeated bursting in the neuron, but not steady firing.
    • What happens as you excited each of the three neurons? Take pictures.
  • Explain the results you have obtained based on the connectivity diagram, and your analysis of the circuit dynamics in Questions 2 and 3.

• Question 5. Determine whether an interneuron is a command neuron.
  • Based on the timing of activation of the interneurons, which one could be a command neuron?
  • To test this idea when the sensory system is still intact, first refer to your answer to Question 3 of the previous problem set. What was the label of this interneuron in the simulation you used then?
  • For the following, open up the simulation from the previous unit (Tritonia swim CPG Simulation: unknown circuit with behavior) and click the "Natural Swim with Current Clamp" button.
    • Describe an experiment to demonstrate that this neuron is sufficient to induce the entire pattern. Use the simulation to carry out this experiment. Take pictures.
    • Describe an experiment to demonstrate that this neuron is necessary to induce the entire pattern. Use the simulation to carry out this experiment. Take pictures.
  • Compare your results to those reported in Single Neuron Control over a Complex Motor Program. Please refer to specific figures in the paper and compare them to the results you obtained.