OVERVIEW  

The aim of this simulated physiological experiment is to demonstrate how simple nerve networks function and how they can be analyzed.  While watching the behavior of an organism, you will be able to record from, stimulate and lesion the eight neurons which are interconnected to form its entire nervous system.  Your aim will be to determine which neurons receive input from the sense organs which neurons control the muscles and which neurons form excitatory or inhibitory synapses on other neurons.  You will have to explain how the properties of the neural network account for the behavior of the animal.

BACKGROUND AND INTRODUCTION  

The idea that we are all automata is an old one.  Democritus proposed that the motion of particles in the brain determines our actions, Descartes pictured the crucial process as a flowing of fluids through nerves and contemporary scientists think of our behavior as being determined by the spread of electrical excitation through networks of synaptically connected neurons.  On one hand, these explanations are quite different, reflecting a steadily increasing knowledge of brain anatomy and physiology.  On the other hand they have in common the conviction that brain function can be regarded as the sum of many small events each of which obeys physical laws.

The idea that the brain is a neuronal machine is plausible because the neuron is a compact and powerful building block for circuits that perform logical functions.  The neuronal cell-type is highly specialized with regard to physiology, structure and chemistry.  (1) Physiology.  The neuronal membrane is excitable it can undergo brief (1 msec) "twitches" (action potentials).  The action, potentials spread rapidly (at a velocity of 1-100 m/sec) outward from their point of origin like waves on a pond. (2) Structure.  Unlike other cells, neurons have long dendritic and axonal processes.  The axons have excitable membranes down which action potentials originating at the cell body travel.  The axons form synaptic contacts on the dendrites and cell bodies of other neurons, (3) Chemistry.  When the action potential reaches the end of the axon it causes the release of transmitter chemicals into the synaptic space.  The chemical released at some synapses is excitatory, i.e., it increases the likelihood that the next cell in line will produce an action potential.  The chemical at other synapses is inhibitory. 

The functional power of neuronal network arises from the fact that whenever a neuron fires a burst of action potentials, it increases or decreases the firing tendency of other neurons on which its axon synapses.  A neuron could form an excitatory synapse with another neuron.  Thus, when the first neuron fires, the second neuron will tend to fire a short timer later.  A neuron could also form an inhibitory synapse with another neuron.  Firing by the first neuron will tend to inhibit the second neuron from firing.  Also, a removal of this inhibitory influence (by the ceasing of the first neuron's firing) will tend to make the second neuron fire as it "rebounds" from the inhibition.

Waves of activity can thus spread from cell to cell in linear, branching or cyclical patterns.  Consider some of the simple functions that can be performed by networks of just a few synaptically connected neurons: 

  1. The monosynaptic reflex.  The simplest of all neural networks consists of two neurons.  The knee-jerk is based on a two-neuron network.  The first neuron is a sensory neuron excited by muscle stretch.  In the spinal cord, it forms an excitatory synapse on the second neuron, a motor neuron whose axon terminates on the muscle.  Stretching the muscle by tendon tap causes cell 1 to fire a burst of action potentials, which in turn causes a release of an excitatory transmitter at the synapse.  This causes cell 2 to fire a burst of action potentials; these propagate down the axon of cell 2 to the muscles where release of excitatory transmitter causes the muscle to contract. 

  2. The polysynaptic reflex.  In circuits producing more complicated reflexes, additional neurons intervene between the sensory neuron and the motor neuron.  These are called interneurons.  The directed withdrawal of a limb from a painful stimulus is mediated by multineuronal circuits in the spinal cord. 

  3. Central pattern generation.  Many processes in the brain wax and wane in a regular rhythmic pattern.  The period may be short, as in breathing, or long as in the circadian cycle of sleeping and waking.  The neural circuits generating rhythmic activity in the vertebrate brain are not yet well understood, but several oscillating systems have been analyzed successfully in invertebrates.  In several cases it has been shown that rhythmic activity arises from networks of neurons connected by inhibitory synapses.  For example, the pattern of alternating contraction and relaxation by which leeches swim is generated by a ring of five neurons each of which inhibits its neighbor on one side.  At any given time, every other cell in this ring is in at state of excitation.  However, the pattern is not stable because there is an odd number of cells in the ring.  If cell 1 is active, then cell 2 is inhibited.  If cell 2 is inhibited, then it is not inhibiting cell 3, which therefore fires.  If cell 3 is firing then it is inhibiting cell 4.  If cell 4 is inhibited, then it in not inhibiting cell 5, which, therefore fires; but, cell 5 inhibits cell 1, which therefore will grow silent, whereupon cell 2 will begin to fire, and so forth.  Do you see why the pattern of activity in this circuit is perpetually oscillatory?  If there were an even number of neurons in this circuit then it would behave like a memory element in a computers it could be set to either of two states. Do you see why?

The basic task of neuroscientists studying any simple neuronal network is to work out its synaptic wiring diagram and then show how the behavior of the system arises naturally from the properties of the individual neurons and their connections.  There are several basic techniques available for this purpose, such as: 

  1. Recording.  Whenever a neuron fires an action potential there is a brief change of voltage in its immediate vicinity.  It is possible to detect this voltage change with a microelectrode wire completely insulated except at its tiny (one thousandth of a millimeters) tip.  The tip of the microelectrode is positioned either within or immediately next to a neuron and the far end of the wire is hooked up to the input of an amplifier.  The output of the amplifier is led to a speaker.  Each action potential is heard as a brief click on the speaker.  This technique is referred to as single-neuron recording.  By recording from a cell, one can determine under what behavioral circumstances it is active.  Exploring the spinal cord with a microelectrode, one would eventually find some neurons which fire action potentials in response to each tap of the knee tendon.

  2. Stimulation.  By use of the same kind of microelectrode, it is possible not only to record the natural activity of a neuron, but also to produce artificial activity.  A small voltage applied to the top of the electrode is conveyed to the tip, where it excites the cell's membrane, causing it to fire a burst of action potentials which continues for as long as the voltage is applied.  By stimulating a neuron, one can ask what behavioral consequences ensue from its activity.  Having found a neuron responsive to tendon-tap, one might stimulate it in order to see whether contraction of a leg muscle occurs (this might not be the case; for example the sole function of the neuron might be to relay sensory information to higher centers in the brain).

  3. Lesioning.  It is also possible to destroy a neuron by beating it with a large electrical current passed through the microelectrode tip.  This technique is referred to as electrocoagulation or lesioning.  By destroying a cell, one can ask whether it is necessary for a certain function.  For example one might destroy a neuron suspected to be a crucial link in the knee-jerk reflex and then test for the continued occurrence of the reflex. 

  4. Combinations.  The techniques described above can be especially powerful when used in combination.  For example consider a simple reflex pathway consisting of cell A and cell B.  Both fire in response to tendon tap; stimulating either of them induces knee jerk; lesioning either of them abolishes the knee-jerk reflex.  How can we tell which is the sensory neuron and which is the motor neuron? (1) We could lesion A and stimulate B, or vice versa: in one case a knee jerk should result; in the other case it should not., (2) We could stimulate A while recording from B and vice versa: in one case the cell being recorded from should fire in response to stimulation of the other cell; in the other case it should not. (3) We could lesion A and record from B or vice versa; in one case, the cell being. recorded from should respond to tendon tap.  In the other case, it should not.  Do you see why these tests distinguish between the motor neuron and the sensory neuron?

In this simulation you will be using these techniques to analyze a neuronal network consisting of eight neurons connected by excitatory and inhibitory synapses.

METHODS

    1.  Once the program is up and running, start a lab session.  You should now see a frog walking across the bottom of the screen.

    2.  Observe the animal.

    3.  Think about the nervous system.  You might infer (correctly) from your observations that the frog’s nervous system consists of the following neurons:

                Sensory neurons

                Motor neurons

You might also infer that there must be some neurons which are connected to each other and to the limb motor neuron to form a circuit that causes rhythmic bursts of activity.  To make your job easier, we’ll tell you that there are:

                Two interneurons

IMPORTANT HINT: In the Rhbit network, each neuron sends only one projection (axon) to only one other neuron, but each neuron can receive more than one input.

    4.  The eight neurons of the frog’s nervous system are represented at the bottom of the screen.  You can select a neuron by moving the mouse over the neuron and pressing the left mouse button.  Then you can perform the following operations on the selected neuron:

        Select the operation by pressing one of the three buttons to the left of the eight neurons.  The button's icon will change as the operation is selected, reflecting that the next time the same button is pressed the corresponding operation will be cleared.  Note that the recording icon is a microphone, the stimulating icon is a battery, and the lesioning icon is a knife.

        You can perform each operation by itself or in combination with one other operation.  You can only perform two different operation simultaneously.  By recording from, stimulating, and lesioning selected neurons, you can determine which neurons are sensory, which are motor, and which are interneurons.  You can also determine the connections between neurons and whether they are excitatory or inhibitory.