Synaptic Physiology I: Postsynaptic Mechanisms

Introduction

• Up until this unit, we have considered nerve cells in isolation.

• We have learned that
  • the passive properties of neurons can filter their responses to inputs;
  • the active properties of neurons are due to a variety of ion channels that can be gated by voltage or calcium, and can provide them with complex properties such as the ability to generate action potentials, to change their responses to steady inputs, or to burst spontaneously; and
  • the complex shapes of neurons not only allow them to send information rapidly over long distances, but allow them to assimilate complex inputs over space and time.

• It's now time to understand how nerve cells connect to one another, because it is when neurons work together in groups - in neural circuits or systems - they are capable of generating far richer behavior than when they work in isolation.

• In this unit, we will briefly describe two types of interactions among neurons - ephaptic transmission and electrical connections - and then provide a qualitative overview of the process of chemical connections between nerve cells.

• The rest of this unit will be devoted to understanding how neurons respond to chemical signals from other neurons. In the next unit, we will explore how neurons generate and release chemical signals to send messages to other neurons. In the units after that, we will focus on how connections can change strength with experience (a process referred to as plasticity), an important physiological basis for learning and memory.

Forms of Communication among Neurons

• Neurons communicate with one another in three ways:
  • Ephaptic transmission
  • Electrical synapses
  • Chemical synapses

Ephaptic Transmission

• Because neurons are sensitive to electrical and magnetic fields, the large scale changes in those fields generated by large groups of neurons - volume potentials - may affect the activity of nearby neurons upon which these fields impinge; this is referred to as ephaptic transmission.
• Although ephaptic transmission is recognized as a potentially important influence on nerve cells, it has been difficult to study. There are claims that it may be present after damage to neurons, and be a basis for chronic pain.

Electrical Synapses

Figure 1: Injecting depolarizing current pulses into one neuron (bottom trace) induces a depolarizing voltage change in a second, electrically coupled neuron (top trace).

Figure 2: Injecting hyperpolarizing current pulses into one neuron (bottom trace) induces a hyperpolarizing voltage change in a second, electrically coupled neuron (top trace).

Figure 3: The electrical equivalent circuit for an electrical synapse.

• Much more study has been devoted to the specialized connections between neurons, their synapses.
• Before it was possible to visualize these specialized connections precisely - which only occurred after the invention of the electron microscope, because they were so small - there was a vigorous debate among researchers. Some insisted that all connections were electrical, and that the neurons formed a continuum of connected cells, known as a syncytium; others insisted that all connections were due to chemical messages sent between neurons, and that the connections should be referred to as synapses, the places where neurons are "held together".
• Like many debates in biology, the question "Are synapses electrical or chemical?" was finally answered unequivocally: Yes, they are. (This is the answer that often occurs in response to these debates, which is worth remembering for the future; in many of these debates, both answers may be correct.)
• We will focus briefly on electrical synapses, whose functional roles are important in many neural circuits.
• Nerve cells may be connected together via tight gap junctions. These are formed by proteins called connexons, creating a barrel-shaped opening in the membrane which is apposed to a similar opening in the membrane of the second neuron. The staves of the barrel are made of a protein called connexin.
• If two cells (let us call them $A$ and $B$) are connected by an electrical synapse, then injecting depolarizing current (positive charges) into $A$ (the bottom trace in Figure 1) could induce a depolarization in neuron $B$.
• In contrast, injecting hyperpolarizing current (negative charges) into $A$ can induce a hyper polarization in neuron $B$, as shown in Figure 2.
• If one neuron can both depolarize and hyperpolarize another neuron, that is usually a very strong indicator that the two neurons are electrically coupled to one another.
• Interactions between electrically coupled cells can be characterized using the electrical equivalent circuit shown in Figure 3.
• If a current is injected into the lefthand neuron ($A$), setting it to voltage $V_1$, then the voltage to which neuron $B$ will be set will be determined by how much current flows directly through the electrical synapse, and how much instead flows out through the membrane of neuron $B$. Once the transients due to the capacitance of the membrane settle out, the voltage $V_2$ to which the righthand neuron ($B$) goes will be $$\frac{V_{2}}{V_{1}} = \frac{R_{2}}{R_{2} + R_{e}} = K_{1,2}, \quad\text{(Equation 1)}$$ where $R_2$ is the membrane resistance of neuron $B$, and $R_e$ is the resistance of the electrical synapse.
• Similarly, if a current is injected into the righthand neuron ($B$), setting it to voltage $V_2$, then the voltage to which neuron $A$ will be set will be determined by how much current flows directly through the electrical synapse, and how much instead flows through the membrane of neuron $A$. The final voltage $V_{1}$ of neuron $A$ will be $$\frac{V_{1}}{V_{2}} = \frac{R_{1}}{R_{1} + R_{e}} = K_{2,1}, \quad\text{(Equation 2)}$$ where $R_1$ is the membrane resistance of neuron $A$.
• The terms for $K_{1,2}$ and $K_{2,1}$ are coupling coefficients.
• Although the picture shows two neurons whose sizes are identical, which would likely lead them to have identical coupling coefficients for currents injected into either neuron, if one neuron is much larger than the other, its membrane resistance is likely to be much lower. As a consequence, current injected from the smaller neuron, which can generate much less current, would have very little effect on the larger neuron; in contrast, the larger process will almost certainly be able to generate a larger current, and because of the larger input resistance of the smaller neuron, its voltage might change much more significantly. So even though electrical connections are bi-directional in principle, they can in practice effectively work primarily in one direction.
• Similarly, although this is less common, the gap junctions can sometimes show voltage dependence, so that the amount of current carried by the electrical synapse will vary with the membrane voltage of the two neurons.
• Electrical synapses have several features that make them especially useful in circuits:
  • They are very fast, because current essentially flows from one neuron to the next. Thus, they are often found in escape systems, where speed is of the essence.
  • If multiple neurons are coupled via electrical synapses, they can act to synchronize one another's activity. As some of the neurons are excited, their excitation spreads to other neurons, until all burst together.
  • Electrical connections can also provide multiple pathways for current to escape from a neuron cell, shunting its response to input until a large enough input is provided.
• Although we will now turn to a consideration of chemical synaptic transmission, it is important to realize that within the nervous system, transmission may not be either electrical or chemical; some neurons form both electrical and chemical synapses with one another.

An Overview of Chemical Synaptic Transmission

Figure 4. A schematic view of chemical synaptic transmission.

• We now turn to chemical transmission of messages, which is both somewhat slower and much more flexible than electrical synaptic transmission.

• Figure 4 provides a schematic overview of the process of chemical transmission, which will allow us to sketch the process, the details of which will be filled in both in this and the next unit.
  • On the left side of Figure 4, we see a schematic action potential (in red) invading the presynaptic terminal.
  • As the action potential depolarizes the presynaptic terminal, it induces voltage-dependent calcium channels to open.
  • Calcium ions are in much higher concentration outside of the nerve cell than they are within the nerve cell, and so calcium ions move along their concentration gradient into the nerve cell.
  • The rise in the concentration of calcium ions within the presynaptic terminal triggers a complex molecular cascade that leads synaptic vesicles to fuse with the membrane of the presynaptic terminal.
  • As the figure shows, once a synaptic vesicle fuses with the membrane, it releases its contents into the cleft between the presynaptic and postsynaptic terminal (this region is referred to as the synaptic cleft).
  • The small triangles schematically represent the contents of the synaptic vesicle: chemicals that are used to transmit messages from one neuron to another, or neurotransmitters.
  • Neurotransmitters diffuse across the synaptic cleft, and some may find receptors that have a complementary shape, allowing the transmitter to bind to the receptor.
  • The ability to release many neurotransmitter molecules from synaptic vesicles makes it possible to greatly amplify the signal from the presynaptic neuron in the postsynaptic neuron, since many receptors may be bound in a short time, and many ion channels in the postsynaptic neuron may respond.

• Receptors fall into two broad categories: ionotropic receptors and metabotropic receptors.
  • Ionotropic receptors are directly linked to ion channels, whose conformation changes once their receptor is bound; this may allow the channels to open or close, depending on how their conformation changes.
    • Thus, in addition to the ungated channels and the voltage-gated channels you have already learned about, there is a third category of ion channel: one that is gated when a chemical binds it, also called a ligand-gated channel.
  • Metabotropic receptors are linked to molecular cascades, and these cascades may in turn lead to the opening or closing of ion channels. Although these cascades are, in general, slower, they can affect larger populations of ion channels, and can thus serve to amplify the signal, as well as to increase its duration in time.

• The effect of the channel on the postsynaptic membrane potential depends on the ions that flow through the channel and whether the channel opens or closes in the presence of neurotransmitter.
  • If the channel opens in the presence of neurotransmitter:
    • ... and if the opening allows ions to flow that depolarize the membrane, this will usually lead to excitation of the postsynaptic neuron.
    • ... and if the opening allows ions to flow that hyperpolarize the membrane, this will lead to inhibition of the postsynaptic neuron.
  • If the channel closes in the presence of neurotransmitter:
    • ... and if the closing prevents ions from flowing that would depolarize the membrane, this will usually lead to inhibition of the postsynaptic neuron.
    • ... and if the closing prevents ions from flowing that would hyperpolarize the membrane, this will lead to excitation of the postsynaptic neuron.

Postsynaptic Mechanisms of Chemical Transmission

• We can now begin to fill in more details about the process of chemical synaptic transmission. We will focus first on the response of the postsynaptic terminal.

• Once receptors have been bound by neurotransmitter, they may induce ion channels to open or close, either directly or indirectly.

• A direct (ionotropic) effect occurs if the binding of the transmitter to the receptor alters the shape of the receptor itself, and the receptor is part of the ion channel; the receptor's change in shape alters the ion channel's shape, and induces it either to open or close.

• An indirect (metabotropic) effect occurs if the binding of the transmitter to the receptor alters a molecule within the nerve cell that triggers a sequence of molecular changes that in turn lead to ion channels opening or closing.

• In a subsequent unit, we will have more to say about neurotransmitters and their receptors.

• For now, let us focus on what happens once the ion channel's ability to conduct ions has been altered.

• For concreteness, let us focus on a particular system: receptor binding induces channel opening, and the channel permeates both sodium and potassium ions, and has a reversal potential of +20 mV.

• This is similar to the properties of the neuromuscular junction, the specialized synapse that transmits messages, using the neurotransmitter acetylcholine, from motor neuron to skeletal muscles, and whose activation can induce a muscle to contract rapidly and strongly.

• What is a reversal potential? If an ion channel permeates more than one ion, then it does not try to move the membrane to a potential corresponding to the equilibrium or Nernst potential of a single ion; rather, it pulls the membrane to a potential that is somewhere between the equilibrium potentials of the ions that it permeates. The exact value of this potential depends on how strongly it permeates the different ions. For example, if a channel is twice as permeable to sodium ions as it is to potassium ions, then its reversal potential will be strongly biased towards the equilibrium or Nernst potential of sodium ions. However, since it also permeates potassium, the actual reversal potential will not be as positive as the sodium ion equilibrium or Nernst potential.

• The reason that this potential is called a reversal potential is illustrated clearly in Figure 5.

Figure 5: Reversal potential for an excitatory synaptic input in current clamp.

• In this experiment, the nerve cell is held in current clamp, and enough current is injected to hold it near different voltages (-80 mV, -40 mV, 0 mV, 20 mV, 40 mV, or 80 mV). Shortly after the recordings begin, the presynaptic neuron is activated, inducing a postsynaptic potential in the nerve cell from which voltage recordings are being made.

• As you can see, when the neuron is held at potentials negative to +20 mV, the resulting voltage change is depolarizing (in a positive direction), as the cell is pulled up to the +20 mV potential of the synapse.

• When the neuron is held exactly at 20 mV, there is no change in voltage. This make sense: since the membrane potential and the reversal potential are equal, the driving force (the difference between these two voltages) has gone to zero, and so the net current through the channels is zero.

• When the neuron is held at potentials more positive than 20 mV, the result is a negative voltage deflection (hyper polarization), as the postsynaptic potential attempts to pull the membrane back down to the reversal potential.

• From this, it is clear that the reversal potential is that potential at which the net ion flow into and out of the channel due to the electrical and chemical gradients are exactly equal and opposite. When the membrane potential is more positive than the reversal potential, a net outward (hyperpolarizing) current develops; when the membrane potential is more negative than the reversal potential, a net inward (depolarizing) current develops. Thus, the current reverses direction when the membrane potential is held at values above or below the reversal potential.

• To understand both the shape and the time course of the postsynaptic potential, it is very helpful to examine the electrical equivalent circuit for the synapse and the postsynaptic membrane.

• As shown in Figure 6, panel 1, we can represent the excitatory postsynaptic ion channels as a conductance ($g_{EPSP}$) in series with a battery at the reversal potential (in this example, 20 mV), and controlled by a switch. In the absence of transmitter, for this example, the switch disconnects the synaptic ion channels from those of the rest of the membrane. They are represented to the right of the synaptic channels as a leak conductance (for this example, at -90 mV) in parallel with the membrane capacitance.

• Before the synapse is activated, the postsynaptic cell is at rest. As shown in Figure 6, panel 1, no net current flows through the synaptic channels ($I_{EPSP} = 0$), no current flows through the ungated leak channels ($I_{L} = 0$), and no current flows through the membrane capacitance ($I_{C} = 0$).

Figure 6: Electrical equivalent circuit for a chemical synapse before activation (1), during the rising phase (2), at the peak of the synaptic potential (3), and during the falling phase (4).

• Once transmitter binds to the postsynaptic membrane, the ion channels for this synapse open, and allow the ions that they permeate to flow through them. The membrane, which was sitting at -90mV, begins to rapidly depolarize towards the reversal potential of the synapse (+20 mV). The activation of the synaptic ion channels is represented in the electrical equivalent circuit by showing that the switch has closed (Figure 6, panel 2). Current flows into the cell through the synaptic ion channels, those gated by binding the transmitter; current flows out of the adjacent non-synaptic membrane through the ungated leak channels and through the capacitor, so that $I_{EPSP} = I_{L} + I_{C}$.

• At the peak of the synaptic potential, the capacitance of the membrane is neither charging nor discharging, so that the membrane is in a steady state (Figure 6, panel 3). At this point, $I_{EPSP} = -I_{L}$.

• After transmitter binds receptors, the receptors and transmitter may unbind, and the transmitter may be rapidly removed from the synaptic cleft by enzymes that inactivate the transmitter, by being re-uptaken into the presynaptic terminal, or by diffusing away from the synaptic cleft. In addition, the receptor may inactivate, becoming unresponsive to the transmitter; or the channel may inactivate.

• The falling phase of the postsynaptic potential (Figure 6, panel 4) is now due entirely to the ungated leak channels and the capacitor, which discharges as the membrane returns to its prior resting potential. Thus, the time course of the falling phase of the postsynaptic potential is set by the passive membrane properties, whereas the rising phase is primarily due to the rate of transmitter release, diffusion across the synaptic cleft, binding to receptors and channel activation, all of which may happen very rapidly.

• If one looks at the level of individual synaptic channels, using a patch clamp electrode, they show the same kind of probabilistic flickering open and closed as do the voltage-gated ion channels. In the presence of transmitter, the probability of their opening or closing is significantly altered. Just as we saw with the voltage-gated ion channels, the sum of the current through the synaptic channels yields the macroscopic current with the correct overall time course.

• Synaptic ion channels are also known as ligand-gated channels, since they must bind to something in order to change their state.

• Ligand-gated channels may be completely insensitive to the voltage across the neuron's membrane. Other ligand-gated channels may also be gated by voltage.

• At the peak of the synaptic potential, the membrane is in a steady state, and $I_{EPSP} + I_{L} = 0,$ which allows us to determine the actual membrane voltage that will be reached by the nerve cell due to the synaptic input.

• From the channel equation, we know that $I_{EPSP} = g_{EPSP}(V - E_{EPSP})$, and $I_{L} = g_{L}(V - E_{L})$. Combining these equations, we can solve for the membrane voltage at the peak of the synaptic potential: $$V_{m} = \frac{g_{EPSP} E_{EPSP} + g_{L} E_{L}}{g_{EPSP} + g_{L}}. \quad\text{(Equation 3)}$$

• For simplicity, we focused on a single kind of postsynaptic potential, one that opened ion channels (induced an increased conductance), and had a reversal potential that was significantly more depolarized than the resting potential of a neuron, and thus would not only depolarize the neuron, but will also tend to increase the likelihood that the postsynaptic neuron will fire an action potential. This is an example of an excitatory postsynaptic potential, usually abbreviated as EPSP.

• If the ion channels that are gated by neurotransmitter and begin to conduct after the transmitter has bound them are permeable to ions whose equilibrium or Nernst potentials are more negative than the resting potential, then the cell will almost certainly hyperpolarize, and become less likely to fire an action potential. This is an example of an inhibitory postsynaptic potential, usually abbreviated as IPSP.

• An important form of inhibition can occur if the transmitter opens channels that increase the permeability of the membrane to ions whose equilibrium or Nernst potential is close to the resting potential, such as chloride ion channels, whose equilibrium potential is often close to or at the neuron's resting potential. Even though the neuron may not hyperpolarize, the opening of the postsynaptic ion channels increases the neuron's conductance, in turn reducing the postsynaptic neuron's response to injected current, and thus reduces the postsynaptic neuron's excitability. Such inhibition is referred to as shunting inhibition (since injected current is shunted by the increased conductance) or silent inhibition (since the membrane potential may not change, even though the neuron is less excitable).

• These excitatory or inhibitory synaptic potentials all result from opening ligand-gated ion channels, which increases the membrane conductance.

• In addition, ligand-gated channels can be closed as a consequence of binding transmitter, decreasing the membrane conductance.

• As a consequence, when these kinds of ligand-gated ion channels are activated, the equilibrium or Nernst potential of the ions to which they are permeable have less effect on the overall membrane potential, and so have the opposite effects of increased conductance postsynaptic potentials.

• Thus, if a ligand-gated channel is open in the resting postsynaptic membrane, and is permeable to sodium ions, then when the channel binds transmitter and closes, the membrane hyperpolarizes, because the membrane is now less permeable to sodium ions.

• Similarly, if a ligand-gated channel is open in the resting postsynaptic membrane, and is permeable to potassium ions, then when the channel binds transmitter and closes, the membrane depolarizes, because the membrane is less permeable to potassium ions.

• Here is a table that summarizes increased and decreased conductance post-synaptic potentials; it is not exhaustive, but should help clarify the concept:

	Increased conductance	Decreased conductance
Sodium ion conductance	Depolarizing	Hyperpolarizing
Potassium ion conductance	Hyperpolarizing	Depolarizing

• A useful way of telling whether a synaptic potential involves an increased or decreased conductance is to inject small hyperpolarizing pulses into a nerve cell during the synaptic potential.

• If the synaptic potential involves an increase in conductance, then a fixed amount of current will generate smaller voltage changes during the synaptic potential.

• Conversely, if the synaptic potential involves a decrease in conductance, then a fixed amount of current will generate larger voltage changes during the synaptic potential.

Analyzing Postsynaptic Membrane Responses

• You now have enough information to explore the properties of the postsynaptic membrane using a simulation. As usual, it may be easier to open the simulation in a separate window as you work through the problems.

• In this simulation, two cells are connected by a single chemical synapse. We have intracellular electrodes in both the presynaptic and postsynaptic cells, and we will use them to explore some of the basic properties of synapses.

• Question 1: First, let's explore the properties of the default synapse, which is similar to the cholinergic synapse one might see at a vertebrate neuromuscular junction. Note that the reversal potential is 0 mV. Run the simulation and observe the results. Try changing the stimulation parameters to get a better feel for what is happening. Is the postsynaptic potential generated by this neuron excitatory or inhibitory? Explain the reasoning behind your answer. To clarify the results, it may be helpful to set the Fast Transient Sodium Conductance and the Delayed Rectifier Potassium Conductance to 0 in the postsynaptic cell, and re-run the simulation. What do you observe? Is the effect excitatory or inhibitory? Explain.

• Question 2: Reset the simulation to its default parameters, and set the conductance of the postsynaptic voltage gated potassium and sodium channels to 0 µS. Are the two postsynaptic potentials the same height? Why or why not? Try extending the simulation time to 80 ms and number of presynaptic pulses to four from two to see if this supports your hypothesis. Explain.

• Question 3: Reset the simulation to its default parameters, and then set the rise time constant to 20 and the decay time constant to 50 ms (this controls how slowly the synapse reacts to the neurotransmitter). How many action potentials are generated in the postsynaptic neuron after this change? To explain what has happened, it will be helpful to see what the synapse is doing if you set the conductances of the postsynaptic gated potassium and sodium channels to 0 µS. Explain what you observe.

• Question 4: What happens on a slightly longer time scale? Leaving the rise and decay time constants at 20 ms and 50 ms, increase the number of presynaptic pulses to 10 and the total duration to 200 ms. Restore the sodium conductance to 120 µS and the potassium conductance to 36 µS. Slightly reduce the maximum synaptic conductance to 0.4. Run the simulation. Why do only some of the presynaptic spikes produce postsynaptic potentials? To answer this question, it will help you to zoom in on the changes in the membrane potential of the postsynaptic neuron before the action potential occurs. Incorporate the importance of sodium inactivation into your answer. Many neurons have these kinds of weaker synaptic connections, where it may take several presynaptic spikes to cause a postsynaptic spike. What is an example where this might be useful to the nervous system?

• Question 5: Next, let's explore the effects of a synapse with a reversal potential more negative than the resting potential. First, we want to make the presynaptic cell fire autonomously so that we can let it be active on its own while we stimulate the postsynaptic cell. One way to do so is to increase the proportion of sodium in the leak conductance. To do this, reset the simulation to its defaults and change the leak potential in the presynaptic cell to -33.2 mV. To simplify the simulation results, set the number of presynaptic pulses to zero. Run the simulation and in a sentence or two, describe why the presynaptic and postsynaptic cells are firing.

• Question 6: Set the maximum conductance of the synapse to 0 µS (while keeping the other changes you made in the previous part of the question). Predict the effects and run the simulation to confirm them. Now change the postsynaptic stimulation current to 10 nA. In a sentence or two describe why the postsynaptic cell is firing. (This and the previous question should be easy, but if you don't have a clear understanding of them, it will be difficult to answer the next question.)

• Question 7: Now set the reversal potential of the synapse to -80 mV and the synaptic conductance to 10 µS (again keeping the other changes from the previous question; turn off the current to the postsynaptic neuron). Is the resulting postsynaptic potential excitatory or inhibitory? Explain.

• Question 8: What are the effects of an increased conductance postsynaptic potential? Reset the simulation. Set the total duration of the simulation to 120 ms. Set the stimulus delay for the presynaptic neuron to 62 ms, and the number of pulses to 4. Under Synapse properties, set the Maximum conductance to 0, and the reversal potential to -40 mV. Under the Postsynaptic Current Clamp menu, set the stimulus delay to 2 ms, the stimulus current to 10 nA, and the number of pulses to 8. Run the simulation. Explain what is causing the presynaptic neuron to fire, and what is causing the postsynaptic neuron to fire.
  • Now change the conductance of the synapse (under Synapse properties) to 100 µS, and run the simulation again. Is the postsynaptic neuron firing action potentials when the presynaptic neuron fires? Explain what is happening. Are the effects excitatory or inhibitory? How can you explain this when the reversal potential of the synapse is at threshold for inducing spiking?

• Question 9: Reset the simulation. To see a postsynaptic potential clearly, set the sodium and potassium conductances in the postsynaptic cell to 0 µS, and set the number of pulses in the presynaptic cell to 1. Use Equation 3 (above) to predict the voltage that you should observe at the peak of the synaptic potential. Now measure the peak voltage you actually observe. How closely do they match? If there is a discrepancy, can you explain what might be responsible for it?
  • To test your hypothesis, change the rise time constant for the synapse to 1 ms, and its decay time constant to 100 ms, and measure the actual voltage reached by the postsynaptic cell after the first postsynaptic potential. How closely does this match the prediction? Explain.

A Summary of Potentials

• By this unit, you have had a chance to learn about and work with many different potentials. The following table should be helpful for comparing them:

Equilibrium (Nernst) potential	"Tug of war" between electrical and chemical gradients across membrane	Ungated ion channels, specific to a single ion
Resting potential	"Tug of war" between equilibrium potentials of different ions to which membrane is permeable	Ungated ion channels, may not be ion specific
Action potential	Sodium ion influx followed by potassium ion efflux creating a nondecrementing voltage	Voltage-gated ion channels permeating either sodium or potassium ions
Other conductances	May depolarize or hyperpolarize nerve cell	Voltage and calcium-gated channels
Synaptic potential	Transient depolarization or hyperpolarization of the postsynaptic membrane	Chemical- (and sometimes voltage-) gated channels
Reversal potential	The potential at which the net current through a set of ionic channels reverses direction	If channel is permeated by more than one ion, not at an equilibrium potential