 Today we're going to continue our theme of studying and describing post-synaptic receptors. This is actually switched from Friday's lecture, and Friday's lecture will return to the theme of development, now that the difference in schedule caused by the Nobel Prize will settle down. So we introduced this slide in the previous lecture. It deals with the superfamilies of ligand-gated channels that are post-synaptic receptors. We talked previously about a large group of ligand-gated channels that are typified by nicotinic acetylcholine receptors, as well as by one certain type of serotonin receptor, the serotonin 5-HT3. I want to caution you that most serotonin receptors are G-protein coupled receptors. And we'll talk about GABA receptors later in this course and several others. All of these receptors have five subunits. Each subunit has four transmembrane domains, as we showed you in the previous lecture. However, most glutamate-gated receptor channels have a different subunit organization and a different topology and are totally different, and of course a different evolutionary history. They have only four subunits per receptor, and each subunit has only three transmembrane domains, plus a itty-bitty loop here called the P-loop, which actually lines the conducting pathway. So we'll talk about that family today, and most glutamate-receptor channels look this way, and most receptors in the vertebrate-central nervous system are glutamate receptors. And then there are the ATP receptor channels, which we will probably not discuss at all in this course, just no time. As usual, we try to pick out a region that is important for understanding the molecules involved. In the previous lecture, we talked about the nerve muscle synapse, in fact, in the previous two lectures, and we also gave you a little bit of a hint about the midbrain. You remember the handlebar mustache, the dopamine-ergic neurons with nicotinic receptors. Well, one could pick various nuclei in the brain to talk about glutamate receptors, because they are so important in so many different brain regions. But the brain region that lies at the heart of much of neuroscience has been the so-called hippocampus. Now, have any of you ever heard of the term hippocampus outside neuroscience? Hippocampus is the Latin word for the species description for the seahorse. Do we have any scuba divers here? Well, you can dive not in Southern California, but in many other regions of the world and see these seahorses. They are really very small, two or three inches typically, and you see them in Aquaria a great deal. They're curved, and so that's why the hippocampus is called a seahorse because it too is curved. A somewhat less glorious name for the hippocampus, however, is that it's curved like a banana. So you can take your pick. The hippocampus is an interesting area of the brain for many reasons. First of all, Ralph Adolfs told you about the cortex when he talked. Cortex is highly evolved and highly involuted. His sultry folds and fissures. The hippocampus is not so evolved, and it is in fact a more primitive form of cortex. It is not neocortex the way most of our brain is. It is archy cortex, older cortex. And instead of having these three beautiful layers that Ralph described in such detail, it has only, I'm sorry, six beautiful layers. Instead of having six beautiful layers, it has only three layers. So in detail, the three regions of the hippocampus, not layers, but regions are called ca1, ca2, which is very small, and ca3. Now, how did the name ca come in? Well, another word for hippocampus is the horn of Ammon. I forget who Ammon is. It may be a biblical figure who blew the ram's horn, I'm not sure. So Cornus Ammon's ca means literally ram's horn region number one, number two, and number three. The hippocampus has these various subdivisions, and it is most famous for its so-called trisynaptic pathway. It comes in from the cortex, the axons come in from the cortex. They form a particular region in the dentate gyrus. This is the first synapse. The second synapse goes over to the ca3 region. And the third synapse is in the ca1 region. The hippocampus has lots of interesting functions. Perhaps the most famous function of the hippocampus is that it tells you where you are. And so animals, hippocampal pyramidal cells fire at a certain place in a maze, and these discoveries were the topic of the 2014 Nobel Prize in Medicine or Physiology. So a bit of an introduction to the hippocampus. In the hippocampus there are many synapses, as there are in many brain regions. Remember there are roughly 10 to the 14 synapses in an adult human brain. 10 to the 11th neurons each has an average of 10 to the 3rd synapses. Typical synapse in the hippocampus would have the presynaptic terminal and the postsynaptic structure. Unlike a muscle fiber where the synapse is right on the fiber, but we do have these intaginations and folds, central neurons grow dendrites out. Those dendrites amplify their surface area not by having folds, but by having spines. And so spines are very famous structures in the central nervous system. Many types of neurons have them. One excellent place to study them is in the hippocampus on the pyramidal neurons. They have a density in the postsynaptic membrane. You can see this. So this is in the electron micrograph and electron micrographs are typically stained with osmium. And so there is a region in the electron micrograph which stains very well with osmium. That's because it has a lot of proteins. And so this so-called postsynaptic density or PSD has been studied wonderfully and systematically by Professor Mary Kennedy here at Caltech. But if we take this fanciful view of a postsynaptic density and fit into the density known characteristics of molecules, obviously this is just not fanciful, but it is a reconstruction. One sees ion channels which are an MDA receptor, a type of glutamate receptor. Here's another type of glutamate receptor. And then there are lots of other organizing molecules in the postsynaptic density which form platforms. Oh, the word escapes me at the moment. To keep the channels all where they ought to be and to keep the signal transduction molecules downstream from those channels where they ought to be. And here in the presynaptic terminal, you also recognize the synaptic vesicles. In this case, the synaptic vesicles are loaded not with acetylcholine as they would be at the nerve muscle synapse, but with glutamate. So there are specialized transporters that load the synaptic vesicle with glutamate. There are other molecules that neutralize the glutamate molecules. And then you can imagine the number of molecules that are involved with fusing the presynaptic vesicle during an action potential to the presynaptic membrane. The glutamate is released into the synaptic cleft. Diffuses along the synaptic cleft as we've discussed for acetylcholine activates the glutamate receptors. Let's stop for a moment and have a little quiz. So here we have quiz three. Could we have the envelope, please? Could we have the file cards, please? We're trying to trick you, and you can probably get all of these questions simply by looking at them. They really are quite obvious if you've been to the lecture. Close book, close notes, close laptop, close iPad, close iPhone. I don't know whether you can put your notes on your Apple Watch. Has anybody tried this? Engaged channels, nicotinic acetylcholine receptors, nicotinic are activated by nicotine. One gigohm is 10 to the ninth ohms. If you were listening last time, then you knew that used dependent blockers are helpful for epilepsy or arrhythmia. Anybody object was this a trick? Be here, but if you were here, you could piece it out. Here's a larger topic of today, which is the beautiful and important set of glutamate receptor channels. We discussed the postsynaptic density and the molecules that scaffold. That was the word I was looking for. Scaffold receptors and other transduction molecules. In terms of glutamate receptors as molecules, we knew this 20 years ago, and we've known this for a couple of years only. We know that the glutamate receptors in each subunit have these three transmembrane domains and this little part here that actually aligns the conducting pathway. We know that agonist actually binds in a venous flytrap arrangement so that the glutamate receptor closes on the ligand on the transmitter. This is where an antagonist might bind. The agonist binds here at the junction of the venous flytrap. How this leads to the conformational changes that open the channel, we are not sure. Here are the transmembrane domains. Here are more details about the extracellular region. The extracellular region of a glutamate receptor thus would be analogous to but does not have the same structure as the extracellular transmitter binding region of a nicotinic acetylcholine receptor. The transmembrane domains then are part of what anchors the receptor into the membrane and protects the moving parts of the channel so that they can open and close in response to the transmitter. Also, G protein coupled receptors, which are not shown here. We'll discuss them in a later talk. And I want to emphasize that among the three classes of glutamate receptors, the one that's most closely involved in memory and learning is the so-called NMDA receptor. And it has a very large cytoplasmic tail, which seems to be important for scaffolding other molecules and for transducing signals. And we will get to that in a moment. So there are these three families of ionotropic, meaning that they are ion channels rather than G protein gated receptors. And although in brains, they are all gated by glutamate, there are selective synthetic agonists for each of those channels. You remember that the word agonist comes from the Greek word to act agon, and that just means a molecule that opens the channel. So the neurotransmitter is an agonist and there are synthetic agonists as well. So the AMPA receptors are opened by, you guessed it, AMPA, the kyanate receptors by kyanate, and the NMDA receptors by NMDA. And the terminology for these channels is becoming standardized by international agreement. So there are various terminologies for them. Today we'll just say they are glue R1, glutamate receptor number one through four, and glutamate receptors number five through seven. And N stands for NMDA receptor number one and number two A or B or C. So as usual, a family has sprung up of these molecules and the family of molecules has various physiological characteristics that seem well suited for their role in cells. But we're not sure yet. Three families. One of the most interesting points is the ion selectivity. As you might expect, and we will treat this in more detail later, glutamate receptors have roughly the same ion permeabilities as nicotinic receptors, sodium and potassium. So they don't have a true Nernst potential. They have a reversal potential. However, there are interesting observations and stories about the calcium permeability. So calcium would be flowing in with sodium. AMPA and kyanate receptors without this magical glue R2 subunit are permeable only to potassium and sodium. And with the glue R2 subunit, they are permeable to calcium as well. NMDA receptors are always permeable to calcium, which is a very large part of the interest of NMDA receptors. So the story of the AMPA receptor permeability to calcium is very interesting because the glue R2 subunit is itself controlled with regard to its calcium permeability by an event at the RNA level. So you remember that DNA leads to RNA, leads to proteins. In the RNA of encoding the glue R2 channel, it is possible for the adenosine to be edited by a special group of enzymes to become an inosine. This changes the amino acid encoded by that codon. When the amino acid is an arginine positively charged side chain, calcium cannot flow through. When it is a glutamine neutral, then calcium can flow through. An example of this sort of thing is voltage clamp currents for the glue R2 with Q that is glutamine in the channel, right in the middle of the conducting pathway, trying to trick the calcium ions into thinking they're still in water. With glutamine there, they are nicely permeable to sodium. So you get a peak of current and then the desensitization that I told you about for nicotinic receptors also happens for glutamate receptors. And then when we switch the outside solution to a pure calcium solution, we also get a current through those Q channels. But when we use the version of the RNA that encodes the arginine side chain, these are the one letter abbreviations for side chains, and either you learn them in by 8 by 9 or we'll learn them in by Chem 170, arginine or glutamine for Q. Again, when we have sodium in the external solution, we get a nice current through them, the desensitization seems to be a bit different, a bit slower. But when there's calcium in the external solution and arginine in the selectivity filter of the ion channel, no calcium flows through. And with regard to activating second messenger pathways, this is probably quite important. However, by far the most important calcium permeable glutamate receptors are the NMDA receptors. They, like all of the glutamate, most of the glutamate receptors, they are tetramers. They are composed always of two NR1 subunits and a pair of NR2 subunits, though the pair can be the 2A, the 2B, the 2C or the 2D. The NR1 subunits are not quite bystanders, bystanders, but they're not so important with regard to the physiological properties as the NR2 subunits. You do need to have a ligand for them, and actually that ligand is not glutamate, it's a co-ligand, either glycine or de-serine. So the non-protein version of serine, the version that goes in a protein is L-serine and this is D-serine. So those ligands are present most of the time in the extracellular solution at concentrations that are high enough so that they don't need to be released from synaptic vesicles. They are constitutively present. We call them co-ligands. The glutamate then gets released from the synaptic vesicles, and the glutamate binds to the NR2 subunits, and the channel can then open, but as we'll see, we need a couple of other points to occur as well. So the NR2 subunits, either A, B, C, or D, bind glutamate, and they have these cytoplasmic tails, which have not been resolved structurally, so I merely presented them as blobs. And they are important for signal transduction and for scaffolding to the post-synaptic density. Now, way back two weeks ago, we talked about this obscure biophysical phenomenon, and I told you that it actually is an important neuroscience phenomenon, so it's time to revisit it. It is the time required to exchange waters of hydration around an ion in solution. We pointed out that sodium and potassium can exchange the waters of hydration in about a nanosecond, so that none of the actions of sodium and potassium are limited by this time to exchange waters of hydration. Almost the same as true for calcium, but as the most charged dense cation, magnesium holds its waters of hydration most tightly for a long time. That is about 10 microseconds, and that is enough to impede flows of magnesium through ion channels. In fact, there are very few, if any, ion channels that are permeable to magnesium. Now, and then we said that because magnesium doesn't lose its waters of hydration, it actually can form a plug in other ion channels. It gets stuck in ion channels, and certainly the most glorious example of that is getting stuck in the NMDA receptor. So here is a diagram of the NMDA receptor that might be open and allow calcium or sodium to flow in, but magnesium gets stuck. Magnesium is present at 2 millimolar, a fairly high concentration in blood and in the extracellular solution. So magnesium has a high propensity for getting stuck in the ion channel in the NMDA receptor. So the NMDA receptor then becomes a coincidence detector. The channel opens only when two events happen concurrently. One is the binding of glutamate, actually not shown in this, there it is, sort of out here in this primitive version of the picture. And the second is that the magnesium needs to leave. When the magnesium leaves, then the NMDA receptor channel can conduct. Now the magnesium is charged, it has 2 plus on it. And so in order for the magnesium to leave the channel, we need to put a positive voltage here to oppose the magnesium ion, to repel the magnesium ion, and to move it out of the channel. And so that can be done by an action potential. So the depolarization due to an action potential removes block by magnesium. Two things need to happen. We need to have the binding of glutamate released from the pre-synaptic terminal and an action potential in the post-synaptic cell. And presumably that occurs because the post-synaptic cell has become excited by other molecules, for instance, amper receptors or nicotinic receptors. This depolarizes enough to relieve the magnesium block, this allows sodium and calcium to flow in. Any questions about that process? So we can actually isolate, a given cell has both amperkinate receptors and NMDA receptors. And we can isolate those two responses electrophysiologically with a patch clamp or a voltage clamp circuit. Here we are holding the cell at a membrane potential, which is more positive than the reversal potential, which is around zero. So as a result, the currents would like to flow outward. So a very interesting point is that when a pre-synaptic, when we apply a pulse of glutamate, either with pre-synaptic stimulation or artificially with a little puff, we find a mixture of NMDA currents and amper currents. The selective blocker antagonist for NMDA currents is a molecule called APV, DAPV. So in the control solution we have a waveform outward currents. When we block the NMDA receptors with APV, we have a current that is nearly as large as the control, but decays a lot more rapidly. That is in fact a characteristic of NMDA receptors, which is that they are a bit slower than the amperkinate receptors. If now we use another blocker, go back to the control and use another blocker, selective for blocking amper receptors, we reveal that the waveform of the NMDA receptors and it is longer, slower than the waveform of the amper receptors. If now we do this experiment not at positive potentials, but at negative potentials, in the presence of the usual extracellular magnesium concentration and we again block the NMDA receptor channels, not much happens because there's magnesium stuck in those channels and we don't get many much inward currents, not around minus 70 millivolts. But when we then block the amper receptors with this drug called CNQX, we see no current left because at minus 70 millivolts, there's no inward NMDA current because of the magnesium block. So all of this is reasonable and holds together and using experiments of this sort, we can isolate the amper component and the NMDA component. And the reason we're interested in the NMDA component is that all of that good calcium stuff is coming in with it, getting ready to serve as a second messenger. So using these subtractions, we can figure out the NMDA receptor component and the amper receptor component. And using these subtractions, we can plot the peak inward current as a function of where we have set the voltage across the membrane. So this is a current voltage plot. If we set the voltage across the membrane at plus 25 millivolts with our voltage clamp, we get these data. If we set it at minus 50 millivolts, we get these data at minus 100 millivolts, et cetera. So these subtraction procedures allow us to show that the peak inward current through non-NMDA receptors, that's another word for the amper or kinate receptors. And NMDA receptors are so important that either we call glutamate responses, NMDA responses, or non-NMDA responses. So here's the nice, linear, boring, electrically understandable response through the non-NMDA receptors. But if we look at the current through the NMDA receptors, sorry, we see that for outward currents, we get normal, nice behavior. But when we start to try to force currents inward through the receptor, the magnesium starts to come in and block. And the more we pull on the magnesium with a negative inward potential, the more it blocks. So we get this strange current voltage relationship, which really tells us about the block by magnesium of these channels. Now, if anybody would like to talk about this again, I'm happy to do so. What causes the magnesium to unblock and empty? What causes the, an excellent question asked by Ellen, is what causes the magnesium to leave the channels? What causes the magnesium to leave the channels is that the inside of the cell becomes more positive. Actually, just less negative, does it? And that more positive internal potential drives the positively charged magnesium ions out of the conducting pathway. It's probably more accurate to say that the magnesium ions are being driven out by the charge than to say that they are swept along in the other ions that leave. Probably the first view is more accurate than the second. So it's simple charge repulsion, electrostatic movements. Any other questions? So as usual, for every macroscopic phenomenon associated with lots of ion channels working together, there is a single channel phenomenon and here is the single channel phenomenon. So these are single NMDA receptor channels. You studied in a patch clamp just the way Feynman told me to do, although I didn't listen. And they are, they too show the blockade by magnesium at negative potentials. So there's a little, we are setting the voltage across the patch. Here are the outward currents. Now, you'll have to take my word for it that the upward transitions here are outward currents. The upward transitions are when the channel opens and the lower levels are when the channel is closed. That's reversed for inward currents. The upper transitions are the channel closed or the upper levels are the channel closed and the lower levels are the channel open because the direction is reversed. But you can see that in normal external magnesium here, 1.2 millimolar, we get robust easily measurable outward currents at positive potentials. But little wimpy ones, very brief at negative potentials because the magnesium is being attracted into the negative cell because magnesium is positively charged as being attracted. And although you cannot tell quantitatively from these particular traces, you can believe that the block and the attraction seems to be greater as you go to more negative potentials. So a channel opens, it doesn't take long until a magnesium ion finds it and then it blocks the cork in the drain or in the wine bottle. Now if we take away the magnesium and do the experiment simply with sodium and potassium, we have channels opening outward very nicely at positive membrane potentials and also at negative membrane potentials, channels open nicely. And you can see now more clearly that an individual ion channel acts like a conductor in the sense that as we increase the driving force across that conductor, the currents get larger. So the currents here at plus 60 are roughly twice as much as the currents at plus 30 and here the currents at minus 60 are roughly twice as great as the currents at minus 30 and of course when you intersect you get zero current. So presumably channels are opening here but there's no current flowing through them. Any questions about this trace or this explanation? So at the level of macroscopic current voltage relationships then we can see this blue trace as a simple multi-channel addition of all of these partially blocked channels here and the block gets worse when we go to negative potentials. And of course right down here at the resting potential is where a cell is usually sitting. So at the resting potential those NMDA channels are usually blocked and no calcium flows in but when another activation of the channel produces an action potential or other severe depolarization then these channels begin to open and they open enough so that sodium and importantly calcium can flow into those channels and act as a second messenger. Any other observations on this? Well obviously it becomes very important in neuroscience to measure that calcium influx directly and in fact as we've said in this course and we'll say again in this course new tools are really what drives neuroscience and you can contribute to neuroscience from a large number of diverse fields. One can make a contribution to neuroscience as a chemist, as a physicist, as a biochemist, mechanical engineer, electrical engineer, statistician, bioinformatician. If I left out anybody in this room, chemical engineer, bioengineer, a computation and dynamical systems person. Oh dear, I left somebody out, anybody else? Okay so the wonderful aspect of neuroscience is that I actually don't consider myself a neuroscientist and most people don't consider, most people who work in neuroscience don't consider themselves neuroscientists. At best they consider themselves hyphenated neuroscientists, a social neuroscientist or a molecular neuroscientist. But I'm just happy to say that I do biophysics or whatever I'm doing these days. So here is a contribution from a chemist named Roger Chin. He's synthesized a molecule whose fluorescence changes when it binds calcium. You'll notice that there are four carboxylic acids in this molecule called fura2. At neutral pH those carboxylate groups are dissociated, they are negative charges and they very much like to fold back and onto a calcium ion. When they fold back and onto a calcium ion, the fluorescence of this group changes very nicely. And so this molecule called fura2 is a fluorescence sensor for intracellular calcium. Most interestingly when it binds calcium it's activation at one wavelength increases and it's activation at another wavelength decreases. So this is the intensity of the fluorescence and not only does its activation change but it shifts wavelengths. So that means that there is a very nice internal control that one can do by measuring its activation at the exciting wavelength of 340 nanometers or at the exciting wavelength of 380 nanometers and the ratio between the fluorescence activated at those two wavelengths tells us pretty much precisely the calcium concentration inside a neuron or another cell. This happens with a time course of microseconds so we can follow the calcium channels transients very nicely. Here is an example of a cultured hippocampal neuron with fura2 in it. Here is a dendrite and this would be a spine. The experimenters have put calcium on that spine, sorry have put glutamate on that spine and arranged it so that they see the fluorescence from fura2 and what we're looking at here is actually the ratio of 340 over 380. It increases when they add glutamate. They then add AP5 which is the selective antagonist for NMDA receptors. They block the receptors, no calcium flows in, and then almost every physiological experiment requires reversal. That is, you put in a drug or you apply a stimulus, you get an effect, then you have to wash it out to show that that effect is reversible. So here is the reversibility. They again add glutamate, they again see the calcium flow in. This is a crude version of the experiment. There are now very much nicer versions of the experiment. Those nicer versions of the experiment depend on GFP. Would anyone like me to say what GFP is? As you'll remember in GFP, the fluorophore is very well protected from the environment by a can of beta strands. So now render this fluorescence susceptible to a molecule. You always tell that typo is better when you're actually giving the lecture. If we actually render this fluorescence sensitive to another molecule or to a condition, we have a very good fluorescent sensor. Now in this particular case, what has been done is to take GFP and make a couple of changes to it. First, GFP is enhanced for being expressed as a protein in mammalian cells. Therefore it's called EGFP, enhanced GFP. Part of the way this is done is to change the codons so that it expresses better. Another way it's done is to change the amino acids. The other cute thing that's been done to this molecule is to make it circularly permuted. What circularly permuted means is that the experimenters have joined the N and the C termini of GFP, but have opened up the molecule in another region, circularly permuted. Opening up the molecule in another region allows one of the beta strands to peel away, no longer protecting the fluorophore against the environment. It does not absorb photons and it does not fluoresce. So the trick now is to couple another protein to putting back that helix, making a fluorescent version of GFP. This is done by coupling Calmodulin to the circularly permuted GFP. Calmodulin, the calcium binding protein, undergoes a fairly dramatic conformational change when it binds calcium. It's one of the most common proteins in cells in all animal kingdom, very well conserved. So what the experimenters have done with lots of trial and error, lots of structure prediction, lots of protein engineering is to couple the conformational change in Calmodulin when it binds calcium ions to linkers that heal the circularly permuted EGFP and make it fluorescent when there are calcium ions nearby. This molecule has undergone a large number of mutations and engineering to make it better and better as a calcium sensor. In fact, this is all GCaMP. Now, do you know what GCaMP stands for? Well, good, GFP Calmodulin protein. Thank you, GFP Calmodulin protein. There is now GCaMP6, which works very well indeed better. But as you can see, the version number three and the version number five still give lots of nice changes in fluorescence when calcium enters a cell. Here the calcium is entering the cell because the experimenters are stimulating the dish electrically on which the cells are sitting. Then there are other versions of the sensor. So what's the advantage of GCaMP over FURA? Would anyone like to speculate about that? The advantage of using GCaMP over FURA? Well, FURA is a wonderful chemical, but you need to put it into cells. You need to load it into cells either by injecting it or by making a version that is permeant through the membrane. But the genetically encoded fluorescent calcium sensors are simply genetically encoded and one uses the techniques of modern gene transfer, transfection, transgenic animals to put those molecules into a cell. And one can now use the techniques of modern gene transfer with cell selective promoters or other engineering techniques to put them only in the cells where one wants them. And so one can optically record then calcium transients in the cells that one wants. And so these days it's becoming almost possible to do electrophysiology without electrical electrodes. So control of synaptic plasticity by NMDA receptors. We'll talk about them later, but we already know because I've told you but not giving you the details that there is a central role for calcium in initiating long-term changes. That's the calcium hypothesis. We can measure cytosolic calcium with fluorescent dyes and biosensors. We will talk in a later lecture about the control of the detailed timing of postsynaptic calcium. We've talked about the structure and behavior of the NMDA receptor, the number of subunits, four of them, the fact that it's a coincidence detector, the fact that it's also permeable to calcium, that it's slower than amper receptors. We talked a little bit about its pharmacology and also that it serves as a scaffold for other molecules. We've talked about the postsynaptic density. We will talk about these potentiations that are triggered by calcium-sensitive signaling machinery right near the blob, which is the mouth of the NMDA receptor. And we will talk in further lectures about the biochemical changes downstream from these calcium fluxes. On Friday we will return to development for a second lecture of development. See you then.