 Good morning. This is the second lecture on the visual system. Ralph and I thoroughly enjoy teaching this part of the course. We enjoy teaching the whole course. We enjoy teaching this part of the course because we get to swap off lectures and to alternate between Ralph and myself. So you get a feeling for the variations in among how neuroscientists approach topics. So on Wednesday he talked about the process from the retina to the visual cortex. Today I'm going to talk about the first part of the process, the retina and the detection of light. And then on Monday Ralph is going to talk about a higher level vision. Then Wednesday I will talk about the motor system and Friday about the olfactory system. And so we have fun switching off topics and giving you interesting perspectives. Today we talk about the visual system and there are actually six topics involved in the visual system. First of all, I'm going to talk with you about lenses and optics, then about the photoreception event, then about the retina itself, connections to the brain and master switch that actually controls a good part of differentiation in the retina. And finally the time frame for evolution of the major features. So the topics for today are actually two. Photo transduction in the retina of course is one and also evolution of the eye is another. The motivation here on evolution is that we actually, except for John Allman's course on the evolution of on evolution undergraduates actually do get only a limited exposure to evolution as a concept. And this is a field that's changed rapidly in the last decade or decade and a half because of the enormous amount of data available from genomes show very clearly that genomes are all related. But we get this countervailing set of arguments from people who don't believe in evolution. And do any of you have friends who didn't do not believe in evolution. As an undergrad, you had friends who did not believe in evolution. One particular Mason, you have friends who don't believe in evolution. You encounter one at a cocktail party. Here are your arguments. Mostly chapter 26 of candle talks about the first topic photo transduction and it was partially written by our own Marcus Meister. In terms of the evolution of the retina, we quote Dubjanski, a great 20th century biologist who said nothing in biology makes sense in except in the light of evolution. There are many ways to solve a problem in biology. And in order to know, to appreciate how an organism does it these days, we ask how it evolved. The key point in evolution is that all modern biological processes evolved from related processes. So we are not talking about the origin of life here. We are talking about evolution after life appeared. A way of saying that in molecular terms is that every modern gene evolved from other genes. Specifying that a little bit more precisely, we say that every gene has an orthologue in related species and most genes have paralogs in the same species. I'm going to explain to you what I mean by orthologs and paralogs, but basically we'll get into it in great detail. But because all vertebrate eyes are quite similar, the hung for orthologs is very straightforward and it's successful in every case. And in fact the two organisms share many orthologs is very powerful evidence for the idea that they evolved from a common ancestor, which is of course a central aspect of evolution. So let's give as an example of orthologs and paralogs the globin genes. These are the proteins that carry oxygen in the blood. We are going to talk about the human on the right and the mouse globin genes on the left. We're going to go back about 700 million years to the point where in the common ancestor of humans and mice, there is a single globin gene. The globin is the protein that eventually binds the heme that carries oxygen. The record suggests that roughly and very roughly 500 million years ago, before the present, there was a split, a transposition separating the alpha and the beta globin genes. Because this happened before humans and mice split, vestiges of that split remain in mouse genomes and in human genomes today. So the orthologs, well let's talk about paralogs first. The paralogs then are related occurred because of the split. So the split produced paralogs, related genes, usually with related function within an organism. Now there was another split between the fetal and the adult beta genes. Again, before the common ancestor of humans and mice gave rise to the two species. And there was another split right here between the delta and the beta genes, again before humans and mice diverged. However, after the divergent, there was a further split only in humans between the gamma G and the gamma A globin, which was not present in the mouse genes. And so we see that if we understand common ancestors and splits and descent, we can explain most of the genes in a complex family like the globin genes. So we have orthologs, which resemble each other across the species, the beta globin, mouse beta versus human beta. Paralogs resemble each other distantly or closely depending on when they diverged within a species. A good example would be the human epsilon versus the human beta. There's also the mouse epsilon versus the mouse beta. And then we have this more recent divergence in the human lineage that produced gamma G versus gamma A. So these are the concepts which we put into molecular terms, into protein terms when we discuss evolution. We discuss orthologs and paralogs. Let's now go to the specific topic for today, which is how the retina works and what this tells us about evolution. So first sub-topic is lenses and optics. The lens, of course, has an index of refraction greater than that of water, and it can therefore bend light. And the reason it has an index of refraction greater than water is that it has a high concentration of protein. Actually remarkably, the high concentration of protein occurs in all animals, but there are various proteins that subserve this function. Many of them are called crystallines because they are in the lens. Some of them are also enzymes. They perform various functions in other tissues. Apparently, to get a crystalline, really the only requirement is that the protein should have good stability so that it can be present at high concentrations. And it also should have no prosthetic groups or attached groups, such as vitamins, that might absorb the light. All we want is a high index of refraction. We don't want a colored lens. To give you an idea of the protein concentration, this cell is 22% protein. And I said in lecture one that it's 4 millimolar protein, and you can work out the idea that a typical protein, you can see it right there on your gel from your biochemistry lab course. It runs at around 50 KD, 55 KD. Average residue has a molecular mass of 110, so we can go back and forth between 4 millimolar and 22%, which is really a remarkably high concentration of protein, and a lot of cells metabolism goes into generating proteins and keeping them fit. And so I told you that crystallines come from various, or are various proteins in various species. If we look at a typical picture of the gene regulation, we see the coding region. We see a promoter. Here is the coding region of a crystalline. Here is the upstream factors, the Tata box, the transcription factors, the RNA polymerase that you learned about and by 8 by 9. There are a lot, and as you know, the non-coding regions of the genome are very much larger than the coding regions. So there are lots of non-coding regulatory regions in most animal genomes. They can be folded in ways that we don't completely understand yet to activate the RNA polymerase to transcribe the crystalline, to transcribe a protein. In the case of the common protein, the common transcription factor that I'm going to talk about later in this lecture, it is one of the transcription factors that binds to an upstream regulatory region of DNA and activates RNA polymerase. So what do we learn? Existing proteins used for various functions, probably the use in the lens came later than the use in other tissues. And it is also a fact that you learned about gene sharing that several distinct transcription factors can share activation of a protein and could activate that protein independently. And here in the case of crystalline is a great example of that. So we've got the lens and we've got the optics. We know how the lens has accomplished its high index of refraction. What about the aperture? That's part of the optics too. So the aperture mechanism is actually controlled by smooth muscles. There is a nerve from the brain, releases acetylcholine, makes a muscarinic synapse. The smooth muscles are around the iris. When acetylcholine is released, the smooth muscle, the acetylcholine activates muscarinic receptors. It contracts the smooth muscle cell, which is connected by non flexible fibers to the aperture itself. So the contraction leads to a smaller pupil. Now, people who don't believe in evolution say, this is just such a wonderful process. It must have been created. I want to remind you that in fact, innervated smooth muscles control many organs in the body. They control the diameter of blood vessels, the peristaltic activity of the intestinal tract, the diameter of the bladder neck. Many of these smooth muscles are controlled by muscarinic receptors. If any of you have friends who get their eyes dilated when they go to the ophthalmologist, the dilation of the eye involves atropine, which comes from the plant atropa belladonna, meaning beautiful woman, because in the old days it was thought that a beautiful person had wide eyes, dilated pupils. So this is where atropine comes from. And so in each case, the nervous system has evolved circuits that extract and integrate information from sensors and employ smooth muscles in a homeostatic loop. So if at this cocktail party, your friend, a new acquaintance says to you, the eye is remarkable because it has evolved to control the aperture. You can say to your new found acquaintance, well, so has this fainter. Okay, so moving on from the lens and the optics to the photoreception event, photoreceptor organs have actually evolved independently. About 40 times. And in each case, they respond to the visible spectrum, which is an eponymous statement that is a photoreceptor response to visible light. But it's always between 400 and 700 and maybe a bit in the near UV. So how do we explain the fact that the limited part of the electromagnetic spectrum, which goes out from, of course, x-rays through radio waves? How do we explain the fact that eyes come back to this limited part of the spectrum? Well, infrared light, remember E equals h nu, infrared light is insufficiently energetic. It cannot provoke the cis trans isomerization that actually is part of the photochemistry of vision. Shorter wavelength UV light is too energetic and it might damage organic molecules. You've seen clothes get bleached. You've seen you've got sunburns, etc. So there is a reasonably friendly part of the electromagnetic spectrum that is energetic enough to do photochemistry, but not energetic enough to do damage. And that's what we call the visible spectrum. As Ralph Adolphs told you the other day, the photoreceptor cells in the rods and cones in the retina actually receive their photons from what appears to be the back. That is, light goes through the entire retina before it arrives at the photoreceptors. Well, that's okay because the retina is transparent, so you don't lose much energy in going through the retina before the light gets to the photoreceptors. There are two types of photoreceptors as Ralph explained the rods and the cones and their distribution varies across the retina between the fovea, which is quite dense and has primarily rods, and the rest of the retina, which is specialized for a larger visual field. The rods have free-floating discs. They're not actually free-floating, but they are separate physically from the plasma membrane. The rhodopsin, the visual pigment, is on those discs. The cones have invaginations, which causes the vast increase in area. But again, there's rhodopsin on these greatly amplified amount of membrane. In each case, of course, the greatly amplified amount of membrane provides room for lots of rhodopsin molecules within a single cell so that each of those photons stands a pretty good chance of being caught absorbed by a rhodopsin molecule. And then the nutrients come from a connecting cilium that goes to the nucleus where the proteins get made. And then one selective advantage of the fact that the rods and cones are on the other side of the retina is that they conduct another layer of cells, epithelia, which keeps the outer parts well nourished. If we look now at the rhodopsin, there are actually, here's a good example of paralogs. So there are four paralogs of rhodopsin in the human genome. They obviously evolved from a common ancestor. And each rhodopsin interacts slightly differently with the common molecule retinol. And these slight differences in the interaction with retinol change the absorption spectrum of the retinol so that we have the blue absorbing, the red absorbing, the green absorbing cone pigments. The cones are specialized for color vision and a much broader absorbing rod rhodopsin, which is specialized for black and white vision. So there are lots of mutations that change the spectrum and these mutations are mostly clustered in the membrane in the region that in fact interacts with and slightly tugs at the retina. The chemistry, the initial event in photoreception is rather simple. The retina, vitamin A, bound to the rhodopsin is in the cis configuration, which is metastapal. And you can see there is this conjugated part of retina so that it absorbs light in the visible part of the spectrum because it's nicely conjugated. When cis retinol absorbs a photon, it undergoes an isomerization, a photoisomerization, changing it to the trans configuration, which is more stable energetically. So each time a rhodopsin molecule absorbs a photon there's a good probability that it will undergo the isomerization. This then leads to spontaneous dissociation of the retina. So now the retina is free and the rhodopin, the activated opsin is also free. Then there are enzymes associated mostly with the epithelium that regenerate the 11 cis retina that requires energy back into the, sorry, that trans back into the 11 cis. This requires energy and then we go back and reload the rhodopsin, get ready for another photoreceptor event. But the key millisecond time scale photoreception event is right here. It's the light induced isomerization. Now we have discussed about a week ago the usual G protein coupled receptor pathway. And I mentioned that rhodopsin is in fact a GPCR, a G protein coupled receptor. The usual pathway begins in the membrane then moves to the cytosol and then if it is activated for long enough actually moves to the nucleus. There are a couple of key differences in the GPCR family that have the selective advantage of being able to respond more rapidly to visual stimuli than other G protein coupled families. First of all, in the membrane we have the receptor, seven trans membrane helices, just like a usual GPCR. It activates a G protein, again, standard GPCR. It activates the one G protein family that we have not discussed in detail, neither I, Q nor S, which we discussed in detail, but GT. And actually this was the first G protein alpha subunit studied carefully. T stands for transducent and it transduces the signal from photons to activating the effector. We discussed the fact that G proteins have effectors, either channels or enzymes. In the case of the vertebrate visual system, the effector is an enzyme. We also discussed the fact that enzymes make intracellular messenger. And in the case of the retina that intracellular messenger is cyclic GMP. Unusually this messenger activates an ion channel. But the first few parts of the GPCR pathway involved for vision should, you should pardon the mixed metaphor, ring a bell. Although they take place more in the cytosol than in the membrane. We talked about the beginning of the GPCR family pathway. We have a neurotransmitter or hormone that binds to the receptor that activates the G protein that activates an effector. And we said it at fastest it's 100 milliseconds. That's about how fast the visual system is also. Now, the difference is that in the usual G protein coupled receptor family, there are lipid tails on the proteins to keep them in the membrane so that they can all encounter each other rather quickly. However, in the rods and cones, there is a limited cytosol between the discs or between the folds in the cones. This limited cytosol allows the rod or the cone to lack the lipid tails on the downstream members of the family. So, for instance, G sub T transducent is a cytosolic G protein. It doesn't have a lipid tail. And the enzyme effector is also cytosolic. It doesn't have a tail. However, because of the fact that the space between discs is so small, there's not many places that these enzymes and proteins can diffuse to. So, they are close enough so that it says though they were in the membrane itself. And then we have the typical GTP cycle. So, they're not membrane bound, but the membranes effectively restrict their motion so that we can get a high density of these proteins and of the rhodopsin and they can still interact within milliseconds. So, less than 100 milliseconds and very close to each other, much less than a micron away. So, that's a slight difference in terms of membrane embedding versus cytosolic arrangement. Now, remember we said also in our G protein lecture that we make intracellular messengers and that this involves transfer of a phosphate group. So, it involves energy and that intracellular messengers get made and bind to proteins. We also said in these intracellular messengers typically are either calcium or cyclic nucleotide. We also said in the previous lecture that these messengers can themselves in a few cases activate ion channels and I promise to get back to you with that. Well, here we are with that example. In a few places intracellular messengers activate ion channels and that's where we're going today. Among vertebrates, the intracellular messenger for vertebrate phototransduction is cyclic GMP. Not cyclic AMP, but cyclic GMP. Just reminding you here's guanosine. Here's guanosine triphosphate. Cyclase is cyclic GMP. There is as we learned in a previous lecture always an enzyme that hydrolyzes cyclic GMP. It's a phosphodiesterase and hydrolyzes it into cyclic, into simply GMP, not cyclic GMP. Now there's a lovely twist here. This lovely twist is the opposite of what we think most of the time in the nervous system. In the photoreceptor we begin with high cyclic GMP. It opens ion channels, keeps the plasma membrane depolarized and keeps glutamate release at the synaptic terminal high. The result of photoisomerizing right now in the activating rhodopsin is that actually we activate the phosphodiesterase. This reduces the cyclic GMP concentration. It results in closing an ion channel in the outer segment of the membrane, the outer segment of the photoreceptor where the disks are. And so this transiently hyperpolarizes, not depolarizes like in a synapse, but hyperpolarizes the entire plasma membrane. So just a reminder that in previous lectures we talked about the effector enzyme, the cyclase, making a cyclic nucleotide, breaking in another enzyme, phosphodiesterase. We talked about the fact in previous lectures that the phosphodiesterase is often inhibited by caffeine, and AMP itself is not very interesting. We also mentioned in a previous lecture that the analogous system, which occurs in the retina, having an enzyme, a cyclase, making cyclic GMP, getting hydrolyzed to an uninteresting molecule. And it is, in fact, the phosphodiesterase, which is the effector for GT for transducing. I also gave you this little puzzle to tell me which commonly used drug can also inhibit a phosphodiesterase. Indeed, a paralog of the phosphodiesterase expressed in the retina is expressed in other places in the body, and that paralog is inhibited by Viagra. And the package insert for Viagra says that Viagra may cause a perception of bluish haze or increased light sensitivity in some patients. So the selectivity of Viagra for a particular phosphodiesterase subtype is not perfect. Okay, so then we also study this system using the excised inside-out patch, that aspect of the patch clamp, which allows us to pull the patch away from the cell while the gigaseal remains mechanically intact. We can still measure channels opening and closing, and now we can ask whether, in fact, those channels are indeed opened by cyclic GMP. Sure enough, when we add cyclic GMP to the bath, the channels open very nicely. So this was really an experiment done in the mid-80s, was really the first hint about the transduction process of photoreceptors. The idea that in a normal photoreceptor, channels are open. When phosphodiesterase gets activated by light, the channels, the cyclic GMP gets hydrolyzed into an uninteresting molecule, GMP, and those channels close. So transduction process receptor going to GT activates the enzyme, which decreases cyclic GMP. So if we look at the membrane potential of a photoreceptor, it looks, as one gives a light pulse, it actually looks electrophysiologically opposite to most neurons. When we activate rodopsin, we actually get a hyperpolarization. That hyperpolarization is caused because channels close. Sodium channels close. This is transient, takes typically a good fraction of a second, sometimes much faster, and then the hyperpolarization goes away and we go back to the resting potential. So, rodopsin absorbs light, cation channels close in the plasma membrane of the outer segment, hyperpolarizes the entire cell. This hyperpolarization relays visual information to the synaptic terminal made by the photoreceptor cell onto subsequent cells, and it slows the ongoing release of glutamate. So we have a challenge for the photoreceptor. It has to release a whole lot of glutamate all the time. And apparently this challenge is subserved by a specialized organelle called the synaptic ribbon. So here is the cytoplasm of the photoreceptor. Here are the post-synaptic processes downstream from the photoreceptor. So it is a photoreceptor to horizontal cell synapse. There are an enormous number of synaptic vesicles in the photoreceptor terminal waiting to be released. They apparently get lined up on the ribbon and they get nicely moved toward the synapse and get released at pretty high efficiency in large numbers. Of course the calcium channels that control the release are right down here. Those calcium channels are continually activated because the cell is continually depolarized. It sounds really funky, but that's the way the cell does it. And it takes so much energy actually that a typical photoreceptor cell turns over its entire sodium concentration every 20 seconds or so. So that photoreceptor cell has to have lots of ATP, lots of nourishment, works very hard. Alright, we have talked about the phototransduction cascade. I want to talk about two aspects of that cascade. Amplification and adaptive and homeostatic mechanisms. We know that the retina is a remarkable organ because it adapts. When it's fully dark adapted, many species can detect one photon hitting per photoreceptor cell. That event registers. Not one photon in the retina, but one photon per photoreceptor cell. On the other hand, when fully light adapted as in bright sunlight, species can accurately analyze light at intensities about 10 to the 10th fold higher. So there is not simply one adaptive homeostatic mechanism that underlie these phenomena. There are several of them at the photoreceptor level, at the circuit level in the retina. And probably the adaptation is complete by the time that impulses get to the central nervous system. Now, I have a few Henryisms. One of my Henryisms is that I don't like the term homeostasis or adaptation. Because we inevitably tend to assign our favorite process such as memory or learning or addiction and say that it occurs because of homeostasis or adaptation. Actually, that tells us nothing. Homeostasis, adaptation, compensation, these are not mechanisms in themselves. They're really adjectives. So the best way to talk about these processes is adjectives. So we talk about homeostatic mechanisms and adaptive mechanisms. And one of the goals of neuroscience at the circuit level or at the cellular or molecular level is to understand the adaptive and homeostatic mechanisms. But if you say, well, this process occurs because of the homeostasis, you've actually sort of defined homeostasis rather than defined how it works. Moving on. So the amplification, first of all, that allows a single photon to generate a perception occurs because after the rhodopsin is activated, the receptor can activate about 500 transducent proteins. So in the space between disks, 500 transducent proteins can be activated by a single activated photoreceptor. So that is amplification. Then the phosphodiesterase, which chews up cyclic GMP, is fairly efficient. It has a turnover number of around 4,000, 5,000 per second, which is about as fast as the cyclic GMP molecule can find the phosphodiesterase. It then gets hydrolyzed while it's activated. So we have another amplification. And the third application is that each millisecond, each time a cyclic GMP molecule binds to a channel, each millisecond that it's open, the cation channel in the rod can allow the flux of about 10,000 ions. So this amplification occurs at several steps. We understand the amplification when the system is at its most sensitive, that allows a single photon to be amplified by a factor of 500 in terms of transducent, a multiplicative factor of 4,200 in terms of cyclic GMP molecules. And then for each of those, we interrupt the flux of about 10,000 ions for every millisecond that the channel closes. What about adaptive and homeostatic mechanisms? Well, those are more complex to describe, and I'm not going to describe them all. But one of them is that transducent itself, which is not shown here, hydrolyzes GTP to GDP and shuts itself off just the way any protein does. So that is, we get some adaptive, we get a shutoff of the transduction right there. Another part of adaptation is that the activated receptor also gets deactivated. And the way this happens is that a kinase called rhodopsin kinase appropriately enough, which phosphorylates proteins, remember a kinase takes ATP, takes the terminal phosphate of ATP and puts it onto a protein, thereby making ADP in a phosphorylated protein. So this phosphorylated C-terminus mostly of rhodopsin binds another protein called apparently appropriately enough arrestin. And a rhodopsin molecule that is capped by arrestin can no longer interact with the g-protein. So rhodopsin kinase helps to terminate the event as well. And then we also remember that guanylate cyclase has to synthesize new cyclic GMP from GTP. It is partially inhibited by calcium at concentrations greater than around 75 nanomolar. So these are cation channels, they flux both sodium and calcium. So typically the tonically open channel sets calcium in the cytosol at around 500 nanomolar. But when the cation channel closes, calcium continues to be pumped out via the usual processes, calcium pumps. And this lowers calcium to around 50 nanomolar, and it activates the denominal cyclase. We start off here at a high concentration of calcium. The channel closes, no more calcium is flowing in, it gets pumped out instead. This activates guanylate cyclase, we make more cyclic GMP. So that's another homeostatic or adaptive mechanism. There are lots of these in the photoreceptor and in the circut downstream. With the result that over the time scale of seconds to minutes, you can adapt to varying light intensities. So we've talked about the photoreception process, let's now go on to the topic of the neurons in the retina. Here is a picture very similar to the one that Ralph showed you. The rods and the cones with their outer segments facing outward. Then there are two layers of synapses, the so-called outer plexiform layer and the inner plexiform layer. Remember we're going in toward the center of the eye here. In between those two layers, there are cell bodies of several types of neurons. There are the horizontal cells and you can see why they're called horizontal. There are the bipolar cells, you can see why they're called bipolar. And after the inner plexiform layer is the ganglion cell. The ganglion cell in the retina is unique until we get to the ganglion cell. We are dealing with cells that do not fire action potentials. They don't need to, everything is within a few hundred microns. They just do graded signal processing. But the ganglion cell feeds the optic nerve millimeter to many millimeters and it does have sodium channels. But in general we've seen these processes before. They have paralogs all over the body. So glutamate, the major transmitter, we've seen it before. Other neurons in the retina make dopamine. We've seen that before. There are paralogs of dopamine allergic cells in the handlebar moustache that I like to tell you about, for instance. Some neurons make acetylcholine. We know about neurons that make acetylcholine. In the retina there are also inhibitory neurons that release GABA. So in short, the person who says to you at this hypothetical cocktail party, the eye is this wonderful complex set of cells and mechanisms that must have been intelligently designed. You reply, no, there are paralogs to genes and processes that occur elsewhere in the body. Channels, receptors, transporters, we've seen them all other places. I won't go in detail into the types of cells in the retina, but if you'd like to learn more about it, you could work in Marcus Meister's lab as a surf, for instance. So we move on now to connections to the brain. You've seen this picture before. They talk about Roger Sperry's experiments in 1948 involving goldfish. You remember that Roger Sperry cut the nerve leading from the optic nerve that those cells that conduct impulses, individual fibers go back to their original destination. Dorsal retina projects to the tectum. The posterior retina projects to the anterior tectum and vice versa. The tectum in higher animals is called the superior colliculus. So when he saw that fibers go back to their correct place, he postulated a so-called chemo affinity between the nerves and their target cells. These were very important experiments, but in terms of what the neuroscience community believes, they may not even have been Roger Sperry's most important experiment, because he then went on to conduct the Nobel Prize-winning split brain experiments that we know very well that form modern ideas about the specializations of the two hemispheres. So these chemo affinity type interactions lead, of course, to the maps. Maps are not unique to the visual system. We see them in the somatic sensory system. We see them in the auditory system. They're probably unique to nervous systems, but they are ancient. In fact, as far as I can tell, visual maps arose at least 500 million years ago. And Ralph has discussed visual maps already. He will discuss them in greater detail. So here is a creature from which we diverged about 500 million years ago. It is Lemulus, the horseshoe crab. You can't find any horseshoe crabs on the west coast of any continent, but horseshoe crabs are on the east coast of many continents. This may be Woods Hole, Massachusetts. It might be Recalbote Beach in Maryland, just hundreds and millions of Lemulus horseshoe crabs. This is about a foot across. The females are larger than the males. They have compound eyes with omatidia, very much like other vertebrates. They find their way around reasonably well. And they have maps from the retina to the brain. Now, remember that Sperry's chemo-affinity encouraged us to find molecules that subserve chemo-affinity. For the first couple of decades, all we found was chemo-repulsion. Ralph has told you about efferens and effkinases, which keep molecules away from each other and help to form maps. But in fact, over more recent years, actual chemo-affinity has been discovered. And I told you about the search for chemo-affinity, for instance, in the fly visual system. The fly is an arthropod, so is Lemulus. So there are maps in Lemulus that are reasonably similar in their molecular basis to the maps in Drosophila. But Drosophila is this wonderful genetic animal that allows us to find many types of molecules. So here would be the Drosophila nervous system, the retina, the lamina, the medulla. There are maps at several places in the Drosophila visual system. And I told you about the experiments from Kaizen's laboratory about a multi-gene family in which gene products can interact selectively with other gene products and form the kinds of chemo-affinity maps that allow us to have a precise visual system. So we've discussed this in our lectures on development. All right, where are we? We are at topic number five, which is master switches for eye development. I mentioned the PAX6 gene earlier in the talk, and this is a transcription factor that produces some of the components of the eye. Markably, one can take, and of course in Drosophila, PAX6 is called EY stands for eyeless, because a Drosophila with an interruption of the PAX6 gene has no eye. Well, you can do the opposite experiment, which is to take the primordial cells that are going to give rise to the adult eye in Drosophila. You move it to the leg, and the Drosophila generates a primitive eye spot. And so this is one of the classic experiments in Drosophila that you can move one organ to another. And in fact, you can simply induce EY, PAX6 or eyeless, in this spot and get a primitive eye. So there clearly is a master switch that goes part of the way toward controlling all of the gene activation in the eye. And remarkably itself, eye formation is so similar among organisms from arthropods to vertebrates that you can take a human PAX6 orthologue. Remember now these are orthologs because they have the same function in two species, not paralogs. And a human PAX6 orthologue will induce a primitive eye in an EY mutant and eyeless mutant Drosophila. So orthologs have an important role here too. Now the final topic is that your mythical friend or your hypothetical friend at the cocktail party will say, alright, that's fine, I believe it, but all of this evolution is so complex that it would take much more than the billion years we have available. Actually, the time frame for evolution of the major structural features, and remember this has happened independently, probably 40 times, might take only a couple of hundred thousand years. And so one can show a, or postulate, a series of intermediate steps in the development of the eye that each have use. So one starts with photoreceptors right here and a little dimple in the skin. They can receive photons. And then every intermediate step in which gives you better photoreception. So put the little photoreceptors in a curved spot, you can get better imaging. More and more curved, better and better imaging. Then a primordial simple lens gives better imaging and so on and so forth. And so we don't have to postulate an intelligent design of a complete eye. We can actually see the selective advantage of individual parts. So you can go to this website, which is now several years ago, about the estimation of the time required for an eye to evolve. And what I've just showed you is a summary of the possible intermediate steps in evolving an eye. I think this is the most pessimistic estimate. So we've gone all the way through the evolution and the parts of the eye. I've showed you orthologs and paralogs with other species. And I'll be at office hours this afternoon. Have fun with the midterm.