Alejandro Sanchez - Alvarado - Stowers institute - Part 3 - Molecular basis of regeneration in planaria

Hello. My name is Alejandro Sanchez Alvarado. I am an investigator with the Howard Hughes Medical Institute and a professor in the Department of Neurobiology and Anatomy at the University of Utah School of Medicine. This is the last part of a three-part series on the topic of regeneration, which I am talking about to you today. This part is going to investigate the experimental paradigms and molecular and cellular approaches that we are undertaking to try to understand, at a less phenomenological level, the basis of regeneration. And so, as you can read from the title, this is going to be essentially a summary of the molecular basis of regeneration using planarians as a model system. So let's start with the first slide. And what I'm going to show you here is one of the reasons why we think that regeneration in planarians is really a good place to actually begin to look for unravelling this problem. Regeneration in these organisms is both fast and robust. In fact, it's almost fast and furious, because you can decapitate an animal, and seven days after such decapitation, that animal has a head that is capable of performing all of the functions that the animal displayed prior to amputation. So, here are what we call a regeneration series of a planarian - cephalic regeneration. So, here's the organism at time 0. This is even before the wound actually contracts. On day 1, we begin to see that the wound is healed. And on day 2, essentially, we begin to notice that this unpigmented area begins to show accumulation of undifferentiated cells. These are the cells that are now directly apposed to the epithelium of the regeneration blastema and will respond to those signals to actually give rise to what we call a regenereation blastema, which is shown right here on day 3. It's this white, unpigmented tissue opposed to the pigmented, pre-existing tissue. On day 4, you might be able to see this... We begin to see what is going to be the beginning of the photoreceptors. Also, notice anlage, shown right here and right here. Those two condensed areas are the areas where the photoreceptor neurons are going to form, the pigment cups are going to form, and eventually the photoreceptors will emerge. On the fifth day, essentially this animal has completed most of its differentiation of the missing structures in this regeneration blastema, this white area in front of the animal, such that by 7 days, the animal now is completely competent to respond to light, to respond to food stimuli, to touch and so forth, and behave appropriately, and you can see that right here. Here's the animal, you can see the two black spots right here are the photoreceptor organs with the pigment cups. And that's essentially on day 7. So, that's one week in which this process occurs. That means that it is amenable to experimentation. You don't have to wait for 30 days or a month or half a year to determine whether or not... what the outcome of your experiment is. If it was performed successfully, is or is not going to be. So, essentially, in 1 week, one can interrogate these animals exhaustively and get an accumulation of sufficient data to begin to increase our complexity of understanding of the problem of regeneration in animals. The organism that we chose for our studies is the genus and the species Schmidtea mediterranea. We chose this as a model planarian for a number of reasons. Number 1: because we want to extrapolate from these organisms into our own biology, we wanted an animal that was bilaterally symmetric, like you and I. So, while that mentioned in one respect that if you were to bisect this animal along the midline of the animal, you essentially produce two mirror images, just as if you were to do the same thing with us, split us in half, our internal organs are slightly asymmetric, but we'll still have one arm here, one arm over here, a leg and a leg... bilateral symmetry and so forth. So, these animals are bilaterally symmetric, and the other thing that's also important for us is that they have derivatives of all three germ layers that form in the embryogenesis of more complex animals like mammals to produce all of the appropriate organs and structures. These germ layers are the ectoderm, the mesoderm, and endoderm of which planarians have all three. In their body plan, even though they are coelomates, meaning there's not a coelum, where the organs reside, they actually contain a series of complex organ systems that essentially sit on top of each other like geological strata. And these are, again, derivatives of the three germ layers. So, you have a situation here, where there is depicted in purple or lilac the gastrovascular system. This is where the food is actually processed to serve as a source of carbon for all of the other cells of the animal. The food enters through a single opening, which is part of the pharynx, and so you can see the pharyngal tube here. Food will enter this way, go into the gut, and then from the gut, get dispersed throughout the rest of the animal. The food that is not digested for the most part is now actually secreted out or excreted out from this same pharyngal opening. There's also a series of excretory cells that regulate osmolarity of the cells and ionic content of the cells, and this is known as the excretory system, composed of protonephridia. And they are again on the dorsal surface, right here, and these are equivalent to our kidneys, for example. And then, they also have a very complex central nervous system consisting of two cephalic ganglia, which are shown right here, or two hemispheres of the brain analogously, and then these two cephalic ganglia sit on top of two ventral cords, which are neurons that traverse all the way to the very, very tip of the animal. Both of these ventral cords are connected by commensal neurons, such that there is neuronal contact between these two ventral cords that traverse the anterior-posterior axis of planarians. And they also come in two flavors... so called *** and asexual biotypes. So, the *** biotypes of these animals, the Schmidtea mediterranea, are hermaphrodites. They are hermaphrodites, and those have both *** and ovaries, and the appropriate reproductive machinery or apparatus to allow for exchange of gametes. Even though they're hermaphrodites, they cannot self-fertilize, so they actually have to copulate with each other to exchange gametes, and this is shown right here. It's a very complex series of structures that actually allow taxonomies to distinguish species from each other. But, there's also an asexual form, and the rest of our conversation today is going to focus essentially on that asexual form. And, the asexual form actually arose as the result of a chromosomal translocation, from chromosome number 1, shown right here, to chromosome number 3, shown right here. So, you can see here, that the chromosome 1 is missing all of the structure right here, but now in chromosome number 3, you can see that there is an excess material of chromatin. In comparison to the *** strain. This translocation is responsible for the complete elimination of all of this *** reproductive machinery. These animals no longer produce gametes, the asexual animals, and the only way they have to reproduce is asexually. And so, how do they do this? Well, the animals essentially... the tail will adhere to the substratum in which the animal resides, and then the head will begin to crawl away from that anchor - from that anchored tail until a very, very thin waist actually forms that just pinches off. And you end up now with a tail that will go on to regenerate a head, and a head that will go on to regenerate a tail. And this particular situation produces essentially two clones or two copies of each other. And this is asexual or progenitive reproduction. Another fine attribute of these animals is that these animals is that they actually have ... besides having traditional looking chromosomes with centromeres and telomeres and so forth ... is that their genome is relatively small. It's about 5 times the size of Drosophila, to a size of approximately 700 Mb of nucleotides that the DNA contains. So, this will allow us, also, or has allowed us, actually, to go on and try to sequence the genome of this organism. So, what I'm going to do in the next slide is to show you a movie of how all of these different organs are distributed along the D-V axis of the animal. I want to give you a little preamble for this movie. What we've done is that we've taken this animal, we have fixed them, so they're no longer alive, and then we label all the proteins that these animals have that have been phosphorylated on their tyrosine residues. So, this essentially means thousands of proteins, so that way we can visualize cells within almost every organ system, if not all organ systems, in these animals. And then by using a methodology known as confocal microscopy, we took optical sections from the dorsal to the ventral area of the animal, and then, I'm going to show you that series of optical sections as they traverse from dorsal to ventral. And I'm showing this movie right here. So, what you see here is essentially the most dorsal part of the animal. You can see now that the photoreceptor pigment cups are illustrated here. This is part of the excretory system. Now you can see some of the branches of the gastrovascular system right there. This is where the food is distributed. Here and here were the connections of the photoreceptor neurons, which are stereotypic to the cephalic ganglia, which is this spongy looking structure right here. This is the brain of the organism. This brain actually sits on top of two ventral cords, which may or may not be visible here, as these two lines right there. And as you traverse more, you see some musculature and then the cilia or the ciliated epithelium on the ventral surface. It is this ciliated epithelium that allows the animals to locomote. Even though they have a very complex web of muscle fibers covering the entire outer circumference of the animal, the animals essentially locomote by cilia which are on the ventral surface that beat essentially in one direction, propelling the animal forward. If the animal runs into an obstacle, in order to negotiate that obstacle, it will use its musculature to lift its head, turn its body, lay the head down, and then the movement of the cilia will now propel the animal in that direction, or whichever direction the head is pointing. So, this is essentially what the general structures of the animals are, and what's truly remarkable is that all of these structures that I show you -- the dorsal epithelium, the excretory cells, body wall musculature, photoreceptors, cephalic ganglia, ventral cord, and so on -- after amputation are regenerated de novo and reintegrated to the pre-existing tissues in the animal. So, there's another attribute that's quite peculiar and interesting to us about these animals. And that is that this regenerative capacity and developmental plasticity that planarians display find their root in a very interesting population of cells of undifferentiated cells that actually populate most of the animal. These cells are known as neoblasts, which is a term that was coined by Harriet Randolph in 1893, when she was studying another worm, Lumbricus in this case. What you see here is where these neoblast or undifferentiated cells are located in the body plan of the animal. What we've done here is to use two molecular markers. The first one is BrdU, which is shown right here. And it's these little green spots that you see distributed throughout the animal. The head of the animal is right here, the tail of the animal is right there at the bottom, the pharynx is right here in the middle, which is devoid of this signal. What we do is that we provide these animals, in their food or by injection, BrdU, which is a thymidine analog that gets incorporated into cells when they are entering the cell cycle, so-called the S phase of the cell cycle. So, this thymidine analog gets incorporated into the newly synthesized DNA, and since it has a halogen on its composition, in this case bromine, there are antibodies that will recognize this bromine, and will allow us to detect where the bromine is, which is attached to this thymidine analog, which in turn is part of DNA. So, we can label the cells that actually took up the BrdU. So, we provide the animals with a single feeding of BrdU, we let some time progress... let's say 12-16 hours, and in that period, we fix the animals. We put them into a fixative, and now we process them to detect with an antibody where the BrdU got incorporated. And what you see here is where that BrdU is, and these are the cells that are dividing. And these are essentially the neoblasts. And what you essentially see here is that they are distributed throughout the body plan of these animals, with the exception of two areas -- the pharynx, which I pointed out to you earlier, which has very little signal to speak of and the area in front of the photoreceptors. Interestingly enough, in 1898, T. H. Morgan demonstrated that these are the only 2 fragments that when removed from a planaria are incapable of regenerating the complete animal. And we think that the reason for that is that both of these tissues are devoid of cell division or mitotic activity, and therefore are incapable of mounting a regenerative response. Even though the animal itself will go on to regenerate the missing pharynx and the area in front of the photoreceptors, these two structures will not do it on their own. To confirm these results, we can look at another independent marker, in this case, a marker that does not depend on the incorporation of an exogenously added analog, for DNA synthesis, but we look at a modification in proteins that takes place during the cell cycle... in this case, an antibody that was developed by David Alice a few years ago for detecting changes in histone phosphorylation in Tetrahymena known as a phosphorylated form of histone H3. This antibody recognizes histones H3 that have become phosphorylated. And what is really neat about this is that phosphorylation occurs during 2 phases of the cell cycle, the G2 and the mitosis transition. So, at that point, the phosphorylation occurs. It's the only time that it occurs, so it's essentially labeling cells that are entering mitosis. And we essentially see the same general distribution of mitotic cells as we see for BrdU, and that is that there are no mitotic figures of H3P signal in front of the photoreceptors, none in the pharynx, but the distribution of mitotic figures is essentially throughout the rest of the body of the animal, just as we saw for BrdU. So, what are these cells that are labeling with BrdU? Well, these cells are the so-called neoblasts that I told you about earlier, a term again coined by Harriet Randolph. Why were these neoblasts easy to recognize as early as 1893? The reason for that essentially is that these are very peculiar cells in an adult animal because these are completely undifferentiated cells. So, initially, what people thought was that these looked like embryonic cells, but in the adult. Why did it look like embryonic cells? Well, they are undifferentiated, they have a very large nucleus, which I have pseudocolored here in green, and very scant (and sometimes highly basophilic) cytoplasm. By basophilic I mean, they'll actually stain really dark purple, and we believe the reason for that is that there's a lot of RNA and ribonucleic acid in this cytosol. They're also small -- they're about 5-10 microns in diameter. And, the most interesting attribute for us is that, in planarians, it is these neoblasts that are the only cells capable of undergoing cell division. All other cell types are essentially post-mitotic. So these cells are the only ones that are dividing, and therefore the ones that are actually producing division progeny to produce all the other differentiated cell types that you find in these animals. There's about 30-40 different cell types, based on morphology alone. So, this is just to give you a general idea of the biology of these animals. So, now we want to go in and try to do the same kinds of things that experimental biology would like to do, which is to actually look further to try to uncover more evidence by increasing the resolution of the tools with which to interrogate the system. So, it is now possible to actually visualize gene expression in the whole animal. And we do this through a process known as whole-mount in situ hybridization. And what this essentially means is that we can synthesize, in a test tube, an RNA molecule that is complimentary to the endogenous messenger RNA, and by complementary, I mean that it will actually recognize the other bases, such that they will hybridize to each other forming a duplex. The synthetic RNA that we synthesize, we tag it, just like the BrdU was tagged with bromine, we tag it with a molecule that can be detected by an antibody, and then we subject these animals to hybridization with this probe. And then we develop it with the antibody and see which cells are expressing the genes that we're interested in looking at. So, here are examples of the complexity of expression patterns that these animals are capable of displaying. So, in this particular case, we see the expression pattern of a major metalloprotease surrounding, essentially, all of the area around the pharynx. So, this is a molecule that is expressed only in these cells. The rest of the animal was exposed to the probe, but there is no positive signal. Here's an example of a gene that is actually expressed exclusively in the gastrovascular system. This is the digestive system of the animal. It's not in the pharynx, but it's in the rest of the animal. You can see the anterior branch that I showed you earlier in that movie with the anti-phospho-tyrosine antibody, and you can also see how all of the branches in the rest of the animal essentially are labeled with this probe. We also find discrete cell types being detected by different probes. We also find probes that actually label the brain or the cephalic ganglia quite nicely. You can see its delineation, right here. And then, we also find probes that actually allow us to even increase the resolution of analysis for organs, such that we can find genes that are expressed in specific regions of those organs, and those are shown in the slides right here. Here are what we think are sensory neurons expressing a sodium channel. Here is the periphery of the cephalic ganglia. So genes that are only expressed in this area, and also the anterior part of the brain shown right here, and so forth and so on, where the different probes will actually produce different marks. So, that's now possible for us to visualize gene expression in these animals. That's a good step forward in trying to understand which molecules might be involved in all of these different cells that make up planarians. So, one of the things that we wanted to do very early on is not just to catalog these genes, and not just to actually take the molecules and see where they are. We want to know what the molecules actually do. And to ask that question, we needed to devise methodologies that will abrogate or eliminate the function of these genes. We call these loss of function assays, where we can specifically go into the cells and eliminate or remove the function of a particular gene by getting rid of its expression or by getting rid of its protein. So, we decided to explore the methodology that was originally devised or described for C. elegans by Andy Fire and his laboratory in 1998, known as double-stranded RNA or RNA interference or RNAi. This methodology essentially targets transcripts or known sequence, and by targeting them, they actually target them for destruction once they go into the cytosol of the cells. In Andy's laboratory, a post-doctoral fellow, Lisa Timmons developed with Andy a methodology that allows the delivery of these double-stranded RNA molecules as little pills in bacteria. So, essentially what we do now is that we clone our gene into a vector that gets transformed into bacteria, such that bacteria actually produce the double-stranded RNA we want to use to silence genes in planarians. We then mix these bacteria that are producing the double-stranded RNA with an artificial food mixture, and then we feed these to the animals. And, the food now is distributed throughout the body of the animal, and with it it distributes the double-stranded RNA. Double-stranded RNA will traverse the membrane of the cells, and then if the gene is expressed in those cells, it'll target it for destruction. If not, nothing happens. The advantage of doing this is that it allows us to generate a permanent reagent because these bacteria with the clones producing the double-stranded RNA can be kept in the freezer indefinitely. It's easy to produce ... all you have to do is just grow the bacteria. And then it's also easy to feed a large number of animals to deliver these RNAs. We need numbers such that we can test whether our findings are reproducible or not. And so essentially, what we do is that we take these animals... maybe 10 or 20, and then we allow them to eat this food. And then the food gets essentially dispersed throughout the body of the animal, and if the gene is being expressed in a specific cell type, it'll be targeted for destruction. So, what we decided to do was the following: We wanted to use this methodology as a way to begin to functionally interrogate the molecular basis of regeneration. So, we wanted to test a relatively large number of genes and then see what their role might or might not be in the process of regeneration. So, we call this an RNAi screen that was essentially performed to uncover genes involved in regeneration. And this work was carried out by several people in my lab: a post-doctoral fellow, Peter Reddien, and two undergraduate students, Kenny and Adam Bermange. So, what we have here is a situation where we have cloned about 1,065 genes into these vectors, created 1,065 bacterial lines that we can feed independently to cohorts of animals, and then ask the following question after they eat: We take the animals, we amputate it, we do it a couple of times to make sure that the effect is complete, and we ask the question: Can the animals regenerate? And if they regenerate, is that regeneration normal, or is it defective? So, this is essentially what we did, and it was published in Developmental Cell in 2005, if you want to consult the details of these experiments. What turned out was the following: We tested 1,065 genes, and we uncovered 240 different phenotypes that were somewhat associated with the process of regeneration. Through a series of secondary screens and further analysis which I'm going to mention to you in a few minutes, we were able to place all of these 240 genes into specific stages of the process of regeneration which I'm showing you right here. So, here's when the animal's amputated. We call that amputation. The first phase is wound healing. The second phase is the ability of the neoblasts to respond to the signals from wound healing and its proliferation. The third phase has to do with the division progeny of these cells, the neoblasts, to give rise to the blastema -- the cells that eventually are actually going to restore the form and function of the animals. The fourth phase has to do with the differentiation of these undifferentiated cells that were produced by the division progeny of the neoblasts and how they're actually going to differentiate and pattern. Then the fifth phase, which is shown right here, actually is how the actual new tissue that was formed out of that regeneration blastema goes on to integrate to the pre-existing structure of the animal, which is called morphallaxis, as discussed in the prior lecture. And the final phase is what happens to the animal once all this is done: the maintenance of all of those tissues that are turning over and the division progeny of the neoblasts are replacing as the animal ages. So, I want to start with the first phase, which is wound healing and initiation of regeneration. So, for this particular experiment, we found several genes, and these are the results of one such gene. Here's a situation where the animals were amputated, as I described to you, after being fed this particular molecule, and what you can see here is a trunk fragment... so the animal was truncated pre- and post-pharyngeally. And, what you see here is that this animal is incapable of producing both anterior and posterior blastemas. An intact animal after regeneration would look like this, suggesting that here was a cohort of genes that was responsible for initiating regeneration. There were also a few other genes that prevented the wound healing proper, and essentially, after amputation, the animals just lysed. They were incapable of healing the amputation wound. The second type of genes that were uncovered in the screen had to do with neoblast maintenance and the response of these neoblasts to the amputation. And these were actually quite interesting, because we already had an example of what a planarian without neoblasts should look like. So, if you remember from what I told you earlier, I said that neoblasts are the only cells capable of undergoing cell division. So, what one can do, in this particular instance, is to subject the animals to treatments that will actually destroy cells that are dividing. Such a treatment, for example, is X-irradiation, which is being used in radiotherapy for treating cancers and so forth, since those cells are rapidly proliferating, and chemotherapy has the same effect. Irradiation was more accessible to us. So, if we subject an animal to large doses of irradiation, what happens is that we kill all the neoblasts, and then we can observe what happens to the animal when the neoblasts are absent. They've been basically eliminated from the animal. And essentially what we see is this: Here's an animal that was decapitated and its tail removed, and after 10,000 rads of gamma irradiation, what you see here is that neither the head nor the tail regenerated. So, there was no regeneration at all. With time, what happens is that this animal that is incapable of regenerating, which you see right here, begins to curl. We call this a curling phenotype which is very stereotypic to this type of consequence. And then animal essentially just lyses in due time and dies. That's also true if you don't cut the animal, if you keep it intact, because the neoblasts in the intact animal are also dividing. And what happens here is that the area in front of the photoreceptors begins to resorb... I told you earlier that there's no mitotic activity in that area. So, what happens is that the division progeny from cells behind the photoreceptors travel to that area to maintain that tissue intact. And in this case, since the neoblasts are eliminated, and there's no mitotic activity in the area in front of the photoreceptors, That is probably the reason why that area begins to resorb first. And then, in due time, the head is completely lost, and then it assumes also this curling phenotype that you see up here in the amputated animal. So, as it turns out, by introducing double-stranded RNA against a single gene, we actually were able to recapitulate this phenotype -- an inability to mount a regenerative response. You can see that all of these animals are devoid of blastemas and a folding of this ventral surface into this curling phenotype. I want to point out something important here, and that is that, if you look at all of these, we know what the names of these genes are, and we know what the names of these genes are because these planarian genes have homologs or cousins so to speak with the genes of other species, in particular, mammals, and humans included. What is truly remarkable about these studies is that about 80% of all the genes that cause a regeneration phenotype in planarians, you and I have. We actually share those genes, so it's important to think that the genes that we're studying to understand regeneration are actually genes that our genome actually contains, and that those functions in all likelihood are going to be conserved -- the molecular functions of those genes. Our question is essentially the following: What is the function of these genes in the context of regeneration? And why is the function of these genes in organisms that cannot regenerate not being deployed to allow regeneration to occur in those systems? To give an example, here's a chondrosarcoma protein, a planarian homolog right here. Chondrosarcoma really says that this is a type of bone cancer. So, we know that planarians have it, even though they don't have bones. Here's another molecule, CDC23, which is a cell cycle regulator, a molecule that is conserved from yeast all the way to humans. Here's a PI transfer protein, for example, that is involved in phosphoinositide metabolism. And, histone deacetylases, and so forth, and so on, suggesting that it's not that these animals can regenerate because they have genes that are unique to them, but they're doing so with genes that are already widely conserved across evolution. So, one of the things we wanted to know was the following: Why do these animals not regenerate? Is it that they don't regenerate because they don't have neoblasts as in the case of an irradiated animal? Or is it because they're not responding to wounding or what is going on? So, we need to increase our powers of observation and analysis by looking at cellular phenotypes. So, this gives us, essentially, an example of morphological phenotypes. Gross anatomical defects that we can look and say "Aha! Something is wrong with this animal." We want to know what is the cellular cause of that defect. So, to do that, we go back and use that anti-histone-H3-phosphorylated form antibody to see what the mitotic figures look like. So, we know that if we irradiate the animals, we get rid of the neoblasts and we throw in this antibody, we detect no mitotic figures, because all of the neoblasts are gone. So, the prediction would have been that well, maybe the reason why they don't form a blastema and they assume this curling phenotype is because there are no neoblasts there. So, I'm going to give you an example to show you that it was not that simple. So, here's the control... here's an animal that will go on to regenerate its tail and its head. We can actually quantify each of these spots, and essentially what we see here is that there's normal distribution of mitotic figures. Here's an animal that was incapable of forming a head and a tail blastema and nevertheless had essentially the same number of mitotic figures that you would find in the wild-type situation, suggesting that the cells are dividing just fine... they're producing division progeny, but for whatever reason, they're incapable of mounting a regenerative response. So, the cells are still there. Here's the other one that we would have predicted would have been the most likely effect or consequence of an inability to regenerate, and that is that the neoblasts are not dividing, that there's no cell proliferation, and by absence of cell proliferation that prevented the formation of the head and the tail. And so, we showed that here, and we call this low mitotic figures. And this is the paradoxical one. Because this one... this animal right here... a gene CDC23, which is a cell cycle regulator... What happened here is that, in fact, there were many more mitotic figures than in the control, right here. So, how could an animal that has more mitotic figures and is making more division progeny still be incapable of mounting a regenerative response? So, these are the types of phenotypes that we found at the cellular level that, nonetheless, as varied as they are essentially produce the same phenotype -- an inability to produce regeneration blastemas. So, we decided to quantify this and visualize this in a better fashion. And what I want to explain to you here is this graph that we use to catalog all of these genes, based on the number of mitotic figures that each of these phenotypes produce, from very low all the way to very high. Normal being right here in the middle. And then we also quantify the size of the blastema, the new tissue that is being regenerated. So, the situation is here, as follows. Let's start with the one that we would predict would happen. Ones that have very low mitotic figures... each of the genes is represented by a single blue square right here that produces 0 blastema. So, those are those genes right here. What are those genes that are low to very low mitotic figures and still produce no blastema at all? Well, there were a number of metabolic genes involved here. These are not specific to regeneration... things that are absolutely required for the survival of all cells, and that's why no regeneration occurred. But, among them, there were some very specific genes that are likely to actually be involved in the initial stages of regeneration. Examples are these chromobox genes, this PI transfer molecule, this structure specific recognition protein, as well as histone deacetylases for example. But, what about the other genes? The genes that, actually when silenced, produce huge amounts of mitotic figures and still are incapable of giving rise to regeneration or regeneration blastema, which are shown right here, this 0 blastema. Well, it turns out that the reason why these animals were incapable of mounting a regenerative response is because all of the cells that were cycling were being arrested in a specific place in the cell cycle. Most of the molecules that produced this phenotype, CDC23, gamma tubulin, proteasome C2, and proteasome B7, they are all part of the anaphase promoting complex. So, essentially what was going on here was that the neoblasts were entering the cell cycle, but they all got stuck in anaphase, and they were unable to exit the cell cycle, and that's why we began to see an accumulation of mitotic figures. This is essentially what told us that the screen had worked, because the screen was done blind, and once we looked at what the function of these genes were, they essentially phenocopied each other and they are known to molecularly interact with each other suggested that, whenever we find a phenotype that's similar to each, they may be somehow related to each other, and they may be pointing to specific molecular or cellular pathways to explain the phenotype. So, the next stage of regeneration actually involves the neoblast progeny's function and blastema formation, proper. And those genes are represented by the group of genes that are shown in this graph right here, in the middle, where we essentially see that they have normal mitotic figures, but they also have a range of blastema sizes, ranging from 0 all the way to almost complete -- 2 to 2.5. And these are the list of genes that I think are actually involved in carrying out the functions of the division progeny of neoblasts to actually regenerate the missing structures. And there is essentially a laundry list... a variety of genes that are probably representing just as many pathways involved in the process. A complex situation that can now begin to be dissected molecularly to try to understand the hierarchies of function. Who is controlling who, and so forth, and so on? The stage 4 of regeneration actually refers to differentiation and patterning. And so, these are animals that are able to mount a regenerative response. They went through the first two phases apparently uninhibited. They form a normal regeneration blastema of size 3, normally indistinguishable from the control, but when they begin to differentiate, it shows some defects. We wanted to see what those defects might be at the cellular level, and for that we used another antibody that recognizes the photoreceptors neurons of these animals. This is an antibody that was given to us by Kiyakazo Agata in Japan. And this marker labels the photoreceptor neurons, which we like because they actually send projections to their own side, as well as across the midline into the other side of the brain. The other thing that is really nice about these neurons is that they are dorsally located, but they are traveling ventrally to send their projections. So that allows us to test D-V patterning, as well as left-right or midline patterning in these animals. And this is what we get. We get a large number of defects that could be detected with this antibody. So, here's the situation for the wild-type control. You can see here, where the photoreceptor neurons are... this is on one side. This is what we call an optic chiasmata, which is the fibers that actually cross the midline of the animal. And here's the other one on the left side of the animal. And here's the midline of the animal. So, when we look at some of these genes that gave us defects in regeneration, we find for example that these photoreceptors are disorganized. So, you can see that they are really protruded out, the chiasmata is retracted, and there are very few projections posteriorly. There are others that gave some type of asymmetry, for example, where only one photoreceptor formed, and the other did not. Others have produced cyclopia, where the actual photoreceptors did not really separate from each other and just form a single photoreceptor. And again, I want to emphasize that the genes that are actually causing these defects are genes that you and I have in our genome. So, there are functions that are evolutionarily conserved, as we can now begin to use planarians to try to understand what their functions are in the context of patterning and differentiation during adult regeneration. And there's a few others right here, on this side of the slide that essentially show the same displays... the same type of defects. For example, in this case of a bone morphogenetic protein 1, the projections were incapable of crossing the midline, for example, where in this case, the optic chiasmata is incredibly long and then sends very short projections posteriorly. So, these allowed us to catalog these defects into differentiation and patterning defects. Then, morphallaxis is the penultimate aspect of the screen that I want to talk to you about. The morphallaxis refers again to the functional integration and anatomical integration of the newly formed parts to the pre-existing parts of the animal. And, for this, the following assay was carried out: We looked at genes that affect the morphallaxis of the tail fragments. That is, the emergence of the pharynx, which actually emerges in the pre-existing tissue rather than in the regeneration blastema. And we essentially found 11 genes that allow more-or-less normal blastema formation, but really affected the remodeling of these particular tail fragments, suggesting that these genes may be playing a role in the remodeling of the pre-existing tissue to give rise to missing body parts. Finally, we wanted to distinguish genes that were specific to regeneration vs. homeostasis. And so, homeostasis refers to the maintenance of all of the structures in our body, but in the absence of injury. In the absence of traumatic injury. So, for example, planarians, when unamputated, as their cells age, they die. And as they die, the division progeny of neoblasts will now go on to replace those dying cells. So these animals essentially escape death by constantly replacing their dying cells from division progeny of the stem cells or the neoblasts. So, many of these genes are involved in regeneration are likely to also be involved in homeostasis, but we also postulated that perhaps, regeneration might have a cohort of genes that are unique to regeneration and are not being used in homeostasis. So, we took 143 genes that we knew were affecting regeneration, and we now did the following experiment: We would take these genes that we knew caused a regeneration defect, and we fed it to animals, and rather than cut the animals, we just leave them alone, and we observed them every 2-3 days for 10 weeks, and then see if a phenotype would emerge. And out of this, we got a total of about 113 genes or so that actually did cause defects in the unamputated animals, suggesting that these genes were not only necessary for regeneration, but also for tissue homeostasis. The converse of that, or the complement of that particular set of genes, are a number of about 30 genes or so that did not severely affect tissue homeostasis, but that actually have consequences in the ability of these animals to regenerate missing body parts after amputation, and we think these genes might actually be involved in regeneration proper. So, I want to summarize what I've told you as follows: We think that we can now begin to use planarians to really delve into the molecular and cellular aspects of regeneration in multicellular organisms like you and I. We can now measure biological processes in planaria, such as quantifying blastema size, quantifying cell numbers, and so forth, and so on. It is also possible to catalog and visualize their expression by whole-mount in situ hybridization, as I showed you earlier. It is also possible to interfere with their function... interfere with gene function by using double-stranded RNA and invoke the mechanisms of RNA interference. So, that allows us then to find out what happens to the animal when a particular gene product is missing and what consequences the absence of that protein may have on the biological processes that we want to study, such as regeneration. And then, finally, I think the most important thing for us is that we can now begin to use all of these accumulated tools and knowledge to begin to systematically dissect the mechanisms of regeneration. What is remarkable in my mind is that many of the genes that are being used by planarians to regenerate their missing body parts are actually shared with other organisms, including mammals. We also know now, by sequencing the genome, that there's a number of genes that have been lost in other invertebrate model systems, but which are present in humans that are shared with planarians as well. (The other model systems being Drosophila and C. elegans.) So, now we can begin to ask questions of genes that are present in humans in an invertebrate model system that has them, such as planarians. So, this brings us to the end of the series of talks, and I hope that you enjoyed it, and please consult the literature that's associated with this series of lectures to go further into the topic of regeneration. Thank you.