Rediscovering Biology: Molecular to Global Perspectives
Genetics of Development Expert Interview Transcript: John Incardona, Ph.D.
Incardona is a senior fellow at the University of Washington in Seattle. He studies the Sonic Hedgehog gene and its role in development. He also studies how the compound cyclopamine affects the hedgehog signaling pathway.
Interview with John Incardona, Ph.D. Icardona is a senior fellow at the University of Washington in Seattle. He studies the Sonic Hedgehog gene and its role in development. He also studies how the compound cyclopamine affects the hedgehog signaling pathway.
Can you describe the sequence that led to the discovery of the sonic hedgehog gene?
Sheep ranching has always been a major activity in the Rocky Mountains in the U.S. From around the turn of the century up until the 1950s there was always a low level of the birth of cyclopic lambs on sheep ranches in the Rocky Mountains. The sheep ranchers actually kept it hidden because it was at very low levels, and they thought it might reflect a genetic problem in their stocks. They didn’t really want anybody to know about it.
In the 1950s there were a couple of ranches, particularly in the Sawtooth Mountain regions of Idaho, where all of a sudden were having 25% of the lambs being born with this malformation. They couldn’t really ignore it anymore so that’s when they got the U.S. Department of Agriculture involved to try to figure out what was causing the malformations.
What did the USDA find?
Again initially when they first looked at what was happening, [they confirmed] that 25% of the lambs were born with cyclopia or a milder form of the malformation. In some senses that could look like a genetic problem, because a lot of times if you have a recessive mutation, a quarter of the prodigy will be homozygous for that mutation and wouldn’t have the malformation.
So they initially did some genetic studies, some crosses of various sheep and so on to try to determine if there was a genetic basis. They found that it wasn’t. So then they suspected one of the ranch plants.
These are sheep that are free-ranging in meadows up in the mountains. These were operations that were associated with production of lambs for consumption of meat. So they were very large lambing operations where they would take large numbers of ewes, mate them to a bunch of rams, and then take them up to the meadows in the mountains and just let them free range. There are a lot of wild plants growing around.
Basically what [the USDA] did then is a very rather involved epidemiologic study of trying to determine which plant it might be that was causing the malformations after consumption by the pregnant ewes.
The way they did that was first a sort of haphazard looking at the most common plants in the meadows where this was happening. It turned out that there was a sheepherder on one of the ranches, a fellow who had immigrated from the Basque region of Spain, who was astute enough to recognize that a lot of the pregnant ewes actually got sick and vomited after eating a particular plant called the corn lily, the Latin name of which is “Veratrum californicum.”
When they tested the Veratrum plants they found that they were now producing the cyclopic lambs. So then they went and did even more careful studies to try to figure out the timing and they found by feeding the pregnant ewes at specific stages during their pregnancy, there was a very narrow window around the 13th day where if the ewes ate the plant the lamb would then be cyclopic.
Would you compare and contrast the testing methods?
So this Basque shepherd suggested that they look at the corn lily because that was a plant that if the ewes ate enough of it would make them vomit, so he knew it did something to them that wasn’t good. The USDA scientists then fed the corn lilies to the sheep and found out that they did indeed cause these malformations.
Of course, plants make a lot of different compounds. Most of our medicines were derived from plants so they figured it must be some particular compound that the corn lily produces that causes this birth defect.
A chemist for the USDA then purified all the alkaloids out of the plant and fractionated them. They took different fractions and then fed these to the sheep, identified which one causes cyclopia, and then did a chemical analysis on it. They found three compounds that were very highly related that were steroidal alkaloids — they look like a steroid hormone — and the one that was most abundant they named “cyclopamine.” There was another compound which had actually been identified already in some other plants and then another which they named “cycopazine.”
They were then able to purify large quantities of cyclopamine in pure form much later-decades later-and this allowed us to do much more precise mechanistic studies to determine how cyclopamine caused these horrendous malformations.
Will you talk more about the recent studies?
It’s not like nobody did any work on this between 1960 and 1996. There actually has been a lot of work done with purified cyclopamine trying to identify its mechanism of action but nobody really could figure out what it was doing. This is where the genetics really helped out because it turned out that genetic studies of cyclopia and holoprosencephaly identified the Sonic hedgehog gene as being important in patterning the forebrain. A mouse mutant was generated in the Sonic hedgehog locus and the embryos produced by those mutant mice had holoprosencephaly and cyclopia.
It was later determined that humans with rare familio forms of holoprosencephaly also had mutations at the Sonic hedgehog locus. So here we have the Sonic hedgehog gene, which when mutated causes holoprosencephaly. We have a naturally produced compound, cyclopamine, from a plant that caused the same birth defect in range animals that ate the plant. When we looked at that we said, that is a very likely target. It’s very likely that cyclopamine could be acting on something about Sonic hedgehog signaling.
What the scientists for the USDA had shown is if they fed that purified compound to a range of animals-mice, rabbits, hamsters, cows, sheep-even if they opened up a chicken egg and put it on a chicken embryo they could get the same defects.
We knew that we could cause those defects in the chicken embryo and it turned out that a lot of the easiest ways to assess Sonic hedgehog activity are with a chick embryo. So we took cyclopamine, treated chicken embryos with it, and then went in with our molecular tools to ask the question: Is Sonic hedgehog signaling disrupted by having cyclopamine around?
With the purified form of cyclopamine, we can easily do more focused mechanistic studies with chicken embryos which are logistically much easier than [working with] a herd of sheep. But you can take large numbers of chicken eggs and work with them in the lab.
What does cyclopamine do to a cell’s Sonic hedgehog product?
Hedgehog is a signaling molecule that’s involved in patterning. The way patterning in the embryo works is different tissues-mesoderm, ectoderm, and so on-talk to each other and one of these ways that they talk to each other is through Hedgehog. The underlying notochord helps pattern the overlying nervous tissue by sending the Hedgehog signal. There are different points where cyclopamine could disrupt the signal-the origin of the Hedgehog molecule, delivery to the responding sign, and then reception of the signal in the responding cells.
It actually turns out that cyclopamine affects the ability of cells to receive and interpret the Hedgehog signal. It was a little bit more complicated to demonstrate that. To do that we had to go to an in vitro system, take little bits of embryonic tissue out, and culture it and do more sophisticated experiments. By doing that, again, using the embryo, but now pieces of it were taken out and cultured, we were able to show that reception of the signal was blocked by cyclopamine.
We found that cyclopamine is actually in pharmacologic terms called a “receptor antagonist,” so basically it’s an antagonist of the Hedgehog receptor. Cyclopamine interferes with the ability of cells that are receiving the Hedgehog signal to see it and interpret it. Hedgehog is derived from the notochord, which is this rod of mesoderm tissue. Hedgehog is sent from the notochord into the overlying neuroepithelium, the neuroepithelial cells, which are the precursors of the nerve of the nervous system, receive the Hedgehog signal. Cyclopamine actually blocks the receptor pathway within the neuroepithelial cells.
What happens in the course of embryonic development when cyclopamine is present?
Normally what is supposed to happen is the neuroepithelium is basically a sheet where all the cells are the same and it overlies the notochord, which is at the midline. In the forebrain regions in the anterior regions, the neuroepithelium needs to be divided into two halves, of which the eye is actually an extension. Even though it’s a peripheral neuroreceptor pathway, the eye is actually an extension of the central nervous system. This field of cells that are all the same needs to be divided in the middle and put into two halves and that’s what the Hedgehog signaling pathway does. It tells the midline cells to be one type and then extending away from the midline, various cell types are determined depending on how much of the Hedgehog signal they see.
Now if the Hedgehog signaling pathway is blocked, for example by cyclopamine, that division at the midline doesn’t happen and instead of having an eye field divided into two, it just stays right in the middle and you get one big eye or cyclopia. Also, you get holoprosencephaly, where in the brain part behind the eye — the forebrain — instead of having two halves there’s one.
In a normal embryo, the patterning of the nervous system happens on a temporal scale from head to tail basically, and it’s almost like a zipper being closed. Hedgehog signaling is coming from the notochord to the overlying neurectoderm pattern: the head, then the trunk, then the tail, and so on. So if Hedgehog signaling is blocked at this very early stage up at the head, you get these midline patterning defects represented by holoprosencephaly and cyclopia.
Specifically what happens in the anterior regions of the neuroepithelial or neuorepiderm, is that you start out with a field of cells where everybody’s kind of all the same, but you have to have cells that are obviously divided into two halves, because the normal situation is two eyes, two halves of the forebrain. So the prosencephalon, your cerebral hemispheres, are divided and you have one on each side. The Hedgehog signaling from the notochord is received by the neuroepithelial cells and it tells the cells right at the midline to be one type, a little bit further away from midline-depending on how much hedgehog signal those cells see-they become different cell types.
In a cyclopic embryo, what happened was the Hedgehog signaling was blocked at the time when this anterior patterning is happening, so the signal that went from the notochord to the neuroepithelium was not received and therefore the eye field didn’t get divided in half so you get one big eye forming instead of two smaller eyes on either side.
Now there are other secondary consequences for formation of the face as well. What’s going to become the nose starts out actually above the eye. It has to migrate down between the eyes to make the nose and if there’s one eye in the middle it can’t migrate. There are problems that happen with the jaw because now all these midline structures are affected by not receiving the Hedgehog signal.
Will you compare the function of the Sonic hedgehog gene for vertebrates and invertebrates?
Hedgehog proteins are involved in patterning in just about all metazoans-multi-cellular animals. Everything from fruit flies on up to humans all have some form of Hedgehog. Drosophila only have one Hedgehog gene. In vertebrates, being a bit more complicated, it’s evolved into a multi-gene family where they have Sonic hedgehog, Indian hedgehog, Desert hedgehog.
In all animals, Hedgehog proteins are involved in pattern formations. The larva of the fruit fly is segmented. It has all these segments with little hooks on it and that’s how it crawls like a little caterpillar. In Drosophila, Hedgehog is involved in setting up the anterior-posterior pattern of all the segments.
In vertebrates, Sonic hedgehog is involved setting up the dorsal/ventral, the back-to-front pattern of the nervous system. So telling one side from another is really what hedgehog signaling does in all different multi-cellular organisms.
Can you explain your research in more detail?
My research is focused on the mechanism involved in cyclopia and holoprosencephaly and more precisely how cyclopamine affects the Hedgehog signaling pathway. The first basic question is: What does cyclopamine do? The second question, once we determined that it was the Hedgehog pathway that was affected, was how does cyclopamine affect the Hedgehog pathway?
Now it’s relatively complicated because there could be multiple steps where a drug or a compound like cyclopamine could affect the pathway. There’s the generation of the Hedgehog protein sending the Hedgehog protein out of tissues like the notochord and delivering the protein to receiving tissues like the neuroepithelium. The signal is transmitted to ultimately affect gene expression, because what Hedgehog signaling does is it acts on cell types that are undifferentiated to turn them into a particular type of neuron. So ultimately its changes in gene expression patterns downstream of receiving Hedgehog signaling that are the ultimate effect of the signal.
So my research is focused on trying to figure out where among all those different steps could cyclopamine be blocking. It turned out we had to do somewhat more sophisticated experiments other than just looking at the embryos or looking at the patterning of the neural tube to really get at mechanisms. That required in vitro studies.
Can you describe the chicken embryo studies in more detail?
Development of the neural tube starts from the neuroepithelium, which is a flat sheet of cells and it has the underlying mesoderm derivative of the notochord. The spinal cord, which ultimately is formed from the neural tube, is a solid rod. So the neuroepithealium folds up, closes at the top, and becomes a neural tube. Hedgehog is a signal that comes in patterns the ventral side, the bottom side. There are other signals that come into pattern the dorsal side.
For example, in the fully formed spinal cord, you have motor neurons that come out of the ventral side and you have sensory neurons that are on the top. That pattern of dorsal sensory neurons and ventral motor neurons is patterned by a competition between the dorsal signals and hedgehog coming from the ventral side.
So one of the ways that we can figure out whether or not Hedgehog signaling was affected by cyclopamine was to look at all these different cell types, the precursors to the sensory neurons and the precursors to the motor neurons, etc. and see the relative proportions and try to figure out where was the problem.
For example, one way to get a ventral problem could be if we have too much signal coming from the dorsal side. There were various ways that we can look at that for example, we have antibodies that can tell the precursors of the motor neurons which start in ventral regions in two little pools on either side of the midline. There’s actually a structure called the “floor” which develops the neural tube immediately adjacent to the notochord called the foreplay. It’s this little triangular region. The foreplay itself becomes the source for Sonic hedgehog.
The midline cells of the neuroepithelium that are closest to the notochord become foreplay and later themselves start making hedgehog. So the notochord plus the foreplay is a little signaling center which is then sending hedgehog up into higher regions of the developing neural tubes –there’s a gradient. There’s a high concentration of hedgehog close to the midline and a lower concentration the further you get away from the foreplay and the neural tube.
The same thing is happening on the dorsal side. There are signals coming down from the dorsal side in a gradient and you can actually do semi-quantitative things where you can count cell types and such to see if you have enough sensory neurons, are there enough motor neurons, and are they in the right place.
However, if Hedgehog signaling is weakened in the foreplay to the notochord, you can get motor neurons developing at the midline because the motor neuron is determined by how much of the signal it sees. If you’re reducing the signal, it’s going to see less and it’s going to develop in the wrong place.
So in an embryo that’s cyclopic, if we section back through its developing neural tube in the regions that are going to become the spinal chord, we can take antibodies to sensory neuron precursors, motor neuron precursors, four point precursors and look and see that, yes, all of these cell types have been shifted down to the ventral side.
Of course that still doesn’t necessarily tell you if you’re blocking the ventral signal or promoting the dorsal signal. We had to do more sophisticated experiments in vitro to determine that and that’s what I want to do in later stages of my project.
Can you explain in more detail that if there is a disruption of Hedgehog signaling why malformations occur?
Disruption of Hedgehog signaling gives holoprosencephaly only because of the timing. The interior patterning happens first in the embryo in its pattern from head to tail so it’s strictly an issue of timing. So if Hedgehog signaling is blocked at that very early stage-which in the sheep is around day 13 and in a chicken embryo is around four hours of development — you get cyclopia and holoprosencephaly. It turns out that Sonic hedgehog is used by just about every organ system in the body during some point of their development. If hedgehog signaling is blocked at later stages you can have malformations associated with these various structures.
For example, limb patterning uses Hedgehog signaling, but it happens at a stage much later than the patterning of the forebrain. The whole axis of the neural tube is set up before the limbs start to be patterned. And actually back in the ’50s and ’60s when the USDA was doing their studies, they found that giving cyclopamine or Veratrum to sheep at later stages induced limb defects and we can do the same thing in a chick embryo. If you treat a chick embryo at later stages when other organ systems are being patterned, you can get defects related to the activity of Hedgehog in these other tissues.
Could you talk about the milder form of holoprosencephaly?
Patterning of the interior neural tube in some ways is a very concise event. Yet if you look at it closely, you can break that down into a temporal sequence. Particularly in higher mammals, it is relatively large so there are still smaller windows of opportunity to block segments of patterning of the forebrain. Holoprosencephaly is really a spectrum of malformations that can exist in very mild forms where, for example, only the olfactory bulb is missing.
Some humans with very mild forms of holoprosencephaly just might have a single central incisor, which is indicative of a very mild ventral midline patterning defect. Complete blocking of the entire series of events that happens in patterning the forebrain gives you complete holoprosencephaly, where the two cerebral hemispheres are fused, you have one midline eye. Cyclopia is actually the most severe form of holoprosencephaly — there’s sort of a little saying where the face predicts the brain. So the severity of the facial malformation can predict to a certain extent the severity of the brain malformation. If there’s one midline eye there’s generally one “holo” sphere-holoprosencephaly-where the cerebral hemispheres are not divided and you have complete holoprosencephaly.
If you break down patterning in the forebrain into even into smaller scale you can get milder forms of holoprosencephaly and as a consequence are actually milder facial defects. Defects that were seen in sheep with the Veratrum studies by the USDA and defects that I’ve seen in chicken embryos by giving various concentrations of cyclopamine get different ranges of these malformations.
For example, you can induce a chick embryo to have eyes that are very close together and instead of two nostrils in the beak, you actually have a single middle one. There are actually humans born with the birth defect called “cebocephaly” where they have a nose with one nostril and their eyes very close together. You can have eyes that are just barely fused at the midline as opposed to a single midline eye. The mildest forms of holoprosencephaly in humans can be represented by the absence of the corpus colussum, which is the big fiber tracks that connection the two hemispheres.
Patients that have milder forms of holoprosencephaly just might have mild developmental disability. Some of the facial characteristics of mild forms again could just be eyes that are rather closely spaced and sometimes a single central incisor so those are very mild forms of holoprosencephaly.
The severe forms where you have cyclopia are incompatible with life, largely because the animals don’t have a pituitary. The pituitary is a midline structure and if that’s gone then you have hormonal problems, plus you don’t have a normal airway for breathing so those are incompatible with life.
How did the Sonic hedgehog gene get its name?
Historically, Drosophila researchers always named their mutations after what they looked like. So, for example, the normal color of the fruit fly eye is red and there’s a spontaneous mutant that makes it white and the name of that gene was given “white.” Even though the normal gene makes the eye red when it’s mutated it makes it white so it’s called the “white gene.”
So when the scientists who first identified these genes involved in pattern formation in the early Drosophila embryo, they subsequently name “hedgehog.” They named it because the larva had bristles all over it. Remember, hedgehog sets up the anterior-posterior pattern of the segments in the fruit fly and these segments have a particular pattern of very fine little hairs which are what the maggot uses to crawl with. If there’s no hedgehog signaling, the whole thing instead of having neat rows of hairs, the whole thing is covered with hairs so it had all these bristles. It looks like a little rolled up hedgehog: plus the embryo was contracted, so it was this little fuzzy ball.
Drosophila has one Hedgehog gene. Scientists who identified the vertebrate hedgehog gene family, found three of them. So now you have three genes that are closely related, they’re expressed at different times and different tissues. They had to give them different names. Of course, the mutations of vertebrate hedgehog don’t make humans look like a hedgehog. It’s just that they had to get that name because they used the Drosophila gene to pull out the human homologue.
At the time when the Hedgehog gene was isolated from vertebrates, “Sonic the Hedgehog” videogame was popular as I suppose one of the postdocs that cloned it had a kid that liked to play Sonic the Hedgehog so that’s how Sonic was given its name. The other two they named “Desert hedgehog” and “Indian hedgehog.”
Can you talk more about Hedgehog signaling?
Hedgehog signaling is involved in determining cell types, so it acts in precursors of a particular tissue to turn them into a differentiated cell type. Some of these cells, if the pathway is constitutively active to where it’s not turned off, can actually turn into tumorous cells. If the pathway is always on in some precursors of brain cell types, they actually cause a brain tumor called “meduloblastoma,” which is a pediatric brain tumor that’s very hard to treat.
There’s actually hedgehog signaling in the hair cycle, in the skin. That’s something that even happens in adults. You’re constantly shedding hair and making new ones. Hedgehog signaling is involved in the development of a hair. If Hedgehog signaling is always on in the follicular cells that make a hair, you get a cancer called “basal cell carcinoma” which is actually the most common tumor in the aging population. If there are mutations in the receptor pathway in the cells that would normally receive the signal (from ultraviolet radiation in the skin) that result in constant activation of the pathway, then you get the tumor.
Since cyclopamine blocks the receptor pathway, cyclopamine can actually be used to potentially to treat some of these tumorous cell types because they arise from continued activation of the receptor. If you potentially apply cyclopamine to tumor cells, you can stop their growth and shut down the development of the tumor.
Would that cover other types of diseases too?
There are generally two types of clinical syndromes associated with the hedgehog pathway-reduced activity of the pathway, which gives you something like holoprosencephaly or over activation of the pathway, which gives you a particular tumor type.
Cyclopamine would really only be useful for those situations in which the pathway is too active, which is going to be the cancers that are associated with activation of pathway. Of course, the one caveat is it would be somewhat like the thalodomide story. Thalodomide was this drug that was banned in the ’50s and ’60s because it caused these horrendous birth defects. Now cyclopamine is going to be used to treat cancers. The thing that always has to be remembered is it has the ability to cause really serious birth defects. So in adults that have tumors that arise from activation of the pathway, it might not be an issue unless you are a pregnant woman. In children that have tumors, there still could be tissues in which Hedgehog signaling is important where you would be concerned about side effects. Even in adults, it’s in the hair.
There are other aspects both related to holoprosencephaly and other aspects of development, other organ systems where the Hedgehog pathway’s still poorly understood. One of those is how the signal itself is delivered to tissues so, for example, how it travels from the notochord to the overlying epithelium. There are in some cases where if you mutate genes that are involved with the transport of Hedgehog you get the same effect. You get holoprosencephaly.
So one of the things that we’re working on now is trying to understand how the protein is transported around. The precise mechanism of action in cyclopamine is still somewhat up in the air. We know that it probably binds to the receptor component that actually transduces the signal, but how it affects that protein we don’t know yet. That’s going to take really sophisticated biochemical experiments to figure that out and there are groups around the country that are working on that, particularly because cyclopamine is going to be used for potentially anti-tumor treatments. There’s going to be a lot of interest and a lot of money being spent — there’s pharmaceutical companies involved-so now there’s a lot of people interested in this compound that came from the Rocky Mountains and the Cascades because of these malformations in sheep.
The precise way that cyclopamine antagonizes the receptor pathway isn’t known yet but scientists are pretty close to understanding it. Now we know it probably binds to one of the receptor components specifically and affects its ability to transmit and transduce the signal from the outside of the cell to the components that ultimately affect the gene expression, so that’s really where the cutting edge of the research is right now.