The cells come from the early embryo of a mouse, three and a half days after conception. It’s when the embryo looks like a beach ball, explains Mitch Weiss, a hollow ball, and inside that ball is a small mass of cells that are destined to become the embryo, the whole embryo, every part of it from the head to the feet.’’ The cells are known, in the technical jargon of biology, as undifferentiated. Weiss, who is a physician and a researcher at a Cambridge, Massachusetts, biotechnology company known as Ontogeny, Inc., uses the word naive to describe them. They haven’t become educated to become heart, liver, lung, blood, whatever, he says. They are identical, at least for the moment.
Before coming to Ontogeny a year ago, Weiss spent a decade at Boston’s Children’s Hospital and the Dana Farber Cancer Institute treating and studying blood diseases and cancer in children. He also studied, he says, the basic biology of how blood is formed. In the human body, red blood cells survive 120 days, white cells anywhere from 24 hours to 10 years; platelets, which help blood clot, will last 5 or 6 days. All of them are constantly being regenerated from a single type of cell in the bone marrow known as a hematopoietic stem cell. If you knew the chemical signals—referred to as growth factors or, more generally, inducing molecules—that your body uses to incite a hematopoietic stem cell to differentiate, Weiss says, you might use them to help your body replenish blood and bone marrow quickly after chemotherapy. You might create red blood cells for patients undergoing bone-marrow transplants, or white blood cells for cancer patients whose immune systems are impaired by chemotherapy. If we could find factors like that, says Weiss, we could help a lot of people.
The catch is that maybe 1 in every 10,000 cells in your bone marrow is a hematopoietic stem cell, which makes it difficult to find and study. Another huge challenge is finding the inducing molecules that turn it on and off. So Weiss is showing off his naive embryonic cells under a microscope, pointing out how they grow into tiny clumps of incipient tissue, composed of a whole gemish of fat cells, muscle cells, blood cells, and nerve cells.
By taking one such gemish, breaking it into single cells, and then seeding those cells onto a plastic laboratory dish with some nutrients and the right growth factors, he can generate pure colonies of blood cells in the laboratory. The first time I did this, says Weiss, I looked in the tissue-culture dish and I said, ‘Whoa, I’m making blood.’
The idea that you can learn how blood or other tissues are formed in the embryo and then apply that knowledge to re-creating them in an adult is one of the hottest new ideas in medicine. Developmental biologists have demonstrated that the same growth-inducing molecules that spur the differentiation of cells, tissues, and organs in the embryo serve critical roles throughout life. These molecules represent a gold mine of pharmaceutical promise: if used correctly, they could make not only blood but also nerve cells, for instance, to replace ones that are dead and dying in people with Parkinson’s disease. They could make the insulin-producing pancreatic cells that are missing in patients with juvenile diabetes. They could make bones grow faster after they’re broken and perhaps someday, although not particularly soon, regenerate entire organs. Liz Wang, who works with Weiss at Ontogeny, calls such plans the Star Trek fantasy: You’ll have this Star Trek library of tissues, so you can say, ‘Well, I need this kind of cell and this cocktail of growth factors, and this will give me this new kind of tissue.’
The Star Trek fantasy, along with the field’s tamer promises, has sparked the creation of a dozen new biotechnology companies, of which Ontogeny is among the first, and perhaps the most extraordinary. Ontogeny—the name means the development of an organism from egg to adult—was founded in August 1994, after the discovery of a group of genes, known as hedgehog genes, which play a crucial role in guiding the development and growth of an embryo. Leaders of five labs at the forefront of developmental biology joined Ontogeny’s scientific advisory board, under Doug Melton, a cellular and molecular biologist at Harvard.
Melton, a Midwesterner, came to Harvard in 1981 and since then has been studying frogs, chicks, and mice to learn how animals develop. He studied philosophy of science before going on to biology; he chose developmental biology because he considers it the most fascinating subject around. To him, even an investment banker or a kindergarten student can look at an egg and then look at an adult and understand the mystery. The basic problem of how an egg knows to make an embryo or an adult, he says, fascinates everyone.
For a human, what begins as a single fertilized egg will become in a few days a minuscule beach ball of several hundred undifferentiated cells. Each of these cells contains an identical genetic blueprint—your dna, stored on 23 pairs of chromosomes, encoding perhaps some 100,000 genes—which will eventually instruct your cells to produce the building blocks and signaling molecules necessary to form your body. Crucial to this story is that while the cells in your body all contain the same dna, they will use it differently, activating different sets of genes.
Development begins quickly. When the embryo is about two weeks old, it initiates the first step in the process of differentiation, of changing identical cells into the myriad types needed to compose a fully functional being. First the embryo decides which end will be up, and which down—that is, which will be mouth and which anus. Then the cells of the embryo develop into three distinct sets, like three mutually exclusive clubs: some become ectoderm, eventually developing into skin and nervous system; others become mesoderm, making bone, blood, and muscle; and the last become endoderm, making, among other things, your internal organs from your mouth on down. The endoderm curls up into a tube, known as the gut tube, one end at your mouth and the other at your anus. Your organs emerge from that gut tube. One region becomes mouth, below that becomes pharynx, then thyroid, then lungs. Your pancreas will grow out from this gut tube, below the bulge of your stomach and the incipient windings of your intestines. And so it goes, nine months from beginning to end.
Developmental biologists now understand in a general way how development happens. Encoded somewhere in the genetic blueprint is a formula that says specific genes must be turned on at specific times in specific cells. What turns them on, and thus leads to the development of every organ, cell, and tissue, is a process called induction. One cell tells another cell what to do by sending a signal, usually a protein—the inducing molecule.
Induction can sometimes involve cell to cell contact, says Melton, but in most cases where it’s been well understood, cell A secretes something to cell B. That then changes the fate of cell B, which might otherwise have been a skin cell and now will be a nerve cell, or something like that. First, however, the cell that receives the signal to become, say, a nerve cell has to be ready to make that change. Biologists call such readiness competence. A cell that’s competent to become bone will not become nerve, even if it gets the nerve signal. Competence is achieved by more protein signals. So, Melton explains, the early embryo is awash with inducing molecules at ever varying concentrations, some making cells competent, some setting them on the path of differentiation, others simply spurring more cells to emit yet other signals.
While this sounds complex—and the creation of a human body undoubtedly is—the number of inducing molecules that tell the body to develop is limited. There may be as few as 100, all of which fall into five or six related families. Melton uses an analogy to explain how it works: When I go to a Chinese restaurant, I’m often amazed that the menu can have 100 dishes, maybe 200. But if you go into the kitchen, you don’t find 100 pots, you find maybe 10 pots, and the way the chef mixes and matches the ingredients gives you a specific result. It’s like that in the embryo. For example, the same kinds of signals will be used to make bone as to make a motor neuron. But they appear at different times and in different combinations. So let’s suppose there are 100 inducing molecules. The body can make 10,000 cell types with these 100 inducing molecules by using numbers 1, 8, 12, 14, 16 in case A, and numbers 2, 4, 9 in case B, and so on.
The philosophy behind Ontogeny and its fellow developmental biology companies, according to Melton, is based on a simple, widely accepted assumption: the body reuses those inducing molecules throughout life. Evolution likes to stick with what works, and so the same molecules that spur the growth of cells and organs in the embryo will serve to maintain those cells and organs in the adult.
Using the embryo to look for those molecules is simply pragmatic. Traditionally, when confronted by an adult problem, the pharmaceutical industry looks in the adult organism for the solution. Say you wanted to find the molecules responsible for mending a broken leg so that you could then put them to work healing a fracture quickly and strongly. One way to do it is to study the leg that isn’t broken, says Ontogeny’s president and ceo, Doros Platika, a former neurologist and gene therapist from Harvard Medical School and the Albert Einstein School of Medicine. Then look at the bone that is broken and see what’s been turned on. What they have in common you eliminate, and you explore the differences, thinking that among the differences will be the molecules that you use to repair the bone. There’s nothing wrong with this, except the molecules that get turned on when you break your leg are many. They include those from inflammation and bleeding; they include pain molecules; and the molecule that triggers the repair is in there, too. In fact, the molecules you’re looking for, the ones that trigger the healing process, may be released at the moment of breaking, or an hour later, or three hours later; they may stick around for ten minutes and then vanish. So if you look a day later, or four hours and 17 minutes later, you may never find them. And even if you look at the right time, the molecules may be in concentrations too small to detect.
Looking for a signal in an adult, says Platika, is like fishing in a very deep ocean for one very small fish. So instead the Ontogeny researchers go to the embryo, where bones are growing furiously in a tiny space. That’s where the molecules that induce the growth of bone should be present in disproportionately large numbers. It’s still not a slam dunk, says Platika. It’s like fishing in a stocked pond, maybe, but it’s still fishing.
But the pond is one that developmental biologists have learned to manipulate easily. A mouse embryo is only about that big, says Melton, holding his thumb and forefinger about an inch apart. In a very quick and trivial way, you can find out where and when a gene is expressed and know immediately whether you should be looking at the adult kidney or the bone or the eye. Also, you can cut little bits and pieces out of embryos and grow them in tissue cultures and ask, ‘Will this molecule make muscle? Will this molecule make bone?’
The bottom line, he says, and the bet on which he founded Ontogeny, is that it is much, much easier to find out how the bone, the pancreas, or the liver is made in a little embryo than it is in an adult, and the molecules that signal at that stage, when they’re just making the organs, are used again and again in the adult.
Take the hedgehog genes, for example, of which three were discovered in vertebrates in 1993. (The name comes from the strange, bristly appearance fruit flies take on when they’re missing one of the genes.) When a cell turns on a hedgehog gene—that is, when the gene is expressed—the gene produces a hedgehog protein, which is an inducing molecule. These three proteins can account for a significant fraction of all the developmental interactions that are known to occur in the vertebrate embryo, says Tom Jessell, a developmental neurobiologist at Columbia whose work on hedgehog genes led him to Ontogeny’s scientific advisory board. You use these proteins again and again to control the development of many, many different tissues.
The most remarkable of the three genes, at least so far, is known as sonic hedgehog, which was named after the popular video game. The sonic hedgehog gene is expressed, at one time or another, in many parts of the developing embryo. According to Harvard biologist Andy McMahon, whose lab played a key role in the discoveries and who is involved with Ontogeny, it is expressed in the brain, the teeth, the tongue, the esophagus, the lungs, and throughout the entire alimentary track. It’s expressed in the male reproductive system, the kidneys, and even the hairs of the skin. In most of these places, we don’t know what it’s doing, he says. We have suspicions based upon the time it is expressed and what would normally be going on in those cell types, and what it might be doing in other places.
But when sonic hedgehog showed up in limbs and spinal cords, two areas that developmental biologists have been studying for decades, they had plenty of experience in figuring out its purpose. In the developing limb, sonic hedgehog is responsible for controlling the shape and placement of the digits. The sonic hedgehog protein is secreted in the embryo in an area of the developing limb that later in life, in a human, will become the fleshy part under your little finger. (It’s where you do a karate chop, Jessell says.) Once secreted, the concentration of sonic hedgehog diminishes as it drifts across the developing limb. This diminishing concentration of the protein, says Jessell, seems to be responsible for the growth of five fingers of similar, but slightly different, form.
In the spinal column, sonic hedgehog spurs the growth of the spinal cord itself, and then at varying concentrations spurs the differentiation of the various types of nerve cells that allow you to sense and move. Modify the concentration of sonic hedgehog protein in a tissue culture in which the embryonic cells have been primed to become nerve cells, and you can get the cells to differentiate into at least four kinds of neurons. It is a very economical way of generating cell diversity while providing one signaling molecule, Jessell says. In the adult, sonic hedgehog continues to be expressed, perhaps to keep the neurons alive and functioning.
Then there are indian and desert hedgehogs, the other two vertebrate hedgehog genes. Like sonic, they are both expressed in many areas of the developing embryo. Desert’s best-understood role is to regulate sperm production in males. Genetically engineer a mouse, for instance, without the desert hedgehog gene (biologists call this a knockout mouse) and you end up with healthy, fertile females and males that don’t produce sperm. Indian plays a key role in the development of cartilage in growing limbs. McMahon predicts that when they knock indian out of a mouse, the result will be an embryo with severely stunted limbs.
The three hedgehogs play crucial roles in the adult as well as the embryo, says Melton. Again, sonic is the most tantalizing. In the embryo, explains Ontogeny neurobiologist Nagesh Mahanthappa, sonic hedgehog is required to produce certain neurons in the developing brain, including those that generate dopamine, a chemical used to communicate signals between brain neurons. In fact, as Mahanthappa and Ontogeny’s Kevin Pang recently showed, sonic hedgehog is required for continued survival of those neurons, at least in embryonic cell cultures. In Parkinson’s disease, those dopamine-producing neurons die off and the victims lose control of their movements, becoming trapped inside their bodies. The currently accepted treatment is to give them a drug known as L-dopa, a chemical that is transformed into dopamine in the brain, but its usefulness decreases over time.
The Ontogeny researchers hope sonic hedgehog can help, either by generating the neurons needed to produce it or by improving the function of remaining neurons. The most straightforward approach, says Mahanthappa, is to inject the sonic hedgehog protein into the area of the brain affected in Parkinson’s and hope it stimulates the production of dopamine or the regrowth of dying neurons. Mahanthappa and his colleagues are trying this approach now, using animals.
If the straightforward method doesn’t work, however, the Ontogeny biologists are prepared to try more ambitious approaches: gene therapy, for example. They would stitch the hedgehog gene into the dna of an otherwise harmless virus and then inject that virus into the area of the brain where the dopamine-producing neurons are dying. The virus would infect local cells, which is what viruses do, and when it forced the cellular machinery to make copies of itself, it would have them pump out dopamine as well. Finally, there’s the most ambitious approach, which is to use sonic hedgehog to create neurons in a test tube, in much the same way that Weiss creates blood cells in a test tube. Then those neurons could be implanted in the brain, where, the researchers hope, they would start pumping out the necessary dopamine.
Even a modest success would help patients. Parkinson’s is a progressing disease, Mahanthappa says, and by the time the patients actually show up in the doctor’s office and indicate that something is wrong, they’ve already lost about 80 percent of their dopaminergic neurons in that population. So conservatively, sonic will just maintain the survival of the remaining 20 percent so it doesn’t dip below that. That would still be therapeutically useful for these patients.
The indian hedgehog gene shows therapeutic promise as well. Ontogeny is looking into using indian hedgehog protein to repair cartilage or strengthen bone. There’s a big market, says Liz Wang, among people who rip the cartilage in their knees during an athletic youth and 30 years later are looking to have their knee joints replaced. Because the torn cartilage has to be removed, the joints are without their full quota of cushioning and lubrication. It would be very good if you could actually get the cartilage to grow and resurface the inside of the knee joint, says Wang. But even if it was done in a temporary manner, it would still be useful. If it’s good for five years or ten years, you won’t have to have this titanium implant replacing your regular knee joint for another decade or so. Cliff Tabin of Harvard Medical School has shown that when an embryo is forced to produce extra indian hedgehog, it makes even more cartilage than usual. Whether indian hedgehog can be used to coax an adult bone to manufacture cartilage remains to be seen. Indian hedgehog protein could be applied directly, or a virus could be genetically engineered to carry the indian hedgehog gene into the bone.
Finally there’s desert hedgehog, whose most obvious developmental role is to generate sperm. In adult males the protein seems necessary to maintain sperm production. Platika says that if researchers can find a way to boost the production of desert hedgehog, they might be able to increase sperm production in men with a low sperm count and infertility. On the other hand, if they could turn desert hedgehog off, the result might be a male contraceptive in which everything of importance works but no sperm is produced.
The ambitions of Ontogeny, however, go far beyond hedgehogs. After all, there are at least 100 other inducing molecules that the body uses to spur embryonic development. The Ontogeny researchers are working to find what roles these play in creating cells or organs that might later be missing or dysfunctional in adult disease.
Diabetes is one of the best examples of how the approach works, and it is the disease that the Ontogeny biologists feel closest to tackling. In juvenile diabetes, the body lacks insulin because cells in the pancreas that normally make insulin, known as pancreatic beta cells, are being destroyed. Insulin injections, the standard treatment for diabetes, are a poor substitute for a functional pancreas that secretes the hormone in balance with the demands of the body.
The Ontogeny researchers are hoping to find the cells from which beta cells are descended and the factors that make them. We know that the pancreas begins to form in the first trimester, says Kevin Pang. But it takes almost all the rest of gestation to develop. So Pang and his colleagues are letting mouse embryos grow in the lab, pulling cells out every day from the area that is becoming the pancreas, and checking to see what genes are turned on and when insulin starts being produced. If they can untangle the signals necessary for insulin-producing beta cells to develop, they might be able to grow the cells in the lab, encapsulate them in a membrane, and inject them into the body.
While the diabetes project at Ontogeny is not as far along as some of the hedgehog work, it has one great advantage: it lacks the complications that abound with other diseases. For instance, even if the Ontogeny biologists make neurons to replace those missing in Parkinson’s patients, which they’ve already done for rodents, that doesn’t tell them how to get the neurons into the brain, hooked up to the proper synapses, and actually working. Even if you know how to make a given neuron, says Jessell, you cannot use that cell type to recover function in cases of neurodegenerative disease, because there are ten subsequent steps. The great attraction of the diabetes project is that none of those secondary constraints apply in quite as acute a way. You don’t need to secrete insulin in any given place. If you can make a beta cell and put it back in the body, it doesn’t matter where. The body and the circulation will take care of that. Those secondary problems are much more solvable.
If the Ontogeny researchers find ways to solve these secondary problems, the next question will be how far they can take the developmental approach. The obvious step, says Melton, is to go from creating new cells—such as nerve, bone, or pancreatic beta cells—to growing tissues and eventually entire organs. With the right progenitor cells and the right cocktail of growth factors, the Ontogeny researchers could theoretically grow any organ they want. Taking knee ligaments from cadavers or waiting for heart transplants would be things of the past. The researchers believe Wang’s Star Trek fantasy is possible, but they also know the obstacles all too well. Melton emphasizes that the idea of growing something like a new heart is decades down the road, not years.
I would start at trying to make heart muscle cells, he says, then see how you would fashion them into heart muscle tissue, and then see about making organs. There are people who like the idea of trying to make organs straightaway. I don’t want to say it’s crazy to think about organ development. I don’t think it is. I just think the challenge now is to start with cells.
