| Today's Chemist at Work | E-Mail Us | Electronic Readers Service
Richard Klausner, director of the National Cancer Institute (NCI), notes the difficulty of making predictions. “It’s always amazing how, for some things, we dramatically underestimate how far they are in the future, and for other things, we dramatically overestimate them. We’re not that bad at predicting things that might be part of the future, but we’re really bad at predicting the timing and kinetics and path to them.”
to a good start
The excitement—and hype—surrounding human genome sequencing stems from hope that the genome can be used to help determine the underlying causes of diseases. Once the genome yields its secrets, the goal is to treat diseases and not merely symptoms, and to provide cures and not just palliative therapies.
The human genome sequence is an important step, but is it the answer to everything? Some people—William Haseltine, chairman and chief executive officer of Human Genome Sciences (Rockville, MD), for one—say that the elation over the raw genome sequence is misplaced. “I distinguish between knowledge of the genome and knowledge of genes,” he says. “We essentially had in our possession all the human genes by 1995. The genome project itself is a relatively meaningless footnote. If you’re to extract information about genes from the genome, you have to know the genes first. However, the fact that our technologies have allowed us to isolate and characterize a useful form of virtually all human genes is very significant.”
Douglas A. Livingston, the director of chemistry at the Genomics Institute of the Novartis Research Foundation (La Jolla, CA), expresses a similar view, although he doesn’t speak quite so harshly as Haseltine. He compares the situation to that in the science fiction cult classic The Hitchhiker’s Guide to the Galaxy, in which the hero hitchhikes through the galaxy trying to find “the answer to life, the universe, and everything.” When he finally gets the answer, it turns out to be, simply and literally, 42. “I think of the sequencing of the human genome in those terms. AGCTTTT is not the answer to life, the universe, and everything,” Livingston says. “In order to make some sense of that, you start to get into the issues of what it means. What are the products? How do they function? How do they interact?”
Kazumi Shiosaki, vice president for drug discovery at Millennium Pharmaceuticals (Cambridge, MA), agrees that sequencing the genome is just the beginning. “That’s just the nucleotide sequence. The next step of what the sequence means—does it represent a membrane protein? does it represent an enzyme? does it represent a structural protein?—is going to consume a lot of people’s time in terms of trying to understand the function of each of these. Layered on top of that, the challenge for the pharmaceutical industry is to say that we know which ones are misbehaving themselves to cause disease.”
One caveat, however, is that knowledge of the human genome cannot lead to a cure for all diseases. “We are in a war with bacteria and viruses, and that war is waged with medicines that are related to the biology of those infective agents,” says Ronald C. Breslow, professor of chemistry at Columbia University (New York). “It is not clear how much an understanding of the human genome can contribute to this part of pharmaceutical science.”
Genomic information has already found its way into the drug discovery process, primarily as a way to identify new targets. However, seeing genomics simply as a way to identify drug targets is a “very limited view of genomics,” says Wolfgang Sadee, professor of biopharmaceutical sciences and pharmaceutical chemistry at the University of California–San Francisco. Sadee, who is also editor-in-chief of PharmSci, the electronic journal of the American Association of Pharmaceutical Scientists, says, “We’re entering a new era in therapy, in medicine, understanding of health, understanding of disease. Just finding the target in fact is a difficult thing. Just because you have a new gene doesn’t mean that you have a new target. Target identification is a science in its infancy because we cannot model these complex systems very well.”
Breslow points out that “even when genes play a major role, we must not ignore the function of proteins in regulating the expression of genes.” For example, Breslow’s group is working on developing anticancer agents that arrest the development of tumors by causing the cells to differentiate. (A hallmark of cancer is cells that proliferate indefinitely without differentiating into a particular cell type.) “Our novel anticancer compounds are targeted to a protein that helps turn genes on by modulating the binding of DNA to histones,” Breslow says. “We understand how our compounds work, and even have an X-ray structure showing one of our compounds bound to the regulatory protein, which is called histone deacetylase. As the result of binding to this protein, our compounds turn on genes that cause cellular differentiation.”
Modern medicinal chemistry is more focused on the interactions of small molecules with proteins than with genes, which code for the synthesis of those proteins. “Many of the medically relevant proteins have already been identified and will continue to be important targets for modern therapies even after the human genome is fully sequenced,” Breslow says. The human genome sequence may help scientists finish the task of identifying these proteins.
Unraveling the human genome will allow “the early phase of drug discovery involving the identification of safe and effective agents that modulate the function of human proteins” to be completed within the next 50 years, predicts Stuart L. Schreiber, scientific director of the Center for Genomics Research at Harvard University. He doesn’t rule out the possibility of a much shorter timeframe. “This logically flows from the appreciation that the human genome is finite, with only 150,000 or so genes. This means we will want roughly 300,000 small molecules, an activator and inactivator of each gene product, with effectively complete specificity within the context of a human patient,” Schreiber says.
Some multifunction proteins will require effectors for each of the individual functions, but “housekeeping” proteins will not need any effectors. Schreiber estimates that about 5000 of these small-molecule effectors already exist and some are comparable to a gene knockout. However, he adds, “If even 3000 existing small molecules is an accurate estimate, we are already 1% toward that goal.”
“When I started in this business,” says Livingston, “you went and made a compound. You gave it to a biologist, and he fed it to an animal. The animal did one of three things: It got better, it stayed the same, or it died. You got the results and went back and made another compound. It was slow, but at the end of the day you had a drug.”
Shiosaki says that early-stage drug discovery has moved away from animal models to in vitro cell assays to identify candidates. “One can go very well using just those tools and come up with compounds that are very potent and very selective. You then have invested many chemists’ time—up to a year in some instances—and come up with one of these exquisitely potent compounds in a biochemical assay,” she says. “Then the first time you take it into an animal model, you find out that the compound has absolutely no oral bioavailability, for instance. Then you’re back to square one.”
However, Shiosaki says that scientists have become clever about engineering cell lines that can be used to assay various metabolic functions. For instance, cell lines can be used to mimic absorption by the gut or metabolism by liver enzymes such as cytochrome P-450. “A chemist might not start on a structure that was the most potent—although traditionally that was what was done, because that was the only data you had. Rather, a chemist would maybe work on a compound that had a fair-to-middling degree of potency but the right physical chemical properties, or at least the desirable ones, that would ensure a higher success in becoming a drug.”
Shiosaki says that finding a potent compound is only a tiny fraction of the making of a drug. “Finding a small-molecule drug that is potent and selective for a given target—say, an enzyme or a receptor—any good medicinal chemist worth their salt can pretty much do that nowadays. That’s only 5 or 10% of the game. The other 90 or 95% is, how do you make a drug out of that?” Shiosaki also says that companies are more concerned about the characteristics that make compounds good drugs rather than simply good pharmacological agents.
Compounds that are “hits” in a primary biochemical assay—say, 1000 compounds out of a 10,000-compound library that can inhibit a particular enzyme—are profiled with other assays. For example, a human colon cell line known as Caco-2 provides an indication of membrane permeability. In addition, the Caco-2 assay has some transporter and efflux mechanisms that mimic some of the pathways in the gut, Shiosaki says. In other tests, compounds are assayed against panels of metabolizing enzymes.
“It becomes a triangulation game,” Shiosaki says. “It’s like a Venn diagram, where you have a subgroup of compounds, the 1000 compounds, that hit with a certain potency. Within this is another Venn diagram of what compounds are permeable across a membrane. And another one in which there’s not significant metabolism by various liver enzymes. And another Venn diagram of compounds that cross the blood–brain barrier effectively. There should be an intersection, hopefully, of a reduced number of compounds that actually have the desired pharmacology, that is, they inhibit that enzyme but also have the desirable physical chemical properties that might make them more likely to succeed as a drug. It’s really helping to reduce the attrition rate in terms of the going-up-the-wrong-tree phenomenon.”
With all the in vitro assays that are now available, Shiosaki says that the “big trophy in the sky” is an assay to help determine toxicity. RNA expression profiling shows promise in this area, although it won’t be entirely predictive. In this technique, DNA arrays are used to detect the mRNA that is expressed in a given tissue before and after being subjected to a particular compound. Some genes are known to be upregulated in response to toxic compounds. “We’re asking for various techniques that would help us prioritize or give us an indication that there may or may not be issues of toxicity associated with a compound,” Shiosaki says, adding that much work is still needed to assess the validity of such technologies.
Haseltine predicts that a “flood” of protein, antibody, and peptide drugs will hit the market in 10 years. A decade after that, he says, such drugs will constitute half of all new drugs introduced to the market. Baker forecasts that more than 80% of those protein drugs will be antibodies. One use of antibodies will be as “molecular sponges” to prevent protein–protein interactions.
Compared with small-molecule drugs, antibodies are very specific and are less likely to cause toxicity based on factors other than the mechanism of action. “All sorts of orally available small molecules target all sorts of things, but they also toxify the liver and have drug–drug interaction problems. They interfere with cytochrome P-450, and that’s not mechanism-based toxicity. It’s like an innocent bystander getting blown away,” comments Baker. “From the point of view of a clean safety profile, antibodies are extremely attractive. You can design them to be incredibly specific with incredibly high affinity.”
molecules as tools
Livingston predicts that one of those powerful tools will be small-molecule chemistry, using vast libraries of compounds. “If you find something that interacts specifically with a target or even with a class of targets, you wind up with something that, based on its structure, not only defines the class of protein that’s critical in a given pathway but also is a tool for target validation. I think there’s huge power in small molecules to apply to this whole critical functional genomics problem,” Livingston says. In addition, a small-molecule approach to functional genomics brings scientists that much closer to identifying potential small-molecule therapies, he says.
Greater integration of chemistry into biological research will allow biology to be studied in a “less reductionist way,” Klausner says. “The next dramatic set of tools that will change biological research will be the development, annotation, and incorporation of small-molecule probes—both as universal perturbational agents to query biology and as imaging agents—so we can, in real time and with some ability to ask meaningful questions, look at biological processes in their natural settings.” Klausner says that it will require a “sociological change” to incorporate chemistry in biological research. It will also require the development of informatics tools.
Molecular analysis will foster better descriptions of disease states. “We still have, in many cases, relatively primitive descriptions of disease states as clinical entities and not molecular processes. Cancer is a great example. The ability to accurately and appropriately molecularly classify and describe diseases has been going on for quite a while, but it’s going to ramp up quite dramatically,” Klausner says.
C. Russell Middaugh, a professor at the University of Kansas (Lawrence, KS), spent 10 years at Merck Pharmaceuticals (three as a member of the company’s gene therapy group). Middaugh expects that the first major success will come in 5 to 10 years. Before that success will be realized, the vectors used to deliver genes must be improved. “Everyone feels that the ideal vector to deliver genes to cells has yet to be identified. There are sort of two views. One is that different targets will require different vectors. Therefore, work on many different systems is appropriate,” Middaugh says. “Probably a less frequently espoused view, but one that I think people secretly hold, is that there will probably be one or two efficacious vectors that will become widely used.”
The vector currently considered the “leading candidate” is the adeno-associated virus, Middaugh says. “There have been some spectacular results in animal models that feature long-term expression of gene products at therapeutic levels. One of the big disappointments has been that the very spectacular success with DNA-based vaccines, in which a wide variety of animals were shown to generate significant immune responses using this technology, has not so far propagated into humans.”
Gene therapy research is being conducted in academic research laboratories, small- to medium-sized biotechnology companies, and large pharmaceutical companies. However, he says, “most of big pharma is taking a wait-and-see attitude toward this. They don’t really want to assume the major risks. They’re willing to fund some of these smaller companies and help them, but they don’t want to make any major commitments.”
What sort of impact will gene therapy have on pharmaceuticals? “In terms of longer time periods in human health, major impacts come from things like vaccines and new preventive strategies, and perhaps from more novel therapies such as gene therapy,” Middaugh says. “We know that you can take small molecules and inhibit enzymes in certain metabolic pathways and receptors on cells, and impact the therapeutic courses of disease. But I don’t think that small molecules have the potential to impact disease the way that genetic therapy does. Genetic therapy is going right inside the cell and producing the natural substances of the cell—the proteins—and intervening in a much more specific and much more potent fashion.”
Middaugh believes that before we will see success with gene therapy, there must be a “mental transition in terms of the treatment of the current vectors from ill-defined biological agents to well-defined pharmaceutical agents—things that we understand at the very level of the molecules that make them up. When that transition occurs, it’s going to dramatically enhance gene therapy.”
“We’re getting there,” he says, “and with new developments in rapid DNA characterization, new developments in mass spectrometry to characterize the vehicles, and the use of combinations of spectroscopic techniques to characterize the size and the shape of the vectors we make, we’re ultimately going to be able to define these vectors in molecular detail and treat them just as we would small molecules. To get the kind of therapeutic effect that you’d need for a commercial drug, [we must] transition from treating these things as viruses or heterogeneous globs to treating them as well-defined biological, biochemical, or biophysical entities.”
Harvard’s Schreiber predicts that gene therapy will be combined with molecule therapy in which the small molecule is used to regulate the therapeutic gene. “Somatic gene therapy will become common and effective in the coming years, but germ-line therapy is an eventual reality as well,” he says. Schreiber expects that germ cells will be equipped with extra copies of cancer-fighting and anti-aging genes that can be turned on with a small-molecule drug when needed.
Shiosaki believes that all patients should have their own “gene on a chip.” Profiling a drug against a patient’s gene on a chip would indicate whether the medication would have an adverse or positive effect.
Although pharmacogenomics will have its biggest effect in the decision to prescribe certain medications, it may also help determine which pharmaceuticals are developed. “If the target you’re looking for has a significant variability in the general population, that may not be a great target to go for,” Shiosaki says. “Already, that information is going to help you preselect what targets to work on.”
the fountain of youth
However, he thinks the second half of the century will be even better. “The most important advance in the 21st century will be the introduction of atomic-scale prostheses to repair and restore human body function,” Haseltine says. “That will come in the second half of the century, and it will be tied to the most important revolution of the next century, which will be atomic-scale engineering.”
Haseltine differentiates between regenerative and rejuvenative therapy in terms of the materials that are used. Regenerative therapy will use naturally occurring biologic substances, whereas rejuvenative therapy will use human stem cells and synthetic engineered substances. “We will replace our body parts with a more durable substance and extend human performance in almost any area,” he says. “The fusion of atomic-scale engineering technology with our bodies will enormously enhance human performance.
“We have already made some of the fundamental advances that are needed, which are the isolation and characterization of a complete set of human genes and discovery of materials that are compatible with our bodies for regrowing organs. It’s more a question of engineering and execution at this point,” Haseltine says.
Tissue engineering—taking cells to restructure or rebuild damaged or congenitally defective tissues—is an important part of Haseltine’s vision of regenerative therapy. “The first tissues are just now being reimplanted,” he notes. “That will grow over the next 5 to 15 years to be a major business. We will begin with blood vessels, cartilage, bone, bladder, trachea, and skin, moving on to more complicated organs like the liver and kidney. Twenty years from now, a number of major organs should be reimplanted, and I think by 30 years from now more complicated organs including the heart and lung can be transplanted.”
From transplants, Haseltine sees medicine moving toward the use of stem cells as medicine to rejuvenate the body. “That requires very fundamental breakthroughs in our understanding of how stem cells arise during embryogenesis, what controls them, and how they can be used as medicine. I don’t see stem cells being used as a major form of medicine for at least 10 years, and I don’t see the main implication for another 30 or 40 years because of a wide variety of difficulties.” However, he admits that stem cell research may progress more rapidly than he suspects.
A particular stumbling block is the ethics of embryonic stem cell research. Human embryos must be destroyed to harvest the cells. There are concerns in some circles that such research could encourage abortions. However, stem cells can be obtained from leftover embryos obtained during fertility treatments.
Other people also see regenerative medicine playing a role in the future. “If you consider heart attacks, there is dying muscle tissue. The extraordinary excitement was about growing new blood vessels that would supply the remaining muscle. What was overlooked was that the muscle has degenerated and these new blood vessels don’t regenerate the muscle tissue,” Sadee says. “In regenerative medicine, we would be able to take stem cells and train them to be cardiac muscle cells and replace the muscle tissue. I think stem cells and anything related to them are going to be of extraordinary importance in the future.”
Schreiber sees stem cells being used in combination with small molecules. “Advances in stem cell research, in combination with advances in our ability to discover small molecules that activate specific differentiation pathways, will allow replacement organs to be grown in culture, beginning with the patient’s own stem cells or genetic material.”
Like Haseltine, Klausner also sees nanotechnology playing a role in medicine. “Ultimately, what I think is a fantastic challenge is to link molecular sensing technologies with nanotechnology for the idea of using such molecular machines to remotely sense molecular changes, know what they are, and know where they are. One can ultimately imagine the incorporation in a single molecular platform of sensing, signal generation, external decision-making, and local therapy. I don’t think what we’re talking about violates laws of physics as we understand them, but it is very far off in terms of what we’re capable of doing.”
to the future
Beyond that, Klausner asserts that palliation and cure lie along a continuum. “Our goal should be to have the most effective approaches to preventing or curing disease that we can. Whether that comes through what we might call palliation or fundamental cure, my feeling is, ‘Whatever works.’ I think you should move toward what’s most effective and affordable and accessible to everyone.”
© 2000 American Chemical Society