|
Almost all the major children’s museums of America have in their respective science sections a year-round exhibit named something to the extent of, “What’s In the Mysterious Black Box?” Curious youngsters, eager to get away from boring exhibits enclosed within detached glass cases, are encouraged to place their hands inside a black box containing an unknown object. Using nothing but their sense of touch, the children feel the hidden exhibit until they believe they have identified it based solely on its shape, size, and texture. After being told that they have guessed the object correctly, most of the young participants (hopefully) come away with a newfound appreciation for their abilities to solve problems without necessarily relying on what they can directly see.
As simple as the basic premise of this “children’s” exhibit may be, the dilemma of the "black box" is analogous to many scenarios of “grown-up” science in important aspects. Coming to a definitive solution on a question one cannot even directly visualize in the first place grows seemingly more and more hopeless as the object in question becomes increasingly complex. Despite what sometimes seem to be unlikely odds for success, scientists are willing to work with what information is available and what data is attainable to come to the best conclusions possible on their field of study. As one fitting example, Dr. James P. Snyder’s laboratory atop the sixth floor of the Emory University Chemistry Building is currently using modern computer technology as its “hands” in feeling around the expansive "black box" containing the receptor site for a couple of promising anti-cancer agents, namely taxol and epothilone.
The drug taxol has been effective in treating human ovarian cancer, as well as certain other cancers originating from the epithelium (e.g., small cell lung cancer, breast disease, etc.). Although its clinical results have been promising, taxol has downsides: (1) a number of undesirable side-effects (e.g., neurotoxicity, etc.), (2) low solubility in water, (3) rising resistance, thus reducing the drug’s effectiveness, and (4) difficulty in mass production. Taxol stops tumor growth by freezing cell division: the drug, upon entering the body, binds to a receptor site known to be within the microtubules of the diseased cells. Microtubules, made up of building blocks known as tubulin, are what constitute a cell’s skeletal structure and thus must be able to dissemble into separate tubulin sections if a cell is to divide in half. Upon binding, taxol causes the microtubules to polymerize (i.e., “stick together”) and therefore prevents microtubule disassembly and rapid cell proliferation.
In an effort to find taxol substitutes, German chemists under the leadership of Gerhard Hšfle have recently discovered promising candidates in class of compounds known as the epothilones. Although the epothilones have yet to be tested clinically, in vitro tests show that they inhibit cell division by the same mechanism as taxol, through tubulin polymerization. The results are exciting, as Hšfle explains: “. . .epothilones are about 2000-5000 times more active than taxol in these experiments. . . .” Furthermore, evidence exists that taxol derivatives and the epothilones dock at the same receptor site within a cell’s tubulin: tests have shown that the two drugs compete with each other for binding when administered in vitro to a single set of test cells.
All of this information has spawned new interest in the structure of the taxol/epothilone receptor site within the microtubules. Knowing the three-dimensional shape and chemical nature of this binding pocket could potentially open the door for the discovery or even design of new compounds that induce tubulin polymerization and thus stop tumor growth. Unfortunately, technology has not found a method yet for taking a detailed picture of this receptor site directly. The usual method for small-scale photography in chemistry, known as X-ray crystallography, is unable to document the makeup of the taxol/epothilone binding site with a resolution suitable for further deductions. Furthermore, both taxol and epothilone are flexible macromolecules (completely different in chemical composition), capable of taking on thousands of different three-dimensional conformations, or “shapes,” through the rotations of their various bonds. Without knowing the exact conformations of the molecules when they bind to the receptor site, deducing the structure of the binding pocket makes for quite an imposing (if not hopeless) problem.
Knowing that the contents of this biochemical “black box” are extremely complex and detailed, Dr. Snyder has decided to use the powerful computing resources available to him presently at Emory to feel around the unknown interior and make some deductions about what is inside. Armed with the evidence that taxol and epothilones both bind to the tubulin receptor site in question, Snyder and his lab mates have tackled the problem of coming up with a receptor model in which new candidates for anti-tumor drugs can be tested for level of polymerization activity.
The simple but crucial idea behind the plan of attack for Snyder’s research has been the reasoning that if taxol and epothilone, two very different compounds structurally, bind to the same microtubule receptor site, then their binding conformations must be similar in some regards. That is, some similarities should exist between the shapes in which epothilones and taxol attach to their shared tubulin pocket. Because the large chemical structures of taxol and epothilones are so different from one another, comparing the various conformational possibilities of the two molecules and looking for similarities would have been impossible without computers. However, Snyder’s lab, with access to state-of-the-art Silicon Graphics computing technology, was able to run huge “conformational searches” on both the taxol and epothilone three-dimensional structures and, furthermore, used a program called APOLLO (Automated PharmacOphore Location through Ligand Overlap) to look for similarities behind between the two sets of molecule shapes. With the reasoning that the conformations from the different sets most similar to one another are the best candidates for binding to the same receptor site, Snyder’s lab deduced strong candidates for the active binding shapes for taxol and epothilone.
This information led directly to the construction of a pharmacophore, formally defined as “the key chemical groups on antagonists in their 3-D forms presented to the target receptor.” In simpler terms, overlapping the active candidates of taxol and epothilone gave Snyder’s research team a sense of what the essential characteristics of the receptor model needed to be. With a sufficient level of confidence in knowing what taxol and epothilone look like three-dimensionally when they bind to a common receptor, Snyder’s lab used this pharmacophore as the basic foundation around which a binding model—the so-called “minireceptor”—could be built. A software package called PrGen (Protein Generator), which Snyder himself helped pioneer, was used to construct amino acids around the pharmacophore and to track the correlations between the new model’s behavior and real-life drug activities.
Of course, in science, checking one’s work is of utmost importance, especially in fields of drug development. Researcher K.C. Nicolaou has laboriously cataloged the in vitro activity of dozens of modified epothilone structures, with several analogs being extremely active while others not binding to the receptor at all. A varied selection of these epothilone derivatives were constructed three-dimensionally by Snyder’s lab and then used to see whether the new minireceptor, minus the pharmacophore backbone, could predict which analogs should be active and which should be inactive. The verdict: although not perfect, the minireceptor was able to distinguish between the active and inactive structures with a reasonable degree of discretion. Given the massive complexity of the problem, Dr. Snyder has taken these positive results to heart and has submitted the current receptor model for scientific peer review.
Critics of Dr. Snyder’s unique research approach see his work as a complete shot in the dark. To be sure, it is most unlikely that his research team has deduced the exact contents of the mysterious "black box." However, since the purpose of creating this taxol/epothilone minireceptor is to speed up the discovery of new compounds which might bind to the same site, it is not so much whether the model is an exact replica of the binding site so much as whether it acts like the real receptor. Or, as Snyder himself states more elegantly, “A minireceptor constructed around a set of superimposed ligands [e.g., taxol, epothilone] need not resemble its natural counterpart but should accommodate a series of ligands in a similar binding sense.” A model that wields this predictive power is exactly what medicinal chemists need to start looking for new drug leads.
A few recent developments have increased the excitement over the minireceptor modeling concept. First, a new version of the PrGen software has recently been released. The new package includes more accurate calculation algorithms, allowing for further refinement of pharmacophore placement and amino acid construction. Even more exciting have been the recent “peeks” inside the "black box" by Kenneth H. Downing at U.C. Berkeley. Using a cutting-edge technology known as electron crystallography, Dr. Downing has been able to take rough pictures of tubulin, the building block of the microtubules containing the taxol/epothilone receptor site. Although these pictures are not of high enough resolution to pinpoint the exact structure of the binding pocket, Dr. Downing’s and Dr. Snyder’s labs are currently collaborating on the question of whether the new rough visualizations of the site and the minireceptor model are agreeable with each other. Already, the joint talks have suggested that a few of Snyder’s candidates for the active conformations of taxol look like promising fits for the rough pocket pictures.
Indeed, two sets of hands feeling around the inside of the "black box" should be more productive than just one. In an area of study as urgent as cancer research—the disease currently claims the lives of 1,500 victims per day—science needs all the help from innovative techniques it can muster. However, unlike the children’s museum exhibits, checking guesses at the contents of the "black box" is not an easy task—most drugs take about 15 years from theoretical conception to a spot on the pharmacy shelf. Hopefully, use of the taxol/epothilone minireceptor will nevertheless speed the process of discovering more effective, safer drugs for cancer patients in desperate need of treatment.
|
