Mention the words "combinatorial chemistry" to any pharmaceutical chemist and you're bound to elicit an energetic response. The term refers to a new drug discovery strategy pioneered by Berkeley chemists that many experts consider the most substantial advance in medicinal chemistry in decades.

Corporate and academic scientists alike are now speeding through once-tedious drug design efforts and finding promising medicinal compounds in a fraction of the time previously required. What makes it all possible is a mass production mentality that is the basis of the combinatorial approach.

Historically, scientists have built and tested drug candidates one at a time, carefully designing molecules from the ground up, piece by piece. With combinatorial chemistry, scientists start by systematically assembling molecular building blocks in all possible combinations, simultaneously generating thousands--even millions--of unique compounds. Researchers then sift through "libraries" of these compounds to find one that best suits their purpose.

Trailblazing scientists, among them Berkeley chemist Peter Schultz, originally used the body's immune system as a general blueprint for building vast libraries of biological molecules.

In early experiments that dramatically demonstrated the capabilities of combinatorial libraries, Schultz produced antibodies with the catalytic power of enzymes--a completely new hybrid molecule with the potential to deliver a powerful one-two punch to toxic molecules. The antibody component can home in on viruses or other harmful agents, and the catalytic component can effectively disable them.

Building on Schultz's success, Assistant Professor Jonathan Ellman published a paper in 1992 demonstrating that the tactic could be used with small organic molecules, the basic elements of drugs. Ellman's paper was a major factor in attracting the attention of the pharmaceutical industry.

"The key challenge was that we had to develop the chemistry to allow us to make many compounds simultaneously. It's different than making compounds one at a time," Ellman says. "You need very general chemistry that's compatible with a lot of different chemical functionality."

Such "generic" chemistry is the bread and butter of academic scientists, who prefer to hand off broad approaches to industry for implementation.

"The strategies and synthesis methods that we have developed are being used extensively in the pharmaceutical and biotechnology industries," says Ellman.

Drug chemists at Berkeley are focusing on generating libraries that are both specific enough to be potential drugs and general enough to be versatile.

"We devised some synthetic schemes that in three or four steps lead to molecules that look `drug-like' and have many possibilities for functionalization," says Professor Paul Bartlett, also an early pioneer of the combinatorial approach. "These are a new class of structures."

Bartlett is particularly interested in devising methods for the discovery of "peptidase inhibitors," a class of molecules that block enzymes responsible for a host of diseases, including AIDS and cancer.

"We haven't designed any compounds that can be used as medicinal agents themselves, but a lot of our strategies have been implemented," Bartlett says.

Just because they are regarded as strategists doesn't mean that College researchers don't put their own ideas to the test, though. Ellman has used combinatorial libraries to discover a potent inhibitor of a degradative enzyme, called Cathepsin D, that is implicated in a number of diseases, most notably breast cancer tumor metastasis. In collaboration with Gary Glick at the University of Michigan, Ellman has also identified molecules that block an antibody-DNA interaction involved in the autoimmune disease systemic lupus erythematosus, which afflicts more than 500,000 Americans.

"Currently, our collaborators are checking these molecules in animal model studies to see if we can block that effect," says Ellman of the latter compound. "Preliminary results look promising, but we need a lot more information."

Pharmaceutical companies are the chief drug discoverers, of course. They are now testing combinatorially-derived drug candidates that target a range of illnesses from cancer to obesity to central nervous system disorders.


With its supercharged take on drug discovery, combinatorial chemistry may have revolutionized medicinal chemistry for good; but it is a mistake to assume that the method has totally supplanted more traditional research.

Tried and true organic synthesis is alive and well in Professor Clayton Heathcock's laboratory, albeit with a modern flavor. Heathcock's goal: to explore interesting ways of synthesizing complex molecules. Though it may sound purely academic, Heathcock's research is providing his colleagues with critical synthetic tools needed to build intricate pharmaceutical molecules from scratch.

"Every time we try something and it works, it builds the platform higher," Heathcock says, elevating his hand. "And everyone is riding on the platform."

Industry drug chemists keep a steady eye on Heathcock's innovations, so that when it comes time to reproduce drug candidates efficiently in the lab, they'll have the means to do so. They also look to Heathcock's group when it comes time to recruit students into their companies.

"The most important product from my lab is really the training of these young folks," Heathcock says with a smile.

Industry is also keeping an eye on Bartlett, who is setting an example for integrating combinatorial libraries with conventional structure/function studies--a particularly powerful pairing, as it turns out. Bartlett uses structural information to start off the design of a molecule and polishes it off by screening combinatorially-generated variants.

"I think there's a much greater degree of overlap between the two approaches than a lot people have appreciated," says Bartlett. "Ultimately, both approaches are about information."

Exemplifying this notion is CAVEAT: a popular computer program Bartlett crafted to help scientists sift through enormous databases of three-dimensional molecular structures. Typically, scientists tap into a database to model an individual molecule. Bartlett is now adapting the program so that it can be used to model an entire database of structures--an innovation that reflects the influence of the combinatorial mode of thought.

"That's exemplary of how we're going to find algorithms and strategies out of the field of structure-based design which are going to be important in the field of combinatorial chemistry," Bartlett predicts.

Other College researchers are following Bartlett's lead, finding that they can blend the combinatorial approach with established projects to venture where they wouldn't previously have dared.

"I think with combinatorial chemistry I can say I might discover a drug...and I won't be committed," jokes Carolyn Bertozzi, one of the newest members of the Chemistry Department's faculty. "I don't think I would have ever said that before."

Bertozzi intends to use the method to target an elusive class of sugar-modifying enzymes, called "sulfotransferases," that are at the heart of chronic inflammation. But first, she will use her training in conventional biology to identify the enzymes.

"The project begins with target discovery and ends with drug discovery," says Bertozzi, adding that she would like to screen Schultz's and Ellman's combinatorial libraries for an inhibitor of the enzymes, once they are isolated.

Biochemical engineers, traditionally concerned with producing pharmaceuticals in large quantities, are now seeing ways to use combinatorial chemistry for new projects, as well. Berkeley Professor Doug Clark is using his expertise in the mechanics of cellular chemistry to put a new twist on the technique, which he calls "combinatorial biocatalysis."

Essentially, the overall concept of synthesizing diverse libraries of compounds remains the same. But Clark uses molecular machinery borrowed from the cell, such as enzymes, to assemble the raw materials of medicinal molecules rather than using the tools of organic synthesis.

"The chemistry that we use is exclusively biological," says Clark. "This enables us to work with much more complex lead compounds than one can typically use in conventional combinatorial chemistry approaches."


In many reaches of the College, researchers are finding that complex chemical problems often have biological
answers. Professor Bertozzi brings to the College particularly practical biological experience that is opening doors to a long-overlooked player in medicinal chemistry: sugar.

"It would be very hard for you to find a disease process that did not somehow involve glycosylation," Bertozzi says, referring to the addition of a sugar to cellular molecules. "Your oligosaccharides [sugar molecules] are intimately involved in almost every biological process."

In addition to being good targets for drugs, sugars are often essential in the production of pharmaceuticals. Many protein-based therapeutics, for instance, require the addition of sugar molecules for activation. Drug companies spend millions trying to find ways to make the modification more efficiently, but according to Bertozzi, the biosynthesis of such products is poorly understood and difficult to control.

"No one to this day has ever made a full length, functionally active glycosylated protein synthetically. The methodology just isn't there; we're going to change that," she says determinedly.

Chemical engineers are also looking for new ways to synthesize complex molecules, and like Bertozzi, they are turning to biology to spark original ideas.

Recently, scientists have found ways to grasp the levers of the cell's machinery, using nature's existing tools to more easily synthesize intricate compounds like proteins and steroids.

"Instead of having chemists come up with multi-step syntheses that are organic all the way, [drug companies] are trying to incorporate simple enzyme-driven steps into the process," says Chemical Engineering Professor Harvey Blanch.

To make enzymes useful in pharmaceutical chemistry, though, scientists must find ways to make them functional outside the watery environment of the cell, within an organic solution.

"Oftentimes, the drug you're interested in modifying isn't very water soluble," Blanch points out.

Blanch has spent the last 15 years devising ingenious methods for outfitting enzymes with the molecular equivalent of scuba gear--"reverse micelles" and microcapsules that keep enzymes safe while submersed in organic solutions.

"The enzyme is protected in sort of a water droplet inside an organic solvent where you can solubilize various substrates that aren't soluble in water," explains Blanch. "The enzyme can work on the small amount of organic material it sees and kick out the product back into the organic phase."

Toward the same end, Clark, who closely collaborates with Blanch, is working with enzymes that are extracted from "extremeophiles"--microorganisms that live in conditions of extreme temperature, pressure, salinity and pH.

"These enzymes tend to be stable in organic solvents," says Clark. "We can now look to these unusual organisms and find enzymes that can function over a much wider range of conditions."

Enzymes are now also helping chemical engineers to overcome a classic challenge in drug production: creating only the active form of a medicinal molecule. Ibuprofen, for instance, is a "chiral" compound--a molecule that is synthesized in two different orientations. Only one orientation, or "enantiomer," is an active drug.

"It's my understanding that the FDA is becoming more and more strict about producing only the active enantiomer; and most drugs are chiral," notes Clark. "Enzymes are very useful for chiral syntheses."


If controlling enzymes and other biological machinery represents a big step for chemists, controlling entire cells is a giant leap. Clark and Blanch are aiming to modify cells to convert them into microscopic factories that produce elaborate medicinal and therapeutic products.

"We want to be able to look inside the cell, study its metabolism, and redirect that metabolism toward the production of a desired product," says Clark.

Clark and Blanch are manipulating such factors as nutrition, temperature and pH to investigate the metabolism of hybridomas, the immune system cells that produce monoclonal antibodies. Such molecules, if brought under the control of scientists, are potentially useful for both diagnosing and treating disease.

Recently, the scientists have applied their knowledge of metabolism toward the study of "transformed" cells that cause breast cancer.

"One of the reasons they are considered `transformed' is that they have unusual metabolism; they keep growing," says Blanch. "What we've found is that pathways we didn't think were operative are in fact operative and vice versa."

Clark and Blanch are studying the response of the cancer cells to a regimen of a drug called Tamoxifen and the female hormone estrogen.

"Through understanding the mechanism of a drug, we're in a much better position to design a better drug or design a better therapy," says Clark.

While Clark and Blanch are looking inside the cell to take control of its behavior, Bertozzi is looking to the cell surface. Biological molecules--typically sugars--integrated into the cell membrane are responsible for cell to cell interactions, and ultimately the cell's role in an organism.

Bertozzi is "feeding" cells synthetic sugars, made with reactive groups that can serve as the foundation for further molecular construction on the cell surface.

"The cells will eat that sugar and convert it into an oligosaccharide that is presented on the surface of the cell," she says.

So far, Bertozzi has succeeded in presenting a ketone, a particularly reactive chemical group that is not normally found on cell membranes. A cell "primed" with a ketone group can be chemically modified to put new biological information on its surface, according to Bertozzi.

"Our technology will hopefully allow us to track cells by labeling them and to target drugs to particular cell types," she says. "Someday, we're hoping we'll be able to direct cells to home to certain tissues."

Eventually, scientists may even be able to selectively modify damaged cells, for instance cancer cells, to present "sticky" molecules on their surfaces. The deadly cells can then be plucked out of the bloodstream using a chemical filter--a simple concept that integrates the most complex aspects of both chemistry and biology.