Playing the numbers

There are two ways to write this profile of Mario Geysen, the new Alfred Burger Professor of Biological and Medicinal Chemistry. The traditional way is to dive in, drafting and redrafting, coming up with new ideas about Geysen along the way, committing them to paper and testing them out.

But if you were a combinatorial chemist like Geysen himself, you might take a different approach. First, you would think carefully about the kind of impression you wanted to make, consider the optimum length of your article and line up the raw materials you would want to use. In this case, they would include impressions of this entrepreneur and Renaissance man who has worked on five continents, started seven companies and served as a chief researcher for one of the largest pharmaceutical companies in the world.

You would then write a computer program that would enable you to produce every possible variation of these elements — perhaps generating tens of thousands of stories. Finally, you would subject them to an automated test to determine which was most effective in creating the desired impression.

Not only would you produce the best possible story given your guidelines, but you would learn a great deal about such issues as grammar, syntax and organization from the vast majority of stories that failed. Next time around, you could write a much better and much more efficient story-writing program and complete the process in half the time.

In effect, instead of writing an article, you would be manufacturing it — which is the same approach that Geysen used 20 years ago to turn the world of medicinal chemistry on its end.

Building a better vaccine

In the early 1980s, Commonwealth Serum Laboratories sent Geysen, an Australian, to Surabaya, Indonesia, to assist in designing and building a plant to produce a vaccine for foot-and-mouth disease. Like some vaccines, foot-and-mouth vaccine uses inactivated particles of virus. The techniques used to neutralize the virus are not foolproof, and, as a result, the vaccine itself occasionally causes outbreaks of the disease. Geysen decided to devise a new vaccine based on a synthetic version of just those virus components — or antigens — that trigger the immune response.

These antigenic regions are called epitopes, and they are composed of peptides, which in turn are composed of amino acids. Geysen had two problems: he didn’t know the identity of the epitopes, and he lacked an efficient way to make the peptides that compose them. He solved both problems by inventing combinatorial chemistry.

Geysen knew he was looking for a combination of peptides that would mimic the epitope region on the foot-and-mouth virus, and he understood the chemical requirements for peptides that could serve this purpose. To automate the process of creating them, he created grids of plastic pins that aligned with the wells of a standard 12-by-8-well microtitre tray. He dispensed amino acids, the building blocks of peptides, into the wells and then dipped the grid of pins into them in sequence, building up unique peptides through a series of chemical reactions. Then the grid was placed in a 12-by-8 tray filled with a substance that would react to the presence of the foot-and-mouth epitope.

“In effect,” Geysen said, “I industrialized the process of creating and testing new peptides.” This process became the basis for Geysen’s first company, Mimotopes, a word coined to refer to synthetic epitopes.

The scale of his work, more than anything, was revolutionary. Before combinatorial chemistry, chemists used to make 50 to 70 compounds in a year; now, thanks to combinatorial chemistry, they can make thousands in a day. Geysen himself ultimately created two 1.28-billion-peptide libraries, one of which is in the Smithsonian. “It was an amazing experience,” Geysen recalled. “In one week, we made more compounds than had ever been made in the entire history of chemistry.”

Breaking through with a new paradigm

Geysen’s breakthrough is a classic case of the right person with the right ideas at just about the right time. In addition to degrees in chemistry, chemical engineering and microbiology, Geysen has a lifelong interest in computers and electronics, hands-on experience with manufacturing and a fascination with complex combinatorial problems embodied in games like chess. His timing was also impeccable. The combinatorial approach he devised would not have been possible without recent breakthroughs in solid phase chemistry and the introduction of the first personal computers, as well as innovations in robotics, in immunological assays and in the ability to modify plastic surfaces.

Gaining acceptance for this new technique with pharmaceutical companies was another matter. Geysen was stymied, at least initially, by the association of his combinatorial approach with peptides, which have a reputation as poor drug candidates. Although he created what was considered a vast chemical library, Geysen did not succeed in developing a replacement for the standard foot-and-mouth vaccine, and he had difficulty persuading organic chemists to adapt his approach for their projects. But by the late 1980s and early 1990s, as Geysen and other researchers developed new techniques, combinatorial chemistry came to the fore as a method of producing compounds for drug discovery. Working at the pharmaceutical giant GlaxoSmithKline, Geysen helped develop a machine capable of creating and testing 1,000 compounds at one time.

A combinatorial approach to creating and spreading knowledge

For a combinatorial thinker like Geysen, coming to the University is a natural extension of the work he has done his entire life. While at GlaxoSmithKline, he realized the power of coupling academic and industrial research. “They’re really quite complementary,” he pointed out. “The idea is to outsource the front end of drug discovery to universities and small start-ups, who are good at developing drug targets and creating potentially useful compounds. At the same time, you take advantage of the strengths of large drug companies, which have the resources and the facilities needed for large-scale testing and clinical trials.”

Geysen’s move to U.Va. is designed to build such a partnership. GlaxoSmithKline has agreed to move the bulk of his equipment and a library of compounds — together valued at $10 million — to U.Va. and to support it under a collaborative agreement that gives the company the right of first refusal to promising compounds developed in his laboratory.

Geysen’s decision to move to Virginia has also solidified the University’s already strong reputation for medicinal chemistry. “I am as confident as I can be,” he said, “that there are more robotics in the Department of Chemistry than at any other university in the world. There is certainly none with this kind of facility for high-throughput synthesis and screening.”

Another combinatorial aspect of Geysen’s decision to come to U.Va. is his determination to share the imperatives of combinatorial chemistry with graduate students. They will in turn share the techniques and mind-set of the discipline with other researchers and, ultimately, with their own students.

“I get a great deal of satisfaction from working with students and getting them to look at the world differently,” Geysen said. His classes are designed to do just that. His students will identify drug targets at laboratories around the University and create actual drug libraries for them. “My students are tackling difficult, real-world problems. It is really crucial for their development as scientists that they have the experience of thinking through the options, understanding the consequences of their choices and determining the desired outcome,” he explained. “Through their work, we hope to create a core facility that will enable us to collaborate with researchers throughout the University.”

Geysen is also interested in exposing students to the excitement — and travails — of entrepreneurship. “When I was a student, you could either pursue a career at a university or join a large company,” he pointed out. “Starting your own company is another, very satisfying alternative.” For one of his classes, Geysen has created a virtual drug discovery program like one that a start-up might pursue, which exposes students to the real-life financial as well as chemical decision-making that goes into creating a library of compounds.

From combinatorial chemistry to combinatorial sciences

For Geysen, the key concept he wishes to convey to his students is something that could be called the “combinatorial mindset.” He is careful to point out he teaches courses in combinatorial sciences, not combinatorial chemistry, and he believes that combinatorial approaches can be applied to a wide variety of difficult problems, from the creation of hard-wired neural networks that are capable of speech recognition (a current project) to investigations of room-temperature semi-conductors. For him, the ultimate goal of his research is not simply to generate millions of chemical compounds but to devise better ways to characterize the rules that govern relationships among great quantities of disparate objects, whether they are compounds, neurons or chess pieces. “Ultimately,” he said, “combinatorial approaches are just smart ways to think about problems involving very large numbers.”