Profile of José N. Onuchic

Modern biology laboratories are filled with devices that have roots in experimental physics, but only in the past few decades have physical theories been applied directly to biological problems at the cellular and subcellular levels. One of the leaders in theoretical biological physics, José Onuchic, was elected to the National Academy of Sciences in 2006 for his contributions to our understanding of protein folding and electron tunneling inside proteins. His Inaugural Article on merging stochastic and deterministic gene networks to describe the decision process that Bacillus subtilis uses to choose between sporulation and competence under stress conditions was published in volume 106, issue 50 of PNAS (1).

Onuchic, 52, is a professor of physics and codirector of the Center for Theoretical Biological Physics at the University of California, San Diego (UCSD). He was born in São Paulo, Brazil, and grew up in São Carlos, where his parents were both professors of mathematics at the local campus of the Universidade de São Paulo. There, he obtained bachelor's degrees in electrical engineering (1980) and physics (1981), as well as his master's degree in applied physics (1982). He is not the only one in his family with a flair for science: his sister and a brother are physicians, and another brother is an engineer. His wife is an elementary school teacher, and they have three college-aged sons.

Even as an undergraduate, Onuchic worked to apply physics solutions to biology (2). He studied for his Ph.D. at the California Institute of Technology (Caltech) with John Hopfield as his supervisor. His thesis topic, electron tunneling in proteins (3), “was a new field and very interesting,” Onuchic recalls. “I had a sense the physics was going to take me toward the physics of complex systems, and biological physics was going to be one of the big items inside the physics of complex systems, emergent phenomena, self-assembly, and physical network problems.”

José N. Onuchic.

After graduating from Caltech in 1987, he returned to Brazil as an assistant professor at the Universidade de São Paulo. However, George Feher was building a biophysics department at UCSD and recruited Onuchic as a theoretician in biological physics. He moved to San Diego in 1990 and has remained there since.

A Tangle with Proteins

Onuchic believed that protein folding and self-assembly was one of the simpler problems to tackle in complex systems. He and his collaborators approached protein folding from a different angle than the typical biochemist: instead of trying to predict a protein's final structure on the basis of its amino acid sequence, they looked at which sequences are able to form a unique structure. A random sequence of amino acids can fold in many different ways, most of which are a tangled mess, but a protein that has survived natural selection can fold only one way.

That insight led to considering protein folding in terms of an energy landscape and led to the concept of the folding funnel. “The idea that we introduced was which kind of sequences are able to fold into a unique structure. By doing that, we changed the entire paradigm” (4).

As a long chain of amino acids folds, it minimizes what physicists call frustration as nonpolar amino acids try to contact each other, while polar ones do the same.

“You try to make as many interactions as possible happy, but you cannot make all of them happy because the chain doesn't allow you to do that,” Onuchic says. “Sequences are able to fold and have unique structures because they minimize the number of frustrated interactions. All the other competing structures will have many unhappy interactions. A sequence that does fold properly will always have the same structure. With a random heteropolymer that is not true.”

Solving the Folding Puzzle

When the energy landscape of a folding protein is represented graphically, it has only one steep depression, called a funnel. “We have an idea that the landscape tells you what the bias toward the native state is. You don't have a single path to fold, you have multiple paths to folding,” Onuchic says. “As the protein folds toward the native state it doesn't have to always get organized the same way; you can have different parts organizing and coming together, but every pathway does not have to be in the same order. You can start at one end, the other end, or in the middle. That's what a lot of people call the new view of protein folding” (5).

Experimentalists agreed this description was much better than the simple state description they had before. The previous view was based on a description of “ensemble” intermediates between the unfolded to native state, but that was a descriptive picture without real theory behind it.

Although the basic “energy landscape” theory is fully developed, the numbers produced by the models still need improvement to allow researchers to understand what differentiates a sequence that is able to fold. “Basically, the models we have for protein folding are not as good as the experimental numbers we observe. That doesn't mean it's the protein's problem—it's our problem,” Onuchic notes (6).

Funnels to Tunnels

During the 1990s, Onuchic was also looking deeper inside proteins, studying how electron transfer and tunneling occurs.

Electron flow inside proteins is essential to understanding bioenergetics, especially for protein complexes involved in photosynthesis and respiration. Electrons often travel long distances inside a protein or between a donor and acceptor, and the best way to explain these large leaps, which can reach up to 20 Å, is through tunneling.

“Going through empty space is possible, but we could understand it as tunneling too,” Onuchic says. He studied how electrons flow via chemical bonds and tunnel through helices and sheets (7).

“We showed that tunneling was not controlled by a typical barrier, but actually by sort of mixing with the bonding orbitals. It's almost like a hole transfer of the electrons, mixing the bond with the orbitals. This idea became a way of understanding how tunneling took place in proteins, and why sometimes you have an electron that took one route and why it sometimes took another” (8).

Onuchic and his collaborators also wrote a widely used computer program that calculated electron tunneling paths in proteins (9).

“The method was powerful because people could not only understand their experiments, but actually make predictions and design proteins and chemical compounds,” he says.

Strategies of Gene Networks

In his Inaugural Article, Onuchic and his colleagues developed a model that combines stochastic and deterministic gene networks that the Bacillus subtilis bacterium uses to sense its environment and decide whether to commit to sporulation, depending on local conditions and what its neighbors are doing.

As stress on bacteria increases, most decide to enter sporulation, an energy-intensive and time-consuming operation, whereas a minority chooses competence, which allows them to feed on DNA released by neighboring bacteria choosing sporulation.

The system will not work if all of the bacteria choose competence, “so the question is why only 10% try that when they are all in the same environment,” Onuchic says.

“It pays for the individual cell to take the risk and escape into competence only if it notices that the majority of the cells decide to sporulate. But if this is the case, it should not take this chance because most of the other cells might reach the same conclusion and escape from sporulation.” He compares it to game theory, in which “defecting” or “cheating” can be the best strategy.

The group found that the decisions bacteria make about competence is a stochastic process. So some of them try, and some of them do not. “The competence process is fully stochastic while the sporulation process is not,” Onuchic says.

“The link between molecular and cellular networks will become a big thing.”

“We can write a full set of equations about the control of these things and how they get information, then go through these stages, called assessment, decision, and commitment.”

Integrating Network Systems

For Onuchic, the next step of his research efforts is developing models that integrate gene networks and cellular networks to understand protein evolution, protein interactions, more complicated cellular systems, and organisms.

“People in systems biology know a lot about what the parts are, but not how a decision is made,” Onuchic says, referencing his work with experimentalists building models with real data (10). “We're very interested in building model circuits that can make a full computation and full decision and come to the answer.”

Building models of networks will be an essential part of future biology, Onuchic thinks. He suggests that it is becoming hard to do biology experiments without a theoretical background. He believes that physicists and those with a mathematical bent will become increasingly important collaborators for biologists.

“Answering a problem is hard to do only one way or another. For our small network, now at least we have a paradigm where our system works. We can start to design things and ask ‘What happens if I touch this part of the network, what's going to happen to the entire decision-making process?’” Onuchic says.

“It's very tough for experimentalists to do that in reality, because you have no idea of the effect. In such complex processes, where there are many competing effects, there is the need of a set of equations,” Onuchic says.

Onuchic sees his future research involving protein structures, including folding, molecular motors, and biomachines. Gene networks, “the new love of his life,” are his other prime focus. “The link between molecular and cellular networks will become a big thing, not just for me but for my center and the scientific community at large.”