Evolution is a powerful creative force—just look at the life all around us for evidence. And humanity has harnessed that creative process to produce tremendous variety in our domesticated animals and crops. But doing so has been a long-term project, involving many generations and the many years those occupy.
This year’s Nobel Prize in Chemistry goes to three researchers who figured out how to get evolutionary processes to work for us on the level of individual molecules and accelerate it to the point where the results were available in weeks or months rather than years.
Half of the award goes to Frances Arnold of Caltech for the development of directed evolution of enzymes. The goal of directed evolution is to create an enzyme, or catalyst, that performs a chemical reaction of our choosing, even if that reaction is completely useless for the organism the enzyme evolves in. Arnold put together a process that in retrospect seems obvious but hadn’t been done systematically prior to her work.
It starts with choosing the appropriate enzyme to use as evolutionary raw material. Arnold recognized that even for cases where the reaction isn’t done by any known living things, there are often related reactions that are catalyzed by known enzymes. In many cases, these enzymes may have a tiny bit of the catalytic activity we want, providing the starting point for evolution.
From there, you have to make random changes to the DNA that encodes the enzyme to provide opportunities for improved function. Arnold’s lab developed two methods for this. One was to make them truly random, which was accomplished by copying the DNA using an enzyme that was prone to mistakes (its ability to recognize when it used the wrong base was deleted). Another was to focus random mutations where they were most likely to have an effect—in the active site that binds to chemicals and catalyzes reactions. Another research group developed a third: mix and match parts from several related enzymes from different species.
These approaches provide the variations that evolution can work on. But the key to evolution is selection—finding those few instances within a sea of variation that do what you want. To demonstrate her approach, Arnold chose a simple goal: digest a protein that is colored. As a result, any bacteria that carried a variant form of the enzyme that digested the protein would end up surrounded by a clear, colorless halo. These bacteria can then be selected as the starting material for another round of evolution, with additional random changes introduced, followed by another round of selection.
Arnold showed this could accomplish a variety of changes. Over repeated rounds of selection, enzymes were evolved that could work at higher or lower temperatures—or in one case, both high low temperatures. By gradually raising the concentrations of a powerful industrial solvent over multiple generations, another enzyme was evolved that could still work in a 60-percent solvent/40-percent water mix. Another enzyme was generated that could form carbon-silicon bonds, something that hasn’t been seen in any living organism.
Thanks to Arnold’s pioneering work, others have figured out how to perform the process without the need for bacteria, accelerating it even further. And while the process still requires that people come up with a clever scheme to select for the catalytic activity they want, this hasn’t proven to be much of a barrier. Artificially involved enzymes are now used to produce biofuels, detergents, chemicals, and pharmaceutical products. In some cases, entire pathways of evolved enzymes have been assembled to go from a simple starting material to a useful end product over several chemical steps.
Selection and display
George Smith of the University of Missouri, Columbia shares the other half of the award for figuring out how to perform a related type of evolutionary selection. His technique is called “phage display” and relies on the viruses (called phages) that infect bacteria. It gets over one of the biggest challenges of the approach pioneered by Arnold: many of the proteins we’d like to work with are by default kept inside the bacteria in which we make them. The chemicals we’d like to interact with, by contrast, are kept outside the bacteria.
Phage display gets over this problem by ensuring both the protein and the genetic material that encodes this get shipped outside the cell and kept together in the same virus. The key breakthrough here was identifying a protein that’s normally incorporated into the coat that covers the virus and finding a spot on it that can tolerate large-scale changes. The spot is essentially a loop on the protein’s surface, and you can greatly expand the loop without interfering with the virus. Smith realized that this allowed researchers to insert nearly any protein they wanted into that spot in the gene, and the protein would be displayed on the surface of the virus that encoded it.
Smith went on to generate a variant of phage display called “biopanning,” named after the pans that prospectors would use to separate the rare bits of gold in a vast collection of pebbles. In biopanning, large pools of phage are made with different proteins inserted and are then passed over a material that has a molecule you’re interested in stuck to it. Any phage displaying a protein that sticks to your molecule will be retained and can be used to infect new cells. After several rounds of this selection (with or without mutation), the resulting phages will all encode proteins with a high affinity for the molecule of your choice.
That can be useful for a variety of things, from research to generating useful proteins. But medicine has found a great deal of utility in using human antibodies to bind to specific proteins, since the immune system views these antibodies as normal and not a foreign protein to attack. Gregory Winter shares this half of the award for figuring out how to modify phage display to work with antibodies.
Antibodies are actually large complexes of four molecules, two light and two heavy proteins. The part that binds to foreign material resides at the interface between one light and one heavy protein, making it difficult to work with, since the key feature of most antibodies—their specific targeting—is typically split across two proteins. Smith figured out how to take the key portions of the genes for these regions, combine them into a single gene, and then insert that gene into a phage protein. The resulting hybrid gene would ensure that the key parts of the antibody were exposed on the outside of the phage, available for the sorts of selection that Winter developed.
Isolating useful antibodies had typically involved working with mice and multiple rounds of immunization over the course of months, followed by lots of culturing of mouse immune cells. Now, an equivalent process could be done in a week using phage display. Once isolated, the key portions of the engineered gene could be cut out and spliced back into a normal human antibody gene, making the entire complex that the immune system worked with.
Naturally, the pharmaceutical quickly recognized the potential of making antibodies that could bind to any molecule that it wanted to target. (Monday’s Medicine Nobel provided some great examples of these sorts of useful targets.) The first antibody-based drug developed using this technique reached the market in 2002 and has been joined by others since.
Reading all of the above about evolution, bacteria, and viruses might make you wonder how this work could possibly have been awarded for chemistry. So it’s worth stepping back and thinking about what’s actually being accomplished through the techniques the winners developed. If you weren’t using evolution to do it, designing a chemical that could bind to an arbitrary molecule of your choosing would typically end up being called “materials science.” And if you offered a chemist the opportunity to develop a catalyst for any of a massive range of chemical reactions—one that would work in the temperature and solvent of their choice—you’d make that chemist extremely happy.
All of this is most decidedly chemistry. It’s just that evolution solved a lot of chemistry problems long before we did.