Evolution’s Test Tube Revolution

Scientists are harnessing the power of evolution in their laboratories to produce promising drugs and industrial chemicals that might never occur in nature, and they are doing it in days rather than eons.

The ingenious process, which creates mutant genes and selects those with desired characteristics, is called “directed evolution.” Some call it evolution in a test tube.

Enthusiasts believe it will revolutionize the biotechnology industry, calling it one of the most important developments since the invention of genetic engineering a quarter of a century ago.

“It is capable of producing an endless variety of new products with extraordinary and in some cases unpredicted properties,” said A. Stephen Dahms, director of California State University’s Program for Education and Research in Biotechnology. “It’s Mother Nature at warp speed going where no man has gone before and where Mother Nature might not go for 100 million years.”

One of the first new chemicals developed this way--an enzyme added to laundry detergents to remove hard-to-dissolve fatty stains--is on the market.

That mundane application is only the beginning. The future may include environmentally clean agents that can replace pollutants used by the chemical industry, genetically altered plants and entire families of medications to fight cancer and other diseases.

At the City of Hope Cancer Center in Duarte, for example, researchers are testing a directed-evolution drug that may be able to destroy leukemia cells while sparing healthy ones. At Ixsys, a San Diego-based biotech firm, scientists are working on a drug to block the growth of blood vessels in tumors. Researchers from Diversa in San Diego and ZymeQuest in Massachusetts are working on proteins that turn type A or B blood into type O--one step toward making everyone a universal donor.

The products developed for use in medicine will require years to reach the market because of lengthy testing and regulatory requirements, Dahms said, but the industrial applications will come much faster.

The scientists who have pioneered this research say it is coming of age after a decade or more of development.

“A few years ago, I was a lonely researcher working on a couple of enzymes,” said Frances H. Arnold, professor of chemical engineering at Caltech. Now scientists from the industry stream through her lab. Industry, she said, “is investing a lot of money. It’s snowballing. . . . It’s not science fiction anymore.”

Screening for Desired Characteristics

Genetic engineering began 25 years ago with the discovery of gene splicing, a method for inserting a gene isolated from one organism into the cellular machinery of another. That breakthrough has enabled scientists to make human insulin in bacteria cells and produce human blood-clotting factors in sheep’s milk, among other advances.

Over the years, biochemists have refined the process by getting genes to produce proteins designed to their specifications--improved versions of natural substances. But because the current understanding of this approach is limited, the success rate is low. When a new product does result, it is often after a long, expensive process.

Directed evolution, however, takes advantage of the same forces that permitted the diversification of the species--using a variety of chemical tricks that generate thousands, millions, even trillions of mutations of any gene in a test tube.

One method is to use a common technique for copying genes quickly but to poison the process so that many of the copies include errors. Another way is to break several related genes into pieces and scramble them together.

Spliced into a bacterium or yeast cell, many of these mutant genes will begin manufacturing new substances.

Then, mimicking the principles of natural selection, the scientists screen for those few genes that make products with the desired characteristics: a laundry detergent that works at low temperatures, an enzyme that chews up toxic PCBs or an antibody that binds tightly to the surface of cancer cells.

Repeating the process can strengthen the desired trait or add new ones.

Such mutations also take place in nature, but at a much slower pace. For example, most bacteria exposed to an antibiotic will die, but a few with the right combination of mutant qualities will survive. Expose these mutated bacteria to even higher doses of the antibiotic, and a few super-resistant bacteria will emerge.

In nature, it might take decades for bacteria to develop resistance to a specific antibiotic. In order to study resistance and discover new antibiotics, scientists have been able to use directed evolution to produce highly resistant strains in test tubes--within weeks.

The method can be used to introduce new genes into plants and animals or, in theory, to correct genetic defects in humans, raising ethical issues that have dogged the field of genetics for generations.

“If you hate genetic engineering, [directed evolution] is a bad thing. If you like genetic engineering, it is a good thing,” said Henry T. Greely, a law professor and co-director of Stanford’s Program in Genomics, Ethics and Society.

Enzymes in the Laundry

Greely doubts that directed evolution--or any other genetic engineering technique--will soon succeed in altering human heredity, in part because it will take a good deal of research to assess the impact of a novel gene on a recipient. “I think it would be very aggressive to think about using this technique for human therapy,” he said.

In a recent essay in Science magazine, Jon W. Gordon of Mt. Sinai School of Medicine in New York agreed, arguing that even the simplest gene-replacement experiments in animals have sometimes had disastrous results. “We clearly do not yet understand how to accomplish controlled genetic modification of even simple phenotypes [observable characteristics],” he wrote.

But at a simpler level, Jeremy Minshull, a biochemist at Santa Clara, Calif.-based Maxygen Inc., compares the process to the traditional breeding of new traits in animals that were not present in either parent.

Charles Darwin, the father of evolutionary theory, “was a pigeon fancier,” Minshull said at a recent conference on enzyme technologies. “He observed that in a few generations, pigeons could be bred with fantastic feathers on their feet,” Minshull said, even though neither parent’s stock had such feathers. And the only purpose was an aesthetic one, to please the breeders.

These scientists say that by imitating nature, directed evolution has become a quick and powerful way to create chemicals that solve problems nature never posed.

Harvey Bialy, editor at large of Nature Biotechnology magazine, traces the technique’s origins to Columbia University researcher Sol Spiegelman, who, more than 30 years ago, was able to evolve a fast-acting bacteria virus enzyme in a test tube. But the idea didn’t catch on as a practical tool until the 1980s and ‘90s.

The technology is already bearing fruit. Danish firm Novo Nordisk, the world’s largest supplier of industrial enzymes, has been producing a variety of proteins that can remove dirt and grease from fabrics. Among the hardest stains to remove are fatty ones, such as lipstick, frying fats and the grime that accumulates on collars.

Since 1988 Novo has been selling a product called Lipolase, a natural enzyme isolated from a fungus, for use in detergents to attack those stains. But the product works best when the humidity is low--when the clothes are drying, said Novo Vice President Soren Carlsen. That means garments have to be dried after treatment, then washed again to get the enzyme out.

To improve the enzyme, the company used directed evolution to create variants that work while the clothes are wet, during the wash cycle. It began marketing the product last year.

Most enzymes, which act as catalysts in biological processes, are proteins, the building blocks of living cells. However, one of the pioneers of directed evolution, Dr. Gerald Joyce at the Scripps Research Institute in La Jolla, has been applying the technique to making enzymes out of DNA--the material that contains the basic genetic code for most living organisms. In nature, DNA is relatively stable, inactive stuff.

But Joyce and Dr. David Snyder at the City of Hope have developed a DNA enzyme that can attach itself to a genetic messenger present in two common forms of leukemia. The enzyme acts as a molecular scissors that cuts through the messenger, killing the leukemia cells while sparing the healthy ones.

Snyder hopes the DNA enzyme will prove useful in bone marrow transplantation, as a way of purging leukemia cells from cells that are collected from patients and then returned to them following chemotherapy and radiation.

The enzyme, Snyder said, appears to be effective in killing most of the leukemia cells, but the method has not yet been tested in transplant patients.

“I consider this the ideal anti-cancer therapy,” Snyder said. “You define a gene and design a specific weapon to target it, and it attacks the gene and leaves normal cells alone.”

Attacking Tumors and Boosting Blood Supply

At Ixsys, scientists are developing antibodies that block a substance needed for blood vessels to grow in tumors. Without blood vessels--and a steady supply of blood-borne oxygen and nutrients--tumors cannot continue to grow and spread.

The company has been testing a first-generation antibody called Vitaxin in patients with an uncommon, difficult-to-treat tumor, leiomyosarcoma. In the initial trial, said Dr. William Huse, Ixsys’ chief executive, one patient was on the drug for two years and saw “an actual decrease in the size of some tumors.” Using directed evolution to produce thousands of variants of the antibody, the company has a second-generation drug that binds more tightly to its target and should prove more effective than Vitaxin. The company hopes to begin human testing of the new version next year.

The same drug may also prove effective in a variety of diseases that affect blood vessels, including rheumatoid arthritis and the reclosing of coronary arteries following angioplasty, Huse said.

Scientists at the University of Washington and Maxygen have taken enzymes from the herpes virus and enhanced their ability to activate chemotherapy drugs such as AZT and acylovir, which are used to treat cancer, making the drugs effective at reduced doses.

The Washington scientists also have been working on a super-enzyme intended to protect normal cells from the effects of anti-cancer drugs, said Dr. Lawrence A. Loeb. Neither drug, he said, has been tested in patients.

At Diversa, researchers have scoured the world looking for natural substances with benefits for industry or medicine. And they’ve been using directed evolution to create variations with improved properties.

Working with ZymeQuest, the company has been developing enzymes that remove blood type factors from the surface of red blood cells. Diversa Chief Executive Jay M. Short said the two private companies hope to create type O blood from types A and B, part of an effort to create universal-donor blood and extend the nation’s blood supply.

Caltech scientist Richard Roberts and Harvard Medical School professor Jack Szostak have invented an even speedier method of test tube evolution, one that can produce trillions of mutations.

A company the two helped launch two years ago, Phylos of Beverly, Mass., is working on automating the process. Phylos is developing libraries of natural substances for use as novel drugs and industrial chemicals.

Among the potential products, said Phylos Vice President Richard W. Wagner, are improved versions of hormones already in use, such as Epogen, the anti-anemia drug from Thousand Oaks-based Amgen. Epogen’s annual sales exceed $1 billion.

Said Maxygen’s Minshull: “All properties are evolvable. . . . The future is limited only by our imaginations.”

(BEGIN TEXT OF INFOBOX / INFOGRAPHIC)

Directed Evolution

Directed evolution is a technique that mimics nature by creating numerous mutated versions of a gene, then selecting the ones that make substances with desired characteristicssuch as an enzyme that can remove lipstick stains from fabric or an antibody that will bind tightly to a cancer cell. This schematic shows how it works:

1. Pick one or more genes.

2. Create a multitude of mutationsby making defective copies or combining pieces from genes in the same family.

3. Insert each of the mutant genes into a bacterium or a yeast cell, where some will manufacture proteins or other substances.

4. Grow clumps, colonies, of the cells.

5. Pick out the clumps that make products of interest. This is done by growing the cells in a solution or gel that contains a substance you want an enzyme to digest or a protein you want an antibody to cling to. Design the system so that clumps light up or change color when the desired reaction takes place. Set the conditions to match your objectivesyou might want low temperatures for a laundry detergent or body temperature for an anti-cancer drug.

6. Repeat the process as many times as necessary to add new characteristics or improve useful traits.

Sources: Frances H. Arnold, Caltech