Member # 1
posted 25. March 2002 22:31
Public release date: 19-Mar-2002
Contact: Jonathan Sherwood
University of Rochester
Darwin's Time Machine: Scientists begin predicting evolution's next step
Untangling the branches of evolution's past is a daunting enough task for researchers, but some scientists are now turning their eyes toward the future in a bid to predict evolution's course. Barry G. Hall, professor of biology at the University of Rochester, has shown how a model of evolution developed in the lab accurately reproduces natural evolution. The research, published in the March issue of Genetics, demonstrates how the model is so accurate that it can be used to predict how a strain of bacteria will become resistant to antibiotics-giving researchers a possible tool to create drugs to which bacteria cannot adapt.
"Antibiotic-resistant bacteria were a perfect target on which to test the model because we have examples of antibiotic resistance genes that first appeared 40 years ago," says Hall. "We know how those genes evolved in nature during the last 40 years, so if we apply the model to those genes and the model predicts those same evolutionary outcomes as happened in nature, we can be confident that the model works."
"Hall's recent work lays an all-important conceptual foundation for understanding the functional evolution of enzymes," says Anthony M. Dean, associate professor of biotechnology at the University of Minnesota. "Not only has he attempted to predict the outcome of adaptive evolution, but he has removed that last vestige of vitalism, which asserts that natural selection can only be studied in the natural environment."
Evolution is usually a very slow process. A mutation in the genetic code of an organism changes the organism's functions, usually for the worse, but sometimes for the better. Those organisms with mutations that allow them to thrive better than their cousins, tend to survive while the cousins die out. A classic example is to remove the gene that allows E. coli bacteria to digest lactose, but then give the colony of bacteria only lactose to eat. Of the millions in a colony, some bacteria re-evolve the ability to hydrolyze lactose and survive.
Thirty years ago, Hall was growing E. coli in his lab, trying to pick out those bacteria that had a mutation significant enough to help them out-compete their cousins. This process of looking for a single mutation with a dramatic survival benefit was slow and didn't accurately mimic natural evolution, which might normally retain mutations with less dramatic benefits than was apparent in a lab setting.
Researchers developed an alternative; instead of growing a culture of cells and then subjecting them to a stress-like lactose that they couldn't metabolize-and waiting to see if any survived, scientists had decided to take a gene or two and mutate it in a test tube.
"You can introduce a lot of mutations in the lab," explains Hall. "In effect, you can take millions of copies of this gene and give each one a different mutation." Those mutated genes are introduced back into the cells, "and then you ask, can you grow on lactose now?"
The mutations that arose in nature were also found in the laboratory cells, but would that always be the case? Hall, knowing that he had essentially bypassed the cell's normal machinery, needed to know if this accelerated process would accurately mimic mutations that would arise in nature. If he could mutate a 40-year-old antibiotic resistance gene, called TEM-1, in the laboratory and match his mutations against the natural mutations that arose as that gene adapted to better antibiotics, then he could see if the model would accurately predict the how genes evolve in nature.
"Antibiotic resistance evolves rapidly enough that we can observe significant increases in resistance profiles in just a decade," says Miriam Barlow, a doctoral student in Hall's laboratory. "This provides the unique opportunity to actually observe the process of evolution as it happens."
The results matched well. Certain mutations developed in the lab improved the resistance gene more than others. And, most importantly, these were the same mutations that arose in nature.
The evolutionary outcome taken by the antibiotic resistance gene in the laboratory gave Hall an excellent tool with which to test the model, but there was no reason that the model couldn't be extrapolated beyond the past evolutionary outcome into the future. Mutating key genes in the bacteria and subjecting the newly evolved bacteria to new antibiotics will show if and how a species of bacteria will likely evolve to circumvent the new antibiotic. This will tell pharmaceutical companies whether their new drug will have a life span of just a year, or a decade or more. The model should reveal what new traits will evolve, so the researchers may be able to synthesize a second drug that kills the bacteria with the new ability-essentially cutting off the bacteria's evolutionary escape route.
"It's an arms race," says Hall. "One of the things that came out of this experiment is that there's a new drug, called cefepime, that is lethal to many bacteria and looks to be a promising antibiotic. But, we found mutations in the TEM-1 gene that allowed bacteria to hydrolyze cefepime, and we think there is a very good chance that this capability will arise in nature soon."
To find which gene needs to be mutated in order to confer resistance, Hall searches either for a protein that is doing something similar to what he is looking for, or for a gene sequence that is similar to one in another organism that does something similar. He then has to ask: Can this sequence evolve, one mutation at a time, into a sequence that confers resistance? Much like D-O-G can evolve into C-A-T one letter at a time-DOG, COG, CAG, and finally, CAT-any sequence can be turned into any other, but in nature an organism must be able to survive at each step of the process or else the process ends. "CAG" is a nonsense word and (if it were a gene) might cost that organism its life. In contrast, the progression of DOG, COG, COT, CAT is made up of single changes, but each change still creates a valid word. Likewise, in evolution, the series of mutations must remain viable at each stage. Hall found that the model could show along what path certain bacteria were likely to evolve to thwart antibiotics that had not yet reached the market.
"Fighting bacteria with antibiotics has always been done the same way," says Hall. "We make a drug and after a while, the bugs adapt to it, so we give them a variant of the drug. But if we can predict how they're going to get around our treatments, we can work out a way to make that route impossible for them. We can cut them off at the pass."