In February last year, The New York Times published an interview with my University of Chicago colleague Janet Rowley. Janet is deservedly famous for finding a repeated chromosome rearrangement in certain types of leukemia. This was one of the earliest indications that genome changes in cancer cells do not occur randomly.
In the interview, Janet explained how she discovered this particular chromosome change, now called the "Philadelphia Chromosome." She was just looking through the microscope, motivated by her curiosity to know more about these tumor cells.
Janet pointed out that she might well not be able to repeat her discovery in today's scientific environment. She was practicing what she called "observationally driven research." Today, she said, granting agencies don't support that kind of work. "That's the kiss of death if you're looking for funding today. We're so fixated now on hypothesis-driven research that if you do what I did, it would be called a 'fishing expedition,' a bad thing."
In other words, you have to know what kind of result to expect before the funding agencies will give you money to look for it. Surprises are not fundable. But "surprise" is just another word for "discovery." As Janet put it, "I keep saying that fishing is good. You're fishing because you want to know what's there."
Let's look at how we would "fish" for complex genomic novelty through natural genetic engineering. I can think of two approaches. There will definitely turn out to be more.
One approach was included in my book. The idea was to do interspecific hybridization with a well-characterized organism, like the mustard weed Arabidopsis, and follow what happens with the genetically unstable hybrid progeny.
We know that interspecific hybridization and genome duplication lead to high levels of genomic and phenotypic variation. DNA sequencing has found evidence of genome duplication at many critical points of evolutionary divergence, especially in plants. There is a fine Scientific American article by the famous 20th-century evolutionist G. Ledyard Stebbins entitled "Cataclysmic Evolution," which describes how hybridization between two wild grasses can recreate the origin of flour wheat.
The hybrid progeny can be followed, and those plants that develop significant new traits, such as flower patterns, can then be analyzed. Sequencing the whole Arabidopsis genome in a short time is now feasible. The sequence data will let the Arabidopsis genome speak for itself in telling us how the new traits evolved.
We can then look for multiple changes that show signs of coordination in the underlying natural genetic engineering events. Such coordinated events might be insertions of the same or related mobile elements at distinct locations in the genome or the addition of the same domains to more than one protein in the network responsible for development of the novel trait.
The second "fishing" approach to asking how a novel feature can evolve would use a microbe, as suggested in the previous blog on experimental evolution. In this case, however, the changes would not be pre-targeted to a number of different sites in the genome.
Instead, the microbe would be challenged to do something that we would not expect it to accomplish with a single mutation. Our expectations would be based on the microbe's genome sequence and its predicted biochemical capabilities. It could, for example, be challenged to utilize a nutrient that could not be metabolized by a variant of any of the existing enzymes deducible from the genome sequence. Or it might be required to synthesize a nutrient that normally has to be supplied in the growth medium.
As I suggested in the previous blog, selection would repeat my colleague Bernhard Hauer's successful approach of waiting a month or longer for a colony to appear on the selective medium. In this case, however, the experimentalists could take advantage of current DNA sequencing technology to find out what genome modifications were necessary.
The evolution of a novel biochemical pathway would be a significant evolutionary accomplishment, different in kind from the tweaking of existing activities detected so far in microbial evolution experiments. We would know from the DNA sequence if truly novel functional entities had been generated, such as proteins with previously nonexistent domain combinations.
While we can only speculate in the most general terms what we might learn from letting the genomes speak for themselves, there is good precedent in following Janet Rowley's observational-driven example. We have only to think of Barbara McClintock's unexpected discovery of mobile genetic elements to see how fruitful unpredicted outcomes can be.
There have always seemed to me to be two strong motivators in scientific research: pride and curiosity. The pride-driven agenda is to show how clever we can be in establishing a scientific principle. Almost always, this involves elaborating on principles that have already been articulated but perhaps not yet adequately documented.
The curiosity-driven agenda is to figure out how things work. That means carrying out experiments with as few preconceptions as possible. The experiments that do not turn out as expected are the most revealing. They tell us something is going on that we do not understand.
The pride-driven researcher tends not to pursue unexpected results. There is no way to predict a successful outcome. But the curiosity-driven scientist often feels the challenge of a result that cannot yet be explained. While success is not assured, persistence can revolutionize our understanding.
Right now, we need more curiosity-driven research and thinking in evolution science. Molecular genetics and genome sequencing have repeatedly provided unexpected results. We only gain in understanding from following the outliers, wherever they may lead us.