« Carnival of evolution | Main | Reversing evolution: crops evolving into weeds »

Directed evolution versus intelligent design of enzymes

"How have all those exquisite adaptations of one part of the organisation to another part, and to the conditions of life, and of one distinct organic being to another being, been perfected? We see these beautiful co-adaptations most plainly in the woodpecker and missletoe; and only a little less plainly in the humblest parasite which clings to the hairs of a quadruped or feathers of a bird; in the structure of the beetle which dives through the water; in the plumed seed which is wafted by the gentlest breeze; in short, we see beautiful adaptations everywhere and in every part of the organic world." -- Charles Darwin
Evolution denialists often claim that these adaptations must have come from an Intelligent Designer, who has apparently been too busy lately (protecting pedophile priests, maybe, or working to block gay marriage?) to come up with any new designs. They claim that natural selection (nonrandom proliferation of random variants) isn't up to the job.

But intelligent human designers are increasingly relying on processes similar to natural selection. The latest issue of Science discusses two examples of the use of selection-like processes for developing useful enzymes. One paper explicitly calls their approach "directed evolution." The other doesn't use the term, but what would you call generating billions of randomly varying designs in a computer and selecting those that meet certain criteria? Sounds like selection to me. In neither case did the researchers rely only on natural-selection-like processes. Instead, they used some combination of intelligent design and selection from among random variants. But nonrandom selection from among random variants was a key contributors in each case, solving problems beyond the reach of human reason or intuition.


As scientists continue to increase the control of organic structures it will be increasingly difficult to argue against evolution. One cannot claim that the world is flat when satellites are sending back beautiful images of a round earth.

I am not a biologist but have a strong interest in evolution and the origins of life. Are you aware of any research into spontaneous generation? It seems to me that it might be possible that simple lifeforms, viral perhaps, are still being created. With modern scientific equipment, perhaps we can find examples. This cannot be an original idea with me. Any help in finding information on research in this direction, confirming or denying the possibility of spontaneous generation, would be greatly appreciated.

Alex Ambrioso

You might think it would be easier for life to evolve today than the first time around, because all of the organic molecules needed are already available, whereas only some of them are known to be produced by nonbiological processes. But the first replicators (perhaps RNA molecules) didn't have to worry about being gobbled up by a hungry bacterium or destroyed by oxygen, which was absent from the early atmosphere. We could presumably evolve life in lab microcosms, under the right conditions, but how many microcosms would it take to have a reasonable chance of getting life in one of them within one human lifetime?

Other readers may have better suggestions, but I would suggest starting with this book: "Major Transitions in Evolution" and/or try Google Scholar (not plain Google, which will bring up lots of crackpot stuff) for "origin of life." Here are a few relevant papers:
Halliday A. N. 2001. Earth science: In the beginning.. Nature 409:144-145.

Kun A., M. Santos, and E. Szathm√°ry. 2005. Real ribozymes suggest a relaxed error threshold. Nature Genetics 37:1008-1011.

Lincoln T. A., G. F. Joyce. 2009. Self-Sustained Replication of an RNA Enzyme. Science 323:1229-1232.

Miller S. L. 1953. A production of amino acids under possible primitive earth conditions. Science 117:528-529.

Robinson R. 2005. Jump-Starting a Cellular World: Investigating the Origin of Life, from Soup to Networks. PLoS Biology 3:1860-1863.

Szostak J. W., D. P. Bartel, and P. L. Luisi. 2001. Synthesizing life. Nature 409:387-390.

Wachtershauser G. 2000. Origin of life: life as we don't know it. Science 289:1307-1308.

Whitfield J. 2006. Origins of DNA: Base invaders. Nature 439:130-131.

The algorithm of evolution has been rigorously applied in the laboratory for the last two decades. The best examples are directed evolution of proteins and the creation/screening of combinatorial libraries of small molecules. The power of these approaches stems from our ability to create (error prone PCR and solid phase synthesis, respectively) and screen (high throughput plate reading, ELISA, or fluorescence activated cell sorting) potential solutions far faster than we can understand the problem. I would argue that a creator that must rely on evolution is far from omniscient. For problems that are simple, evolution isn't needed. Imagine the foolishness of calculating 1+1 by listing all numbers between 1 and 1000000, picking them randomly, and selecting against X-1 =/= 1. For any being that can fully understand the complexities of life well enough to create life, using evolution to create and optimize life is pathetic.

The creation of synthetic life was presented earlier this year by Venter et al. Sure, most of the gene sequences were found in nature. However, combining this technology with directed evolution and the Shultz technology for expanding the genetic code, one can envision truly unique microorganisms customized for our every industrial or scientific need. We're a couple of decades away, but the progress being made in the fields of chemical biology and molecular biotechnology imply a strong future for synthetic life.

I agree that, depending on the problem, it often true that "our ability to create (error prone PCR and solid phase synthesis, respectively) and screen (high throughput plate reading, ELISA, or fluorescence activated cell sorting) potential solutions far faster than we can understand the problem", at least at present. Will we ever develop software that expands our omniscience enough that we can design solutions faster than we can evolve them?

"Will we ever develop software that expands our omniscience enough that we can design solutions faster than we can evolve them?" Maybe, but the two processes are not completely independent. The beautiful idea behind the Science papers in your original post is that computational analysis gives us a good starting point for evolution. For example, a 200-mer peptide sequence would have 20^200 - 1 mutants. However, by identifying 5 amino acid residues that are most influential for stability and activity, one only needs to screen a library of 20^5 (3.2x10^6). This sort of work was previously done by crystal structure analysis, but computational analysis is much faster.

The same principles are true for computational drug design. Software can predict the framework of a small molecule that is best suited for inhibiting a protein. That framework can be synthesized, bound to a solid phase, and decorated combinatorially.

Post a comment

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)

Type the characters you see in the picture above.