The 2009 PNAS paper “In the light of directed evolution” by Nobel laureate Frances Arnold and her colleague Jesse Bloom is frequently presented as a powerful demonstration of evolutionary principles at work. By actively guiding the evolution of proteins in the laboratory, scientists appear to showcase how complex new functions can arise through mutation and selection. However, a critical analysis of the paper’s methods and results reveals the opposite. Far from providing evidence for unguided, molecules-to-man evolution, these experiments meticulously document the indispensable role of intelligent agency, foresight, and goal-directed intervention in achieving even modest functional modifications. The paper is not a confirmation of Darwinism; it is a textbook case study in intelligent design.
A Fair Summary of the Research
Bloom and Arnold review the lessons learned from “directed evolution,” an engineering strategy used to improve proteins. The process involves repeated rounds of intentionally mutating a gene, then screening or selecting the resulting protein variants for a desired property. The authors highlight several key empirical findings:
- Incremental Improvement: Single amino acid mutations can and do enhance protein properties like stability or catalytic activity. Adaptive pathways can be built one beneficial mutation at a time.
- The Crucial Role of Neutral Mutations: When no single mutation improves a function, experiments always find a wealth of “neutral” mutations that don’t harm the protein’s primary function.
- Opening New Pathways: These neutral mutations are vital. They can increase a protein’s overall stability, making it robust enough to tolerate a future mutation that is functionally beneficial but structurally destabilizing. They can also alter a protein’s secondary, “promiscuous” activities, which can then be optimized by selection to become a new primary function.
As a prime example, the authors describe how their lab successfully re-engineered a cytochrome P450 enzyme. The starting enzyme naturally hydroxylates fatty acids. The goal was to make it hydroxylate propane—a radical change in function. This leap was too large to achieve directly. Instead, the scientists had to guide the evolution through intermediate steps. At one point, the evolving protein became too unstable to continue. The researchers had to pause their search for better function and intentionally select for stabilizing mutations to rescue the project, after which they could resume selecting for the desired activity. This process, they argue, shows how the interplay of adaptive and neutral evolution can navigate the complexities of “fitness landscapes.”
The Core Analysis: An Unbridgeable Gulf Between Lab and Nature
The authors’ attempt to extrapolate these laboratory results to explain the origin of biological information in nature fails for several fundamental reasons. The experiment does not simulate unguided evolution; it simulates intelligent engineering.
The Investigator Interference Fallacy
The most glaring flaw in using this paper as evidence for Darwinism is the constant, necessary, and intelligent intervention of the researchers. The process is “directed” by a mind, not by a blind process.
- A Pre-defined Goal: Natural selection has no foresight or objective. Arnold’s team had a precise, pre-determined engineering goal: converting a fatty acid hydroxylase into a propane hydroxylase.
- Intelligent Selection: The “selection” performed in the lab is nothing like natural selection. The researchers created a highly specific, artificial screening environment designed to isolate and identify only those proteins that had moved closer to their goal. In nature, an organism’s survival depends on the integrated function of thousands of systems, not a single property measured on a lab plate.
- Foresight and Rescue: The P450 experiment got stuck on a “local peak” of instability. In nature, this lineage would simply die out or stagnate. In the lab, the intelligent scientists diagnosed the problem (instability), conceived of a solution (select for stability), implemented that solution, and then resumed their original goal. This is not an example of evolution overcoming a barrier; it is an example of an engineer overcoming a barrier with foresight and corrective action.
The “Assume a Gene” Fallacy
These experiments begin with a fully-formed, highly complex, and information-rich protein (cytochrome P450). They do not explain the origin of the P450 protein fold itself, which represents an island of astronomical improbability in the vast sea of non-functional polypeptide sequences. Douglas Axe’s work has shown that the odds of finding a functional protein fold of just 150 amino acids by chance are a staggering 1 in 10^77. Bloom and Arnold’s work begins with this miracle of pre-existing specified information already in hand, and then proceeds to tinker with it. This is akin to claiming you can explain the origin of a software program by demonstrating your ability to change the color of a few icons. It explains the modification of existing information, not its origin.
The Inevitability of Genetic Entropy
The directed evolution experiment is a hyper-artificial environment, selecting for a single trait to the exclusion of all others. In a real organism, the 23 mutations required to change the P450’s function would be subject to a host of pleiotropic effects. A mutation that enhances one specific catalytic activity will almost certainly degrade other subtle functions or interactions governed by the same gene. According to biochemist Michael Behe’s “First Rule of Adaptive Evolution,” the fastest way for an organism to adapt is to break or blunt existing genes. The specialization of this P450 enzyme is a form of “adaptive degeneration”—it gains a new, narrow function at the likely cost of its original, broader functions and its integration into the cell’s metabolic network.
More fundamentally, the entire neo-Darwinian framework of mutation/selection as a creative engine is false. The genome is subject to relentless decay via the accumulation of nearly-neutral deleterious mutations. As geneticist John Sanford has demonstrated in his work on “genetic entropy,” natural selection is powerless to stop this genome-wide degradation. These laboratory experiments, which use intense artificial selection to find a rare beneficial change, create a deeply misleading picture of the net process in nature, which is overwhelmingly degenerative.
An Alternative Explanation: Engineering Principles on Display
When we strip away the Darwinian narrative, the results of directed evolution are perfectly explained by the principles of intelligent design. The experiment is not an analogy for unguided evolution; it is a direct example of engineering.
- Inference to the Best Explanation: What is the vera causa—the true, observed cause—of the new propane-hydroxylating P450 enzyme? It is the intelligence and ingenuity of Frances Arnold’s research team. We know from our uniform and repeated experience that intelligent agents can set goals, build complex machinery, and modify existing systems for new purposes. We have no such experience of blind, material processes generating high levels of specified information. Therefore, Intelligent Design is the best and only causally adequate explanation for the results observed in the laboratory.
- Front-Loaded Design: The paper’s findings on “neutral networks” and “promiscuous functions” are not evidence of the creative power of random drift. On the contrary, they are hallmarks of robust engineering. A good engineer designs systems with built-in tolerance for minor errors (neutrality) and with the capacity to perform secondary functions (promiscuity). These are not lucky accidents for evolution to exploit; they are design features that allow for limited, rapid, and pre-programmed adaptation within a stable framework. This aligns with the “created heterozygosity” and “nonrandom evolutionary hypothesis” (NREH) models, which propose that organisms were created with pre-loaded genetic diversity and built-in adaptive systems to allow for rapid diversification within their created kinds.
Conclusion: The Signature of Mind
“In the light of directed evolution” is a landmark paper, but not for the reasons its proponents believe. It does not illuminate a pathway for unguided evolution. Instead, it shines a spotlight on the absolute necessity of a guiding intelligence to achieve even minor modifications of pre-existing biological function. By demonstrating that progress requires a pre-defined goal, intelligent selection, and foresight to overcome inevitable barriers, Bloom and Arnold have unintentionally provided a powerful empirical case for the core tenets of intelligent design. The experiment shows that creating new function requires a mind. The signature in the cell, and the modifications made to it in the lab, both point to the same causally adequate source: intelligence.
Leave a Reply