“Natural” is a curious word. It has many meanings. The design theorist distinguishes between natural and intelligent causes. But to a Darwinist, everything must be natural, because scientific materialists reduce everything (including their minds) to matter and energy. So on what basis can Hendrik F.T. Klare and Martin Oestreich offer the following title: “Teaching nature the unnatural”? Writing in Science, they describe a fascinating experiment in the same issue of the journal by Kan et al.

Silicon is found in nature in many inorganic forms, some of which are constructed by living organisms. Yet, no known biological molecule contains a carbon-silicon (C-Si) bond, and no biological processes to form C-Si bonds have been identified. On page 1048 of this issue, Kan et al. show that natural as well as reengineered enzymes can promote C-Si bond formation. The resulting chiral compounds mostly consist of one of the two possible stereoisomers (enantiomers). [Emphasis added.]

Chemists are interested in carbon-silicon bonds for semiconductors, polymers, and pharmaceuticals. Over the past few decades, they have achieved some success with inorganic catalysts, but since cells are so good at evolving new functions, why not teach a cell to do it? It would be an added plus to get homochiral (one-handed) compounds. Traditional methods require isolating one-handed compounds through the use of “elaborate chiral ligands or small peptides,” Klare and Oestreich point out. But “Kan et al. performed their experiments in vitro with non-natural, tailor-made substrates that are not likely to survive in a real biological system.”

The four Caltech scientists taught “nature” to do the “unnatural” by first isolating a microbe that lives in the hot springs of Iceland. The microbe contains a heat-tolerant version of the well-known enzyme cytochrome c, which New Scientist says “typically transports electrons around the cell.” Noticing that this version of cytochrome c was robust, and having a good map of its active site, they taught this enzyme to create carbon-silicon bonds through repeated rounds of “directed evolution” (which New Scientist accurately points out is really “artificial selection” as opposed to natural selection). Looking into the active site, they figured out what they needed “evolution” to do:

Following this approach, Kan et al. selected methionine residue M100, which needs to dissociate to make room for iron-carbenoid formation, as well as valine V75 and methionine M103, which are in close proximity to the active iron heme center (see the figure). They prepared mutants of the wild-type protein with sequential site-saturation mutagenesis.

Even though they left the candidates up to chance, their criteria for success were known in advance. They selected the mutants that did the job they wanted: creating a carbon-silicon bond. They watched and selected until they had a version with good productivity and good enantioselectivity (ability to isolate homochiral products). When the work was done, they had an enzyme that could make C-Si bonds 15 times more efficiently than the best inorganic catalyst.

The beauty and value of Kan et al.’s accomplishment lie in the enzyme-promoted formation of an unnatural bond. This closes a crucial gap between biological and chemical catalysis. The impact is unforeseeable, but it seems that we are a big step closer to potentially facilitating industrially relevant reactions such as alkene hydrosilylation with biomolecules.

It’s a long, long way from silicon-based life portrayed on Star Trek. They only got a C-Si bond, not a version of life that replaces carbon with silicon entirely. It works in a dish and in a bacterium, but getting the product out of a cell membrane without it hydrolyzing on the way will be a big challenge. But it’s a start with a bright future. New Scientist quotes Caltech co-author Frances Arnold:

“It’s a wonderful demonstration of how rapidly nature can adapt to solve problems,” says Arnold. “All of this diversity in the natural world is poised to do entirely new chemistry if you provide these new niches, so to speak.”

“So to speak.” To speak so is to confuse two very different kinds of causation: natural and rational. The first they attribute to blind chance; the second they intuitively know requires intelligence. Notice the cognitive dissonance; their worldview does not allow the intrusion of anything unnatural within the closed system of undirected causes, yet to jump between the two kinds of causation, they either have to deny their own rationality or attribute rationality to “nature.” Arnold did it in the quote above. Watch how Kan et al. do it in this paragraph, paying attention to the word “natural” and all the instances of purposeful, goal-directed activity:

Because of their ability to accelerate chemical transformations with exquisite specificity and selectivity, enzymes are increasingly sought-after complements to, or even replacements for, chemical synthesis methods. Biocatalysts that are fully genetically encoded and assembled inside of cells are readily tunable with molecular biology techniques. They can be produced at low cost from renewable resources in microbial systems and perform catalysis under mild conditions. Although nature does not use enzymes to form carbon-silicon bonds, the protein machineries of living systems are often “promiscuous” — that is, capable of catalyzing reactions distinct from their biological functions. Evolution, natural or in the laboratory, can use these promiscuous functions to generate catalytic novelty. For example, heme proteins can catalyze a variety of non-natural carbene-transfer reactions in aqueous media, including N-H and S-H insertions, which can be greatly enhanced and made exquisitely selective by directed evolution.

The only way this paragraph makes sense is by attributing rational thought to cells and molecules. Cells can be taught. They can be trained. They are ready to learn. Give them the opportunity, and they will generate novelty, using their machineries that are encoded and assembled. They have purpose that can be repurposed. They explore.

These in vitro and in vivo examples of carbon-silicon bond formation using an enzyme and Earth-abundant iron affirm the notion that nature’s protein repertoire is highly evolvable and poised for adaptation: With only a few mutations, existing proteins can be repurposed to efficiently forge chemical bonds not found in biology and grant access to areas of chemical space that living systems have not explored.

Progress in the Darwin-vs-design debate will require clarity. When we find our worthy opponents equivocating about evolution, equating natural evolution with directed evolution, we should stop and ask for clarification. “By directed evolution, are you talking about intelligent design?” When we find them equivocating about nature, using the words unnatural or non-natural, we should stop and ask for clarification. “I thought nature was all there is. What do you mean by unnatural?” These questions might steer the discussion into more productive paths than quibbling about a particular enzyme’s function or speculating about silicon-based Hortas evolving on planet Janus VI.

Image credit: Organosilicon-based life as envisioned by artist, Lei Chen and Yan Liang (BeautyOfScience.com) via Caltech.