Complex I of the respiratory chain is one of the crucial enzymes that ultimately helps you use the oxygen you breathe. It is the first funnel through which electrons are poured on their path to combining with oxygen. Part of this funnel is shaped by oxygen's unique chemical properties.
Enzymes like Complex I, called hydrogenases, first show up around Chapter 6 of A World From Dust, well before oxygen fills the atmosphere in Chapter 8. Still, if any oxygen is around at all, they are shaped by it, because they must avoid oxygen's negative power. Oxygen reacts with stray electrons to form Reactive Oxygen Species (ROS) that shatter the insides of a cell.
If a hydrogenase drops too many electrons out of its "funnel", they are picked up by oxygen and make havoc-wreaking ROS inside the cell. Avoiding the negative consequences of oxygen shapes life as much as running toward the positive consequences of oxygen's energy.
This was shown in the recent study "Reactive Oxygen Species Production by Escherichia coli Respiratory Complex I" in Biochemistry. This study is built on a previous experiment in that increased the amount of electron-carrying "electron boxes" inside the cell called NADPH. Normally Complex I only gathers electrons from the NADH electron box through its funnel, but when there's high amounts of NADPH, it will evolve to accept NADPH as well. It does so precisely at the green sticks at the bottom of this figure:
(Figure provided in supplemental materials to the paper cited above)
The orange sticks are NADH, and the green sticks are placed exactly where the "P" is that makes NADPH different from NADH. If this enzyme is to bind NADPH, those green sticks must get out of the way. In the previous experiment, they did, evolving to alanine (A) and glycine (G).
The key to the new paper is where the green sticks didn't evolve to. Other green sticks still bind NADPH just as well as alanine and glycine. In particular, histidine (H) and glutamine (Q) were not observed, although they interact well with the P in NADPH and can even increase binding. So why were these perfectly capable mutations not observed?
The answer provided in this new paper is that, with histidine and glutamine, too many electrons fall off the NADPH, out of the funnel, and onto oxygen, making too many Reactive Oxygen Species (ROS). Because these are dangerous to the cell, the protein does not evolve in that direction, but rather evolves in the direction of alanine and glycine.
This can be summarized as the letters for these particular positions. In the "word" that is the enzyme, at this one position, we don't see "H" or "Q", but we do see "A" and "G", not because the enzyme works better with those two letters, but because the whole cell works better with those two letters.
Instead of four possibilities, evolution chooses two, and in this very small way, it is constrained by the need to avoid oxygen stress. This is a biochemical example of two important points of the book:
1.) Because of oxygen's chemical tendency to form ROS, the protein has fewer options at that position than it would otherwise. (It is constrained by oxygen's chemistry.)
2.) To understand why it's restricted, we must account for the oxygen stress on the whole cell, not just the efficiency of the one enzyme or the NADPH-binding properties of the one residue. (We must look at the higher level of the cell biology rather than the lower level of the biochemistry; the higher level constrains the lower.)
To cite the central metaphor motivating the book, if this tiny motif in the "tape of life" were replayed, we would still "hear" A and G, not H and Q. There is freedom for the system to select A or G, but not H or Q. The possibilities are constrained by the double-edged sword of oxygen stress, and the river flows in one direction, but not the other.