It was a Eureka! moment. When Michael Behe saw that diagram of a bacterial flagellum in his biochemistry textbook, he says, “That’s an outboard motor! That’s designed. That’s no chance assemblage of parts.” Thousands of other students must have ignored the obvious, or else meekly accepted what their professors told them, that natural selection was capable of creating such things. But Dr. Behe self-admittedly has a stubborn streak that wouldn’t tolerate simplistic answers. From such qualities a Revolutionary was born. This year we’ve been celebrating the 20th year since he published Darwin’s Black Box, the book that first put intelligent design on the map.

Twenty years ago another rotary engine was to earn its discoverers the 1997 Nobel Prize in Chemistry. ATP synthase was known previously by its chemical function, but its mode of action was a black box. When Paul Boyer and John E. Walker opened the box and saw a rotary engine one tenth the size of the bacterial flagellum, they were astonished. In the years since, more details have come to light. It’s hard to find any paper describing this machine that doesn’t call it “amazing” or “remarkable.” (Readers may wish to view our animation of ATP synthase, above.) And as Behe discovered in researching the flagellum, hardly anyone tries to explain how it could have come about by neo-Darwinian processes.

John Walker has continued research on this molecular motor since earning the Nobel. Now at age 75, he must be gratified that all the parts have been described, although details about the extraction of torque from a flow of protons remain to be understood. His latest paper, just published by the Proceedings of the National Academy of Sciences (PNAS) with three colleagues from Cambridge, shows that the surprises keep coming.

Living cells need fuel in the form of adenosine triphosphate, or ATP, to stay alive. This fuel is generated by a molecular machine made of two motors joined by a rotor. One generates rotation by using energy provided by oxidative metabolism or photosynthesis; the other uses energy transmitted by the rotor to make ATP molecules from its building blocks, adenosine diphosphate, or ADP, and inorganic phosphate. The structure has been determined of a fungal machine, isolated from its cellular power stations, the mitochondria, where the machine operates. It provides unsuspected details of the blueprint of the machine and how it works. The working principles of the fungal machine apply to similar machines in all species.

Before now, only about 85 percent of the structure had been determined. That last 15 percent was hidden inside the membrane that anchors the machine. By extracting working motors from a fungus and combining it with counterparts from a cow’s mitochondrial version, the team of Vinothkumar, Montgomery, Liu, and Walker was able to glimpse the machinery with unprecedented clarity:

As described here by cryo-EM, we have determined the structure of the entire monomeric F-ATPase complex from the mitochondria of the moderately thermophilic fungus, Pichia angusta, bound to the inhibitory region of the natural bovine inhibitor protein, IF1 (inhibitor of F1-ATPase). This structure fills significant gaps in our knowledge of the mechanism of the F-ATPase. First, it contributes to our understanding of the coupling mechanism by providing unsuspected details of the peripheral stalk and how it is attached to both the catalytic F1-domain and the membrane sector of the enzyme. Second, it provides an independent description of the transmembrane proton pathway for protons, helping both to define common features in the proton translocation mechanism in bacterial and mitochondrial F-ATPases and to explain human pathological mutations in the subunit.

That last clause reinforces Behe’s concept of irreducible complexity. Mutations break things. If one of the parts is broken or missing, people get sick or die. Although there are variations in ATP synthase between bacteria and eukaryotes, particularly in the number of c-subunits in the rotor, most parts are “highly conserved” between kingdoms and species. The authors speak of parts that are “strictly conserved” or “absolutely conserved.” There are so many essential parts in this machine, it cannot function unless everything is in place and matched for functional interaction.

The authors note that these motors spin at up to 350 Hz. That’s a whopping 21,000 rpm! Think about that. Every rotation yields 3 ATP molecules, which are used for most energy-consuming processes in the cell. Every step of a kinesin machine walking down a microtubule, every motion of actin in muscle, every advance of a polymerase translating DNA consumes ATP. If one ATP synthase motor generates 63,000 ATP molecules a minute, imagine how many ATP molecules are being created by thousands of the motors in a single cell. No wonder you can create up to your body weight in ATP in a single busy day.

Walk up to one of these machines in your mind’s eye. It’s rotating so fast it’s a blur. The bottom half spins like a turbine in a hydroelectric power plant, turned not by water but by protons. The upper half assembles the ATP in three pairs of lobes that continuously run “a cycle of substrate binding, ATP formation, and product release. Thus, each 360° rotation produces three molecules of ATP.” How could the scientists study something moving so fast? They couldn’t. They had to slow it down with an “inhibitor protein” so they could get a good look at it.

They realized that the catalytic upper part rumbles. At top speed, the constant insertion and ejection of substrates generates lateral forces. Without a really steady brace, the machine would vibrate itself out of commission. And a steady brace is just what they found:

Although the relevant interactions are still not fully resolved, the current structure demonstrates that the attachment to the catalytic domain is much more extensive and robust than had been thought, probably involving both the N- and C-terminal domains of the OSCP [oligomycin sensitivity conferral protein] and the N-terminal regions of all three α-subunits. The structure has also provided evidence that, at the lower end of the peripheral stalk (distal from the F1-domain), the hydrophobic N-terminal region of the b-subunit is, as predicted, folded into two transmembrane α-helices, bH1 and part of bH2. It interacts with aH1 and aH2, and bH2 with aH5 and aH6, and most likely, in addition, with the loop between aH3 and aH4. It has also confirmed that the middle part of the peripheral stalk, consisting of the central α-helical pillar that is about 160 Å long and provided by bH2, is augmented by roughly parallel α-helices from subunit d and most likely from subunit h (as in the related bovine F6). Thus, it has the characteristics of a seemingly rigid and inflexible structure.

We include those details to emphasize the complexity of this machine. Without the robust stator and all its interactions, the machine could not handle the lateral forces of its high-speed operation.

But wait; there’s more! The central stalk that confers stability also flexes! Why?

The second role of the peripheral stalk is to help keep the a-subunit in contact with the rotating c-ring, possibly by exerting lateral pressure toward the central axis, thereby ensuring the integrity of the transmembrane proton pathway. Subunit ATP8 may contribute here via its C-terminal region. It is known that the longer C-terminal region of bovine ATP8 extends from the membrane into the peripheral stalk, where it interacts with subunits b and d, thereby providing another brace in addition to subunit b to hold subunit a against the rotating c-ring.

In other words, as protons enter the spinning rotor on the bottom half, the stator provides lateral pressure to keep the protons from leaking out. How well-designed must that be?

Another finding concerns “supernumerary subunits” unrelated to the catalytic activity of the machine. Some of these appear to have a role in arranging multiple motors into efficient rows in the mitochondria:

The supernumerary subunits have no known direct role in ATP synthesis, but some of them mediate interactions between monomeric F-ATPase complexes in dimers [pairs of motors] of the complex that are associated in rows along the edges of the mitochondrial cristae.

The authors found another aspect that challenges Darwinian theory. Different versions of the machine work similarly despite different amino-acid sequences (and thus DNA codes) for particular parts. This was unexpected, they say:

A comparison of the current structures of the bacterial and mitochondrial enzymes (Fig. 4) illustrates their known similarities in overall architecture and in the detailed structures of the catalytic and proton translocating regions. Somewhat unexpectedly, the peripheral stalk regions of mitochondrial and bacterial enzymes are also similar, despite significant differences in subunit composition and lack of similarity in sequence of their constituent subunits…. Thus, peripheral stalks from bacteria, chloroplasts, and eukaryotes have similar designs, and presumably similar physical properties, to allow them to perform their roles in ensuring the maintenance of an intact proton pathway in the interface between rotor and stator and in keeping the stator together.

This means that function is conserved even when the sequence is not. If natural selection could not build one irreducibly complex version of a machine, how much less could it build multiple versions?

If there is any aspect of nature that should arouse our awe, it should be to find rotary engines operating with multiple well-matched parts at high speeds at fantastically tiny scales. The details of what Walker’s team found can be explored in this open-access paper in PNAS. It doesn’t mention “evolution” anywhere, but it does mention design (see above). We hope you will look at the pictures, read the descriptions, and notice their emotional response as they say, “The structure provides more insights into the workings of this amazing machine.”