How to Explain Irreducible Complexity

The journal Science published a paper today (hyped in both the New York Times and the Wall Street Journal) purporting to show the Darwinian evolution of an irreducibly complex biological system. Three University of Oregon researchers, Jamie Bridgham, Sean Carroll and Joe Thornton (hereafter Bridgham et al.) claim to have shown how an irreducibly complex system might have arisen as the result of gene duplication and a few point mutational changes — a process they call “molecular exploitation.” In particular, they claim to have shown how the “functional interaction of the steroid hormone aldosterone and its partner, the mineralcorticoid receptor [MR], evolved in a stepwise Darwinian fashion.” Hormones are small molecules (ligands) that bind to receptors. Receptors are proteins that receive hormones and trigger signaling cascades within the cell, thereby regulating specific physiological or genetic processes.

Since hormones and receptors must work in concert with one another, the authors suggest that their work provides an example of undirected mutation and natural selection producing an “irreducibly complex” system, of the type described by Michael Behe in his book Darwin’s Black Box. 

We think this breakthrough paper should be used as a model for other such efforts in the future. Here, then, is an easy-to-follow lab manual, drawing on the example of Bridgham et al., for explaining irreducible complexity.

Step One. Find something that is not irreducibly complex, and explain that, sort of.

Let’s begin with Mike Behe’s original definition of an irreducibly complex system:

By irreducibly complex I mean a single system composed of several well-matched, interacting parts that contribute to the basic function, wherein the removal of any one of the parts causes the system to effectively cease functioning.

Darwin’s Black Box [1996], p. 39

What is the irreducibly complex system that Bridgham et al. wish to explain? In particular,

(a) What is the system’s function?
(b) What are the system’s “several well-matched, interacting parts?”
(c) What happens when one of those parts is removed?

A logical point: since irreducible complexity [IC] as a biological phenomenon is defined by criteria a, b, and c, any claim to have demonstrated the stepwise (Darwinian) pathway to an IC system must begin by describing that system in terms of (a) its function, (b) its parts list, and (c) the consequences when one of the parts is removed.

Do Bridgham et al. provide answers to (a), (b), or (c)?

No — most tellingly, not even on Darwinian grounds. What they do say, however, is biologically meaningless.

A Tutorial in Evolutionary Theory

To understand why, we need a brief primer in fundamental evolutionary theory. Natural selection preserves randomly-arising variations only if those variations cause functional differences affecting reproductive output. Since Bridgham et al. tell their story by invoking natural selection (see below), the system whose origin they claim to explain must have a selectable function for it to qualify as irreducibly complex. Indeed, given that natural selection favors only functionally advantageous variations, Behe has made clear that “function” in a biological context necessarily means a selectable functional advantage, for an obvious reason: a system of well-matched parts that performs a function can’t lose that function unless it possesses one to begin with. Unfortunately, these receptor-ligand pairs do not meet Behe’s definition of irreducible complexity for an equally obvious reason: receptor-ligand pairs do not by themselves confer any selective functional advantage.

Indeed, in Bridgham et al.’s scenario, the function undergoing natural selection is not simply MR-aldosterone binding, but electrolyte homeostasis, the complex physiological regulation of essential cellular ions such as potassium or calcium. The novel receptor MR evolved, they write, “because it allowed electrolyte homeostasis to be controlled” (p. 100).

Natural selection is acting, therefore, not on MR-aldosterone binding alone. Indeed, it cannot, because unspecified binding confers no functional advantage.

But that is what Bridgham et al. do not seem to understand. They think they are explaining the origin of a single receptor-ligand pair, the mineralocorticoid receptor (MR) protein and the steroid hormone aldosterone. But that is biological nonsense. It is nonsense, moreover, strictly on the grounds of evolutionary theory itself.

Let’s suppose the newly-evolved cellular receptor, MR, interacts with a hormone ligand, aldosterone. This is a novel relationship. Now, will natural selection preserve it?

Who knows? Without more information — that is, without more details about the cellular or organismal effect of that novel binding — the bare function “aldosterone binds to MR” is biologically vacuous.

Compare: Pound a nail, we tell you. Where and why? you ask. Never mind that, we say, just go pound a nail. So you hammer a threepenny nail through the power supply of this blog’s server.

In any case, the receptor-ligand pair by itself is certainly not irreducibly complex. These pairs represent only small components of complex physiological processes such as metabolism, inflammation, immunity, and electrolyte homeostasis. For such pairs to have any selective advantage as part of the regulation of larger physiological processes, many other protein components have to be present. In particular, all the other components of a complete signal transduction circuit have to be present, as well as the component parts of the physiological process that such circuits regulate. (Even the ligand aldosterone itself doesn’t exist apart from a separate enzyme that produces it, and Bridgham et al.’s gene duplication scenario does not account for the origin of this necessary component either.) 

Bridgham et al. appear to grasp the need for more details (albeit in a distressingly loose way) because both early and late in their paper they specify the functional role of MR. The receptor “is activated by aldosterone to control electrolyte homeostasis” (p. 97) they note, and evolved “because it allowed electrolyte homeostasis to be controlled” (p. 100).

Thus, in Bridgham et al.’s scenario, the actual system undergoing natural selection is electrolyte homeostasis, not simply MR-aldosterone binding. There’s a good reason for that: as noted, the function “aldosterone-MR binding,” considered in isolation, cannot be a target for natural selection. Try it, if you think it can. You’ll quickly find that you are floating in biological limbo. Aldosterone binds to MR…MR interacts with aldosterone…MR and aldosterone…OK, enough of that. Why does MR interact with aldosterone? Hello? Can we get an organism here?

Back to Biological Reality

So — is the physiological system of electrolyte homeostasis, of which both MR and aldosterone are small parts, irreducibly complex? Maybe. Take a look at a physiology textbook, or even any review paper on steroid or receptor biochemistry. Bridgham et al. don’t say much about the complexity of electrolyte homeostasis, however, because they are unaware that they have completely misunderstood the relevant unit of selection in their scenario. They write (p. 98):

It is not obvious how the tight aldosterone-MR partnership could have evolved. If the hormone is not yet present, how can selection drive the receptor’s affinity for it? Conversely, without the receptor, what selection pressure could guide the evolution of the ligand?

By Bridgham et al.’s own account, however — although they don’t realize it — natural selection is not acting at this level (the MR-aldosterone relationship alone) at all. To have any selectable function, many more components need to brought into the story. Genuine irreducible complexity re-emerges, and will be quite unexplained by the Bridgham et al. scenario.

For more information go here.