Behe and the Blood Clotting Cascade
Post of the Month: February 1997
by George Acton
re-read Behe's presentation of clotting a few days ago, and found it even more irritating than the first time. It breaks down into (1) a presentation of the clotting mechanism, (2) a critique of a long quote from Doolittle, and (3) a conclusion.
If you try to explain the clottting system to people unfamiliar with it, you begin with something like:
contact factors | XII ---> XIIa | XI ---> XIa EXTRINSIC PATHWAY | IX ----> IXa, V, VIII VII + TF (tissue factor) \ / \ / X -----------------> Xa prothr. ---------> thrombin | fibrinogen -----> fibrin
This leaves out many important facts, but it shows that the essence of the system is a sequence of protease reactions. It reflects what happens in vitro in the two common clinical assays the "PT" which measures the extrinsic pathway, and the "PTT" which measures the intrinsic pathway. The components of the "intrinsic pathway" are all in plasma; the extinsic pathway requires TF = "tissue factor" which is extravascular.
A more current view of the physiology emphasizes the importance of TF in initiating the cascade (Furie & Furie, NEJM 326:801, 1992). This makes sense, because traumatic rupture of a blood vessel exposes the plasma to tissue factor.
TF T | XI ---> XIa VII ---> VIIa ---------------- VIII ---> VIIIa | | V ---> Va XIa | | | | | IX ---------> IXa + VIIIa | | | X -----------------> Xa + Va | prothrombin ---------> thrombin (T) | fibrinogen -----> fibrin
Note that in this scheme, there is a single cascade of 4 or 5 serine protease enzymes, two non-protease cofactors (VIII and V) homologous with each other and the final substrate, fibrinogen. The only complications are that Factor VII may act downstream to activate X, and thrombin acts to activate 3 earlier factors. A low basal level of activated thrombin keeps the system primed for exposure to TF.
In this view, 3 factors included in the older scheme (Factor XII = Hageman, prekallikrein and HMK) are now excluded, since deficiencies do not cause clinical disease, although they are associated with long clotting times in vitro. The role of XI isn't clear, since deficieny isn't invariably associated with disease.
The current scheme of the clotting system raises questions about Behe's concept of an "irreducibly complex system", which obviously rests on a reliably objective way to define the "system". Yet Behe flatly declares that the clotting system is irreducibly complex. One wonders *which* "clotting system" he's talking about.
Behe's approach of using the earlier nomenclature for the clotting factors is gratuitously forbidding to people unfamiliar with the subject. It has the effect of telling them that a moderately complex system is totally impenetrable, rather than that it is approachable and comprehsnsible. It's as annoying as watching someone try to present arithmetic to schoolchildren using Roman numeral notation.
The NEJM review has schematics of the factors, showing the many homologies of domains. The 5 proteases are each composed of a single peptide chain which is almost exactly the same length in all, and the point at which the chain is cut in the activation process is in exactly the same relative position in all 5. Eight defined domains or motifs, e.g. the serine protease domain, are located in exactly the same positions in 4 of the 5 protease factors. In other words, at this low level of detail all four are 100% homologous.
This immediately suggests to someone with an exposure to introductory molecular biology an explanatory mechanism (gene duplication) and this in turn predicts a huge number of other homologies. Some known or predicted homologies include:
(1) neutral point mutations at the level of sequence data,
(2) arrangement of domains,
(3) 3D shape from crystallographic data,
(4) intron-exon organization of the genes,
(5) conserved chromosomal location of the genes in related species,
(6) homologies to related proteins outside the clotting system,
e.g. serine proteases,
serine protease inhibitors,
proteins modified by vitamin K-dependant enzymes.
The evolutionary hypothesis is highly restrictive and parsimonious in what the data must show. For instance, each and every mammalian sequence for each and every gene in the system must fit into a cladistic tree, and the tree must be the same for all the genes. The homologous genes must fit into a tree based on the four reduplication events that gave rise to them. In other words, of all the possible sequences we might find, only a tiny proportion can occur. An unexplained deviation from these sequence predictions has the same significance as an unexplained human footprint in the Jurassic.
Behe incorrectly implies that the homologies are few in number and confined to primary structure. He dismisses the vast amount of homology data by saying that it has to be meaningless, since the system is obviously "irreducibly complex" and that homology data is useless for telling us how the system arose. This is really no different from a Creationist rejecting all sequence homology data for evolution on the a priori grounds that it's illegitimate to compare different "kinds".
Note the brittleness of Behe's position. Once it is admitted that gene duplication, recombination and specialization can explain one elaboration of the system -- say two sequential proteases -- it's difficult not to apply them to others, like the obvious similarities between Factor V and Factor VIII. There's no rational place to draw the line and say that we've reduced the complexity of the system to an irreducible level.
It's puzzling why Behe leaves out the obvious homology with the serine protease digestive enzymes. Standard texts such as that of Lubert Stryer have stressed this point. Surely someone with Behe's academic credentials knows about this material. The effect of Behe's presentation after using Stryer is like reading a discussion of the Shroud of Turin that totaly omits any mention of the radiocarbon dating studies.
His use of the Doolittle quote is inappropriate. Doolittle was sketching a scheme by which the inhibitory components of the clotting system could evolve in parallel with the cascade. When he said that something "arose", he meant "arose by a process of gene duplicaton and specialization", but Behe implies that this means "arose ex vacuo". The effect and intent is the same as the Darwin quote on the eye so beloved of Creationists.
His conclusion that "nobody knows how it arose" is highly reminiscent of the standard ending of a tabloid piece on sensational "discoveries". The ending is never that a new theory of space-time dislocation is being developed to explain the B-52 on the moon. It's that "scientists are baffled". It's reasonable to suppose that Behe is writing for a National Enquirer audience, since he's left behind anyone who knows something about the details of the clotting system or what Doolittle is talking about. It's highly disingenuous of him to violate norms of presentation to a scientific audience and protest when serious scientists reject his work. In one admittedly personal view, it is doing a deep disservice to a National Enquirer audience to write an account of science for them that encourages National Enquirer attitudes, and that is what Behe has done.
This isn't a respectable performance. The only question I had after reading it was whether Behe really believes that the clotting system offers support for "irreducible complexity" and is kidding himself, or whether it was a deliberate effort to ignore and distort the evidence. And on the latter assumption, if he wants to promote an ideology or sell copies of his book and increase his speaking fees. With Johnson I think it's conscious dishonesty and a mixture of commercial and ideological motives. But I didn't read enough of Behe to be able to decide.
Article originally posted February 18, 1997
Revision posted March 5, 1997
Home Page | Browse | Search | Feedback | Links