More Articles on Evolution
A Delicate Balance
Russell F. Doolittle
Many years ago, when I was a graduate student in biochemistry at Harvard,
I entered an essay in a prize competition in which submissions were made anonymously
by pseudonym. The essay was entitled "The Evolution of a Unique Enzyme System:
The Comparative Physiology of Blood Coagulation," and the immodest pseudonym
I used was Charles Darwin.
The gist of the essay was that, whereas vertebrate blood coagulation is an
extremely complex process, and although at first glance no part of the system
ought to be viable without the entire ensemble, it nonetheless ought to prove
understandable in terms of natural selection. I pointed out that it was unlikely
that the entire melange of enzymes and protein substrates evolved in one fell
swoop. Instead, three processes had been at work. First, there was a series
of gene duplications of the sort that had recently been observed for hemoglobins.
Second, there were the simple point mutations we know today as amino acid
replacements. Finally, mechanisms were brought into play that controlled the
amounts of the various homologous factors. I suggested that the presence and
role of these three mechanisms could be evaluated by comparing blood clotting
in various organisms, particularly earlier diverging animals that might have
simpler systems. To this end, I began an experimental program dealing with
blood clotting in all sorts of creatures, wrote my Ph.D. dissertation on the
subject,1 and, indeed, have devoted the intervening 35
years to the general subject of proteins and their evolution.
Now it appears that I have wasted my career. In Darwin's Black Box,
Michael Behe has concluded that blood clotting--Behe's "favorite pathway,"
as Allen Orr puts it--is simply "too complex to have evolved."2 Worse, he has taken one of my own articles to illustrate his
view. The article was the text of a 1993 lecture presented at an international
conference on blood clotting.3 It was one of a series of
talks that were billed as "state of the art" and presented to an audience
of mainly clinicians and biotechnologists. Because the audience was hardly
conversant in the facts of evolution, my tone was intentionally light and
breezy, and my language was casual. The main point was to demonstrate that
the delicate balance of forward and reverse reactions that regulate blood
clotting came about in a step-by-step fashion. I summarized events with the
metaphor of Yin and Yang, and emphasized that other similar point-and-counterpoint
comparisons could be made.
But Behe had a ball with Yin and Yang. While reminding readers over and again
that this was a "state of the art" article, he accuses me of "imagining" the
evolution of blood clotting and "papering over the dilemma [of the irreducibly
complex] with a hail of metaphorical references to yin and yang." He ridicules
the whole thing as a "Calvin and Hobbes" creation. He concludes that "no one
on earth has the vaguest idea how the coagulation cascade came to be."
I beg to differ. In recent years an enormous amount of evidence has been
accumulated about the evolution of blood clotting, and it overwhelmingly supports
the suggestions made in my graduate essay. In this brief comment, I propose
to sketch the basic story.
To begin, we need some basic concepts from molecular biology. Thus, DNA is
composed of very long strings of four biochemical unit called "bases" (abbreviated
A, G, C and T). The linear order ("sequence") of these bases encodes, in a
rather indirect way, the order of another kind of unit in other kinds of molecular
chains called "proteins." The fundamental units in proteins are amino acids,
of which there are twenty. We can determine the sequence of amino acids in
any protein, either directly or by decoding the DNA sequence of its gene,
and compare it with any other. Then, we can arrange proteins into big family
trees according to how similar their amino acid sequences are. Generally speaking,
the closer two organisms are, the more similar are the amino acid sequences
of their proteins. For example, most human and chimpanzee proteins are between
99 and 100 percent identical, but the same proteins from bacteria may range
from 30 to 60 percent identical with ours. We should also take note of the
fact that humans and chimpanzees have a great deal more DNA per cell than
bacteria, and many more genes.
Drawing on these fundamentals of protein chemistry, we can see how the inventory
of genes (and the proteins they encode) has expanded over the eons. Briefly
put, the genes for new proteins come from the genes for old ones by gene duplication,
a process I like to call "biochemical xerox." (Behe will doubtless find that
metaphor charming but simplistic.) These new proteins, in turn, are especially
useful for adapting to new situations: but that's getting ahead of the story.
In his book, Behe notes that it has been "theorized" that similar amino acid
sequences in different proteins might be due to gene duplications, but--as
Allen Orr points out--he refers to this as "a hypothesis," and implies that
such interpretations of events are "just so stories," made up after the fact
to rationalize the observations.
Actually, the process of gene duplication can occur in a number of ways,
and the most common mechanisms are well understood. Sexual organisms, for
example, have two sets of chromosomes (one from each parent) which line up
during the cell division process called meiosis. As it happens, the very long
DNA threads are constantly breaking and being rejoined. The rejoining process
is not 100 percent accurate, however, and often one of the chromosomes comes
away with a little more of the DNA than its pairee, which will have correspondingly
less. The lucky gametes that come away with the more are said to have had
a "gene duplication," although the amount of DNA may amount to only a part
of a gene or maybe a whole string of genes. The process can be seen in action
in that there are people who have certain diseases as a result of having pieces
of genes missing and other people, usually healthy, who have exactly the missing
Thus, the result of such gene duplication is that a creature may have an
old gene that encodes some protein and a new one that, under normal circumstances,
has nothing much to do. Most of the time one of the duplicates will simply
wither away as a result of the relentless rain of amino acid replacements
that are constantly being inflicted on all proteins; natural selection cannot,
after all, operate on idle proteins, but only on those that are being used.
Occasionally, however, the occurrence of a new protein can be of fortuitous
advantage, and it is preserved: we already have a very long list of proteins
that are clearly the products of gene duplications. Indeed, one of the major
goals of molecular evolutionists is to trace the family pedigree of proteins
back through time in an effort to identify the small number of genes that
must have been present in the earliest organisms.
Consider hemoglobin, a protein Dr. Behe has worked on professionally, and
that he discusses in his book. Almost everybody knows that hemoglobin is the
protein packed into red blood cells that carries oxygen around to the tissues.
Behe notes that it consists of two different types of protein chain. He calls
them "analogous," steadfastly refusing to call them "homologous"--a term that
indicates common ancestry, and that everyone else uses. Certainly no thinking
biochemist doubts that these two chains, referred to as "alpha" and "beta,"
are the results of a gene duplication. They are composed of 141 and 146 amino
acid units, respectively, and 63 of them are exactly the same, which is to
say their amino acid sequences are about 45 percent identical.
It is also well known that the foetus has a different hemoglobin in its red
cells. The alpha chains are the same as the "adult" kind, but the other chain
comes from another duplicated gene called "gamma." The gamma chain is also
45 percent identical with the alpha, but 70 percent identical to the beta
(they share 107 amino acid units). Clearly, the gamma chain has shared ancestry
more recently with the beta than it has with the alpha. It also has one very
advantageous physiological property: when combined with the alpha chain it
binds oxygen more tightly than does the adult hemoglobin. As a result, the
foetus, which won't breathe on its own until birth, is insured of the flow
of oxygen moving in its direction from the maternal circulation. As it happens,
humans actually have several genes for hemoglobins, some being expressed only
at embryonic stages, and one only in tissues.
We can make another family tree from hemoglobin sequences by using species
comparisons instead of the duplicated genes. The tree could be based on alpha
or beta hemoglobins, for example. And when we do that something interesting
is observed. Because the rates of change in sequence are fairly uniform, we
can gauge when the gene duplications occurred that gave rise to the alpha,
beta, and gamma chains, as well as the others. It is apparent that earlier
diverging animals ought not to have all the hemoglobin genes that humans have,
because they diverged before particular duplications occurred. In fact, we
know that jawless fish, which are the most primitive vertebrates extant, have
single-chained hemoglobins in their red blood cells, because they diverged
before the pivotal duplication that separated the alpha and beta chains.
This same kind of scenario can be reconstructed for a host of other physiological
processes, including blood clotting. The availability of amino acid sequence
data from all the different clotting factors from various species lets us
gauge when these duplications took place. Unlike the hemoglobins, however,
many blood clotting proteins are also embellished as a result of a process
called "exon shuffling." This is a phenomenon by which structurally stable
parts of proteins are genetically swapped around at the DNA level. The mechanism
is similar to those that occur during ordinary gene duplications. The result
is that many different proteins may have some parts that are similar even
while other parts are not. Because of their intrinsic affinities, such mosaic
proteins are especially useful in creating networks of interactions or "cascades."
Historically, many important "theories" or generalizations have been accepted
only after certain predictions were made and then fulfilled. For example,
when Mendeleef proposed the Periodic Table, he predicted the existence of
two missing elements, germanium and gallium; those elements were in fact discovered
several years later. Einstein's 1915 Theory of General Relativity predicted
the degree to which light waves should be attracted by massive bodies, but
it wasn't until 1919, when light from appropriately situated stars was visualized
during a total solar eclipse, that the prediction could be tested.
On a more modest plane, about ten years ago we predicted that certain of
the genes encoding the blood clotting cascade would be absent in jawless fish.5
This prediction was made on the basis of comparing the sequences of blood
clotting factors in mammals and estimating how long it had been since the
gene duplications leading to their existence. In particular, we noted that
fish should not have both Hageman Factor and prekallekrin, two of the factors
described in Behe's outline of blood clotting in his book.
As far as I know, a study aimed at establishing whether these clotting proteins
are present in lampreys and hagfish has not yet been undertaken, but I am
willing to wager a goodly sum about the outcome. What I want to know, however,
is whether Behe will accept such a result as a proof of the concept, or whether
he will--in typical creationist style--simply try to find a way out.
In fact, Behe employs a variety of creationist arguments that have been abused
in the past. Of all these, the "improbability argument" is the most fallacious
and most misunderstood. Thus, Behe writes about the likelihood of assembling
appropriate combinations of parts of the proteins that function in a clotting
scheme: "Doolittle apparently needs to shuffle and deal himself a number of
perfect bridge hands to win the game. Unfortunately, the universe doesn't
have time to wait." This statement follows upon some absurd arithmetic about
possible combinations of shuffled units and comparisons to the Irish sweepstakes.
His argument omits many relevant considerations: for example, most of the
observed duplications and exon shuffling are localized to specific regions
of specific chromosomes, so the number of combinations is not as large as
he supposes. Its major fallacy, however, lies in the presumption that some
special combination must be achieved. When we play bridge, any specified hand
is just as unlikely as a perfect hand. But someone wins every time, whether
or not they have a perfect hand.
This point about "perfect hands" brings me to what annoyed me most in Behe's
book; his use of Rube Goldberg cartoons. Ironically, I have often used Goldberg's
contrived linkages as examples of how evolution works! In fact, I have used
them in teaching medical students about how macromolecular cascades function.
I have also used the same cartoons in debating creationists, pointing out
that no Creator would have designed such a circuitous and contrived system.
Instead, this is how the opportunistic hand of natural selection works, using
whatever happens to be available at the moment (the resources that result
from such processes as gene duplication and exon shuffling).
Let me conclude by mentioning that support for the Yin and Yang scenario
is now coming from another quarter. Thus, it has become possible during the
last decade to "knock out" genes in experimental organisms. "Knockout mice"
are now a common (but expensive) tool in the armamentarium of those scientists
anxious to cure the world's ills. Recently the gene for plaminogen was knocked
out of mice, and, predictably, those mice had thrombotic complications because
fibrin clots could not be cleared away. Not long after that, the same workers
knocked out the gene for fibrinogen in another line of mice. Again, predictably,
these mice were ailing, although in this case hemorrhage was the problem.
And what do you think happened when these two lines of mice were crossed?
For all practical purposes, the mice lacking both genes were normal!6
Contrary to claims about irreducible complexity, the entire ensemble of proteins
is not needed. Music and harmony can arise from a smaller orchestra.
No one doubts that mice deprived of these two genes would be compromised in
the wild, but the mere fact that they appear normal in the laboratory setting
is a striking example of the point and counterpoint, step-by-step scenario
1 R. F. Doolittle, "The Comparative Biochemistry of Blood
Coagulation" (Ph.D. thesis, Harvard University, 1961).
2 Michael J. Behe, Darwin's Black Box: The Biochemical
Challenge to Evolution (New York: The Free Press, 1996).
3 R. F. Doolittle, "The Evolution of Vertebrate Blood Coagulation:
A Case of Yin and Yang," Thrombosis Haemostasis 70 (1993): 24-28.
4 See, for example, H. Lehmann and D. Charlesworth, "Observations
on Haemoglobin P," Biochemical Journal 119 (1970): 43.
5 R. F. Doolittle and D. F. Feng, "Reconstructing the History
of Vertebrate Blood Coagulation from a Consideration of the Amino Acid Sequences
of Clotting Proteins," Cold Spring Harbor Symposium on Quantitative Biology
52 (1987): 869-74.
6 Bugge et al., "Loss of Fibrinogen Rescues Mice from the
Pleiotropic Effects of Plasminogen Deficiency," Cell 87 (1996): 709-19.