Hen’s Teeth and Horse’s Toes (14 page)

The body of this protist contains an axostyle (a kind of axis running the length of its body) that rotates continuously in one direction. The organelles of the anterior end (including the nucleus) are attached to the axostyle and rotate with it—“much like turning a lollipop by the stick,” as Tamm notes. But, and we now encounter the more curious and wheel-like point, the entire anterior end, including the cell surface, rotates along with the axostyle relative to the rest of the body.

Tamm demonstrated this peculiar motion with an ingenious experiment in which he attached small bacteria all over the cell’s outer surface. Those attached to the front end rotated continuously with respect to those adhering to the back end. But bacteria did not attach to a narrow band between front and back, and this band must therefore represent a zone of shear. Tamm then studied the structure of the cell-membrane by freeze-fracture electron microscopy and found it to be continuous across the shear zone. Tamm concludes that the entire surface must be fluid and that shear zones could, in theory, form anywhere upon it. A very strange creature! “Prais’d be the fathomless universe,” Whitman wrote, “for life and joy, and for objects and knowledge curious.”

THE UNCOMMON
good prose of scientists is more often spare than flowery. In my favorite example, James D. Watson and Francis Crick used less than a page to announce their structure of DNA in 1953. They began with the sparsest announcement: “We wish to suggest a structure for the salt of deoxyribose nucleic acid (D.N.A.). This structure has novel features which are of considerable biological interest.” And they ended with a reminder that they had not overlooked a major point just because they had chosen to defer its discussion: “It has not escaped our notice that the specific pairing we have postulated immediately suggests a possible copying mechanism for the genetic material” (that is, the two strands of the double helix would pull apart and each then act as a template for the reconstitution of its partner).

Francis Crick, now a professor at the Salk Institute in southern California, has continued to generate controversial, challenging hypotheses (and he has often been right). In late 1981, he published a book,
Life Itself
, advocating a theory of “directed panspermia”—the idea that Earth’s original life arrived as microorganisms dispatched by intelligent beings who chose not to make the long journey themselves. (Ten will get you fifty that he’s wrong this time—but only fifty; he’s been right too often.)

Crick has also not lost his gift for a well-turned phrase. In the presentation of his latest controversial hypothesis, published in
Nature
(April 17, 1980) with Salk colleague Leslie Orgel as first author, he outdid the last line of his 1953 paper with Watson. Orgel and Crick conclude: “The main facts are, at first sight, so odd that only a somewhat unconventional idea is likely to explain them.” Indeed, the facts are so interesting, and the wondering about them so intense, that the same issue of
Nature
carried an accompanying article by Dalhousie University biologists W. Ford Doolittle and Carmen Sapienza, who had, quite independently, devised the same explanation and argued the case, in many ways, more forcefully.

What, then, are these disturbing facts? When a younger Crick determined the structure of DNA in 1953, and others cracked the genetic code a few years later, everything seemed momentarily to fall into order. The old idea of genes as beads on a string (the chromosome) seemed to gain its vindication from the Watson-Crick model. Each of the three nucleotides in DNA codes for an amino acid (via an RNA intermediary); a string of amino acids makes a protein. Perhaps we could simply read down a chromosome to find genes lined up, one after the other, each ready to begin the assembly of its essential part.

It was not to be so. Is it ever? We now know that the genetic material of higher organisms is vastly more complex. Many genes come in pieces, separated in DNA by sequences of nucleotides that are not transcribed into RNA. Many proteins are coded by partial sequences on two or more chromosomes. What controls regulate their assembly? (Human globin, the protein component of hemoglobin, contains alpha and beta chains—and the genes for each chain are on separate chromosomes.)

Even more disturbing (and exhilarating) is the discovery, made more than a decade ago but gathering intensity ever since, that only a small percentage of DNA codes for proteins in higher organisms—and that these are the only bits of DNA whose function we may truly understand at the moment. In humans, somewhat more than 1 percent, but not as much as 2 percent, of DNA codes for proteins. Much of the rest contains sequences that are repeated over and over again—hundreds or thousands of identical (or nearly identical) beads, sometimes following one after the other, but sometimes dispersed widely over several chromosomes. Why so many copies? What do they do? The “selfish DNA” hypothesis of Doolittle, Sapienza, Orgel, and Crick provides an unusual answer to the puzzling question of why so much DNA exists in repeated copies (but I will keep you in suspense for a bit and discuss the conventional answers first).

Higher organisms contain different classes of repeated DNA. One type, called highly repeated or satellite DNA, contains short and simple sequences repeated hundreds of thousands or millions of times; 5 percent or so of human DNA falls into this class. We hardly have a clue about the origin and function of satellite DNA; neither the selfish DNA hypothesis nor the conventional hypotheses can explain it. Satellite DNA is, as they say, a “whole ’nother” story waiting to be told.

The current debate over the conventional and selfish DNA hypotheses centers upon the so-called intermediate or middle-repetitive DNA, some 15 to 30 percent of both the human and the fruit fly genome. Middle-repetitive DNA exists in tens to a few hundred copies per sequence; the copies are often widely dispersed on several chromosomes.

I have said nothing, so far, about the DNA of simpler organisms—the prokaryotic bacteria and blue-green algae, which have no nucleus and carry their DNA in a single chromosome. The DNA of prokaryote (prenucleate) organisms is “better behaved” with reference to the original hopes of the Watson-Crick model. Most bacterial DNA is single copy and protein coding, almost those beads on a string after all. But even prokaryotes are not immune to repetition. A hot topic of late concerns the presence in prokaryotes of so-called transposons, transposable elements, or more colorfully, jumping genes. These sequences of DNA, as their various names proclaim, can repeat themselves and then autonomously move about to other positions on the bacterial chromosome. They often exist in about as many copies as middle-repetitive DNA in eukaryotes (higher organisms with a nucleus and paired chromosomes). This has led many biologists to propose that at least some of the middle-repetitive DNA in higher organisms amplifies itself by the same mechanism of transposition. (The selfish DNA hypothesis assumes a correspondence between prokaryote transposons and the source of middle-repetitive DNA in eukaryotes. Some middle-repetitive DNA probably arises in other ways, and selfish DNA will therefore not explain all of it.)

Conventional arguments for the existence of middle-repetitive DNA follow the usual Darwinian perspective. Evolution is about the struggle of organisms to leave more surviving offspring in future generations. This struggle operates by natural selection and selection is a potent editor. Major features of organisms—and some 25 percent of the genetic material cannot be minor—must exist because they provide some advantages to organisms in the struggle for life. We must, in other words, find a function for middle-repetitive DNA in terms of advantages to the bodies that carry it.

Rumblings of claims for nonadaptive and nonfunctional status have been heard from time to time (selfish DNA is the first, and more subtle, explosion for this perspective). Still, as Doolittle and Sapienza detail in their article, the overwhelming majority of proposals have hewed to Darwinian orthodoxy: they assume that middle-repetitive DNA cannot exist in such amounts unless it confers direct adaptive benefits upon organisms. (I will save myself some words from now on by simply writing “repetitive DNA” when I mean “middle-repetitive DNA” only.)

The conventional adaptationist hypotheses have fallen into two classes: one, I believe, obviously wrong on (unrecognized) principle; the other undoubtedly correct in part (I do not believe that all repetitive DNA is selfish DNA). The unreasonable arguments postulate what I like to call a “retrospective significance” for repetitive DNA—that is, they justify its existence by discussing the benefits it may confer upon distant evolutionary futures.

Suppose all working genes could only exist in one copy that coded for an essential protein. How then could substantial evolutionary change ever occur? What will supply the essential protein while evolution monkeys about with the only coding sequence that produces it? But if a gene can repeat itself, then one copy might continue to code for the essential protein, leaving the other free to change. Thus, potential flexibility for evolutionary change has often been cited as the primary significance of repetitive DNA.

I have no quarrel with the idea that redundancy may supply the flexibility that evolution requires for initiating major changes. Susumu Ohno, who first popularized this idea in 1970 in a brilliant book
(Evolution by Gene Duplication)
, argued that, without redundancy “from a bacterium only numerous forms of bacteria would have emerged.” Duplication supplies the raw material of major evolutionary change: “The creation of a new gene from a redundant copy of an old gene is the most important role that gene duplication played in evolution.”

But think about it for a moment. The argument is sound and may represent, in fact, the major
effect
of gene duplication for evolution. Yet unless our usual ideas about causality are running in the wrong direction, this flexibility simply cannot be the adaptive explanation for why repetitive DNA exists. Selection works for the moment. It cannot sense what may be of use ten million years hence in a distant descendant. The duplicated gene may make future evolutionary change possible, but selection cannot preserve it unless it confers an “immediate significance.” Future utility is an important consideration in evolution, but it cannot be the explanation for current preservation. Future utilities can only be the
fortuitous effects
of other direct reasons for immediate favor.

(The confusion of
current utility
with
reasons for past historical origin
is a logical trap that has plagued evolutionary thinking from the start—see essay 11. Feathers work beautifully in flight, but the ancestors of birds must have evolved them for another reason—probably for thermoregulation—since a few feathers on the arm of a small running reptile will not induce takeoff. Our brains enlarged for a set of complex reasons, but surely not so that some of us could write essays about it. Interested readers may wish to consult a technical article that Elisabeth Vrba and I have written about this subject—see bibliography. We wish to restrict the term
adaptation
only to those structures that evolved for their current utility; those useful structures that arose for other reasons or for no conventional reason at all, and were then fortuitously available for other usages, we call exaptations. New and important genes that evolved from a repeated copy of an ancestral gene are partial exaptations, for their new usage cannot be the reason for the original duplication.)

The second set of adaptive arguments is legitimate in proposing an immediate selective benefit for repeated DNA. If genes move about and insert themselves on different chromosomes, for example, they may occasionally link up with other segments of DNA to form advantageous new combinations. More importantly, much DNA, while not coding for protein itself, may play a role in regulating the DNA that does. This regulatory DNA may turn other genes on and off and may determine the sequence and location of expression for the genes that do code for proteins. If repetitive DNA performs these regulatory functions, then its dispersal throughout the genome can have profound immediate effects. Inserted into a new chromosome, it may turn adjacent genes on and off in new ways and sequences. It may, for example, bring together the products of two genes that had never been in proximity. This new combination may benefit an organism (see the classic article of Roy Britten and Eric Davidson, 1971).

Yet, for all these efforts, the nagging suspicion remains that these adaptive explanations cannot account for all repetitive DNA. There is simply too much of it, too randomly dispersed, too apparently nonsensical in its construction, to argue that each item perseveres because natural selection has favored it in a regulatory role. The selfish DNA hypothesis proposes a fundamentally different explanation for much of this repetition. It is radical in that literal sense of getting to the roots, for it demands that we reassess some basic and usually unquestioned assumptions of evolutionary argument—what Orgel and Crick meant when they spoke of facts “so odd that only a somewhat unconventional idea is likely to explain them.”

The argument is simplicity itself once you establish the frame of mind to permit it: if repetitive DNA is transposable, then why do we need an adaptive explanation for it at all (at least in conventional terms of benefits to bodies)? It may simply spread of its own accord from chromosome to chromosome, making more copies of itself while other “sedentary” genes cannot. These extra copies may persist, not because they confer advantages upon bodies, but for precisely the opposite reason—because bodies do not notice them. If they have no effect upon bodies, if they are (in this sense) “junk,” then what is to stop their spread? They are merely playing Darwin’s game, but at the “wrong” level. We usually think of natural selection as a struggle among bodies to leave more surviving offspring. Here certain genes have found a way, through transposability, or “jumping,” to leave more copies of themselves
within
a body. Is any other explanation required? Orgel and Crick’s title reflects this reversed perspective: “Selfish DNA: The Ultimate Parasite.”

I can now almost hear the disappointment and anger of some readers: “That bastard Gould. He led us along for pages, and now he gives an explanation that is no explanation at all. It just plain happens, and that’s all there is to it. Is this a joke or a counsel of despair?” I beg to differ from this not entirely hypothetical adversary (a composite constructed from several real responses I have received to verbal descriptions of the selfish DNA hypothesis). The explanation seems hokey only in the context of adherence to traditional views that all important features must be adaptations and that bodies are
the
agent of Darwinian processes. The radical content of selfish DNA is not the explanation itself, but the reformulated perspective that must be assimilated before the explanation confers any satisfaction.

If bodies are the only “individuals” that count in evolution, then selfish DNA is unsatisfying because it does nothing for bodies and can only be seen as random with respect to bodies. But why should bodies occupy such a central and privileged position in evolutionary theory? To be sure, selection can only work on discrete individuals with inherited continuity from ancestor to descendant. But are bodies the only kind of legitimate individuals in biology? Might there not be a hierarchy of individuals, with legitimate categories both above and below bodies: genes below, species above? (I confess to what evolutionists call a “preadaptation” for favorable response to the selfish DNA hypothesis. I have long argued that species must be viewed as true evolutionary units and that macroevolutionary trends are often powered by a “species selection” that is analogous to, but not identical with, natural selection acting upon bodies.) Selfish DNA may do nothing for bodies, but bodies are the wrong level of analysis. From a gene’s point of view, transposable elements have developed a great Darwinian innovation: they have found a way to make more surviving copies of themselves (by repetition and transposition), and this, in itself, is the evolutionary
summum bonum
. If bodies don’t notice this repetition, and therefore cannot suppress it by dying or failing to reproduce, then so much the better for repeating genes.