The Edge of Evolution
Page 9
Because resistance to [pyrimethamine] can be conferred by a single point mutation, it was assumed that resistance could occur frequently. However, a recent population survey demonstrated a single origin of [resistant genes] in five countries: Thailand, Myanmar, PDR Lao, Cambodia, and Vietnam.16
In other words, even though initial resistance springs up quickly and easily, and therefore mutant genes from many different malarial cells might be expected to be present in a country, only one gene from one original cell dominates a region up to a thousand miles across. How could that be?
Although the first mutation (at position 108 of the protein, as it happens) grants some resistance to the drug, the malaria is still vulnerable to larger doses. Adding more mutations (at positions 51, 59, and a few others) can increase the level of resistance. However, as usual there’s a hitch. Some of those extra mutations (but not the first one) seem to interfere with the normal work of the protein. Perhaps, though, if other mutations in other genes could compensate for these harmful effects, greater resistance could be acquired without causing harm in the process. In other words, to move to the next level of resistance after the first mutation, two further, simultaneous mutations seem to be necessary. As the scientists point out, “Because concurrent mutations in two different genes occur at reduced frequency, this would help explain the rarity with which resistance has evolved.”17 Nonetheless, because malaria grows to huge population numbers—numbers that are much greater than those of mammals or other vertebrates—it can overcome poor odds. Apparently, as for the case of chloroquine resistance, a very lucky malarial cell in one infected person acquired the several changes that gave it greater resistance to pyrimethamine while compensating for any bad side effects. That rare mutant then spread quickly through the population. That double mutant is, it seems, roughly as rare as a CCC.
A second example of what natural selection can do comes from the poor, hijacked mosquito, which involuntarily carries malaria from human to human. In 1946 the insecticide DDT was first turned against the mosquito in order to fight the disease. Taking a page from Sickle Eve’s book, mutant mosquitoes resistant to the chemical first showed up promptly in 1947. Mosquitos can resist DDT if they have mutations in their genes for enzymes whose normal job is to detoxify chemicals.
So, in the wake of the failure of DDT to control mosquitoes, other insecticides have been developed. One kind of insecticide targets an enzyme that is needed for the insect’s nervous system to work. Although the chemical had previously been used on flies, which eventually developed resistance, it hadn’t been widely used on mosquitoes, and no resistant mosquitoes had yet been discovered. To see if mosquitoes might develop resistance, some researchers deliberately altered the mosquito gene in the lab with the same mutations that made flies resistant. Sure enough, the altered mosquito gene became resistant, too. What’s more, the workers showed that only one amino acid change was needed to achieve resistance, and that adding other mutations in the right places could increase that resistance.18
Although it hasn’t yet occurred in nature, we shouldn’t be at all surprised to see resistance of mosquitoes to the new insecticides arise and spread by Darwinian processes. The necessary preconditions are all there: tiny, incremental steps—amino acid by amino acid—leading from one biological level to another.
There’s another very important lesson to be drawn from the fly/mosquito reaction to insecticides, a lesson pointing strongly to the limitations on Darwinian evolution. Mutation has to work with the pre-existing cellular machinery, so there is a very limited number of things it can do.19 Even though there are trillions upon trillions of possible simple mutations to an insect’s genome, all but a handful are irrelevant. The same few mutations pop up in organisms as disparate as mosquito and fly because no others work.
This limitation compounds the limitation noted earlier, that most mutations decrease an organism’s overall functioning—they are destructive, not constructive, even among the tiny fraction of mutations that “work.” Consider the example of the rat poison known as warfarin. It was developed in the 1950s. Warfarin interferes with the function of the blood-clotting system of mammals, so that a rat who eats it bleeds to death. Soon after warfarin was introduced, it lost effectiveness. It turns out that a change of any one of several amino acids in a certain rat protein is enough to confer resistance.20 The likelihood of one of those particular amino acids mutating is on the order of a paltry one in a hundred million. However, since there are probably at least ten times that many rats in the world, the odds of some rat somewhere having the alteration are actually very good. In fact, the resistance mutation has arisen independently about seven times in the same protein.21
Looked at a different way, however, warfarin resistance points not to the strength of random mutation, but to its limitations. Since the same mutation has been selected a number of times, even though the worldwide population of rats contains much variation in all rat proteins, this strongly suggests that the only effective mutations are ones to that single protein. What’s more, although they confer resistance to warfarin, the mutations also decrease the effectiveness of the enzyme, so it only works about half as well as the normal protein. In other words, as with many other mutations we’ve seen, the change is a net benefit only in desperate times.
FROZEN FISH
The examples of Darwinian natural selection discussed so far have all been relatively recent. Resistance to modern pesticides such as rat poison and chloroquine developed in just the past few decades. Even the mutations that first led to Sickle Eve and thalassemia occurred no more than ten thousand years ago. There are two reasons for concentrating on relatively recent examples: First, our information about them is pretty solid, and much less tainted by the flights of imagination that plague most Darwinian storytelling; and second, the recent examples are widely touted by fans of Darwin as our best examples of natural selection in action. My final example of what Darwin can do, however, is much older, and so is a lot fuzzier. We can’t easily determine the steps along the older example’s pathway or measure the advantage of each in a laboratory. Nonetheless, it seems reasonably convincing.
Over ten million years ago currents in the waters around Antarctica began to form a closed loop, circling around and around the southernmost continent. With no warmer water from other parts of the globe flowing through, the temperature of the Antarctic Ocean slowly decreased until ice formed. Because the ocean contains salt, which lowers the freezing point of water, the temperature of the liquid sea decreased below the freezing point of pure water, and then decreased below the freezing point of bodily fluids. Since fish are cold-blooded animals whose body temperature is the same as that of the water they swim in, they were in danger of freezing solid as the environment changed.
Fast forward ten million years. One group of fish, called notothenioids, flourish in the Antarctic ocean, even though the ocean temperature is below the freezing point of their blood. How can they apparently defy the laws of physics? Why aren’t they naturally frozen filets by now?
Notothenioids can flout the ice because they make some amazing proteins that literally stop water from freezing. When pure water is cooled below the freezing point it doesn’t solidify right away. That’s because a large number of water molecules first have to stick together to form tiny seed crystals. Once formed, the tiny crystals rapidly grow larger until all of the water has solidified. But if no seed crystals form, the water can stay liquid indefinitely, even below the freezing point. To make a long story short, antifreeze proteins stick to ice crystal seeds and stop them from growing. No seeds, no ice growth.
In 1997 a group of scientists at the University of Illinois sequenced the gene for an antifreeze protein from Antarctic fish. They were startled to discover so-called control regions to the left and the right of the portion of the gene that coded for the antifreeze protein that were very similar to control regions for another protein, a digestive enzyme.22 Both portions had a certain nine-letter sequence, but in the antifreeze gene th
e nine-nucleotide region was repeated many times. This gave the protein a simple sequence that consisted of three amino acids repeated many times over.23
The scientists proposed that the antifreeze protein evolved in a Darwinian fashion, by random mutations and natural selection, beginning with a duplicate copy of the digestive-enzyme gene. A probable scenario goes something like the following: The first copy of that gene simply continued its normal job. But by chance, in one of the fish in the ancient Antarctic regions, the cell’s machinery stuttered when copying the second, extra gene. That stutter gave the mutant fish several copies of the nine-nucleotide region. The altered protein serendipitously protected the fish a bit from ice crystals, and so its progeny became more numerous in the frigid ocean.
FIGURE 4.2
Schematic illustration of a possible evolutionary pathway of a simple antifreeze protein by small random mutations. A) A second copy of a digestive protein gene is produced by gene duplication. B) A nine-nucleotide region of the gene is accidentally duplicated (the small bump on the line), yielding a simple three amino acid repeat in the protein that has some antifreeze activity. C) “Stuttering” during DNA replication produces many more copies of the simple nine-nucleotide repeat (the enlarged bump), improving antifreeze properties of the protein. D) Regions of the gene that don’t contribute to antifreeze activity are accidentally deleted.
In one of the fish descendants of the original lucky mutant, presumably, the copying machinery stuttered again, adding even more nine-nucleotide repeats and further improving the antifreeze protein. (Tandemly repeated sequences in DNA are particularly prone to being copied extra times.) The progeny of that second mutant were even more fit—they could survive in water that was marginally colder—so they quickly dominated the population. Then a deletion mutation removed the original coding region, perhaps making the antifreeze protein more stable. One or two more mutations, each of which improved it, and we’ve reached the modern version of the protein.
Even though we haven’t directly observed it, the scenario seems pretty convincing as an example of Darwinian evolution by natural selection. It’s convincing because each of the steps is tiny—no bigger than the step that yielded the sickle mutation in humans—and each step is an improvement. The original duplication that started the process happens pretty frequently. The next mutation—the stuttering that led to extra copies of the nine-nucleotide sequence—also is a type that happens relatively often (remember, stuttering is the kind of mutation that leads to Huntington’s disease in humans). The next step, the deletion of the original sequence, is also not uncommon.
The likelihood of the scenario is bolstered by two other discoveries by the Illinois scientists. An unrelated group of fish from the Arctic Ocean—halfway around the world—have a gene that makes a very similar antifreeze protein, with the same repeating three amino acid residues, but which has different control regions to the left and right of the gene.24 This suggests (but of course doesn’t prove) that antifreeze proteins with the same simple repetitive sequence aren’t improbable. Even more striking is that the workers found a hybrid gene from Antarctic fish that contains both the antifreeze sequence and the digestive-enzyme sequence, which they earlier had postulated was deleted in the first gene they found.25 With the hybrid gene it really seems they had caught evolution in the act. The very kind of evolution Darwin anticipated.
SO FAR, BUT NO FURTHER
As we’ll see in the next chapter, complex interactive machinery—whether in our everyday world or in a cell—can’t be put together gradually. But some simple structures can. One example from our large world is a primitive dam. Because gunk accumulates, the drain in my family’s kitchen sink slows and stops every so often. It doesn’t much matter what makes up the garbage—bits of food, paper, big pieces and small. The gutters on our home are like that, too—pieces of different size leaves, twigs, seeds, and so on regularly plug them up. Even large rivers can get clogged by the gradual accumulation of debris. Depending on your circumstances, that might be a favorable development. Sometimes a clogged river or stream might accidentally do some animals some good if, say, it forms a reservoir. Slowing and eventually damming the flow of water doesn’t require sophisticated structures—just a lot of debris. Genetic debris can accumulate in the cell, too. If it accidentally does some good, then it can be favored by natural selection. In a sense, that’s what happened in the case of the Antarctic fish.
Rare examples such as the Antarctic fish set Darwinian pulses racing. But to more skeptical observers, they underscore the limits of random mutation rather than its potential. It turns out that the antifreeze protein in Antarctic fish is not really a discrete structure comparable to, say, hemoglobin. Hemoglobin and almost all other proteins are coded by single genes that produce proteins of definite length. They resemble precisely engineered dams. But the antifreeze protein is coded by multiple genes of different lengths, all of which produce amino acid chains that get chopped into smaller fragments of differing lengths—very much like the junk in my gutter. In fact, the Antarctic protein appears not to have any definitive structure. Its amino acid chain is floppy and unfolded, unlike the very precisely folded shapes of most proteins (such as hemoglobin). Nor does the antifreeze protein interact with other proteins like those found in real molecular machines discussed in the next chapter.
Like a dam across a stream, which can be made more and more effective by adding one stick or leaf or stone at a time, the job of the antifreeze protein is a very simple one, and it is relatively easy to improve the protein incrementally. It doesn’t much matter whether the sticks in a dam are larger or smaller, of many different types or intermixed; as long as there are enough of them, they can block the river. And just as there are many ways to dam a river, there are many ways to make antifreeze proteins. As one group of researchers points out, “A number of dissimilar proteins have adapted to the task of binding ice. This is atypical of protein evolution [my emphasis].”26
The antifreeze protein discovered in Antarctic fish is not so much a molecular machine as it is a blood additive. Another analogy might be to a machine and the lubricant that allows it to keep running. The antifreeze protein is akin to the lubricant, which, although it might be needed for the machinery of cells to work, does not have anywhere near the complexity the machines have. In fact, to survive in the cold, plants and animals frequently add simple chemicals to their fluids similar to automobile antifreeze.27 That works well, too.
Despite ten million years of evolution with quadrillions of fish under relentless, life-and-death selective pressure, the Antarctic antifreeze protein does not have anything like the sophistication and complexity even of such a simple protein as hemoglobin, let alone that of the stupendous, multiprotein systems that are plentiful in nature. Instead of pointing to greater things, as Darwinists hoped, the antifreeze protein likely marks the far border of what we can expect of random mutation in vertebrates.
To put matters in perspective, consider a related problem that has stumped malaria. Although malaria is a ferocious parasite, quite willing to eat anything that gets in its path, P. falciparum needs a warm climate to reproduce. If the temperature falls below about 65°F, the parasite slows down. When the temperature gets to 61°F, it can’t reproduce. It’s stymied.28 If a mutant parasite appeared that was tolerant to somewhat lower temperatures—not to freezing conditions, just to cool temperatures—it would be able to invade regions that are now closed to it. Despite the huge number of P. falciparum available to mutate over thousands of years, that hasn’t happened. Not all seemingly simple problems can be overcome easily, or perhaps at all.
KUDOS
Charles Darwin deserves a lot of credit. Although it had been proposed before him, he championed the idea of common descent and gathered a lot of evidence to support it. Despite some puzzles, much evidence from sequencing projects and other work points very strongly to common ancestry. Darwin also proposed the concept of random variation/natural selection. Selection does e
xplain a number of important details of life—including the development of sickle hemoglobin, drug and insecticide resistance, and cold tolerance in fish—where progress can come in tiny steps.
But, although Darwin hoped otherwise, random variation doesn’t explain the most basic features of biology. It doesn’t explain the elegant, sophisticated molecular machinery that undergirds life. To account for that—and to account for the root and thick branches of the tree of common descent—multiple coherent genetic mutations are needed. Now that we know what sorts of mutations can happen to DNA, and what random changes can produce, we can begin to do the math to find the edge of evolution with some precision.
What we’ll discover is something quite basic, yet heresy to Darwinists: Most mutations that built the great structures of life must have been nonrandom.
5
WHAT DARWINISM CAN’T DO
Debris clogging a stream shares a few things in common with the Hoover Dam. Both slow the flow of water and create large pools. Yet few people would have trouble distinguishing the two, or realizing that only one is the result of the random accumulation of twigs and leaves and mud over time. In the last chapter we looked at mutational twigs that can accumulate into a clog of biological debris. In this chapter we’ll consider molecular Hoover dams.
But first, a word about complexity. In my previous book, Darwin’s Black Box, I described certain intricate biochemical structures as “irreducibly complex” and argued that step-by-step Darwinian processes could not explain them, because they depended upon multiple parts. Critics claimed that I was simply throwing up my hands at a difficult problem, and that it would eventually be solved. They may say it again, regarding this chapter. But the discoveries of the past decade have made the problem worse, not better, both at the level of protein machinery and at the level of DNA instructions. This chapter illustrates some of the new challenges, and in the following chapters I will explain how we can generalize from them.