Rapidly accumulating data from genome sequencing projects have allowed scientists to look at the many different ways DNA can change. In other words, only in recent decades have we been able to examine the kinds of variations—mutations—that can spring up in a genome. The cellular machinery that replicates DNA is extremely faithful. In people and other multicellular organisms it makes only about one mistake in every hundred million nucleotides of DNA it copies in a generation. Yet since the number of nucleotides in a cell’s genome is on the order of millions to billions, on a per-cell basis, mistakes actually happen pretty often. On average, depending on the kind of organism and how much DNA it has, a mutation happens at a rate from about once every hundred cells to ten mutations per cell. If DNA were exactly like a blueprint, with no wasted space, and every line and curve representing a point of building, then this mutation rate would be fatal. After all, one critical mistake is all it takes to kill (or cause the building to collapse). But in fact, DNA isn’t exactly like a blueprint. Only a fraction of its sections are directly involved in creating proteins and building life. Most of it seems to be excess DNA, where mutations can occur harmlessly.
Mutations come in different flavors. When Sickle Eve was conceived, one copy of the DNA section that served as a blueprint for the beta chain of her hemoglobin was altered, so that a single amino acid was substituted for another. That, unsurprisingly, is called a substitution mutation—a straightforward switch, where one single letter of the billions in DNA is traded for another. A single-letter substitution often leads to a change in a protein amino acid sequence, as it did with Sickle Eve, but not always.3
There are other kinds of mutations, too. One class is called deletion mutations. As the name implies, deletion mutations occur when a portion of DNA, ranging from a single letter to a large chunk of the genome, is accidentally left out when the DNA is duplicated. For example, some people have thirty-two nucleotides (letters) deleted in a gene for a protein called CCR5. Blessedly, the mutant gene confers resistance to HIV, the virus that causes AIDS in humans. The opposite of a deletion mutation is an insertion mutation. This happens when extra DNA is accidentally placed into a region. People who suffer from Huntington’s disease (such as 1930s folk singer Woody Guthrie) have many extra copies of a particular three-nucleotide segment (C-A-G) in the gene for a protein called huntingtin.4 Like deletions, insertions can range from just one letter to many. Sometimes the insertion happens because the molecular machinery copying DNA “stutters,” backs up, and recopies a region it has just copied, so that a piece of DNA is copied twice. Other times a large piece of DNA (thousands of nucleotides in length) from an active element from one region of DNA copies itself into the genome where it hadn’t been before.
A special kind of insertion occurs when the extra DNA comes from a different organism. Viruses are small scraps of genetic material, either DNA or RNA, that invade cells and use the cells’ resources to copy themselves. Sometimes they insert their own DNA into the host genome, where it can remain indefinitely. Other times, while a virus is replicating its own genome, a piece of the cell’s genetic material accidentally gets picked up and added to that of the virus. If the extra material does the virus more good than harm, it can become a permanent part of the viral genome.
Another kind of mutation is called an inversion. When some of the normal machinery of the cell goes slightly awry, a piece from the DNA double helix can be cut out, flipped over, and stitched back in. This sort of mutation is thought to help divide one species into two species. Organisms with inverted regions in their DNA can mate with each other, but they often cannot mate as successfully with their “unflipped” cousins. One species of mosquito that carries malaria in west Africa seems to be dividing into several separate species because of large genomic inversions.5
Another type of mutation, thought by Darwinists to be especially consequential, is gene duplication. Occasionally an entire gene or set of genes gets copied twice on a chromosome, so the mutant organism now has two or more copies of a gene where its kin have only one. For example, laboratory resistance to chloroquine was seen in some malarial cells that mutated extra copies of segments of the parasite’s chromosome 3.6 When genes accidentally duplicate, evolution has a golden opportunity. Now one copy of the gene can continue to take care of its original job, while the second, spare copy of the gene is free to be used for a different job. We’ll see later on that, although gene duplication can help in limited circumstances, like Darwinian processes in general it doesn’t take us very far.
How often do mutations occur? Any one particular nucleotide (like, say, the one that will give the sickle mutation) is freshly substituted about once every hundred million births.7 Small insertion and deletion mutations pop up roughly at the same rate. Gene duplications also seem to occur at about the same frequency.8 So if the population size of a species is a hundred million, then on average each and every nucleotide is substituted in some youngster in each generation, and each gene is also duplicated in someone, somewhere. And so on. On the other hand, if the population size is only a hundred thousand, it would take a thousand generations for a duplicate of a particular gene or a particular nucleotide substitution to arise (on average)—because that’s how long it would take to reproduce a hundred million organisms.
A word of caution. Although substitutions, insertions, deletions, and duplications all happen roughly at the same rate, there is a critical distinction between breaking something old and building something new. It’s always easier and faster to blow up a bridge than to build one. For example, in human history a new sickle mutation (not one that was just inherited from a parent who had it) has freshly arisen at most only a few times, perhaps just once. Yet thalassemia has popped up hundreds of times. The reason for the difference in the numbers is easy to see. To get sickle, one particular nucleotide has to be substituted. To get thalassemia (which breaks a hemoglobin gene), on the other hand, any of hundreds of nucleotides can be substituted or deleted and the gene will no longer produce a working protein. Any of a large number of substitutions or deletions will suffice. In general, then, mutations that help in trench warfare by breaking something will appear at a rate hundreds of times faster than ones that help by doing something new.
YOU CAN PICK YOUR FRIENDS, BUT…
Scientific work in earlier centuries first noted the remarkable anatomical similarities between humans and other primate species. With the advent of modern biology, the sequences of their protein and DNA could also be compared.
One of the side benefits of our new understanding of DNA is that scientists can often use it to figure out who is related to whom. For example, DNA tests can establish paternity in disputed cases, or determine which side of the family a genetic disease has come from. This can infer relationships not only among modern humans, but with ancient ones, too. By comparing protein and DNA sequences, the origin of Sickle Eve can be pinpointed with reasonable accuracy. In the 1980s scientists compared data from modern humans and proposed the hypothesis of “Mitochondrial Eve”—that all modern humans are descended from a single woman who lived perhaps a hundred thousand years ago.
Although it is trickier and depends on more assumptions, the same general sorts of methods and reasoning that establish relationships among modern humans, and between modern and ancient humans, are also used to figure out how different species are related to each other. If two kinds of organisms share what seems to be a common mutation or set of mutations in their DNA, it can be assumed that a common ancestor of the two species originally suffered the mutation, and the descendants simply inherited it. Admittedly, assumptions are involved, but they strike many people as reasonable.
In the early 1960s the first sequences of proteins became available. Scientists were shocked. Many had expected the biological molecules of different organisms to be completely different. But the molecules often turned out to be similar in a very suggestive way. For example, one of the first proteins to be sequenced from a wide variety of organisms wa
s hemoglobin. The sequence of hemoglobin in various species reflected the biological classification system that had been set up centuries earlier. The amino acid sequence of the beta chain of human hemoglobin was much different from that of fish, somewhat different from that of kangaroo (a marsupial mammal), pretty similar to that of dog (a placental mammal), and identical to that of chimpanzee.9 The protein pattern fit wonderfully with Darwin’s image of a branching tree of life. Not only hemoglobin, but many other molecular similarities were discovered between humans and other primates and, more broadly, underlying all of life.
One serious objection might be raised. Perhaps the different animals all had similar hemoglobin because that’s the only protein that could really work to carry oxygen efficiently. Just as all organisms have to be based on carbon, because carbon is the only element versatile enough for life, perhaps all animals simply have to have certain similarities in their molecular machinery. So by necessity any large animal would have to have a protein similar to hemoglobin, even if it arose separately.
That objection, however, doesn’t hold for a feature shared between two organisms that has no functional role to play. When two lineages share what appears to be an arbitrary genetic accident, the case for common descent becomes compelling, just as the case for plagiarism becomes overpowering when one writer makes the same unusual misspellings of another, within a copy of the same words. That sort of evidence is seen in the genomes of humans and chimpanzees. For example, both humans and chimps have a broken copy of a gene that in other mammals helps make vitamin C. As a result, neither humans nor chimps can make their own vitamin C. If an ancestor of the two species originally sustained the mutation and then passed it to both descendant species, that would neatly explain the situation.
More compelling evidence for the shared ancestry of humans and other primates comes from their hemoglobin—not just their working hemoglobin, but a broken hemoglobin gene, too.10 In one region of our genomes humans have five genes for proteins that act at various stages of development (from embryo through adult) as the second (betalike) chain of hemoglobin. This includes the gene for the beta chain itself, two almost identical copies of a gamma chain (which occurs in fetal hemoglobin), and several others. Chimpanzees have the very same genes in the very same order. In the region between the two gamma genes and a gene that works after birth, human DNA contains a broken gene (called a “pseudogene”) that closely resembles a working gene for a beta chain, but has features in its sequence that preclude it from coding successfully for a protein.
Chimp DNA has a very similar pseudogene at the same position. The beginning of the human pseudogene has two particular changes in two nucleotide letters that seem to deactivate the gene. The chimp pseudogene has the exact same changes. A bit further down in the human pseudogene is a deletion mutation, where one particular letter is missing. For technical reasons, the deletion irrevocably messes up the gene’s coding. The very same letter is missing in the chimp gene. Toward the end of the human pseudogene another letter is missing. The chimp pseudogene is missing it, too.
The same mistakes in the same gene in the same positions of both human and chimp DNA. If a common ancestor first sustained the mutational mistakes and subsequently gave rise to those two modern species, that would very readily account for why both species have them now. It’s hard to imagine how there could be stronger evidence for common ancestry of chimps and humans.
FIGURE 4.1
Human and chimp hemoglobin genes are very similar. The top bar is a schematic illustration of the region of the primate genomes that contain genes for the betalike chains of hemoglobin, including the pseudo-beta gene (in gray), which cannot produce a functional protein. The arrangement is identical for both humans and chimps. The bottom bar is an expanded view of the pseudo-beta gene. Gray regions correspond to regions of functional genes that code for part of the protein. Both human and chimp pseudo-beta genes contain the same mistakes that preclude making a working protein.
That strong evidence from the pseudogene points well beyond the ancestry of humans. Despite some remaining puzzles,11 there’s no reason to doubt that Darwin had this point right, that all creatures on earth are biological relatives.
The bottom line is this. Common descent is true; yet the explanation of common descent—even the common descent of humans and chimps—although fascinating, is in a profound sense trivial. It says merely that commonalities were there from the start, present in a common ancestor. It does not even begin to explain where those commonalities came from, or how humans subsequently acquired remarkable differences. Something that is nonrandom must account for the common descent of life.
HE’S TWICE THE FUNGUS HIS DADDY WAS
The work on the hemoglobin genes of humans and chimps was done several decades ago. More recent work on whole genomes of yeast species further shows the power of the idea of common descent. Even better, this line of analysis has produced some of those eureka moments that make science so exciting—moments when newly accessible data suddenly illuminate a murky landscape like a flare in the night. It also points to the limits of random mutation.
Although most people think of yeast as the active agent that leavens bread or gives beer its zip, biologists classify yeasts as fungi—distant relatives of animals and plants. Scientists who work on yeast had long been suspicious of some features of the DNA of baker’s yeast (whose scientific name is Saccharomyces cerevisiae).12 It contains a number of genes that code for very similar proteins that seem to have almost redundant roles in the cell. The odd arrangement of genes led a couple of groups of scientists to hypothesize that, sometime in the misty past, perhaps a baker’s yeast cell was born with the mother of all gene duplications. Instead of just one gene, or a chunk of the genome, the entire DNA of the yeast was duplicated! Instead of the roughly 12 million nucleotides that its brothers and sisters had, the prodigy had 24 million. At one stroke the offspring was literally twice the fungus his daddy was. Over time, however, much of the duplicated DNA was lost by deletion mutations.
That was the hypothesis—but how to test it? With just the sequence of baker’s yeast (S. cerevisiae) DNA to go on, the suspicions couldn’t be confirmed. So a French group sequenced the entire genomes—tens of millions of nucleotides—of four other diverse kinds of yeasts. The researchers saw that duplicate genes in baker’s yeast could be lined up with their counterparts in the other yeasts. When they were aligned, one copy of a duplicated baker’s yeast gene would sometimes be next to the left half of some genes that formed a single group in another yeast species, while the second baker’s yeast gene copy would be next to the right half of the group in a separate region of the baker’s yeast genome. That arrangement is consistent with the hypothesis—made years before the genomes were sequenced—that the whole yeast genome duplicated and then many duplicate genes were deleted over time.13
This is yet more evidence for common descent. On the other hand, the genome duplication seems not to have done a whole lot for its recipients. All five yeasts have similar cell shapes and lifestyles.14 The duplicated baker’s yeast has the ability to make alcohol, but one unduplicated yeast can eat petroleum, arguably a trickier business. Another yeast species, containing more duplicated DNA than baker’s yeast, avoided whole-genome duplication; it apparently duplicated genes the old-fashioned way—one by one (or in blocks). Darwinists like to think that genome duplication is one of the magic bullets of random mutation—it suddenly granted vast new possibilities to the genome. Yet genome duplication—a spare copy of each and every gene to play with—and a hundred million years of time seem not to have given baker’s yeast any advantage it wouldn’t otherwise have had.15 This leads to a very important point. Randomly duplicating a single gene, or even the entire genome, does not yield new complex machinery; it only gives a copy of what was already present. Although duplicated genes can be used to trace common ancestry, neither individual gene duplications nor whole genome duplications by themselves explain novel, complex forms of life.
INCH BY INCH
If genetics has supported common descent, what of the usefulness of random mutation? It has fared decidedly less well, but still has some victories to boast of. Darwin argued that evolution had to work by tiny, random, incremental changes that improved the likelihood that a mutant organism would survive and prosper. So whenever we see such small beneficial changes or series of such changes, we should tip our hat to the sage of Down House. Sickle Eve was one example, as were the mutations that confer chloroquine resistance on malaria. To drive home the point that Darwinian random mutation can certainly explain some simple features of life, in the rest of the chapter I’ll recount several more cases, beginning with a few malaria-related examples.
As malaria developed resistance to the wonder drug chloroquine, scientists rushed to develop new treatments. One successor drug is called pyrimethamine. Interestingly, malaria can counter it with a single amino acid substitution. That single amino acid change makes malaria one hundred times more resistant to the drug. Malarial DNA has only about 23 million nucleotides. A sick person can be burdened with as many as a trillion parasite cells. If you do the math, the resistance mutation should occur by chance in at least one parasitic cell in almost every sick person. Looked at another way, resistance should develop independently many times over in a large group of patients treated with the drug. But a recent report by scientists at the National Institutes of Health pointed out a conundrum.
The Edge of Evolution Page 8