The Tangled Tree

by David Quammen

  Martin’s reticulated tree, as drawn by Martin, 1999.

  Yes, it was a defining phenomenon back then, more than three billion years ago, Woese conceded, with genes or gene-like information leaping sideways between one sort of living entity and another. But that era of rampant HGT, in Woese’s view, occurred at a time before the tree of life began to rise and branch. It happened before species existed, and therefore before limbs could diverge from a trunk, before branches could diverge from limbs. It happened prior to the origin of cellular creatures, as we understand cells today. Woese was alluding to another important idea, also unprovable but probably correct: that the origin of life, in some form, occurred a billion years or so before the origin of the first recognizable cell.

  What did that early, inchoate period look like? Using a phrase coined by another scientist, he called it the RNA-world. In that world, RNA, not DNA, was the repository of complexity (nearly random complexity, at first); and RNA, not cellularity, was the basic structural unit of whatever passed as life. Certain chemical components, catalyzed somehow by energy and physical circumstances and chance, had assembled into “aggregates” of interactive molecules, with a capacity to replicate themselves. The crucial element of those aggregates was RNA, single stranded and not very stable. DNA, double stranded and stable, didn’t yet exist. The aggregates differed from one another, there was a diversity of elaboration, and so as they self-replicated, drawing on environmental materials, they also began to compete. Some of them developed an inclination to translate their linear sequences of nucleotide bases (strings of A, C, G, and U) into another medium by linking amino acids together. Those linked amino acids constituted the earliest form of what a biochemist would now call peptides—short chains of amino acids—and then, as the chains grew longer and more intricate, proteins. These “aggregates” described by Woese in 2002 resemble the progenote concept he had offered back in 1987, with at least one signal distinction: in 1987 he hadn’t yet become attuned to the importance of horizontal gene transfer, especially during the earlier stages of evolution.

  Now he emphasized that phenomenon. As the molecular aggregates attained lives of their own, replicating, competing for survival, fragments of RNA from one aggregate would break off and attach to another aggregate. This was HGT in its earliest form. It happened often. The RNA-world was a great orgy of promiscuously shared materials. As time passed and possibilities burbled, some of the proteins proved useful in helping some of the aggregates replicate. At that point, RNA came to represent not just randomly assembled molecules but also potentially valuable information. But there were still no cells in this soup—nothing with a nice wall or a membrane bounding its internal substance, separating inside from outside, separating self from other—and certainly no division into groups that could be called species.

  Complexity increased, under the impetus of competitive replication. Some of the aggregates brought proteins to their aid for slightly advanced structural purposes, and one possibility thus opened was self-containment. Packaging, with special advantages. Boundary protection. Bingo: the first cells.

  These were primitive things compared with modern cells as we know them, probably leaky and clumsy and unstable; but cellular in some sense nonetheless. Each now enclosed in a wall or at least a membrane, they found ways to replicate themselves using the RNA (or maybe, by now, DNA) instructions they contained. The cell wall shielded them, to some degree, from the distractions and confusions of HGT—noisy information arriving from outside. Primitive though they were, the early cells replicated themselves more faithfully than could be done by a naked strand of RNA. They generated lineages. HGT became less rampant. Eventually the lineages yielded whole populations of cells that resembled one another. At this point, you might call those cells by a different name: organisms. The cells constituting a population weren’t all identical, they encompassed some variation, but there was more similarity among them than between them and the cells of other lineages. Bingo again: species. A major boundary in the history of life had been crossed. Woese called that boundary the Darwinian Threshold.

  “As a cell design becomes more complex and interconnected,” he wrote, “a critical point is reached where a more integrated cellular organization emerges, and vertically generated novelty”—meaning parent-offspring inheritance, with modest variation, as distinct from horizontal gene transfer—“can and does assume greater importance.” At that critical borderline, the Darwinian Threshold, evolution as Darwin understood it begins. This progression is all stated in present tense, you’ll notice, as though Woese is sitting there, three billion years ago, and watching these things happen. In transcendent moments, he may almost have felt that he was.

  None of it altered the shape of his preferred tree of life. In his 2000 paper on what he called the “universal” tree, and in his 2002 treatise on the evolution of cells, Woese included tree figures almost identical to what he had offered with Kandler and Wheelis in 1990. He ignored a decade’s worth of discoveries in the realm of horizontal transfer. He omitted contortions of the sort shown in Ford Doolittle’s drawings. His tree still had three major limbs. His branches diverged but did not inosculate. Woese seems to have felt that he had dismissed the challenge of HGT by pushing it far backward in time, down among the roots of the tree, before the rise and divergence of “modern” cellular life. Others disagreed.

  67

  A half dozen years passed. Newer data, still more HGT. Eugene Koonin and two colleagues looked for horizontal transfers between major bacterial lineages and found plenty. The percentage of genes traceable to sideways transfer, they discovered, ranged from a small fraction in some bacteria up to a third of the entire genome in Treponema pallidum, the nasty spiral bacterium that causes syphilis in humans. As additional genomes were sequenced, Bill Martin and a pair of collaborators found even more HGT. Comparing a half million genes distributed among 181 different genomes of bacteria and archaea, Martin’s team concluded that roughly 80 percent of the genes in each genome had arrived by horizontal transfer at some point in evolutionary history. Peter Gogarten and a colleague detected HGT in plants, other scientists showed that it was important among fungi, and Julie Dunning Hotopp’s work (which I mentioned earlier) revealed its widespread occurrence among insects. You could almost say that HGT finding became a scientific craze—except that craze doesn’t capture the difficulty of such work, nor the depth of the implications. And few people outside the rarified realm of molecular evolutionary biology had heard about this phenomenon or the hubbub it was causing among experts.

  Then, in January 2009, the British magazine New Scientist appeared with a tree on its cover and a big, incendiary headline: “Darwin Was Wrong.” Beneath those words was a subhead in smaller type: “Cutting Down the Tree of Life.”

  The cover story, written by the magazine’s features editor, Graham Lawton, began with Charles Darwin and his little sketch in 1837. A fruitful idea, depicting evolutionary relationships as a tree, Lawton noted, and by 1859, in The Origin, that spindly sketch had grown woody and tall, to a “great tree,” Darwin’s preferred simile. The article leapt from there to Ford Doolittle, citing his affirmation that the tree image was “absolutely central” to Darwin’s thinking about evolution. It became the unifying principle for understanding life’s history and, throughout the 150 years since first publication of The Origin, evolutionary biology had been largely devoted to learning details of the tree. “But today the project lies in tatters,” Lawton wrote, “torn to pieces by an onslaught of negative evidence.” He was talking about horizontal gene transfer.

  Lawton’s article was a fair introduction to the subject, despite the misleading provocations of the cover and the fact that those provocations were echoed inside by the very title of his piece: “Axing Darwin’s Tree.” Journalists don’t often get the privilege of titling their own magazine stories (though a features editor might), so it’s hard to know who was responsible. All that cutting down and axing—lumberjack imagery—and the clai
m of Darwin’s wrongness were pointedly extreme. They may have helped to sell magazines and enticed readers to turn pages, but they caricatured the genuine challenge to Darwinian orthodoxy that the new discoveries raised. The article itself was more nuanced, but still drastic enough. Lawton quoted Doolittle’s 1999 paper, the one that had splashed this debate onto the pages of Science, and also a more recent Doolittle comment, presumably gathered by email or phone: “The tree of life is not something that exists in nature, it’s a way that humans classify nature.”

  Lawton cited Bill Martin’s work—his critique of the tree of 1 percent—and mentioned endosymbiosis, but noted that, for a while, evolutionary biologists had taken HGT for a marginal phenomenon. No more. One researcher, Michael Rose at the University of California, Irvine, told Lawton flatly: “There’s a promiscuous exchange of genetic information across diverse groups.” Lawton mentioned endosymbiotic gene transfer from organelles as another version of that, making the point that HGT occurs in eukaryotes as well as in bacteria and archaea. He added a statement by John Dupré, an English philosopher of biology: “If there is a tree of life, it’s a small anomalous structure growing out of the web of life.” Finally, the article came back to Ford Doolittle, who seemed uncomfortable with the vehemence of the attack he had helped launch. “We should relax a bit on this,” Doolittle told Lawton. “We understand evolution pretty well—it’s just that it is more complex than Darwin imagined. The tree isn’t the only pattern.”

  Other scientists didn’t want to relax. They wanted, Lawton wrote, mixing his metaphors, to see “the uprooting of the tree” as the start of something bigger and more radical. But wait, was the tree being uprooted or cut down? Never mind. “It’s part of a revolutionary change in biology,” said Dupré. Evolution as understood in the future would be much more about mergers and acquisitions and collaboration than about change within isolated lineages. It would be about not just the divergence of branches but also their inosculation. Another scientist, a one-time postdoc and coauthor of Doolittle’s named Eric Bapteste, noted: “The tree of life was useful.” In Darwin’s hands, that image had helped people picture evolution, when evolution was such a startling new idea. “But now we know more about evolution,” Bapteste said, and “it’s time to move on.”

  The cover date of this New Scientist was January 24, 2009. The timing was significant: two weeks later came Charles Darwin’s two hundredth birthday, on which biologists and historians around the world would be celebrating his life, his work, and his theory. That entire year, 2009, in fact, would be filled with Darwin events and retrospectives, from Cambridge to Mumbai to Albuquerque. The editors of New Scientist had planned their own observance, and this tree-cutting issue was it: an equivocal sort of birthday card for Mr. Darwin. Seizing the opportunity of the bicentennial, with its broad (if brief) public attention to all things Darwinian, they offered Graham Lawton’s article as an alert that new discoveries—genes moving sideways, who knew?—continue to reshape our understanding of life’s history. That was useful. But their cover headline was pure provocation: Darwin at two hundred, wrong! Promptly and stridently, some of the big names in evolutionary biology disagreed.

  Daniel Dennett is a distinguished philosopher, author of Darwin’s Dangerous Idea, among other writings. Jerry Coyne is an evolutionary biologist at the University of Chicago, known for his work on the speciation process and his strong voice as a defender of evolutionary science against creationism. Richard Dawkins is author of The Selfish Gene and many other books, including The God Delusion, and probably the world’s best known and most vociferously atheistic Darwinist. He’s also an extremely bright man with a dangerous and confident wit. These three, along with one other biologist, Paul Myers, an influential blogger, coauthored an outraged letter to New Scientist, which the magazine published a month later. It began: “What on earth were you thinking . . . ?”

  Their complaint was with the garish negativity of the tree of life cover, not the more nuanced explanations of the article. Announcing that “Darwin Was Wrong,” Dennett and company argued, would give aid and comfort to the enemy, “handing the creationists a golden opportunity to mislead school boards, students, and the general public about the status of evolutionary biology.”

  On that point, they were probably right. Creationists did seize the opportunity, golden or otherwise. One instance turned up immediately at a creationist website called Apologetics Press, under the headline: “Startling Admission: ‘Darwin Was Wrong.’ ” This piece, by one Eric Lyons, sketched the main points of Lawton’s article, quoted Doolittle and others, and called the tree image just one more “iconic concept of evolution” that had “fallen on hard times.” As decades passed and more data became available, Lyons claimed, yesterday’s proofs of evolution were turning out to be mistaken. “One wonders what it will take,” he wrote, “to convince evolutionists that it is not just Darwin’s tree of life that needs to be cast aside, but the entire theory of evolution.” Lyons’s use of the New Scientist headline was exactly what Dennett and his letter coauthors had feared.

  Amid the stink over the magazine’s cover, and the essential or nonessential significance of the tree, a quieter comment from New Scientist—in the form of an unsigned editorial—went largely overlooked. This introductory note was meant to put Lawton’s article in perspective and to negate, in advance, the sort of gleeful creationist misconstrual that came from Lyons. Biology, like physics, the editorial said, is a work in progress. Big upheavals in evolutionary thinking have occurred before. There was the original one, triggered by Darwin and Wallace in the nineteenth century. There was a second revolution in the 1930s and 1940s, when Mendelian genetics was integrated with Darwinian natural selection and mathematics. That came to be known as the Modern Synthesis of evolutionary theory, also known as neo-Darwinism. And now, in the era of molecular biology and genome sequencing, important biological topics were “turning out to be much more involved than we ever imagined.” That was putting it mildly. “As we celebrate the 200th anniversary of Darwin’s birth,” said the editorial, “we await a third revolution that will see biology changed and strengthened.”

  Is that third revolution happening? If so, you could argue that it began with Carl Woese taking up a vague but astute suggestion by Francis Crick about doing “protein taxonomy” as a gauge of relatedness among different creatures. Woese’s lonely work led to the now-crowded field of molecular phylogenetics and to the recognition that, against even so strong a signal as 16S rRNA, horizontal gene transfer has made the history of life unimaginably more complicated than Charles Darwin could have guessed. It also brought a dawning curiosity about what extraordinary things occurred during that distant time before what Woese called the Darwinian Threshold. It was all about evolution, and it went beyond Darwin’s thinking without negating that thinking, just as Einstein and quantum mechanics went beyond Isaac Newton’s. “None of this should give succour to creationists,” the New Scientist editorial added. Right again, though, of course, it has.

  68

  By the time that ruckus broke in the pages of New Scientist, Ford Doolittle had begun what he half seriously calls “my retreat into philosophy.” His talks at meetings and his publications, especially the 1999 paper in Science, had made him a leading spokesman for the view that horizontal gene transfer is vastly significant and that “the history of life cannot be properly represented as a tree.” Tree or not tree, that was the question. It seemed very important to many biologists—not just Dawkins and Dennett and Coyne—because of their engagement in two parallel struggles: the struggle to understand life’s history, yes, and also the struggle to defend the teaching of evolutionary theory against creationist rhetorical and political attacks. Now the tree image itself had become controversial, with “tree huggers” and “tree cutters” arguing pro and con.

  “I’m not unhappy at having started that polarizing debate,” Doolittle told me, “because I think it’s been useful.” The debate stimulated fresh thinking and work
. It drove efforts to gather more data, and it generated new hypotheses, “some of which have turned out to be interesting and probably true, and some of which”—he chuckled—“have just turned out to be interesting. I guess I’ve always believed that if you’re not wrong half the time,” he said, “then you’re not being brave enough.”

  But there came a point of diminishing returns. “I kind of got tired of opening every journal and saying, ‘Oh, is this paper for me or against me?’ ” He realized, with increasing clarity, that this wasn’t really a scientific issue. It was philosophical and representational and semantic. “Whether or not there is a tree of life,” Doolittle told me, “depends a whole lot on what we mean by a tree of life.”

  It reminded him of a classic paradox that philosophers call the Ship of Theseus problem, recorded by Plutarch two millennia ago in his Life of Theseus. The legendary hero Theseus arrives back in Athens after an epic adventure in Crete, and the Athenians evermore preserve his ship as a historical relic. But it’s a working relic. They don’t mothball the thing, they don’t put it on a pedestal; they use it for important ceremonial voyages. Some of the planks rot over time and must be replaced. Then more rotten planks, more replacement. At what point is the “Ship of Theseus” no longer the ship of Theseus? Hard to say. Well, the same problem exists, according to Doolittle, with the identity of an organism and its lineage after billions of years of horizontal gene transfer. If its genes are half bacterial and half archaeal, does that organism belong to the Bacteria or to the Archaea—or is the question impossible or meaningless to answer?

  The best way to understand the tree, Doolittle decided, is that it represents a hypothesis about the history and relatedness of all life. He had said this back in 2000 in the Scientific American article, but passingly, without developing the thought. Now he did that. It isn’t just an “attractive hypothesis” arisen from nowhere or by consensus; it’s Charles Darwin’s hypothesis, formulated in 1837, when Darwin drew that little sketch in his B notebook and scribbled above it: “I think.” The hypothesis was meant to explain patterns Darwin had seen among living creatures and fossils, leading him to the conviction that evolution had occurred. If so, how had it occurred? From common ancestry, Darwin argued, beginning with a few original forms or maybe just one, passing slowly through changes, some forms diverging from others, differentiating into many species, those changes shaped primarily by the process he called natural selection. If all that were correct, what would life’s history look like? Darwin hypothesized: it would look like a tree.