The Tangled Tree

by David Quammen

  Although many bacteria live as solitary cells, taking their chances and meeting their needs independently, others aggregate into pairs, clusters, little scrums, chains, and colonies. The coccoid cells of Neisseria gonorrhoeae, which cause gonorrhea, lump together by twos, forming bilobed units resembling coffee beans. The genus Staphylococcus gets its name from Greek words for “granule” (kókkos, the spheroid aspect) and “a bunch of grapes” (staphylè), because staph cells tend to bunch. Most of the forty staph species are harmless, but Staphylococcus aureus can inflict skin infections, sinus infections, wound infections, blood infections, meningitis, toxic shock syndrome, plus other nasty conditions, and if you’re so unlucky as to pick up a dose of those little grapes in one of their antibiotic-resistant forms, such as MRSA, a monstrous product of horizontal gene transfer (as I’ve mentioned, and to which I’ll return), you could be in a world of hurt. Cells of Streptococcus species, including those that cause impetigo and rheumatic fever, stick together like beads on a chain.

  Bacteria can also form stubborn, complex films on certain surfaces—the rocks of a sea floor, the glass wall of an aquarium, the metal ball of your new artificial hip—where they may cooperate together in exuding a slimy extracellular substance that helps nurture them collectively, maintain the stability of their little environment, serve as a sort of communications matrix among them, and even protect them from antibiotics. These living slicks, known as biofilms, can be thinner than tissue paper or as thick as a good dump of snow, and may incorporate multiple species. The little rods of Acinetobacter baumannii are infamous for their ability to lay down persistent biofilms on dry, seemingly clean surfaces in hospitals.

  Cyanobacteria, including that monumentally abundant Prochlorococcus, convert light to energy and deliver, as byproduct, a large share of Earth’s free atmospheric oxygen. Purple bacteria photosynthesize too, but do it by drawing upon sulfur or hydrogen instead of water as fuel for the process, and they don’t produce oxygen. Lithotrophic bacteria, the rock eaters, deriving their energy from iron, sulfur, and other inorganic compounds, exist in more ingenious variants than you care to know. Japanese researchers have recently discovered a new bacterium, Ideonella sakaiensis, that digests plastic. Certain enterprising ocean bacteria, such as Marinobacter salarius, have risen to the challenge of degrading hydrocarbons from the Deepwater Horizon oil spill. Other bacteria are quite capable, in the presence of oxygen or without it, of feasting on garbage, sewage, various inorganic compounds, plants, fungi, and animal tissue, including human flesh. Lactic acid bacteria, which may be rod shaped or spherical, turn up in milk products, busy at their task of carbohydrate fermentation and resistant to the acid they create. Many of them also like beer.

  Not all such particulars were known to Carl Woese in 1977 as he examined the fingerprints from his first few methanogens. But the vast scope, ubiquity, and multifariousness of bacteria certainly were. The terrain of bacteriology was known even better to Ralph Wolfe, who had trained in the classic fundamentals under van Niel and others. Woese’s reaction to his own preliminary results must have seemed all the more radical, then, all the more shocking, as he shared it not just with George Fox and members of his own lab but also with Wolfe, just after they repeated the rRNA analysis of delta H, the first methanogen. “Carl’s voice was full of disbelief,” Wolfe wrote in a memoir, “when he said, ‘Wolfe, these things aren’t even bacteria.’ ”

  22

  Ralph Wolfe told me the same story, with some elaboration, thirty-nine years later when I called on him in Urbana. By then, he was an emeritus professor of microbiology, ninety-three years old, a frail and slender gentleman with a quick smile, still maintaining his office and coming to it, as though retirement were not an entirely satisfying option. On the wall behind his desk hung a replica of Alessandro Volta’s pistola, a gunlike device invented by Volta in the late 1770s for testing the flammability of swamp gases, including methane. On the desk itself were papers and books and a computer.

  Woese’s lab back in the day had been in Morrill Hall, on South Goodwin Avenue, and Wolfe’s was in an adjacent building, connected by a walkway. Woese would occasionally trundle over on various business. “He came down the hall and happened to see me,” Wolfe recalled, “and says, ‘Wolfe, these things aren’t even bacteria!’ ” Wolfe laughed gently and, for my benefit, continued reenacting the scene.

  “ ‘Of course they are, Carl.’ ” They look like bacteria in the microscope, Wolfe had told him. But Woese wasn’t using a microscope. He never did. He was using ribosomal RNA fingerprints.

  “ ‘Well, they’re not related to anything I’ve seen.’ ” Coming back to the present, Wolfe said: “That was the pivotal statement that changed everything.”

  23

  We went into fast-forward mode,” Woese recalled in his account of these events. By the end of 1976, his team had done fingerprints and catalogs on five additional methanogens, with more in the pipeline. And sure enough, he wrote, none of the new catalogs was prokaryotic, not in the prevailing sense of that word, which meant bacteria and only bacteria. None of the organisms was eukaryotic, either. But “they were all of a kind!”—a third kind, something else, something anomalous, something hitherto unsuspected to exist. Woese started thinking that he would need to declare a new kingdom of life—create a new name, invent a huge new category—to recognize their uniqueness and contain them. It wasn’t really a new kingdom, of course. It was a newly discovered natural grouping of life-forms, which had existed apart for a long time, unrecognized, and which might be called a “kingdom” or an “urkingdom” or a “domain,” according to preferred human convention.

  Woese believed that this discovery, still unannounced, offered “a rare opportunity to put the theory of evolution to serious predictive test.” He meant Darwin’s theory of evolution, as opposed to any others—the one that recognized hereditary continuity plus a degree of random variation over long stretches of time, and explained the shaping of that variation, to yield adaptation and diversity, mainly by way of natural selection. If Woese’s preliminary findings were correct, he noted, those findings should serve as a guide for predicting roughly what further data and discoveries would appear. From the premise that 16S rRNA represented a very slow-ticking molecular clock, with a minimum of selected variation, he deduced that his newly found kingdom must represent a very old division. Very old—having originated near the beginning of cellular life, maybe three and a half billion years ago. Now he would try to sketch its boundaries and its characteristics. As he and his team added more microbes to its membership—more methanogens and maybe other creatures too, each known by its catalog of RNA fragments—Woese expected two things: that this unnamed kingdom would remain dramatically distinct from the rest of the living world and that it would nonetheless encompass great diversity. “Testing these two main evolutionary predictions,” he wrote, “drove our work from that point on.”

  Three domains and (within the eukaryotes) four kingdoms, four types of cell.

  In August the team published a carefully limited paper, just a hint of what was coming, in the Journal of Molecular Evolution, the same journal at which Emile Zuckerkandl continued to serve as editor. It was a logical match of subject and outlet because Zuckerkandl, back in his days as Linus Pauling’s sidekick, had helped articulate the very premise that Carl Woese was now putting dramatically to use: that the branching of lineages “should in principle be definable in terms of molecular information alone.” The molecular information at issue in this case consisted of ribosomal RNA sequences from the first two methanogens Woese’s team had characterized. One of those methanogens was a strain of M. ruminantium, isolated from rumen fluid (from the paunch of a cow) donated by a friendly contact in the university’s Department of Dairy Science. The other was delta H, the conveniently nicknamed strain of the fourteen-syllable monstrosity, M. thermoautotrophicum, known to live at high temperatures and metabolize hydrogen. I asked Ralph Wolfe where they had gotten their starter sa
mple of that exotic beast, delta H.

  “It was isolated here from the sewage.” More specifically, from a sewage sludge digester.

  “In Urbana?”

  “Yeah.”

  The first author on this discreet paper was Bill Balch, Wolfe’s graduate student, who had earned his authorship priority by developing the sealed-tube technique of growing and labeling methanogens. “It was because of that technique,” Wolfe told me, “that we could now do these experiments with Carl. Because everything was sealed, and you could now inject the P-32 into the culture.” P-32, remember, was the radioactive phosphorus. “Whereas the previous techniques, you had to keep opening the stopper and flushing it out, and it would have been a radioactive nightmare to do it that way.” Balch’s system allowed for injecting the P-32 by syringe through the black rubber stopper. Balch grew the microbes, George Fox extracted the RNA, and Woese’s trusted lab technician at the time, a young woman named Linda Magrum (she had replaced the earlier Linda in that role, Linda Bonen), prepared the fingerprint films for Woese to analyze. All three of them, plus Ralph Wolfe himself, appeared as coauthors, with Woese’s name last, reflecting his role as senior author. Besides describing the methodology, this paper noted drily that the two methanogens didn’t look much like “typical” bacteria. It mentioned that the divergence might represent “the most ancient phylogenetic event yet detected”—a big claim, vague enough as stated to pass almost unnoticed.

  In October the team published a second paper, in a more far-reaching journal, the Proceedings of the National Academy of Sciences (known as PNAS). This time George Fox was first author, and the data covered ten species of methanogen, each one assessed for similarity to the other nine and to three species of what the authors still cautiously called “typical bacteria.” Fox had created a simple measurement system by which the catalog of one microbe could be compared with the catalog of another, yielding a decimal number—a coefficient—representing degree of similarity. Comparing each of these thirteen microbes with all the others gave an overall picture of which were how closely similar to which others. The data could be arranged in a rectangular table, names down the left margin, names again across the top, numbers at each cross point, as in a chart showing the various mileage distances between all pairs in a list of cities. Instead of mileage: a similarity coefficient. From those numbers, and the premise that similarity reflected relatedness, Fox generated a dendrogram, a branching figure, showing nodes of divergence between major lineages and a branch for each organism. Although they printed this dendrogram sideways—like a bracket for the NCAA basketball tournament—rather than vertically, it was, in fact, a tree: the first of the new trees of life in the era of Carl Woese. There would be many more.

  This one showed the “typical bacteria” occupying one major limb. The ten methanogens all branched from a second major limb. “These organisms,” said the paper, “appear to be only distantly related to typical bacteria.” Again the five authors were saying less than what they believed. The phrase “typical bacteria” was an interim delicacy that would soon disappear.

  A third paper, the most bold and dramatic, appeared in PNAS a month later under the authorship of Woese and Fox alone. Its title hinted only obliquely at its intent: to reorganize “the primary kingdoms” of life. Again using Fox’s similarity coefficients, it compared methanogens against one another and against “typical bacteria,” and each of those also against several eukaryotic organisms, including a plant and a fungus. Its conclusion was radical: there are three major limbs on the tree of life, not two. The prokaryote-eukaryote dichotomy, as proposed by Stanier and van Niel, as generally accepted throughout biology, is invalid. “There exists a third kingdom,” Woese and Fox wrote, and it includes—but may not be limited to—the methanogens. It isn’t the bacteria, and it isn’t the eukaryotes, they explained. It’s a separate form of life.

  The two authors gave their kingdom a tentative name: archaebacteria. Archae- seemed apt, suggesting archaic, because the methanogens appeared so ancient, and their metabolism might have been well suited to early environments on Earth, about four billion years ago, before the onset of an oxygen-rich atmosphere. Woese had made that very point in an interview with the Washington Post. “These organisms love an atmosphere of hydrogen and carbon dioxide,” he said (or at least, so he was quoted). “Just like the primitive earth was thought to be,” he said, adding, “No oxygen and very warm.” But the other half of that compound label, archaebacteria, tended to blur the central point of the discovery: that, as Woese had announced to Wolfe, these things aren’t even bacterial forms of life. They’re quite different. Wolfe himself told Woese that archaebacteria was a terrible choice. If they’re not bacteria, why retain that word at all? The provisional name stuck for about a dozen years, before being emended to something better, something that stood by itself: the archaea.

  24

  George Fox was no longer a rangy young man when I sat with him in a nondescript pizza parlor near the campus in Urbana, after the opening session of a Carl Woese memorial symposium, and watched him eat a nondescript little pizza. Fox is a man who prefers simple, plain food, and he had cringed when I ordered pepperoni and mushrooms on my own. At age sixty-nine, he carried the full body and slight jowls of a lifetime spent in laboratories and classrooms; wire-rim spectacles had replaced the dark horn-rimmed glasses he had worn in the 1970s photos, and his brown hair was graying at the temples, but his eyes still shined brightly blue as he recalled the days and years with Woese. Now a professor at the University of Houston, Fox had flown up for the Woese meeting, which was hosted by the Carl R. Woese Institute for Genomic Biology (its name reflecting the fact that Woese has become a venerated brand at the University of Illinois). Fox would give one of the invited talks.

  He had spent his academic career at three institutions: Houston, for almost three decades; preceded by Illinois, as a postdoc with Woese; and before that it was Syracuse University, as an undergraduate and PhD student. The circumstances of Fox’s arrival in Urbana were haphazard, beginning from a coincidence in Syracuse, where Woese himself grew up. There at the university, Fox belonged to a professional engineering fraternity, Theta Tau, of which Carl Woese’s father—also named Carl Woese—was a founder, and so Fox was required to know the name. As he shifted interest from chemical engineering to theoretical biology, he noticed and became fascinated by some of the early work of Carl Woese the son. In particular, there was a paper on what Woese called a “ratchet” mechanism of protein production by ribosomes—a risky proposal, a wild and interesting idea (later proven wrong in its details), published in 1970. So Fox wrote to this ratchet guy asking for a postdoc fellowship, and Woese seemed to see the Syracuse connection as karma. He had a position to fill, yes, with the departure of Mitch Sogin, the ultimate handyman grad student, and he offered that to Fox.

  “We did not discuss salary,” Fox said over his pizza and Coke. “He never sent me a letter offering the position. It was all completely verbal.” On such assurance, Fox got married and showed up in Urbana that autumn with his wife. Arriving unannounced, he encountered a man at the lab door, an unprepossessing figure in jeans and a drab shirt, with a chain holding a huge bunch of keys. “He looked like the goddamn janitor.” Fox gave his name and prepared to talk his way in. “No!? Welcome!” It was Woese.

  “He sat me down in his office and . . .” Fox hesitated. “You got a piece of paper?” On a yellow sheet from my legal pad, he began sketching the layout of the lab. He drew a long rectangle and subdivided it. There were three major rooms, he explained, and the middle room, here, held the light table, where Carl usually worked. Linda Magrum and Ken Luehrsen were here, in the left room. Over here on the right side of the center room was Carl’s little personal office and the electrophoresis room. The radiation room and the darkroom were across the hall, and then storage, three more spaces barely bigger than closets. Woese gave Fox a table in his office, Fox said, with a door that stayed open, “so he could see me.” Like th
e young Luehrsen, only more so, as a postdoc, Fox was on probation.

  At the beginning, Woese assigned him to assembling sequences from 5S rRNA, the shortest and least informative of the ribosomal RNA molecules, as a way of getting up to speed on what the lab was doing. That project yielded some unexpected results, impelling Woese to try to make Fox an experimentalist. But it wasn’t his forte, and he knew that. He wanted to do the sort of “theoretical stuff,” the deep evolutionary analysis of molecular data—what would now be called bioinformatics—that Woese himself did. Reading the code, drawing conclusions that went back three billion years and more. Woese, on the other hand, wanted him to generate data. “I was under a lot of pressure,” Fox recalled—the pressure of Woese’s expectations versus his own interests and skills. “What I had to do was, every other day, come up with a novel insight, so that he would continue to allow me to work on the sequence comparison project.” Failing that impossible standard, he was banished back to the lab, set to the tasks of growing hot cells and extracting their ribosomal RNA. But Fox continued, in flashes, to show his value to Woese as a thinker. Gradually he proved himself, not just sufficiently to work on sequence comparisons but well enough to become Woese’s trusted partner, as well as the sole coauthor on the culminating paper in 1977, with its announcement of a third kingdom of life.

  25

  Wondering how that announcement was greeted by the scientific community at the time, I had put the question to Ralph Wolfe, several months before the pizza with George Fox.

  “It was a disaster,” Wolfe said mildly. Then he explained, with the sympathy of friendship, why Woese’s declaration of a third kingdom—the substance of the claim, and the manner in which Woese made it—had sounded discordantly to many of their peers. The crux of the problem was a press release.