The Tangled Tree

by David Quammen

  “Where do cyanobacteria live in the wild?” I asked, trying to get a grip on these invisibilities as living creatures.

  “Everywhere. I mean, in water—a lot of pond scum is cyanobacteria. Greenish tinge on the side of an old building in Oxford would be cyanobacteria, probably.” He knew I had attended Oxford and seen plenty of greenish tinge on old buildings. “They’re common. They have a lot of interesting, different colors. So I had gotten into cyanobacterial molecular biology as a kind of . . .” He paused to consider: why exactly? Not for medicine; not just toward evolutionary questions. Doolittle fancied the subject, and he was a scientist. “For its own sake,” he said bravely.

  Furthermore, there happened to be a good international community of scientists doing similar work—on the molecular biology of cyanobacteria, these things that until recently had been classified with algae—of whom the leader in Doolittle’s time was still Roger Stanier. Leaving Berkeley in 1971 with some distaste for American politics, Stanier had accepted a position at the Pasteur Institute, in Paris, on the condition that he be allowed to work exclusively on cyanobacteria. But none of those cyanobacteria experts, not even Stanier, had studied RNA maturation (that process of cutting longer molecules to fit function) for the production of ribosomes, Doolittle’s own little specialty since his time in Norman Pace’s lab. “So I thought I could just kind of repeat what I did with Norm”—repeat it but now on cyanobacteria—“and get at some low-hanging fruit.”

  Doolittle began working with various kinds of cyanobacteria. He published a couple papers on the RNA maturation topic and other aspects of their biology, and in response to one of those papers he received a congratulatory note from Stanier himself. Nice piece of work, it has given me some ideas, Stanier told him; would you please send a few reprints for my group? That was heady praise from a grand man in the field, gratifying for an assistant professor. (Carl Woese would tell him some time later that Woese was jealous of the note, having never gotten any praise from Stanier.) But still Doolittle knew that his line of research was unambitious, that he hadn’t reached very high. Around that time, he read the Margulis book.

  He found her endosymbiosis theory enticing and was struck also by the book’s illustrations—of single-celled organisms, symbiosis events, even a tree of life—which had been drawn and labeled freehand by an illustrator named Laszlo Meszoly, in a style that Doolittle saw as “funky.” That was a compliment. The 1960s had ended, but its cultural vapors remained and, if R. Crumb had sketched an amoeba swallowing a bacterium for some hallucinogenic report in Rolling Stone, it might have resembled Meszoly’s work. Doolittle, with an itch for drawing himself, liked that cartoonish touch applied to serious science. “I think it was part of the inspiration for me to do figures,” he said—a consequential inspiration, given that his own drawn figures would later help communicate a radical new vision of the tree of life.

  The Margulis book also caught his attention because it was evolutionary, not just biochemical. And it tracked evolution far into the past, deep to the base of the tree. “You know, for somebody like me,” Doolittle said, “I don’t really care whether elephants are related to hippopotamuses or whales. Those kinds of things didn’t concern us.” The details of mammal phylogeny seemed like small beans, he meant, for microbiologists who were wondering how one vast kingdom of life had diverged from another. “But the relationship between eukaryotes and prokaryotes seemed like a pretty big question.” The endosymbiosis theory, heterodox but concrete, went to that question. Doolittle began thinking: Hmm, we could test this.

  Then one day, in about 1973, Linda Bonen walked into his lab. It was the same Linda Bonen who had worked as a technical assistant for Woese in Urbana, helping him use electrophoresis and X-ray films to sequence ribosomal RNA. She was now living in Halifax, having come with her husband, an exercise physiologist who had accepted a job at Dalhousie within the Phys. Ed. Department. Bonen wanted some interesting work. She was well qualified for certain onerous, difficult tasks along the borderline of biochemistry and molecular biology, and some of those skills might be useful to other researchers. It was Woese himself who had said to her, when he heard she was moving to Halifax, “I know who you want to work with.” He meant Doolittle. Woese and Doolittle were well acquainted from the time period, just a few years earlier, when Doolittle had done his first postdoc in Urbana. They kept in touch intermittently by letter and phone. Woese may even have written to him on Bonen’s behalf, or Doolittle got wind of her skills somehow; he doesn’t recall. “But when she came to me,” he said, “I knew who she was and what she could do.”

  So the little group constituting Doolittle’s lab, now with Linda Bonen as technical assistant, took up a new question: the origin of chloroplasts in complex cells. Among their first efforts was a comparison of ribosomal RNA samples from five sources: the cytoplasm of a red alga (a eukaryote), the chloroplasts of that alga, and several kinds of bacteria, including the familiar E. coli. If the Margulis theory was correct, Doolittle knew, chloroplasts contained in the complex cell should resemble the bacteria because they had originated from bacteria themselves.

  Doolittle and Bonen reenacted the same messy, dangerous preparations of material that Bonen had learned in Urbana from Woese. They grew their organisms in nutrient lacking phosphorus, then added P-32 (the radioactive isotope) so that the bugs would build it into their molecules, including their ribosomal RNA. Then they broke open the cells, extracted the rRNA, selected out the subunits they wanted (16S from the bacteria and 18S, its eukaryotic equivalent, from the alga), cut those into short fragments with enzymes, and ran the fragments in races against one another by electrophoresis. From the electrophoresis runs, they printed images on X-ray films. Like Woese, they called each image of spread fragments a fingerprint, although it looked more like a herd of stampeding amoebae. From the fingerprints, they deduced the base sequences of the fragments and compiled them into catalogs.

  Margulis’s tree of eukaryotes, drawn by Laszlo Meszoly, 1970.

  Bonen did most of the wet work, with Doolittle helping. He was the boss, but she knew the techniques. When they raced the fragments in a second dimension, pulling them sideways to learn more about their composition, the electrophoresis was powered by 5,000 volts and sizable amperage. To cool the paper racetrack, they kept each end submerged in a tank of Varsol, the same flammable solvent that Mitch Sogin had used. “We had a special room built here,” Doolittle told me, “and the room had CO2 tanks, huge CO2 tanks.” What was the need for carbon dioxide? “To put out the fire in the Varsol that might have been started,” he said and then laughed. Of course, the CO2 itself would be poisonous to a human if, in response to a fire signal, it automatically flooded the room’s atmosphere. “You had, like, thirty seconds, between when the alarm rang and when the room filled up with CO2, to get the hell out of there.” He laughed again at the absurdity of the old methods.

  Bonen, with less relish for the absurd, told me simply: “I was in a small lab that had all these special precautions.” Instead of blowing herself up, getting suffocated by a safety system, or succumbing to radiation poisoning, she produced the fingerprints as desired. Those images were printed on chest X-ray films, big rectangles, each exposure done within a shallow plastic box known as a cassette, which gave lightproof protection while the radioactive fragments burned their images onto the film. The cassettes came as hand-me-downs from a local hospital. Bonen helped Doolittle learn to read the fingerprints and assemble catalogs of fragments for comparing one to another.

  Their results were clear and dramatic. They found that chloroplasts within their red alga differed drastically, by this rRNA measure, from ribosomal RNA in the alga’s own cytoplasm. It was almost as if Doolittle and Bonen were looking at two distinct creatures, from two different biological kingdoms—which in effect they were. If you had a kidney transplant and the donor was a stranger, the ribosomes in your new kidney wouldn’t differ from your other ribosomes by nearly so much as these chloroplasts diffe
red from their alga. Why not? Because your kidney would come from another human (or, if your doctors resorted to xenotransplantation, maybe from a baboon or a genetically engineered pig—in either case, a mammal). But these chloroplasts were xenotransplants from a completely different kingdom of life. They matched far more closely with the bacteria, those outsiders chosen for comparison, than with the red alga of which they were functional parts. What it meant, as Bonen and Doolittle noted quietly in their published paper, was that one cardinal point of the endosymbiosis theory had been confirmed: yes, chloroplasts in plants are descended from captured bacteria.

  At the end of the paper, they cited Lynn Margulis. They cited Merezhkowsky. They thanked Carl Woese for “advice, encouragement, and much unpublished data.” Doolittle might also have thanked Woese—and probably did, more privately—for exporting his methodology to Halifax in the person of Linda Bonen.

  33

  Ford Doolittle and Carl Woese had a longstanding friendship, founded in Urbana during Doolittle’s postdoctoral fellowship of the late 1960s. It was fueled by common interests, strengthened occasionally by collaboration, stressed occasionally by competition, and became really complicated only in Woese’s later years, when they disagreed about the tree of life.

  Urbana was their crossroads by double happenstance. Doolittle, fourteen years younger than Woese, grew up in that little Illinois town long before Carl Woese arrived. His father was on the art faculty of the university, as I’ve mentioned. Doolittle had left for Harvard College and then gone on to graduate school at Stanford by the time Woese came in 1964. Four years later, now with his doctorate, Doolittle returned to the University of Illinois as a postdoc in the lab of Sol Spiegelman, who had gained fame from some intriguing, spooky experiments on RNA replication in vitro. These experiments involved creating what Spiegelman himself called “the little monster”: a self-replicating molecule of synthetic RNA that produced unlimited generations of itself in a beaker. Sol Spiegelman was also the man who had recruited Woese to Illinois, and who had brought Sanger sequencing in the person of Dave Bishop, who had trained Mitch Sogin, who had preceded Linda Bonen as Woese’s key technician. It’s a smallish world these scientist live in, much interconnected.

  Doolittle’s relationship with his postdoc boss, upon his return to Urbana in 1968, was complicated by overfamiliarity: Spiegelman’s son Will had been Doolittle’s closest pal in high school, and young Ford had worked summers as a junior assistant in Dr. Spiegelman’s lab, washing glassware and doing other lowly chores. Spiegelman had given him his first brush with science, to that degree, and later advised him about graduate school. But Spiegelman was an intimidating figure, not a genial, avuncular mentor. “He had this nasty habit of wearing crepe-soled shoes and sneaking up behind you when you were pipetting P-32,” Doolittle told me. Laboratory pipetting, in those days, involved using your breath to siphon liquid into a glass straw. Having snuck up behind his lab tech, Spiegelman would start to hum. “So, half the time, you’d swallow the P-32.” Doolittle laughed still again, and, at that bizarre memory, I did too.

  “People were afraid of him,” Doolittle added. “I wasn’t afraid of him.” The word afraid didn’t capture it. After all, this was just Will Spiegelman’s brilliant, perverse dad. But when Doolittle returned as a twenty-six-year-old postdoc in the Spiegelman lab—probably reflecting a bad decision by each of them—the disparity of the old relationship never gave way to a more equal collaboration between two scientists. It lingered in stunted form, more of a barrier than a bond.

  Scientific relationships are always shaped by personal chemistry as well as by work and ideas. Sol Spiegelman wasn’t a fellow with whom you could enjoy a relaxed collegiality away from the lab, Doolittle told me. Not if you were Ford Doolittle, anyway, and Spiegelman had known you since you were a scrawny adolescent. But there was this younger professor in the same department, Carl Woese, less formidable, less remote. He wasn’t lofty and gruff—not in those days, anyway. “You could go out and have a beer with Woese,” Doolittle said. So by default and inclination, Woese became a kind of “social and emotional mentor” to Spiegelman’s students and postdocs, including Doolittle.

  That friendship endured through the 1970s, with the Woese and the Doolittle labs working sometimes on parallel projects, in a competitive spirit but also, at times, sharing ideas and unpublished data. In the very same issue of the Proceedings of the National Academy of Sciences in which Doolittle and Bonen published their chloroplast study, for instance, Woese’s group published a similar piece of work, likewise confirming that Lynn Margulis had been right about the bacterial origin of chloroplasts. Just as Bonen and Doolittle had thanked Woese for the courtesy of sharing unpublished data, so he thanked them. Overlapping and reciprocal collegiality; science as it should be.

  One year later, Doolittle and Bonen published another paper, this one in Nature, offering evidence for two significant claims: that blue-green algae are not in fact algae but bacteria (hence they became known as cyanobacteria); and that chloroplasts, at least in some complex organisms, had originated not just as bacteria but specifically as this sort, cyanobacteria. The evidence had been gathered again with Woesean methods, and the paper was filled with respectful nods to Woese’s work.

  Then, in 1978, a new paper of high interest to both men appeared in a relatively obscure European journal. This paper, from a French team in Strasbourg, offered something never yet seen: the complete sequence of bases (not just a sampling of fragments) for Woese’s molecule of record, 16S rRNA, from a single bacterium. The bacterium was that familiar bug E. coli. The method of sequencing was essentially Fred Sanger’s, modified by use of a new ingredient: extract of cobra venom, to help cut the molecule. The information value of the full sequence, to researchers such as Woese and Doolittle, was huge. Until this point, they had been comparing fragments—those short blurts of letters—without knowing how the fragments might fit together. The French team, with their cobra venom, had revealed the fit.

  But that European journal was slow in reaching the University of Illinois. Woese, having caught wind of the paper, was impatient. So he called Doolittle in Halifax, where the journal had arrived, and asked a favor: “Would you read me the sequence?” Doolittle obliged. Sitting in his office, with the October issue of the journal, he recited all 1,542 letters to Woese by telephone. He found that reading them in triplets gave a natural rhythm, making it easier to avoid omissions or duplications.

  That is, he said: “AAA, UUG, AAG, AGU, UUG, AUC,” and so on. He said, “AUG, GCU, CAG, AUU, GAA, CGU,” and so on. He said, “UGG, GAU, UAG.” UAG is a stop codon, a signal to terminate, when it appears within messenger RNA; but this wasn’t messenger RNA, it was structural, and when Doolittle came to a UAG, he didn’t stop. He said: “CUA, GUA, GGU, GGG, GUA,” and furthermore, “ACG.” He read the whole damn thing, the entire eye-crossing stream of letters, while Woese in Urbana copied them down carefully. Finally Doolittle said, “GGU, UGG, AUC, ACC, UCC, UUA,” and they were done.

  Ladies and gentlemen, this wasn’t the Stone Age. It wasn’t the era of campfire incantation by Celtic druids. It was 1978, and a former biochemist was helping a former biophysicist, both of them keen for using molecular methods to plumb evolutionary mysteries, by sharing the latest and hottest scientific data.

  34

  Linda Bonen and her Woesean skills also helped confirm the second major postulate of the endosymbiosis theory: that mitochondria, as well as chloroplasts, are descended from captured bacteria.

  The big point of debate on mitochondria in the 1970s was whether these crucial organelles had arisen from increasing complexity within the eukaryotic cell or, alternatively, had come originally from outside the cell in the form of a captured bacterium. The first view embraced conventional wisdom: somehow the cell innards had evolved toward greater complexity by gradual differentiation, assembling new structures, including a nucleus, chloroplasts in plants, and mitochondria for energy packaging. Maybe these organelles coal
esced from ambient materials, like stardust forming planets. Or maybe they pinched off from some other internal organ, like an appendix, then floated free. Nobody could say. The second view, proposing external origin, echoed Lynn Margulis and her assertions of endosymbiosis. By comparing catalogs of DNA fragments, Bonen and her colleagues in Halifax gave new support to what Margulis and Ivan Wallin had proposed: that mitochondria originated when some sort of single-celled organism (the pre-eukaryote host cell, whatever it was) swallowed a bacterium and then failed to digest it, or became infected by a bacterium and then failed to cure itself, or was otherwise entered by a bacterium and allowed the thing to stay. This signal event happened only once. Descendants of that internalized bacterium became the first mitochondria.

  Among the biochemists at Dalhousie University was another young assistant professor, Michael W. Gray, lately arrived from western Canada by way of a postdoc at Stanford. Gray had grown up in Medicine Hat, Alberta, a “prairie kid,” by his own later description, and then gone to the big city, Edmonton, for undergrad work and a PhD. He trained as an RNA biochemist at a time when that subject seemed entirely detached from evolution. He had never taken a course in evolutionary biology, and his work concerned cellular systems as they function in the present, not their deepest origins. His dissertation topic involved transfer RNA (tRNA), which carries amino acids to the ribosomes for building proteins. He had never heard of Carl Woese or Lynn Margulis. Then he happened to read a journal paper about transfer RNA in a certain fungus, and the paper noted that this tRNA seemed to be traceable to the fungus’s mitochondria. Hmm, he wondered, what’s transfer RNA doing in mitochondria? Since when does protein building have any role in their known functions? Ribosomes make proteins. Mitochondria make ATP. The presence of transfer RNA suggested something else about mitochondria, something that Gray and other RNA chemists hadn’t known: mitochondria contain ribosomes of their own, almost as though they were (or had once been) independent cells. Why?