The Tangled Tree

by David Quammen

  That little riddle led Gray to focus his next major research effort on mitochondria—more specifically, on the DNA and RNA found in mitochondria of plants. This seemed important because it might illuminate mitochondrial origins. He looked in particular at wheat. During his dissertation research, he had turned to commercial wheat germ as a source of the tRNA he needed, and now it occurred to him that wheat germ again might be a good source of mitochondria and the genetic material they contain. As Gray snooped through the literature, having set off on that project, he came upon the work of Margulis. He had missed her 1967 paper at the time, but now he read her 1970 book, Origin of Eukaryotic Cells, offering a fully elaborated endosymbiotic theory, including mitochondria origins. Conventional wisdom be damned.

  “Was she perceived back then as being radical or flakey?” I asked Mike Gray on an afternoon forty years later.

  “Uh-huh,” he said. “Right from the beginning, I think.”

  We had shared lunch at a Turkish restaurant around the corner from Dalhousie, then returned to a small office at the back of Gray’s laboratory. He was retired now, and just closing down lab operations after a quiet but distinguished career devoted largely to the study of mitochondria. He recalled the negative reception to Margulis’s theory, the scientists who “pooh-poohed the idea and had their own scenarios” of nonendosymbiotic origin. There were debates. She was marginalized. But to Gray, her theory suggested that “maybe this was a remnant of a bacterial system that did protein synthesis inside mitochondria. Which I thought was pretty cool.” Soon he too, like his colleague Ford Doolittle, found himself coming to her support.

  It began as a spin-off from the work Linda Bonen did with Doolittle. Gray himself had a grad student—Scott Cunningham, the very first grad student he supervised—who happened to be talking with Bonen one day, and they hit upon the idea of applying the Woesean methods to mitochondria from wheat. They would test that pillar of the endosymbiotic theory—the bacterial origin of mitochondria in all eukaryotes—by looking at ribosomal RNA in the mitochondria of this particular eukaryote, an agricultural plant, familiar to Gray since the PhD work in Edmonton. Now he and his partners would extract rRNA from wheat mitochondria, chop the molecules into fragments, sequence the fragments, and see how their catalogs of short sequences compared with catalogs from other organisms, including bacteria. One problem: working with any extract of mitochondria was notoriously difficult, because mitochondria themselves are not very abundant, compared with chloroplasts within plant cells, and harder to isolate. But that was the beauty of wheat, as Gray understood. The germ of the wheat kernel—that little nubbin inside the grain, constituting the embryo from which a new plant would germinate—contains enough mitochondrial rRNA for such an experiment.

  “Where would you get the raw wheat?”

  “From western Canada,” Gray said. A kid from Medicine Hat knows wheat. “I had it shipped down here.” He had previously gotten his wheat germ from a local flour mill in Halifax, finding it very useful for some experimental purposes. But that wheat germ, having been processed, wouldn’t germinate. You couldn’t grow new plants from it, and if you couldn’t grow new plants, you couldn’t feed in the radioactive phosphate for labeling molecules. Gray and Cunningham wanted wheat seed, as it would come fresh from the field, or at least fresh from the grain elevator standing above a railroad spur on the western plains.

  So Gray ordered it by the bag from a supplier back in Alberta. Cunningham worked out the laborious mechanics of extracting the viable wheat embryos from the seed wheat, using a kitchen blender and sieves, until they had a small, precious quantity of the stuff: 16 grams (about a half ounce). They sprouted those wholesome Canadian wheat embryos on filter paper, adding water and, oh yes, “a horrendous amount of radioactivity,” for labeling that would show on the films. From the wheat embryos, they isolated mitochondria and extracted mitochondrial rRNA. Then, with Linda Bonen leading this part of the process, they cut the rRNA into fragments, exposed the films, read the fingerprints, and made catalogs.

  Gray had chosen well when he picked wheat, because rRNA within plant mitochondria tends to mutate more slowly than the equivalent rRNA within animals. The similarity to its bacterial counterparts and their possible shared ancestors is therefore more pronounced. The work yielded clear findings, as reported in another journal paper, with Gray and Cunningham now added as coauthors alongside Bonen and Doolittle. The Margulis hypothesis was confirmed. Based on their ribosomal RNA, the mitochondria of wheat do not resemble wheat. They are alien little beings, co-opted into wheatly service. They came from elsewhere. They resemble bacteria.

  35

  For almost ten years after that publication on wheat mitochondria, a sizable (or at least, vocal) minority of biologists in the field continued to reject the endosymbiosis theory. Part of their resistance was to swallowing it whole, in the three-part version proposed and promoted by Margulis: chloroplasts derived from cyanobacteria, mitochondria derived from another sort of bacteria, flagella derived from spirochetes or some similar bacteria. That was too much.

  There may also have been subliminal distaste for its implications, including one that still seems almost too personal for comfort: that all animal cells, all human cells, all your cells and my cells and the cells of the doubters, are powered by captured bacteria functioning as organelles. These are not the many teeming bacteria that live in your stomach or your armpits, remember; your gut microbes and other little passengers have been internalized at the body level but generally not within your cells. I’m talking about all those captured-and-transmogrified bacteria that are more fully integrated into your being. These captives have, over the course of maybe two billion years, become part of the machinery inside the cells from which your cells are descended. Their DNA is part of your DNA—the part received exclusively from your mother. Why mother only? Because mitochondrial DNA is passed along via eggs and not via sperm.

  In plainer words: the endosymbiosis theory stipulates that we are all composite creatures, not purely and unambiguously individuals. Small wonder that it was slow to take hold. Nobody likes that Jason Bourne feeling of being told that you aren’t who you thought you were.

  Another challenge to the theory arose when researchers, using electron microscopes, began looking closely at the physical structure, and the amount, of the genomic material in mitochondria and chloroplasts. Both mitochondria and chloroplasts are organelles (that is, organ-like subsidiary units within a eukaryotic cell), yet they have their own genomes, distinct from the genome tucked away on chromosomes within a eukaryotic nucleus. Under microscopic inspection, the DNA in at least some mitochondria and chloroplasts appeared as circular chromosomes, ring shaped, rather than linear chromosomes, as found in the nuclei. Spinach, for instance, held chloroplasts containing tiny rings of DNA. That much supported the theory, because bacteria too carry their genomes on circular chromosomes. If mitochondria and chloroplasts were bacteria, or derived from bacteria, it stood to reason that their chromosomes would have the same shape.

  But these circular chromosomes seemed much too small; far smaller than the chromosomes of any known bacteria. The amount of information they could carry—the number of base pairs, the number of genes—was minuscule compared with a bacterial genome. Again the example of spinach: scientists who scrutinized its chloroplasts, at the University of Düsseldorf in Germany, found teeny little rings of DNA, only one-thirtieth the size of a typical bacterial chromosome. So a new mystery had to be solved: if mitochondria and chloroplasts had indeed originated as bacteria, where were all their missing genes? Gone entirely, shucked off, withered away? Or had those genes somehow been transferred elsewhere within the host cell—maybe into the nucleus itself—from which they could still contribute their gene products to the life of the cell? This little conundrum hinted toward one of the most unexpected revelations (I’ll explore it later) in the saga of the tangled tree: the importance, throughout life’s history, of sideways gene transfer from one organism to an
other.

  Still another reason for resistance to the endosymbiosis theory, back in the early 1980s, was an absence of evidence linking mitochondria with one specific kind of bacteria. If all mitochondria had originated from a single captured bacterium, okay, but which one? There was no prime candidate. That sort of match had been made for chloroplasts—aha, they were cyanobacteria—but not for mitochondria, no. The bacterial precursor hadn’t been identified. Mike Gray and Ford Doolittle published a long review paper on the whole subject in 1982, titled “Has the Endosymbiont Hypothesis Been Proven?” Somewhat surprisingly, given their own 1977 paper on wheat embryos, they conceded that—with regard to mitochondria, at least—it had not.

  Gray and Doolittle were fastidious about this matter of proof. Their paper began with a two-page introduction stipulating what sort of empirical data would potentially confirm the theory, and distinguishing such data from circumstantial evidence. Some forms of “proof” were more proovy than others. Mere correlations could have alternate explanations. This section was a display of thinking and writing that bore the philosophical style of Ford Doolittle. Then followed a section on chloroplasts and cyanobacteria, drawing on Doolittle’s research focus. After that came a disquisition on RNA and mitochondria, played from the strengths of Mike Gray.

  The gist of the whole paper was that, yes, the chloroplast hypothesis stands as proven but, no, the mitochondria hypothesis doesn’t—not quite, not yet. Near the end, Gray and Doolittle mentioned their own work on wheat mitochondria as suggestive but inconclusive. Those wheat mitochondria did resemble (based on rRNA sequences) the bacterium E. coli. Anyway, no better match had been found among other bacteria, by other researchers, in the five years since their earlier paper. But that was just wheat, one plant. And when it came to the mitochondria of animals and fungi—two other big branches of the eukaryotic limb of life—there was no good match at all. There was no persuasive proof of endosymbiosis. There was no favored candidate for the role of the Bacterium That Became Mitochondria. (Conclusive evidence wouldn’t be offered until fifteen years later, when another paper coauthored by Mike Gray matched a full mitochondrial genome with a particular bacterial group.) For the meantime, that question was still in play.

  Carl Woese reenters the story here. In 1985 his lab published a paper titled “Mitochondrial Origins,” announcing that the link had been found. They could place the progenitor of all mitochondria, Woese’s team claimed, within a group of bacteria that still thrive on planet Earth. Turns out the progenitor’s modern relatives are all around us, quietly parasitic, causing galls in walnut trees, in grape vines, and in other plants. They belong to the alpha subdivision of the purple bacteria, an unusual group now known as proteobacteria.

  Woese and his colleagues had again used 16S rRNA, his preferred molecule, as their standard of evolutionary relatedness. Mitochondria have an rRNA molecule closely equivalent to 16S in bacteria, but how close, and to which bacteria? By now, sequencing techniques had improved. Fred Sanger had made still another important contribution, devising a new method of sequencing DNA (not RNA) that was faster and more accurate than anything done earlier. Woese adopted it for his own effort, and thereby made a change that may seem subtle or obscure but was important: instead of extracting the rRNA molecule from ribosomes (in tiny quantities) and sequencing portions of it, his team extracted DNA from genomes and (after multiplying the quantities by DNA cloning) sequenced portions of that master molecule—the portions that coded for 16S rRNA in bacteria and its mitochondrial equivalent. They read the blueprint, that is, instead of measuring the house. It was a methodological adjustment yielding the same sort of information as the old way, but better and quicker. Carl Woese had earned the right to such a shortcut—sequencing DNA, which could be had in largish quantities, rather than rRNA, eked out in tiny smidgens—after all the mind-numbing efforts he had made for a dozen years. This was the beginning of the new information flow, and there again he was, an early adopter of novel methods, but asking deeper questions than most other researchers.

  First author on the 1985 paper was Decheng Yang, a doctoral student who had come to Illinois from a university in northeastern China, eventually joining the Woese lab. He and Woese and their collaborators compared rRNA genes from seven different organisms to see which of six prokaryotes (five kinds of bacteria and one representative of the Archaea, Woese’s new kingdom) best matched the mitochondria from a eukaryote. The archaeon was included as a sort of outlier, thrown into the lineup for breadth. Bacteria were the real suspects. For their eukaryote, the team chose wheat, using the 16S rRNA sequences produced from wheat mitochondria by Mike Gray’s group. The five bacteria they picked for comparison included the ever-available E. coli, the cyanobacterium Anacystis nidulans, two others you can forget about, and a microbe known as Agrobacterium tumefaciens, named for its troublesome habit of causing tumors in agricultural plants. The last of those, A. tumefaciens, was an alpha-proteobacterium.

  Many alpha-proteobacteria have evolved lifestyles that entail living inside eukaryotic cells. The pathogens that cause typhus and Rocky Mountain spotted fever in humans, for instance, are alpha-proteobacteria, doing their damage as parasites within their victims’ cells. Woese’s bug, A. tumefaciens, is innocent of causing human disease but plays hell on plants. And the comparative study showed that it’s the one: the closest match to wheat mitochondria. Mike Gray was on sabbatical back at Stanford at the time. Woese called him to share the news.

  This finding was more vast and dramatic than it might sound. It involved more than wheat. Because other research had already established that mitochondria originated just once in all the history of earthly life, from a single instance of a captured bacterium, Woese and his young team could claim that they had narrowed the search for the ancestor of all mitochondria in complex cells. That ancestor was an alpha-proteobacterium of some sort. Its descendants exist in all of us, powering our cells, making complexity possible.

  On the fourth page of their paper, Woese and his group offered a simple figure: another tree. It was a tree of mitochondria and bacteria. It showed wheat mitochondria and mouse mitochondria and fungal mitochondria as twigs, all clustered closely on the same little branch with the twig representing that bacterium, the tumor maker, A. tumefaciens. If they had sequenced the equivalent rRNA gene in human mitochondria, that would have sprouted here as well. Everything else stood apart—other branches, other limbs, other parts of our genome, our being, our identity (whatever that is), all of them too distant to fit within the figure. Arboreal illustration was starting to get complicated.

  36

  By the time this vindication arrived, in 1985, Lynn Margulis was a full professor at Boston University (where she had begun as an adjunct assistant professor) and an elected member of the National Academy of Sciences, America’s foremost scientific advisory body. Her marriage to Nick Margulis had ended, and her youngest of four children was sixteen. Three years later, she moved to Amherst, as distinguished university professor at the University of Massachusetts, and remained there to the end of her life. She taught. She advised graduate students. She made instructional films and videos. She welcomed houseguests and cooked meals. She seemed to love people, conversation, engagement with humans as much as she loved engagement with ideas.

  She continued to publish journal papers, on an increasingly wide range of topics, some provocative, some technical, as well as popular articles and books, promoting her ideas about endosymbiosis and embracing certain other notions, theories, and viewpoints that seemed even more far-fetched than endosymbiosis once had. She became an AIDS skeptic, challenging the reality that an agent known as human immunodeficiency virus (HIV) is the cause of the syndrome. She became a 9/11 doubter, calling it a “false-flag operation” contrived by unknown parties for dark political purposes. She had long since espoused the Gaia hypothesis, developed by her and the English chemist James Lovelock, which views planet Earth as a self-maintaining system that regulates its own biochemistry, analogo
us to a single living organism. That brought her a lot of acclaim from people who took the living-organism part literally (she didn’t) and viewed Gaia as an almost mystical insight. Margulis herself wasn’t mystical. She favored evidence, argument, and the material world of nature (especially as observed through microscopes), even when she followed those leads into strange terrain.

  She embraced the medically controversial idea of chronic Lyme disease: that the Lyme pathogen, a spirochete, can hide in the human body as a chronic infection, unconquerable by normal doses of antibiotic. She endorsed for publication, in the Proceedings of the National Academy of Sciences, an extraordinarily odd and unpersuasive paper by a retired British zoologist named Donald I. Williamson, who argued that butterflies and their larval form, caterpillars, evolved as separate species, which had later become linked into one life-history form by some ineffable process of hybridization. Caterpillars and butterflies as different creatures conjoined, yes: like a tadpole hybridized with bird, to yield a single new beast with two stages of life. It was analogous, Williamson argued, to the symbiogenesis of eukaryotic cells.