sign in sign up
<<
Home , Mark Wheelis, Monera, Phylogenetic tree, Sol Spiegelman

NSCASCPERSPECTIVE CLASSIC PNAS

Phylogeny and beyond: Scientific, historical, and conceptual significance of the first of

Norman R. Pacea,1, Jan Sappb, and Nigel Goldenfeldc aDepartment of Molecular, Cellular and Developmental , University of Colorado, Boulder, CO 80309; bBiology Department, York University, Toronto, ON, Canada M3J 1P3; and cInstitute for Genomic Biology and Department of Physics, University of Illinois, Urbana, IL 61801

Edited by Edward F. DeLong, Massachusetts Institute of Technology, Cambridge, MA, and approved December 20, 2011 (received for review November 29, 2011)

In 1977, and George Fox published a brief paper in PNAS that established, for the first time, that the overall phylogenetic structure of the living world is tripartite. We describe the way in which this monumental discovery was made, its context within the historical development of evolutionary thought, and how it has impacted our understanding of the emergence of life and the characterization of the evolutionary process in its most general form.

fundamental breakthrough in Several aspects of the paper by Woese and and bringing to the biological occurred in and Fox (1) sparked skepticism. One fore the deep limitations of earlier ac- A 1977, and most did was the arcane nature of the molecular counts of the evolutionary process. not notice. The paper by Woese data, which few could appreciate. The re- and Fox (1) in 1977 was 2.5 pages in length liance on a single to trace major Lead Up to the Paper and contained a single table of numbers trends in was an equally alien The 1977 paper by Woese and Fox (1) was that compared sequence snippets derived concept. Some quibbled about the name an early example of what we would today from small subunit rRNAs of different archaebacteria; others objected to The call molecular —the com- . The table provided the first Times publicity. Most impor- parison of macromolecular sequences to gene sequence-based quantitative assess- tantly, however, the conclusions of the infer genealogical and thereby, evolution- ment of phylogenetic (evolutionary) rela- paper flew into the face of the common ary relationships. The notion of comparing tionships between representatives of the wisdom of the time regarding the basic sequences to infer relationships was put major known kinds of organisms (1). The divisions of biology and the nature of early forward in 1958 by Francis Crick (3) and paper showed that all cellular life falls into evolution. It was generally believed—and more formally, by Emil Zuckerkandl and one of three large relatedness groups: still is taught in our textbooks—that life Linus Pauling in 1965 (4). This was a (our kind of cells, which con- is of two kinds: on one hand, eukaryotes time when determination of se- tain a ), eubacteria and on the other hand, , which quences had become, to some extent, [Woese and Fox (1) termed the group and lack nuclear membranes and as the name tractable with protocols developed by Fred fi this group is where classically studied implies, were supposed to have preceded Sanger (5), who received his rst Nobel fit], and archaebacteria [an un- and evolved into eukaryotes. However, Prize for that development and the de- usual group of recently described organ- eubacteria and the newly discovered group termination of the amino acid sequence of isms named by Woese and Fox (1) to of archaebacteria both lacked nuclear insulin in the 1950s (5). Protein bio- distinguish the group from eubacteria]. In membranes. Eukaryotes seemed not de- chemists began to develop phylogenetic describing the phylogenetic relationships, rived from either bacteria or arch- relatedness maps, phylogenetic trees, based on amino acid sequences derived the results also charted the first scientific aebacteria; all three kinds of organisms from various organisms, mainly . view of deep evolutionary history. Both seemed to represent aboriginal lines Russell Doolittle sketched out these fundamental aspects of biology, the of descent. evolution using blood-clotting fibrinopep- phylogenetic structure of life and the In this retrospective, we view the 1977 tides in the work by Doolittle and Feng course of early evolution, previously were paper by Woese and Fox (1) from three (6); the work by Fitch and Margoliash (7) standpoints. First, we discuss the specific only realms of speculation. used the mitochondrial protein cyto- However, the methods and data used in accomplishments of this landmark paper chrome C to relate animals and some the work by Woese and Fox (1) were and how the program of research initiated fungi. However, not all organisms possess unfamiliar to most biologists, even mo- and led by Woese from the late 1960s to cytochrome C, and for that reason alone, lecular biologists. Traditional biologists, the present day has spawned a revolution fi its amino acid sequence could not be used students of and animals, paid little in and other elds contin- to infer the patterns of relationships attention, because the results had little gent on microbiology, including among all of life. bearing on their interests. Because of and the health . Second, we dis- Carl Woese came to the study of evo- a joint press release by the National cuss the place of that paper in the history lution from a background in biophysics and Aeronautics and Space Administration of , where its un- and the National Science Foundation that precedented use of molecular sequences ’ fi supported Woese s research, the paper associated with rRNA provided the rst Author contributions: N.R.P., J.S., and N.G. wrote the paper. was heralded on the front page of The New window into the deep timeline of life, one The authors declare no conflict of interest. York Times for discovery of “a third form independent of theoretical prejudices that This article is a PNAS Direct Submission. of life” (2). However, the few biologists had flawed earlier efforts to classify life. See Classic Article “Phylogenetic structure of the prokary- who noticed sometimes reacted negatively, Finally, we describe how the under- otic : The primary kingdoms” on page 5088 in issue and articles denouncing the claims were standings sparked by the paper are bring- 11 of volume 74. published. Subsequent developments ing a new face to the study of evolution by See Classic Profile on page 1019. showed that the methods and conclusions compelling biologists to address founda- 1To whom correspondence should be addressed. E-mail: of the paper were sound. tional issues related to the very concepts of [email protected].

www.pnas.org/cgi/doi/10.1073/pnas.1109716109 PNAS | January 24, 2012 | vol. 109 | no. 4 | 1011–1018 Downloaded by guest on September 28, 2021 an interest in the genetic code and its RNases, and the sequences of the resulting origins. During the early 1960s, the nature oligonucleotides were determined by di- of protein synthesis and the makeup of the gestion with other nucleases. Next, frag- genetic code were just being worked ments of incomplete digestion of the RNA out (8). Woese was a contributor to early were isolated and digested completely, thought on the genetic code and had and the digestion products were analyzed; conducted experimental studies to try to eventually, the sequence could be inferred understand the chemical basis of the ca- from the oligonucleotide contents of nonical assignments of different amino overlapping fragments. Mitchell Sogin, acids to particular codons (the DNA or then a graduate student with Woese, RNA base triplets that specify the amino learned the techniques from David H. L. acid sequence of a protein during protein Bishop, a postdoctoral student from synthesis) (9). His 1967 book The Genetic Fred Sanger’s laboratory who was then Code: The Molecular Basis for Genetic working in the Sol Spiegelman laboratory Expression focused prescient attention on at the University of Illinois. Sogin set up the RNA elements of the protein synthe- the necessary facility for Woese’s group. sizing machinery (10). Woese, along with Woese and his students determined sev- Francis Crick (11) and Leslie Orgel (12) eral bacterial 5S rRNA sequences. They after him, are considered founding cham- showed that the rRNA sequences could be pions of the idea that nucleic acids played used as phylogenetic markers for bacteria more than template roles in the origin of (15). They also showed that evolutionary biological systems, thus giving rise to the variation in sequences could be used to notion of a hypothetical prebiotic RNA determine how the RNAs fold into sec- world in which nucleic acids served as both ondary structure (so-called phylogenetic catalytic entities and genetic templates comparative RNA structure analysis) (16). (13). Woese was concerned that the However, it soon became clear that 5S Fig. 1. Ribonuclease T1 oligonucleotide finger- emerging paradigm for the mechanism of rRNA, at only 120 nt in length, was too print. As outlined in the text, data reported in the protein synthesis was too static and had small in size and hence, information con- paper by Woese and Fox (1) in 1977 consisted of no evolutionary dimension. As early as tent to provide for accurate phylogenetic catalogs of oligonucleotide sequences derived 1969, as articulated in a letter to Francis assessments. from RNase T1 digestion of small subunit (SSU) Crick, he understood that the only way to The SSU rRNA, at 1,500–2,000 nt, was rRNAs. The first step of the analysis involved res- reveal the essence of the process was to information-rich, but because of its rela- olution of RNase T1 digests of 32P-labeled rRNA study its conservation and variation in tively large size, it was practically impos- by 2D electrophoresis and locating labeled oligo- different organisms—its evolution—in sible to determine the entire sequence nucleotides on the 80 × 100-cm sheet of electro- a phylogenetic framework. He set out to using the Sanger method. Woese posited, phoresis paper by autoradiography. The result, an RNase T1 fingerprint of the specific RNA, is shown; do that study by comparing rRNA se- however, that the full sequence of the notations were made by Woese during the anal- quences. This task was daunting at that RNA was not necessary for phylogenetic ysis of archaeal rRNAs. Larger phylogenetically time before the development of rapid se- comparisons and that sufficient infor- informative oligonucleotides are in the upper quencing protocols, but it was the only way mation was available in the collection of one-half of the pattern. to quantify evolutionary change. oligonucleotide fragments that result from All cells contain ribosomes, which carry specific nuclease digestions of SSU out protein synthesis and typically are rRNA. Woese argued that any oligonu- paper for secondary and sometimes, ter- composed of ∼50% RNA and 50% pro- cleotide six residues or longer had a low tiary digestions with other ribonucleases to tein. The ribosome consists of two sub- probability of random occurrence in determine the sequences. Preparation and units: a small subunit (SSU), which a molecule the size of the SSU rRNA; analysis of any particular SSU rRNA contains the 1,500- to 2,000-nt-long SSU therefore, they could be assumed to be were somewhat risky processes. They in- rRNA (also called 16S or 18S rRNA to homologous in different organisms and volved labeling cultures with many denote size), and a large subunit (LSU), have common ancestry, and they could be millicuries of 32P-orthophosphate, pro- which has two RNA molecules (a 120-nt used in phylogenetic assessments. He and cessing the highly labeled RNA, and con- 5S rRNA and a 3,000- to 5,000-nt LSU his colleagues began to generate catalogs ducting up to 5- to 8-kV electrophoresis rRNA). [Most eukaryotes contain a fourth of sequences of oligonucleotides that re- in 100-L tanks filled with refined kerosene as LSU rRNA (5.8S rRNA), but this RNA sulted from digestion of the RNA with a coolant. The analysis of RNA sequences in corresponds to one end of the bacterial, ribonuclease T1, which cleaves at G resi- these ways probably could not be con- archaeal, and some eukaryotic LSU dues and therefore, produces oligonu- ducted today because of safety regulations. rRNAs.] Woese began with 5S rRNA, cleotides that are comprised of some because its small size rendered it amena- collection of U, C, and A with a single G. Three Kinds of Life ble to the sequencing technology of the The resulting oligonucleotides were re- George Fox joined Woese at the University time. He studied bacteria because of his solved by 2D electrophoresis, first on cel- of Illinois as a postdoctoral fellow. When background in working with lulose acetate at pH 3.5 and then on DEAE the catalogs of SSU sequences proved subtilis and the practical requirement to cellulose paper at pH 1.9 using Sanger’s informative, Fox worked with William prepare highly radioactive RNA for protocols. The autoradiogram shown in Balch, then a graduate student in the sequence analysis. Fig. 1 is an example of such an RNase T1 laboratory of Ralph Wolfe at the Univer- RNA sequencing developed in the mid- fingerprint of a 32P-labeled SSU rRNA. sity of Illinois, to prepare 32P-labeled SSU 1960s mainly through the efforts of Fred The positions of the different oligonu- rRNAs from methanogens, organisms Sanger and his colleagues using the general cleotides in the electropherogram reflect that produce methane (natural gas) as approach used for Sanger’s protein size, base composition, and sequence of a metabolic product. Although environ- sequencing protocol (14). A 32P-labeled the particular oligonucleotides. Longer mentally important, methanogens were RNA was digested with base-specific oligonucleotides were excised from the little studied because of their requirement

1012 | www.pnas.org/cgi/doi/10.1073/pnas.1109716109 Pace et al. Downloaded by guest on September 28, 2021 for extremely anoxic conditions for the third urkingdom, or did the three a contentious one, but the large-scale growth, and as a result, there was no branches spring independently from some organization shown in Fig. 2 is generally classification system of these organisms. universal ancestor? accepted. Labeled RNAs went to technician Linda These questions could not be answered Magrum for fingerprinting and then to from the necessarily limited 1977 data Revolution in Microbiology Woese for work up of oligonucleotide and indeed, would not be settled for more The development of a sequence-based sequences and compilation of the catalogs. than another decade when Iwabe et al. (18) phylogenetic framework for the identifi- The SSU catalogs provided the first and Gogarten et al. (19) used paralogous cation of microbes would revolutionize phylogeny and classification of those rooting, a phylogenetic technique de- microbial ecology (24) and more generally, organisms, but more importantly, the veloped by Margaret Dayhoff, a pioneer in bring together evolution and ecological methanogen oligonucleotide catalogs dif- (20), to establish that the studies under a single empirically based fered markedly from catalogs of any bac- origin was deep on the eubacterial line. framework. It had long been appreciated teria that had so far been examined (17). The eukaryotes and archaebacteria that microbial ecology is the critical driver Today, we would compare DNA or RNA seemed to constitute a to the of the global biosphere. However, progress sequences from different organisms ex- exclusion of bacteria. The result came as to achieving some understanding of envi- plicitly in terms of percent identity, an no surprise, because by this time (1989), ronmental microbes had been hampered intuitively meaningful comparison even to the phylogenetic work of Woese’s group by the general necessity to culture micro- a nonspecialist. Comparison of oligonu- had stimulated a large-scale effort, par- organisms, even for identification. This cleotide catalogs was not so straightfor- ticularly in Germany, to determine the need was a crippling constraint, because ward, which possibly contributed to the and biochemistry of few environmental microbes, perhaps lack of comprehension and resulting archaebacteria. Many aspects of the fun- much less than 1%, can be cultured using skepticism that greeted the paper by Woese damental molecular biology of eukaryotes standard methods (25). Consequently, and Fox (1). The single table in the work and archaebacteria proved to resemble the microbial ecologist had little access by Woese and Fox (1) was a matrix each other more than their bacterial to environmental organisms and little that compared association coefficients, counterparts. For instance, during tran- knowledge of even the kinds of microbes SAB values, of different SSU oligonucleo- scription, eukaryotes and archaebacteria that occur in the environment. tide catalogs. (SAB values for catalogs of were both known to use TATA binding With the phylogenetic framework for organisms A and B are calculated as in promoter selection, whereas classification in place and the subsequent SAB =2NAB/(NA +NB), where NA and bacteria use σ-factors, a different basic development of recombinant DNA and NB are the numbers of nucleotides in se- mechanism (21); the DNA replication sequencing technology, rRNA gene quences of hexamers or larger in the enzymologies of eukaryotes and archae- sequences began to be used to sidestep the catalogs of organisms A and B and NAB is bacteria resemble one another far more requirement for culture to identify envi- the number in common to the two so than either resembles the eubacterial ronmental organisms (26). rRNA catalogs.) The higher the SAB, the more version (22). The biochemical and molec- could be isolated directly from environ- similar the sequences and the more closely ular criteria generally supported the root- mental DNA as recombinant clones or the related must be the organisms repre- ing of the tree and a deep, at least partial products of the PCR and sequenced to sented by the sequences. The single table relationship of eukaryotes and archae- identify environmental organisms phylo- in the paper by Woese and Fox (1), the bacteria. The rudiments of a scientifically genetically. Such sequences are incisive matrix of SAB numbers, showed clearly grounded universal were identifiers for microbes, and therefore, that life—at least SSU rRNA sequences— in place. correlations between organisms and envi- fell into three distinct relatedness groups In 1990, Woese, with colleagues Otto ronments became possible. Moreover, or urkingdoms as Woese and Fox termed Kandler and Mark Wheelis, proposed the the rRNA or other gene sequences also them. Each urkingdom was further char- formal designation domains to denote the could be used for development of molec- acterized by collections of signature oli- three major phylogenetic groups, which ular tools, such as in situ hybridization gonucleotides found only in that group they proposed be named Eucarya, Bacteria, probes, for the study of environmental and not the other groups. In addition to and (formerly archaebacteria) organisms in their native environments. these idiosyncratic markers, there were (23). The diagrammatic Surveys of rRNA gene sequences from universal signatures, oligonucleotide se- that they used to support their classifica- different environments continue to expand quences found in all organisms examined. tion is reproduced as Fig. 2. For the first knowledge of microbial life in all of the For the first time, it was actually shown time, a universal tree of life had been phylogenetic domains, and it is clear that all life is related phylogenetically. determined in a scientifically rigorous way. that, so far, we are only scratching the This finding was a seminal finding not The field of molecular phylogeny is still surface of a vast reservoir of microbial previously established but widely and im- plicitly assumed. However, more strikingly and wholly unexpectedly was their data indicating that there were at least three, not two, primary lineages of life. The results of the paper showed that a large-scale map of life’s diversity could be seen as three branches corresponding to eukaryotes, eubacteria, and archae- bacteria. However, where was the root of the tree (assuming there was one)? Where was the origin of it all or some place in the tree that would correspond to a hy- pothetical last universal common ances- tor? Were two of the urkingdoms more Fig. 2. Diagrammatic phylogenetic tree used to illustrate the proposal of three phylogenetic domains in closely related to one another than to 1990. Modified from Woese et al. (23).

Pace et al. PNAS | January 24, 2012 | vol. 109 | no. 4 | 1013 Downloaded by guest on September 28, 2021 diversity. Currently, about 2 million rRNA the course of evolution, one needed to provide the easiest possible identification sequences are held by the sequence data- distinguish trivial characteristics from the of species and genera” (36). bases, and more than 99% of these se- “essential characteristics” or “organs of Even distinguishing different microbes quences are environmental sequences (27). high vital or physiological importance” from one another was troublesome for The 1977 paper (1) and its aftermath (30). As he explained in On the Origin microbiologists. For medical researchers, transformed microbiology by introducing of Species, they were all simply germs as Joseph Lister a phylogenetic framework for exploring called them in 1874; microbe was in- life’s diversity and developing a universal [i]t might have been thought ...that those troduced 2 y later (37). Some thought they tree of life, a natural system of classification parts of the structure which determine the were little animals, and others thought based on genealogy. This phylogenetic habits of life, and the general place of each that they were little plants. Louis Pasteur framework spans most of evolutionary time being in the economy of nature, would be spoke sometimes of microscopic plants, of very high importance in classification. on Earth as a microbial world, hitherto ... sometimes as animalcules and sometimes absent from classical evolutionary biology. Nothing can be more false It may as virus (the Latin word for poison). In even be given as a general rule that the less any part of the organization is concerned the early 1870s, the term bacteria (Greek Evolution Without Microbes for little staff or rod) also began to be with special habits, the more important it “ The neo-Darwinian evolutionary synthesis becomes for classification (30). used for the smallest of germs: for all that emerged in the 1930s and 1940s was those minute, rounded, ellipsoid, rod sha- a conception void of microbial foundation. “Embryological characters,” he said, ped, thread-like or spiral forms” (38, 39). It was formed in a world of two kingdoms: “are the most valuable of all” (30). Throughout most of the 20th century, plants and animals. Microbiology largely Comparative morphology, , bacteria were understood to be plants; remained a world apart from evolutionary embryology, and the record were the their classification was the domain of biology until the development of new ways to distinguish from analogy botanists who referred to them as the fis- concepts and methods for determining the and thus, order organisms according to sion fungi or Schizomycetes as Carl Nägeli phylogenetic relationships based on uni- their evolutionary relatedness. However, had named them in 1857. This notion versal characteristics at the core of the bacteria lacked both morphological com- persists in our age today with common molecular genetic system of all organisms. plexity and a fossil record. One could not use of microflora to describe microbes. The concept of highly conserved char- tell which characteristics were ancient Breaking out of the – dualism acteristics, far removed from the vicissi- and of common ancestry, which charac- would prove to be difficult. tudes of life, from which one could teristics were more recent , In his speculative phylogenetic tree of reconstruct the main course of evolution which characteristics were homologous, 1866, , who coined the had been central to evolutionary biology and which characteristics were analogous. terms , phylogeny, and phylum, from its beginnings. In Philosophie placed bacteria and blue-green Zoologique, Jean Baptiste de Lamarck Microbiology’s Scandal (now ) in a division that he (28) argued that, to classify organisms In his great work , Carl called , and he christened another correctly, one had to distinguish relatively Linnaeus (31) had placed all Infusoria the Protista. The Monera were trivial characteristics that were modified supposed to lack the fundamental division fl (microbes) in one species that he pre- through the in uence of environmental sciently baptized Chaos infusoria. Bacterial of labor of nucleus and cytoplasm exhibited conditions from the essential system of classification remained in chaos for the next in true cells and bridge the gap between organs. The former adaptive traits repre- 200 y. Bacteria were not put into groups the living and the nonliving. Whether the sented the branchings of the tree of life. organization of the bacterium and the based on principles of homology and evo- The latter represented the course of in- blue-green alga fit Haeckel’sdefinition was lution but based on as many characteristics creasing complexity of organization. hotly contested throughout the early as possible and principles of utility for in- Comparisons of the essential system of 20th century (32, 40, 41). A few micro- dustry and medicine. Bacteria were ar- organs, he said, could be made only be- biologists proposed that bacteria and the tween the higher groupings of animals and ranged into a nested of orders, blue-green algae should be granted their not between species or genera; they families, genera, and species based on cell own kingdom, Monera (34, 42). However, shape, plane of cell division, ability to form were less conspicuous in plants and not fl others doubted that the grouping was conspicuous at all in the Infusoria (28). spores and/or colonies, possession of a- monophyletic. Additionally, it was far The existence of highly preserved char- gella, whether cells were connected, from clear that the blue-green algae acteristics through which one could trace whether they were branching, staining re- really lacked a nucleus and how smaller the tree of life was equally central to actions, relation to temperature and oxy- bacteria such as Rickettsia and Chlamydia ’s theorizing in regard to gen, pigment production, pathogenicity, could be distinguished from viruses (36). descent with modification. As Darwin and a broad diversity of biochemical prop- In a famed paper titled “The Concept of wrote to his friend Thomas Henry Huxley erties (32). a Virus,” André Lwoff (43) drew what in 1857, the “time will come I believe, Although some microbiologists held he considered to be a unequivocal dis- though I shall not live to see it, when we out hope for a natural classification, one tinction between a virus and a bacterium shall have very fairly true genealogical based on evolutionary relationships (33, on the basis of EM and chemistry. The trees of each great kingdom of nature” 34), by the middle of the 20th century, they virus contained only one kind of nucleic (29). Darwin (30) wrote in On the Origin of were forced to admit that their efforts acid, either RNA or DNA, enclosed in Species that all “true classification is ge- were futile (35). Admitting defeat, Roger a coat of protein; it possessed few, if any, nealogical; that community of descent is Stanier, Michael Doudoriff, and Edward enzymes, and it did not reproduce by di- the hidden bond which naturalists have Adelberg (36) declared in the first edition vision like a cell. In 1962, Stanier and been unconsciously seeking, and not some of their popular text The Microbial World Van Niel (44) wrote a sister paper, “The unknown plan of creation, or ... the mere that “it is a waste of time to attempt Concept of a Bacterium”, that aimed to putting together and separating objects a natural system of classification for bac- resolve issues about the anatomy of the more or less alike” (30). teria ...bacteriologists should concentrate bacterium once and for all. “Any good For Darwin no less than for Lamarck, to instead on the more humble practical ,” they wrote, “finds it intel- have a natural classification that reflected task of devising determinative keys to lectually distressing to devote his life to

1014 | www.pnas.org/cgi/doi/10.1073/pnas.1109716109 Pace et al. Downloaded by guest on September 28, 2021 the study of a group that cannot be readily Clash However, this publication was well and satisfactorily defined in biological The paper by Woese and Fox (1), “Phy- after Zuckerkandl and Pauling (4) had terms; and the abiding intellectual scandal logenetic Structure of the Prokaryotic written “Molecules as Documents of of bacteriology has been the absence of Domain: The Primary Kingdoms,” clash- Evolutionary History,” and Francis Crick a clear concept of a bacterium” (44). ed with the doctrines and methods of (3) had written as far back as 1958 that the “ Borrowing terms from Lwoff’s former classical microbiology in several ways. amino acid sequences of proteins are mentor Edouard Chatton (45, 46), Stanier First and foremost was the – the most delicate expression possible of and Van Niel (44) distinguished pro- dualism. Phylogenetically, the phenotype of an organism and vast karyotic cells (Greek for before karyon or Woese and Fox (1) wrote, life is not amounts of evolutionary information may ” nucleus) from eukaryotic cells (Greek “structured in a bipartite way along the be hidden away within them (3). There for true nucleus). The latter divided by lines of organizationally dissimilar pro- was, indeed, a strange disconnect between and possessed a membrane-bound karyote and eukaryote. Rather it is (at microbiology and molecular biology, one that was finally resolved through the nucleus, an intricate cytoskeleton, mito- least) tripartite” (ref. 1, p. 5090). Second, SSU rRNA sequencing method reported in chondria, and in the case of plants cells, the eukaryotic cell organelles did not the paper by Woese and Fox (1) in 1977. chloroplasts. Prokaryotic cells were small- originate solely in the gradual neo-Dar- winian manner by gene and se- The SSU rRNA sequencing method er and lacked those structures. Stanier and had remarkable predictive success. Most of Van Niel (44) wrote that the lection but rather, saltationally through symbiosis as long had been suggested. the higher taxa above the genus would have to be reclassified on the basis of SSU principle distinguishing features of the The SSU rRNA data left “no doubt that rRNA phylogenies. When that technology procaryotic cell are: 1 absence of internal the chloroplast is of specificeubacterial surprised microbiologists by predicting membranes which separate the resting origin” (1). The case was not yet as cer- unexpected relationships, it was corrobo- nucleus from the cytoplasm, and isolate the tain for mitochondria. The nature of the rated by other data. Nothing was more enzymatic machinery of engulfingspecies(thepureurcaryote) fi striking than the phenotypically diverse and of respiration in speci c organelles; 2 that would have lacked eukaryotic or- fi organisms that came to be included in the nuclear division by ssion, not by mitosis, ganelles was unknown. a character possibly related to the presence archaebacteria urkingdom by 1980 (do- The manuscript received severe criti- of a single structure which carries all the main archaea). It comprised methanogens, cisms when it was submitted to PNAS in the genetic information of the cell; and 3 the found in the guts of rumens, extreme presence of a which contains summer of 1977 (1). One reviewer rec- halophiles known for rotting salted fish, a specific mucopeptide as its strengthening ommended that it not be published on and extreme thermoacidophiles that live in element (44). methodological grounds that their conditions that would cook other organ- claim for a tripartite division of the mi- fi isms (53). The group was found to have The prokaryote was, thus, de ned crobial world was as unfounded as their many highly conserved characteristics in largely negatively. Although they had for- claims in regard to symbiosis and the ori- fi common: their walls lacked murein pepti- saken a natural classi cation for bacteria, gin of eukaryotic organelles. For Woese doglycan (the only positive characteristic they asserted nonetheless that the pro- and Fox, rRNA was a highly conserved, of the prokaryotes), their tRNAs were karyote was a monophyletic group of nonadaptive structure at the core of all unique, the lipids in their membranes were common origin (44). “All these organisms organisms. Nested deep in the center of ether- and not ester-linked, their tran- share the distinctive structural properties essential cellular functions, the SSU rRNA scription enzymes were unlike the enzymes “ associated with the procaryotic cell ...and was, as Woese (50) later put it, the ulti- of the classic bacteria, and their viruses we can therefore safely infer a common mate molecular chronometer.” However, were unlike the viruses of the classic origin for the whole group in the remote for critics, an rRNA oligonucleotide cata- bacteria. evolutionary past” (47). “In fact,” they log was merely a trait like any other, and The SSU rRNA phylogenies also de- fi wrote, “this basic divergence in cellular classi cation on its basis held no more finitively resolved the venerable question of structure, which separates the bacteria and validity than classifying birds, , and whether chloroplasts and mitochondria blue-green algae from all other cellular on the basis of their possession of originated as symbionts (1, 16, 54). Con- organisms, probably represents the great- wings. That one molecule or many could jectures in regard to the symbiotic origin est single evolutionary discontinuity to be be used to discern phylogenetic relation- of cell organelles had been discussed found in the present-day world” (47). ships was more than many classical mi- throughout the 20th century, but that The kingdom Monera, thus, rose like crobiologists could accept; the molecular question remained far outside the realm of approach was pronounced by some as a phoenix from the ashes. It was to be one empirical inquiry (55). As Wilson (56) doomed to failure (51). The Cell in of five kingdoms, along with animals, commented in his famed book The belief that it was impossible to know Development and Heredity, “[t]o many, plants, fungi, and (microbial eu- the phylogenetic relationships among no doubt, such speculations [symbiotic karyotes) (48, 49). microbes because of a lack of fossil record origin of organelles] may appear too fan- The prokaryote–eukaryote dichotomy was well-entrenched in the minds of tastic for present mention in polite bi- for the description of life’s diversity was microbiologists. As Stanier et al. (52) wrote ological society; nevertheless it is within quickly instated among the canons of bi- in The Microbial World in 1970, the range of possibility that they may ology. Born in a time when the search someday call for more serious consider- for a natural system for [r]eflection and experience have shown, ation” (56). had been abandoned, the prokaryote however, that the goal of a phylogenetic Such speculation did, indeed, call for fi concept contained just as many untested classi cation can seldom be realized. The more serious consideration in the early assumptions as the old taxonomic con- course that evolution has actually followed 1960s with evidence that mitochondria and jectures: the prokaryote was a mono- can be ascertained only from direct his- torical evidence contained in the fossil chloroplasts each possessed their own phyletic grouping that preceded and gave record. This record is at best fragmentary DNA and translation apparatus. That these rise to eukaryotes. Indeed, nothing was and becomes almost completely illegible in organelles might have originated as sym- known of prokaryote or eukaryote origins; Precambrian rocks more than 400 million bionts was the conclusion of virtually every they could have arisen one or many times. years old (52). paper, showing that they possessed their

Pace et al. PNAS | January 24, 2012 | vol. 109 | no. 4 | 1015 Downloaded by guest on September 28, 2021 own DNA (57–59). However, proof of mid-1960s, one with the goal of under- path represented by the on origin was lacking, and the debates were standing the evolution of the complex Earth, whose dynamics were imperfectly sterile. As Woolhouse (60) remarked, “the structure of the modern cell from the captured by the slogan . time has come to bury this kind of specu- origin of life (32). In this goal, he was Woese and Fox (1), thus, approached lation with, by way of an epitaph, a parody arguably more ambitious than Darwin’s their work in an open-ended way, one not of Wittgenstein’s well-known remark, statement of the problem, which was re- constrained by theoretical prejudices ‘Whereof one cannot know, thereof one flectedinthetitleofDarwin’s great work, about the particular dynamics of the should not speak’” (60). Stanier (61) de- (30). Indeed, evolutionary process. This strategy was creed that the problem of eukaryotic cell Woese’s concept of the evolution prob- brilliant; not only did they uncover an origins would always remain in the realm lem greatly surpassed prevailing views on unanticipated new domain of life, but of metascience: “[i]t might have happened what at that time passed for the deepest also, they pointed out that there were thus; but we shall surely never know thinking on the question. Woese would many possible modes and tempos of with certainty,” Stanier (61) quipped. not find it necessary to cite On the Origin evolution, of which only one—long of Species until 1992 (63), reflecting nei- time-scale, core cellular mechanisms— Evolutionary speculation constitutes a kind ther a stubbornness of character nor could be properly probed by molecular of metascience, which has the same in- a contrived striving for originality. The phylogeny. tellectual fascination for some biologists fact of the matter is that Woese had Woese and Fox’s approach yielded that metaphysical speculation possessed not even read Darwin’s original work a major surprise that could be interpreted for some medieval scholastics. It can be until around 2000, because its relevance directly and plainly without contamination considered a relatively harmless habit, like ’ eating peanuts, unless it assumes the form to Woese sprogramhadseemedremote from theory. The phylogenetic tree im- of an obsession; then it becomes a vice (61). until that time. plied by table 1 in ref. 1 clearly showed So what was Woese’s conception of the three fundamental groupings, results that SSU rRNA phylogenetics belied that evolution problem? First and foremost, it were subsequently confirmed and elabo- statement. revitalized explicitly drew a connection between the rated in compelling detail by Woese and evolutionary biology and provided a new origin of the genetic code, the translation his collaborators (63, 65). At a stroke, basis for empirical investigations of pro- mechanism, and the emergence of cells those data demolished the conception of found new questions concerning the evo- and their organization. For Woese, speci- the prokaryote as a monophyletic group lution of the cell and its components. ation was an epiphenomenon of the evo- that preceded and gave rise to the eu- Molecular phylogenetic methods and lutionary process; the more important karyotic cell. The draft of the phylogenetic concepts constituted a paradigm apart question was that of degree of organiza- tree that became available during the from classical evolutionary biology. Mo- tion. Woese and Fox (1) wrote that “[e] 1990s provided evidence for a last uni- lecular methods for classification based on volution seems to progress in a ‘quantized’ versal common ancestor, which is sum- GC content, DNA–RNA hybridization, fashion. One level or domain of organi- marized above. and amino acid sequencing of proteins had zation gives rise ultimately to a higher begun to revitalize the aim for a phyloge- (more complex) one ... Ideally one would Progenote netic classification of bacteria in the 1960s like to know whether this is a frequent or Woese’s conception of the evolution and 1970s. However, those approaches a rare (unique) evolutionary event” (1). question had already led him to surmise by could not offer universal and quantitative Even earlier, in 1972, Woese (64) had at least 1971 that life, as we know it today, methods for a universal tree of life. argued that descended from an earlier, radically dif- ferent , which lacked the trans- Influence on Evolutionary Biology there are no such things as ‘special’ evo- lation mechanisms of cells (64). His The work by Woese and Fox (1) and its lutionary problems. Evolution is not rationale for this conclusion is important a mixed bag of various ‘historical acci- continuation within the Woese group have ’ ... to appreciate, because the paper by Woese had an enormous impact (e.g., roughly dents All evolutionary problems ap- pear to have important common aspects— and Fox (1) in 1977 bore directly on one-quarter of the 40,000 total citations to the properties of that hypothetical primi- ’ whether they are intracellular problems or Woese s work originate just from refer- problems on higher levels of biological tive entity. ences) (1, 23, 50, 62), and its methods are organization. Thus in ostensibly addressing Woese and Fox (66) had, in parallel to revolutionizing disciplines as varied as ourselves to of genetic codes, their experimental work on molecular marine science and the study of the human translating mechanisms, etc., we are actu- phylogeny, published a conceptual paper microbiome for medicine. This success has ally discovering and defining the basic that introduced the concept of “aprimi- perhaps overshadowed the original moti- principles for all evolution ... The nature ... ‘ ’ tive entity called [the] progenote, to vations and may even have stimulated so of the elements involved changes from recognize the possibility that it had not one example to another, but the patterning much activity in genomics per se that ‘ yet completed evolving the link between ’ of events in time, the principles of evolu- ” Woese s overarching interests in funda- tionary construction,’ seem to remain in- and phenotype (66). The pro- mental evolutionary questions have been variant (64). genotes were hypothetical ancient life relatively neglected. We close out this forms in the throes of developing the re- retrospective with a brief account of how Woese understood that the goal of ex- lationship between nucleic acid and pro- the work by Woese and Fox (1) began, ploring evolution on the longest possible tein. Woese had suggested the existence the extent to which their work has influ- time scales was not to elucidate the idio- of the progenote from rather detailed enced evolutionary thought, and the way syncratic history of genes. “It is too easy to arguments based on his unrivalled in which their ideas are still unfolding. think of evolution solely in terms of gene knowledge of the translation apparatus, Woese and Fox’s discovery that all living , flow in gene pools, and, of whichwasarticulatedinhisbookThe systems are representatives of one of three course, some vague ‘selection’ parameters. Genetic Code: The Molecular Basis for “aboriginal lines of descent” did not Too much emphasis is placed on the Genetic Expression in 1967 (10); he con- emerge in a vacuum (1). It was a water- micro-changes at the expense of the sidered these arguments to indicate that shed event (but not the culmination) in macro-ordering” (64).Hewantedtoun- a general principle of evolution was that an iconoclastic program of research that derstand universal aspects of the process complexity emerged through (i) a process Woese had been pursuing since the of evolution, not the particular sample of refinement and (ii) a two-step cyclical

1016 | www.pnas.org/cgi/doi/10.1073/pnas.1109716109 Pace et al. Downloaded by guest on September 28, 2021 dynamic process that he proposed to Is There Really a Tree of Life? phological diversity so readily observed in explain the quaternary structure of pro- HGT is the capability that organisms the eukaryotic world, and the measure- teins and that turned out to have re- possess to transfer genes to a non- ment of that difference using molecular markable predictive power (ref. 67, p. genealogically related recipient (70). Once sequences was anathema compared with 388). These same arguments suggested thought to be an exclusive aspect of mi- the more descriptive phenotypic dis- “ ” a resolution of the old chicken-and-egg crobial life, HGT has now been docu- tinctions that he favored. paradox of whether translation preceded mented to occur between bacteria and A second issue concerned the purpose of fi the gene or vice versa. The resolution is multicellular eukaryotes (71) as well as a classi cation system. Mayr (78, 79) that both emerged from what we would between eukaryotes (72). With genomic privileged the facility of information re- today call a co-evolutionary process of evidence for the widespread occurrence of trieval over a system based on evolutionary fi fi re nement of translation, in which the lateral gene transfer, the notion of a relationships. In his view, the classi - earliest proteins would have been statisti- universal phylogenetic tree came under cation system should enable a biologist to “ cal in character. In other words, early severe scrutiny and criticism (73). How- quickly identify an organism with a min- translation produced not a single protein imum of effort and loss of time” (79) ever, among Bacteria and Archaea, the fl but a family of similar proteins, any one of genes that are horizontally transferred are rather than re ect the genealogy of the which was adequate for the task at hand. typically metabolic genes or genes that organism. In his response, Woese (80) Thus, early life was a rickety, free-wheel- confer such traits as antibiotic resistance broadened the discussion to include the ing construct whose lack of complexity and not the nonadaptive informational foundational evolutionary questions that could tolerate imprecision at the level of genes that are involved in transcription we have presented above and that, by amino acids. The complex modern trans- and translation. that time, had occupied his thoughts for lation apparatus would have evolved in nearly three decades. It is now widely recognized that there – a progenote evolutionary era, with a can be wide compositional variations in the Interestingly enough, the Mayr Woese tempo and mode based on a high muta- debate did not address one of the most of organisms that are nominally tion rate of the primitive error-prone contentious issues arising from the paper classified together on the basis of their translation system and a pervasive hori- by Woese and Fox (1): the definition of SSU rRNA phylogeny. That within-group zontal exchange of genes within the pop- species in the microbial world. Mayr variation reflects the presence of cosmo- ulation. During this era, there were no (78, 79) had famously defined the Bi- politan genes associated more closely with lineages as such. ological Species Concept in 1942 as groups a particular environment than with any one Woese and Fox (1) noticed that their of interbreeding natural populations that particular organism (74, 75). Woese’s work already implied something funda- are reproductively isolated from each mentally interesting about the progenote original motivation for constructing a deep other, but clearly, this concept did not state. Table 1 in ref. 1 established that the universal phylogenetic tree based on SSU apply to Bacteria, Archaea, or even mi- three urkingdoms had to have been at rRNA sequences to understand the evo- crobial eukaryotes. With the advent of least 3 billion y old each and therefore, lution of the molecular genetic system re- SSU rRNA phylogeny, it became common that the mains largely untouched by HGT. Even practice to define operational taxonomic those elements of the translational appa- units by the requirement of close sequence time available to form each phenotype ratus that are most susceptible to HGT similarity (typically 97%) (24). However, (from their common ancestor) is then short (the aminoacyl-tRNA synthetases) show ... as more and more sequencing capability by comparison we think that this im- that the canonical pattern of the rRNA became available, it became clear that the plies that the common ancestor ...was not tree is largely preserved (76). Moreover, a prokaryote. It was a far simpler entity; spectrum of microbial life was continuous ‘ ’ the frequency of HGT is known to depend rather than discrete because of cosmo- it probably did not evolve at the slow rate on the evolutionary distance between the characteristic of prokaryotes (1). politan genes and HGT. Bacteria seem to lineages concerned, and thus, even if there form a radiation, with no specific species fi In short, Woese and Fox’s early data is signi cant HGT, the impact on the boundaries (27, 81). According to some already suggested to them that life evolved rRNA phylogenetic tree would be ex- estimates, only ∼40% of the genes in from at least a two-step process: a late pected to be rather small: the tree can a particular bacterium typically occur in all stage that they characterized as having reflect both the evolutionary history of the of the genomes of that particular named a slow rate of evolution, one which could as well as the predominant pat- species; the other ∼60% of genes in the be measured using the slowly changing terns of gene transfer into and out of the typical bacterial occur only spo- sequence of rRNA, and an earlier stage lineage (77). radically in other representatives of the refractory to molecular phylogenetic species or not at all (81). More than 30 y analysis, where complexity evolved much Molecular Phylogeny vs. Morphological after the paper by Woese and Fox (1) was more rapidly, leading to the emergence of published, the fundamental biological genotype and phenotype as separate cel- When Woese, Otto Kandler, and Mark concept of species remains unresolved. lular features along with a fully functional Wheelis (23) formally proposed the three translation mechanism. domains as the natural classification of Legacy Woese would return to the progenote taxa in 1990, a direct confrontation with Woese and Fox set out to determine the speculations in the succeeding years, em- the status quo ensued. The issues were degrees of relatedness between all living phasizing especially the characteristics of well-revealed in Woese’s debates with his organisms using rRNA sequences as high gene mutation and horizontal gene most distinguished opponent, Ernst Mayr a marker of cellular evolution on slow transfer (HGT) (68), but by 1998 (69), the (78, 79). First, there was the question of time scales subsequent to the emergence of context had become genomics and the the great molecular and biochemical di- the modern lines of descent. They discov- ever-increasing evidence for the wide- versity of the microbial world compared ered that there are three domains of life, spread occurrence of HGT. Its occurrence with the great morphological diversity ex- not two domains as had been previously was interpreted by some as being anti- hibited by multicellular eukaryotes. For believed. Their work also strongly con- thetical to the entire program of building Mayr (78, 79), the degree of differences strained the nature of life for times shorter a tree of life—all which he had addressed exhibited within the prokaryotic world than about 1 billion y, indicating that so many years earlier. simply could not compare with the mor- before the emergence of a strong phy-

Pace et al. PNAS | January 24, 2012 | vol. 109 | no. 4 | 1017 Downloaded by guest on September 28, 2021 logenetic signal of vertical descent, early life namics. The results challenged the biology has left a richer legacy of accom- had to evolve rapidly; we now suspect that foundations of classical evolutionary the- plishments and promise for the future. its evolution was reticulate in nature. ory, requiring new modes of evolution to be Modern versions of the techniques used by considered, indicating the presence of an ACKNOWLEDGMENTS. We thank Professors Carl Woese and Fox (1) are now routinely used unexpectedly large microbial pangenome Woese and George Fox for reviewing the manu- to sample environments as varied as geo- (“a field of genes” to use Woese’sfavorite script for accuracy. N.R.P. is supported by the Alfred P. Sloan Foundation, J.S. is supported by thermal hot springs and gastrointestinal phrase) (1), and forcing us to reconsider the Social Sciences and Humanities Research microbiomes, providing unprecedented in- basic concepts such as the nature of species. Council of Canada, and N.G. is supported by sight into community structure and dy- Perhaps no other paper in evolutionary National Science Foundation Grant EF-0526747.

1. Woese CR, Fox GE (1977) Phylogenetic structure of the 26. Olsen GJ, Lane DJ, Giovannoni SJ, Pace NR, Stahl DA 55. Sapp J (1994) Evolution by Association: A History of prokaryotic domain: The primary kingdoms. Proc Natl (1986) Microbial ecology and evolution: A ribosomal Symbiosis (Oxford University Press, New York). Acad Sci USA 74:5088–5090. RNA approach. Annu Rev Microbiol 40:337–365. 56. Wilson EB (1925) The Cell in Development and Hered- 2. Lyons RD (November 3, 1977) Scientists discover a form 27. Pace NR (2009) Mapping the tree of life: Progress and ity (Macmillan, New York). of life that predates higher organisms. NY Times, Sec- prospects. Microbiol Mol Biol Rev 73:565–576. 57. Ris H, Plaut W (1962) Ultrastructure of DNA-containing tion A, p 1. 28. de Lamarck JBM, Zoological Philosophy: An Exposition areas in the chloroplast of Chlamydomonas. J Cell Biol 3. Crick FHC (1958) The biological replication of macro- with Regard to the Natural History of Animals, trans 13:383–391. molecules. Symp Soc Exp Biol 12:138–163. Elliot H (1984) (University of Chicago Press, Chicago). 58. Nass MMK, Nass S (1963) Intramitochondrial fibers 4. Zuckerkandl E, Pauling L (1965) Molecules as docu- 29. Darwin C (1857) Darwin to Thomas Henry Huxley: Dar- with DNA characteristics. I. Fixation and electron stain- ments of evolutionary history. J Theor Biol 8:357–366. win Correspondence Project. Letter 2143. Available at ing reactions. J Cell Biol 19:593–611. 5. Sanger F (1959) Chemistry of insulin; determination of http://www.darwinproject.ac.uk/entry-2143. 59. Nass S (1969) The significance of the structural and the structure of insulin opens the way to greater un- 30. Darwin C (1969) On the Origin of Species: Facsimile functional similarities of bacteria and mitochondria. derstanding of life processes. Science 129:1340–1344. Edition (Harvard University Press, Cambridge, MA). Int Rev Cytol 25:55–129. 6. Doolittle RF, Feng DF (1987) Reconstructing the evolu- 31. Linnaeus C (1774) Systema Naturae (Laurentius Salvius, 60. Woolhouse WH (1967) A review of the plastids by JTO – tion of vertebrate blood coagulation from a consider- Stockholm), 10th Ed. Kirk and RAE Tilney-Bassett. New Phytol 66:832 833. ation of the amino acid sequences of clotting proteins. 32. Sapp J (2009) The New Foundations of Evolution: On 61. Stanier RY (1970) Some aspects of the biology of cells fi Cold Spring Harb Symp Quant Biol 52:869–874. the Tree of Life (Oxford University Press, New York). and their possible evolutionary signi cance. Organiza- 7. Fitch WM, Margoliash E (1967) Construction of phylo- 33. Kluyver AJ, Van Niel CB (1936) Prospects for a natural tion and Control in Prokaryotic and Eukaryotic Cells, genetic trees. Science 155:279–284. system of classification of bacteria. Zentralbl Bakteriol eds Charles HP, Knight BCJG (Cambridge University – 8. Judson HF (1996) The Eighth Day of Creation: Makers Parasitenkd Infektionskr Hyg Abt II 94:369–403. Press, Cambridge, UK), Vol 20, pp 1 38. of the Revolution in Biology: Expanded Edition (Cold 34. Stanier RY, Van Niel CB (1941) The main outlines of 62. Balch WE, Fox GE, Magrum LJ, Woese CR, Wolfe RS Spring Harbor Laboratory Press, Plainview, NY). bacterial classification. J Bacteriol 42:437–466. (1979) Methanogens: Reevaluation of a unique biolog- – 9. Woese CR, Dugre DH, Saxinger WC, Dugre SA (1966) 35. Van Niel CB (1955) Natural selection in the microbial ical group. Microbiol Rev 43:260 296. The molecular basis for the genetic code. Proc Natl world. J Gen Microbiol 13:201–217. 63. Wheelis ML, Kandler O, Woese CR (1992) On the na- fi Acad Sci USA 55:966–974. 36. Stanier RY, Doudoriff M, Adelberg EA (1957) The Mi- ture of global classi cation. Proc Natl Acad Sci USA 89: – 10. Woese CR (1967) The Genetic Code: The Molecular Ba- crobial World (Prentice Hall, Englewood Cliffs, NJ). 2930 2934. 64. Woese CR (1972) The emergence of genetic organiza- sis for Genetic Expression (Harper and Row, New York). 37. Carter KC (1991) The development of Pasteur’s concept tion. Exobiology, ed Ponnamperuma C (North-Holland, 11. Crick FHC (1968) The origin of the genetic code. J Mol of disease causation and the emergence of specific Amsterdam), pp 301–341. Biol 38:367–379. causes in nineteenth-century medicine. Bull Hist Med 65. Cavicchioli R (2007) Archaea: Molecular and Cellular 12. Orgel LE (1968) Evolution of the genetic apparatus. J 65:528–548. Biology (Wiley-Blackwell, New York). Mol Biol 38:381–393. 38. Cohn F (1872) Uber die Bacterien, die Kleinstein Leb- 66. Woese CR, Fox GE (1977) The concept of cellular evo- 13. Gilbert W (1986) Origin of life: The RNA world. Nature enden Wesen. Samsung Gemeinverständlicher Wissen- lution. J Mol Evol 10:1–6. 319:618. schaftlicher Vorträge 7:1–35. 67. Goldenfeld N, Woese C (2011) Life is physics: Evolution 14. Sanger F, Brownlee GG, Barrell BG (1965) A two-di- 39. Woodhead GS (1892) Bacteria and Their Products (Wal- as a collective phenomenon far from equilibrium. Ann mensional fractionation procedure for radioactive nu- ter Scott, London). Rev Condensed Matter Physics 2:375–399. cleotides. J Mol Biol 13:373–398. 40. Sapp J (2009) Transcending Darwinism thinking later- 68. Woese CR (1982) Archaebacteria and cellular origins: 15. Rogers MJ, et al. (1985) Construction of the myco- ally on the tree of life. Hist Philos Life Sci 31:161–181. An overview. Zbl Bakt Hyg I Abt Orig C 3:1–17. plasma evolutionary tree from 5S rRNA sequence data. 41. Sapp J (2005) The prokaryote-eukaryote dichotomy: 69. Woese C (1998) The universal ancestor. Proc Natl Acad Proc Natl Acad Sci USA 82:1160–1164. Meanings and mythology. Microbiol Mol Biol Rev 69: Sci USA 95:6854–6859. 16. Fox GW, Woese CR (1975) 5S RNA secondary structure. 292–305. 70. Chia N, Goldenfeld N (2011) Statistical mechanics of Nature 256:505–507. 42. Copeland HF (1938) The kingdoms of organisms. Q Rev in . J Stat 17. Fox GE, Magrum LJ, Balch WE, Wolfe RS, Woese CR Biol 13:383–420. Phys 142:1287–1301. (1977) Classification of methanogenic bacteria by 16S 43. Lwoff A (1957) The concept of virus. J Gen Microbiol 71. Dunning Hotopp JC (2011) Horizontal gene transfer be- ribosomal RNA characterization. Proc Natl Acad Sci – 17:239 253. tween bacteria and animals. Trends Genet 27:157–163. USA 74:4537–4541. 44. Stanier RY, Van Niel CB (1962) The concept of a bacte- 72. Gladyshev EA, Meselson M, Arkhipova IR (2008) Mas- 18.IwabeN,KumaKI,HasegawaM,OsawaS,MiyataT – rium. Arch Mikrobiol 42:17 35. sive horizontal gene transfer in bdelloid . Sci- fi – (1989) Evolutionary relationship of archaebacteria, eubac- 45. Chatton E (1938) Titres et travaux scienti ques (1906 ence 320:1210–1213. teria, and eukaryotes inferred from phylogenetic trees of 1937) (Sottano, Italy). 73. Doolittle WF (1999) Phylogenetic classification and the – duplicated genes. Proc Natl Acad Sci USA 86:9355 9359. 46. Chatton E (1925) Pansporella perplexa, Amoebien á universal tree. Science 284:2124–2129. fl 19. Gogarten JP, et al. (1989) Evolution of the vacuolar spores protégées parasite des Daphnies. Re exions 74. Woese CR (2004) A new biology for a new century. H+-ATPase: Implications for the origin of eukaryotes. sur la biologie et la phylogenie des Protozoairies. Microbiol Mol Biol Rev 68:173–186. – – Proc Natl Acad Sci USA 86:6661 6665. Ann Sci Nat (Zool) 8:5 84. 75. Frigaard NU, Martinez A, Mincer TJ, DeLong EF (2006) 20. Schwartz RM, Dayhoff MO (1978) Origins of prokar- 47. Stanier RY, Doudoroff M, Adelberg EA (1963) The Micro- Proteorhodopsin lateral gene transfer between marine yotes, eukaryotes, mitochondria, and chloroplasts. Sci- bial World (Prentice Hall, Englewood Cliffs, NJ), 2nd Ed. planktonic Bacteria and Archaea. Nature 439:847–850. – ence 199:395 403. 48. Whittaker RH (1969) New concepts of kingdoms or or- 76. Woese CR, Olsen GJ, Ibba M, Söll D (2000) Aminoacyl- 21. Thomm M (2007) Transcription: Mechanism and regu- ganisms. Evolutionary relations are better represented tRNA synthetases, the genetic code, and the evolution- lation. Archaea: Molecular and Cellular Biology,ed by new classifications than by the traditional two king- ary process. Microbiol Mol Biol Rev 64:202–236. Cavicchioli R (Wiley-Blackwell, New York), pp 139–157. doms. Science 163:150–160. 77. Andam CP, Williams D, Gogarten JP (2010) Natural tax- 22. Lao-Sirieix S-B, Marsh VL, Bell SD (2007) DNA replication 49. Whittaker RH, Margulis L (1978) classification onomy in light of horizontal gene transfer. Biol Philos and cell cycle. Archaea: Molecular and Cellular Biology, and the kingdoms of organisms. Biosystems 10:3–18. 25:589–602. ed Cavicchioli R (Wiley-Blackwell, New York), pp 93–109. 50. Woese CR (1987) Bacterial evolution. Microbiol Rev 51: 78. Mayr E (1990) A natural system of organisms. Nature 23. Woese CR, Kandler O, Wheelis ML (1990) Towards 221–271. 348:491. a natural system of organisms: Proposal for the do- 51. Margulis L, Guerrero R (1991) Kingdoms in turmoil. 79. Mayr E (1998) Two empires or three? Proc Natl Acad Sci mains Archaea, Bacteria, and Eucarya. Proc Natl Acad New Sci 1761:46–50. USA 95:9720–9723. Sci USA 87:4576–4579. 52. Stanier RY, Doudoroff M, Adelberg EA (1970) The Micro- 80. Woese CR (1998) Default taxonomy: Ernst Mayr’s view 24. Pace NR (1997) A molecular view of microbial diversity bial World (Prentice Hall, Englewood Cliffs, NJ), 3rd Ed. of the microbial world. Proc Natl Acad Sci USA 95: and the biosphere. Science 276:734–740. 53. Fox GE, et al. (1980) The phylogeny of prokaryotes. 11043–11046. 25. Amann RI, Ludwig W, Schleifer KH (1995) Phylogenetic Science 209:457–463. 81. Fraser C, Alm EJ, Polz MF, Spratt BG, Hanage WP (2009) identification and in situ detection of individual micro- 54. Gray MW, Doolittle WF (1982) Has the The bacterial species challenge: Making sense of ge- bial cells without cultivation. Microbiol Rev 59:143–169. hypothesis been proven? Microbiol Rev 46:1–42. netic and ecological diversity. Science 323:741–746.

1018 | www.pnas.org/cgi/doi/10.1073/pnas.1109716109 Pace et al. Downloaded by guest on September 28, 2021