A Molecular View of Microbial Diversity and the Biosphere Norman R

Home , Organotroph

A Molecular View of Microbial Diversity and the Biosphere Norman R. Pace, et al. Science 276, 734 (1997); DOI: 10.1126/science.276.5313.734

The following resources related to this article are available online at www.sciencemag.org (this information is current as of September 6, 2007 ):

Updated information and services, including high-resolution figures, can be found in the online version of this article at: http://www.sciencemag.org/cgi/content/full/276/5313/734

This article cites 20 articles, 19 of which can be accessed for free: http://www.sciencemag.org/cgi/content/full/276/5313/734#otherarticles

This article has been cited by 97 articles hosted by HighWire Press; see: http://www.sciencemag.org/cgi/content/full/276/5313/734#otherarticles

This article appears in the following subject collections: Microbiology http://www.sciencemag.org/cgi/collection/microbio on September 6, 2007 Information about obtaining reprints of this article or about obtaining permission to reproduce this article in whole or in part can be found at: http://www.sciencemag.org/about/permissions.dtl www.sciencemag.org Downloaded from

Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the American Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. Copyright 1997 by the American Association for the Advancement of Science; all rights reserved. The title Science is a registered trademark of AAAS. ticularly the multicellular forms. A Molecular View of Microbial The breakthrough that called to question many previous beliefs and brought order to microbial, indeed biological, diversity Diversity and the Biosphere emerged with the determination of molecular sequences and the concept that se- Norman R. Pace quences could be used to relate organisms (5). The incisive formulation was reached Over three decades of molecular-phylogenetic studies, researchers have compiled an by Carl Woese who, by comparison of ribo- increasingly robust map of evolutionary diversification showing that the main diversity somal RNA (rRNA) sequences, established of life is microbial, distributed among three primary relatedness groups or domains: a molecular sequence–based phylogenetic Archaea, Bacteria, and Eucarya. The general properties of representatives of the three tree that could be used to relate all organ- domains indicate that the earliest life was based on inorganic nutrition and that pho- isms and reconstruct the history of life (6, tosynthesis and use of organic compounds for carbon and energy metabolism came 7). Woese articulated the now-recognized comparatively later. The application of molecular-phylogenetic methods to study natural three primary lines of evolutionary descent, microbial ecosystems without the traditional requirement for cultivation has resulted in termed “urkingdoms” or “domains”: Eucarya the discovery of many unexpected evolutionary lineages; members of some of these (eukaryotes), Bacteria (initially called eu- lineages are only distantly related to known organisms but are sufficiently abundant that bacteria), and Archaea (initially called ar- they are likely to have impact on the chemistry of the biosphere. chaebacteria) (8). Figure 1 is a current phylogenetic tree based on small-subunit (SSU) rRNA sequences of the organisms represented. The Microbial organisms occupy a peculiar was not until the late 19th century and the construction of such a tree is conceptually place in the human view of life. Microbes development of pure-culture techniques simple (9). Pairs of rRNA sequences from receive little attention in our general texts that microbial organisms could be studied as different organisms are aligned, and the dif- of biology. They are largely ignored by most individual types and characterized to some ferences are counted and considered to be professional biologists and are virtually un- extent, mainly by nutritional criteria. How- some measure of “evolutionary distance” be- known to the public except in the contexts ever, the pure-culture approach to the study tween the organisms. There is no consider-

of disease and rot. Yet, the workings of the of the microbial world seriously constrained ation of the passage of time, only of change on September 6, 2007 biosphere depend absolutely on the activi- the view of microbial diversity because most in nucleotide sequence. Pair-wise differences ties of the microbial world (1). Our texts microbes defy cultivation by standard meth- between many organisms can then be used articulate biodiversity in terms of large or- ods. Moreover, the morphological and nu- to infer phylogenetic trees, maps that repre- ganisms: insects usually top the count of tritional criteria used to describe microbes sent the evolutionary paths leading to the species. Yet, if we squeeze out any one of failed to provide a natural taxonomy, or- modern-day sequences. The tree in Fig. 1 is these insects and examine its contents un- dered according to evolutionary relation- largely congruent with trees made using any der the microscope, we find hundreds or ships. Molecular tools and perspective based molecule in the nucleic acid–based, infor- thousands of distinct microbial species. A on gene sequences are now alleviating these mation-processing system of cells. On the

handful of soil contains billions of microbial constraints to some extent. Even the early other hand, phylogenetic trees based on www.sciencemag.org organisms, so many different types that ac- results are changing our perception of mi- metabolic genes, those involved in the ma- curate numbers remain unknown. At most crobial diversity. nipulation of small molecules and in inter- only a few of these microbes would be action with the environment, commonly do known to us; only about 5000 noneukary- A Sequence-Based Map not concur with the rRNA-based version otic organisms have been formally described of Biodiversity [see (10, 11) for reviews and discussions of (2) (in contrast to the half-million de- phylogenetic results with different mole- scribed insect species). We know so little Before the development of sequence-based cules]. Incongruities in phylogenetic trees about microbial biology, despite it being a methods, it was impossible to know the made with different molecules may reflect Downloaded from part of biology that looms so large in the evolutionary relationships connecting all of lateral transfers or even the intermixings of sustenance of life on this planet. life and thereby to draw a universal evolu- genomes in the course of evolution. Some The reason for our poor understanding tionary tree. Whittaker, in 1969, just as the metabolic archaeal genes, for instance, ap- of the microbial world lies, of course, in the molecular methods began to develop, sum- pear much more highly related to specific fact that microbes are tiny, individually in- marized evolutionary thought in the con- bacterial versions than to their eucaryal ho- visible to the eye. The mere existence of text of the “Five Kingdoms” of life: animals, mologs; other archaeal genes seem decidedly microbial life was recognized only relatively plants, fungi, protists (“protozoa”), and eukaryotic in nature; still other archaeal recently in history, about 300 years ago, monera (bacteria) (3). There also was rec- genes are unique. Nonetheless, the recently with Leeuwenhoek’s invention of the mi- ognized a higher, seemingly more funda- determined sequence of the archaeon Meth- croscope. Even under the microscope, how- mental taxonomic distinction between eu- anococcus jannaschii shows that the evolu- ever, the simple morphologies of most mi- karyotes, organisms that contain nuclear tionary lineage Archaea is independent of crobes, usually nondescript rods and membranes, and prokaryotes, predecessors both Eucarya and Bacteria (12). spheres, prevented their classification by of eukaryotes that lack nuclear membranes morphology, the way that large organisms (4). These two categories of organisms were Interpreting the Molecular had always been related to one another. It considered independent and coherent relat- Tree of Life edness groups. The main evolutionary di- The author is in the Departments of Plant and Microbial versity of life on Earth, four of the “Evolutionary distance” in this type of phy- Biology and Molecular and Cell Biology, University of California, Berkeley, CA 94720–3102, USA. E-mail: five traditional taxonomic kingdoms, was logenetic tree (Fig. 1), the extent of se- [email protected] thought to lie among the eukaryotes, par- quence change, is read along line segments. 734 SCIENCE ⅐ VOL. 276 ⅐ 2 MAY 1997 ⅐ www.sciencemag.org ARTICLES The tree can be considered a rough map of from free-living organisms that branch pe- ample of similarities and differences is in the the evolution of the genetic core of the ripherally in molecular trees. Moreover, the nature of the transcription machinery. The cellular lineages that led to the modern most deeply divergent eukaryotes even lack RNA polymerases of Eucarya and Archaea organisms (sequences) included in the tree. mitochondria (15). These latter organisms, resemble each other in subunit composition The time of occurrence of evolutionary little studied but sometimes troublesome and sequence far more than either resembles events cannot be extracted reliably from creatures such as Giardia, Trichomonas, and the bacterial type of polymerase. Moreover, phylogenetic trees, despite common at- Vairimorpha, nonetheless contain at least a whereas all bacterial cells use sigma factors to tempts to do so. Time cannot be accurately few bacterial-type genes (16). These genes regulate the initiation of transcription, euca- correlated with sequence change because may be evidence of an earlier mitochondrial ryal and archaeal cells use TATA-binding the evolutionary clock is not constant in symbiosis with Eucarya that was lost (11)or proteins (17, 18). different lineages (7). This disparity is evi- perhaps other symbiotic or gene-transfer Because of the shared history of Eucarya denced in Fig. 1 by the fact that lines events between the evolutionary domains. and Archaea, we should, perhaps, look to leading to the different reference organisms The root of the universal tree in Fig. 1, the Archaea to identify fundamental prop- are not all the same length; these different the point of origin of the modern lineages, erties of far more complex cells such as our lineages have experienced different extents cannot be established using sequences of own. The eukaryotic nuclear membrane, for of sequence change. Nonetheless, the order only one type of molecule. However, recent instance, is considered by cell biologists to of occurrence of branchings in the trees can phylogenetic studies of gene families that be an intrinsic component of the nucleus, be interpreted as a genealogy, and intrigu- originated before the last common ancestor somehow responsible for its integrity. The ing insights into the evolution of cells are of the three domains have positioned the fact that Archaea remained “prokaryotic,” emerging. root of the universal tree deep on the bacte- that is, did not develop a nuclear mem- A sobering aspect of large-scale phyloge- rial line (10). Therefore, Eucarya and Ar- brane, indicates that a membrane is not netic trees such as that shown in Fig. 1 is chaea had a common history that excluded required for nuclear function, which Ar- the graphical realization that most of our the descendants of the bacterial line. This chaea certainly achieve (as do Bacteria, for legacy in biological science, historically period of evolutionary history shared by that matter). Indeed, the archaeal nuclear based on large organisms, has focused on a Eucarya and Archaea was an important time zone even seems to exclude ribosomes (19), narrow slice of biological diversity. Thus, in the evolution of cells, during which the and the genome of M. jannaschii is sprinkled we see that animals (represented in Fig. 1 by refinement of the primordial information- with homologs of eucaryal nuclear and nu- Homo), plants (Zea), and fungi (Coprinus) processing mechanisms occurred. Thus, cleolar structural genes (12). What consti-

constitute small and peripheral branches of modern representatives of Eucarya and Ar- tutes a “nucleus?” Certainly the acquisition on September 6, 2007 even eukaryotic cellular diversity. If the chaea share many properties that differ from of the nuclear membrane was a relatively animals, plants, and fungi are taken to com- bacterial cells in fundamental ways. One ex- late event in the establishment of the eu- prise taxonomic “kingdoms,” then we must recognize as kingdoms at least a dozen other eucaryotic groups, all microbial, with as Fig. 1. Universal phylo- much or more independent evolutionary genetic tree based on history than that which separates the three SSU rRNA sequences. traditional eukaryotic kingdoms (13). Sixty-four rRNA sequences representative

The rRNA and other molecular data www.sciencemag.org solidly confirm the notion stemming from of all known phylogenetic domains were the last century that the major organelles of aligned, and a tree was eukaryotes—mitochondria and chloro- produced using FASTD- plasts—are derived from bacterial symbi- NAML (43, 52). That tree onts that have undergone specialization was modified, resulting through coevolution with the host cell. Se- in the composite one quence comparisons establish mitochondria shown, by trimming lin- as representatives of Proteobacteria (the eages and adjusting Downloaded from group in Fig. 1 including Escherichia and branch points to incor- Agrobacterium) and chloroplasts as derived porate results of other analyses. The scale bar from cyanobacteria (Synechococcus and corresponds to 0.1 Gloeobacter in Fig. 1) (14). Thus, all respi- changes per nucleotide. ratory and photosynthetic capacity of eukaryotic cells was obtained from bacterial symbionts; the “endosymbiont hypothesis” for the origin of organelles is no longer hypothesis but well-grounded fact. The nuclear component of the modern eukaryotic cell did not derive from one of the pro- karoytic lineages, however. The rRNA and other molecular trees show that the eukaryotic nuclear line of descent extends as deeply into the history of life as do the bacterial and archaeal lineages. The mitochondrion and chloroplast came in relatively late. This late evolution is evidenced by the fact that mitochondria and chloroplasts diverged www.sciencemag.org ⅐ SCIENCE ⅐ VOL. 276 ⅐ 2 MAY 1997 735 caryal line of descent, occurring only after lism, which it supports (a principle long ide–fixing bacterial symbionts (24). This the separation from Archaea. Perhaps the appreciated by ecologists). Energy for fixing invertebrate and metabolically similar ones nuclear membrane is after all not funda- CO2 is gathered in two ways: by phototrophy may contribute significantly to primary pro- mental to the function of the nucleus but (photosynthesis) or lithotrophy (coupling ductivity in the ocean (25). It is not nec- rather is a relatively late-arriving embellish- the oxidation of reduced inorganic com- essary to go to unusual (from our perspec- ment. One hypothesis would be that the pounds such as hydrogen, hydrogen sulfide, tive) places such as ocean-floor vents to nuclear membrane was an invention de- or ferrous iron to the reduction of a chemical encounter other equally fascinating hydro- rived from the Golgi apparatus to serve as a oxidant, a terminal electron acceptor such as gen sulfide–dependent eukaryotes (26). gathering basket for nuclear products, for oxygen, nitrate, sulfate, sulfur, or carbon di- Underfoot at the ocean beach, for example, distribution by the Golgi throughout the oxide). Thus, metabolic diversity can be microbial respiration of seawater sulfate cre- cell. The properties of nuclear pores would generalized in terms of organotroph or au- ates a hydrogen sulfide–rich ecosystem pop- be consistent with this hypothesis; they are totroph, phototroph or lithotroph, and the ulated by little-known creatures such as large orifices, typically Ͼ10 nm in diameter, nature of the electron donor and acceptor. Kentrophoros, a flat, gulletless ciliate that unlikely to gate anything except large mol- The phylogenetic distributions of differ- under the microscope appears fuzzy because ecules (20). The evolutionary record sug- ent types of carbon and energy metabolism it cultivates on its outer surface a crop of gests, then, that we look to something more among different organisms do not necessar- sulfide-oxidizing bacteria (27). These bac- fundamental than the nuclear membrane ily follow the evolutionary pattern of rRNA teria are ingested by endocytosis and there- for the integrity of the nucleus and by (Fig. 1). Presumably, this lack of correspon- by provide nutrition for Kentrophoros.In which to define the essential quality of the dence is because of past lateral transfers of other anaerobic environments, methano- eukaryotic cell. The shared evolutionary those metabolic genes and larger scale sym- gens, members of Archaea, live intracellu- history of Eucarya and Archaea suggests biotic fusions. Nonetheless, there are do- larly with eukaryotes and serve as metabolic that we may be able to recognize the most main-level tendencies that may speak to hydrogen sinks (28). Still other symbioses fundamental elements of our own nucleus the ancestral nature of the three domains of based on inorganic energy sources are all through study of the archaeal version. life (21). The perspective here is currently around us and are little explored for their limited mainly to Archaea and Bacteria. diversity of microbial life (26). The Metabolic Diversity of Life Such broad generalities cannot yet be as- Many lithotrophic, but comparatively few sessed for the Eucarya because so little is organotrophic, representatives of Archaea The molecular-phylogenetic perspective known about the metabolic breadth of the have been obtained in pure culture (29).

(Fig. 1) is a reference framework within domain, the properties of the most deeply There are primarily two metabolic themes, on September 6, 2007 which to describe microbial diversity; the divergent lineages. There is considerable both relying on hydrogen as a main source of sequences of genes can be used to identify information about one pole of eukaryotic energy. Among the known Euryarchaeota— organisms. This capability is an important diversity, that represented by animals, one of the two archaeal kingdoms known concept for microbial biology. It is not pos- plants, and fungi. We know little about the through cultivated organisms—the main sible to describe microorganisms as tradi- other pole, the amitochondriate organisms electron acceptor is carbon dioxide, and the tionally done with large organisms, through that spun off of the main eucaryal line early product, methane—“natural gas.” Most of their morphological properties. To be sure, in evolution (22). The known instances of the methane encountered in the outer few some microbes are intricate and beautiful in such lineages, represented by Trichomonas, kilometers of Earth’s crust or on the surface is

the microscope, but they are mainly rela- Giardia, and Vairimorpha in Fig. 1, are pri- determined by isotopic analysis to be the www.sciencemag.org tively unfeatured at the resolution of rou- marily pathogens. Pathogenicity to humans product of methanogenic Archaea, past and tine microscopy. Therefore, in order to dis- is a rare trait among the rest of eucaryotes present. Such organisms probably constitute tinguish different types of microbes, micro- and bacteria, and no archaeal pathogen is a large component of global biomass. They biologists early turned to metabolic proper- known. This correlation may indicate that certainly offer an inexhaustible source of ties such as utilizable sources of nutrition, for nonpathogenic, deeply divergent eucaryotes renewable energy to humankind. instance, sources of carbon, nitrogen, and are abundant in the environment but not The general metabolic theme of the oth- energy. Microbial taxonomy accumulated as yet detected. They should be sought in er established kingdom of Archaea, Crenar- anecdotal descriptions of metabolically and anaerobic ecosystems, possibly coupled chaeota, is also the oxidation of molecular Downloaded from morphologically distinct types of organisms metabolically to other organisms. A driv- hydrogen, but with a sulfur compound as the that were essentially unrelatable. Molecular ing theme of the eucaryal line seems to be terminal electron acceptor. All of the culti- phylogeny now provides a framework within the establishment of physical symbiosis vated representatives of Crenarchaeota also which we can relate organisms objectively, with other organisms. Beyond that, the are thermophiles. Consequently, such or- and also through which we can interpret the general metabolism of the rudimentary eu- ganisms have been referred to as “ther- evolutionary flow of the metabolic machin- karyotic cell seems simple, based on fer- moacidophilic” or “hyperthermophilic” Ar- eries that constitute microbial diversity. mentative organotrophy. By virtue of sym- chaea; some grow at the highest known Laboratory studies of microbial metabo- biotic partners, however, eukaryotes are temperatures for life, up to 113°C in the case lism have focused mainly on organisms such able to take on phototrophic or lithotro- of Pyrolobus fumaris (30). These crenarchae- as Escherichia coli and Bacillus subtilis.Inthe phic life-styles and to use the electron- otes might seem bizarre, capable of thriving broad sense, such organisms metabolize acceptor oxygen (23). at temperatures above the usual boiling much as animals do; we are all “organo- Symbiotic microbes commonly confer point of water on a diet of H2,CO2, and trophs,” using reduced organic compounds the lithotrophic way of life even on ani- elemental sulfur and exhaling hydrogen sul- for energy and carbon. Organotrophy is not mals, although this was only recently rec- fide. Yet, in terms of the molecular struc- the prevalent form of metabolism in the ognized. The 2-m-long tubeworm Riftia tures of the basic cellular machineries, these environment, however. Autotrophic metab- pachyptila, for instance, lives in the vicinity creatures resemble eukaryotes far more olism, fixation of CO2 to reduced organic of sea-floor hydrothermal vents and metab- closely than either resembles the bacterium compounds, must necessarily contribute to a olizes hydrogen sulfide and carbon dioxide E. coli (17). greater biomass than organotrophic metabo- by means of sulfide-oxidizing, carbon diox- The metabolic diversity of microbes is 736 SCIENCE ⅐ VOL. 276 ⅐ 2 MAY 1997 ⅐ www.sciencemag.org ARTICLES usually couched in terms of the utilization bon and energy metabolism. In the relative- DNA fragments are a source of rRNA, as of complex organic compounds. From that ly closely related “gamma subgroup” of the well as other genes, but require sorting of standpoint, metabolic diversity seems, on kingdom of Proteobacteria (delineated by rRNA genes from the others. The quickest the basis of cultivated instances of organ- the genus Escherichia in Fig. 1), for instance, way to survey the constituents of microbial isms, to have flowered mainly among the we find the phenotypically disparate orga- ecosystems is through the use of the poly- Bacteria. Even here, however, reliance on nisms E. coli (organotroph), Chromatium merase chain reaction (PCR) (36). The organic nutrients probably was not ances- vinosum (hydrogen sulfide–based pho- highly conserved nature of rRNA allows for tral. The most deeply branching of the cul- totroph), and the symbiont of the tube- the synthesis of “universal” PCR primers tured bacterial lineages, represented by worm R. pachyptila (hydrogen sulfide–based that can anneal to sequences conserved in Aquifex and Thermotoga in Fig. 1, are basi- symbiont). The superficial metabolic diver- the rRNA genes from all three phylogenetic cally lithotrophs that use hydrogen as an sity of these types of Bacteria belies their domains. In principle, PCR carried out with energy source and electron acceptors such underlying close evolutionary relatedness, these primers amplifies the rRNA genes of as sulfur compounds (Thermotoga)orlow giving no hint of the close similarities of all types of organisms present in an envi- levels of oxygen (Aquifex)(31). Cultivated their basic machineries. The versatility of ronmental sample. Individual types of genes instances of these deeply branching bacte- Bacteria makes the metabolic machineries in the mixture are separated by a cloning rial lineages also are all thermophilic and of Archaea and Eucarya seem comparative- step and then sequenced. thus share two important physiological at- ly more monotonous. As the sequences of A molecular-phylogenetic assessment of tributes with the deeply branching and diverse genomes are compared, it will be an uncultivated organism can provide in- slowly evolving Archaea: a hydrogen-based possible to map the flow of metabolic genes sight into many of the properties of the energy source and growth at high tempera- onto the rRNA-based tree and thereby see organism through comparison with its stud- tures. This coincidence suggests that the how metabolic diversity has been molded ied relatives. One example of the perspec- last common ancestor of all life also metab- through evolution. tive that phylogeny can offer on an other- olized hydrogen for energy at high temper- The molecular perspective gives us more wise unknown organism is seen with the atures. This inference is consistent with than just a glimpse of the evolutionary past; sulfur-oxidizing microorganisms that pro- current notions regarding the origin of life, it also brings a new future to the discipline vide nutrition to symbiotic invertebrates that it came to be in a geothermal setting at of microbial biology. Because the molecu- such as the vent tubeworm R. pachyptila high temperature (32). lar-phylogenetic identifications are based (24). Although many attempts to cultivate Chlorophyll-based photosynthesis was a on sequence, as opposed to metabolic prop- the symbionts for phenotypic characteriza-

bacterial invention. It seems to have ap- erties, microbes can be identified without tion failed, rRNA analyses revealed that on September 6, 2007 peared well after the establishment of the being cultivated. Consequently, all the se- many of the basic cellular properties of the bacterial line of descent, at or before the quence-based techniques of molecular biol- symbionts were already familiar to us. The divergence of the line in Fig. 1 leading to ogy can be applied to the study of natural Riftia symbiont and a number of other sul- Chloroflexus, a photosynthetic genus (33), microbial ecosystems, heretofore little fur-oxidizing symbionts associated with in- and after the deeper divergences such as known with regard to organismal makeup. vertebrate animals all proved to be fairly those leading to Aquifex and Thermotoga, closely related to one another, close rela- genera that are not known to have photo- A Sequence-Based Glimpse of tives also to the intensively studied organ- synthetic representatives. Most bacterial Biodiversity in the Environment isms E. coli and Pseudomonas aeruginosa

photosynthesis is anaerobic, however. Oxy- (37). Because of their phylogenetic proxim- www.sciencemag.org genic photosynthesis, the water-based pho- Knowledge of microorganisms in the envi- ity, many of the properties of the symbionts tosynthetic mechanism that produces the ronment has depended in the past mainly can be inferred from those of the well- powerful electron acceptor oxygen, arose on studies of pure cultures in the laboratory. studied organisms. For instance, we can pre- only in the kingdom-level lineage of cya- Rarely are microbes so captured, however. dict with good confidence the nature of the nobacteria. This invention changed the sur- Studies of several types of environments ribosome and antibiotic-susceptibility pat- face of Earth profoundly and conventional- estimate that more than 99% of organisms terns, the nature of the DNA-replicative ly is thought to be the basis, directly or seen microscopically are not cultivated by machinery, the character of the RNA poly- indirectly, of most present-day biomass. routine techniques (34). With the se- merase complex, the character of biosyn- Downloaded from Anaerobic photosynthesis is widely dis- quence-based taxonomic framework of mo- thetic pathways and their regulatory mech- tributed in the late-branching bacterial lecular trees, only a gene sequence, not a anisms, the nature of the cell envelope and kingdoms. The more ancient theme of functioning cell, is required to identify the energy transduction schemes, and many lithotrophy, metabolism of inorganic com- organism in terms of its phylogenetic type. other cellular properties of the symbionts. pounds, is also widely distributed phyloge- The occurrence of phylogenetic types of On the other hand, because E. coli and P. netically, intermixed with organotrophic organisms, “phylotypes,” and their distribu- aeruginosa do not oxidize sulfur, these rela- organisms. The pattern suggests that organ- tions in natural communities can be sur- tions cannot provide insight into the sulfur- otrophy arose many times from otherwise veyed by sequencing rRNA genes obtained oxidative pathways of the symbionts. The photosynthetic or lithotrophic organisms. from DNA isolated directly from the envi- rRNA sequence does, however, identify Indeed, many instances of Bacteria can ronment. Analysis of microbial ecosystems free-living and cultivated (but less-studied) switch between these modes of nutrition, in this way is more than a taxonomic exer- close relatives of the symbionts—for exam- carrying out photosynthesis in the light and cise because the sequences provide experi- ple Thiomicrospira sp. L-12 ( a hydrother- lithotrophy or organotrophy in the dark. mental tools—for instance, molecular hy- mal-vent isolate) and Thiothrix sp.—that Particularly among Bacteria, this type of bridization probes—that can be used to also rely on sulfur oxidation and so are energy metabolism seems highly volatile in identify, monitor, and study the microbial likely to provide good models for this pro- evolution: Bacteria that are closely related inhabitants of natural ecosystems (35). cess in the symbionts. by molecular criteria can display strikingly Ribosomal RNA genes are obtained by Every nucleic acids–based study of nat- different phenotypes when assessed in the cloning DNA isolated directly from the en- ural microbial ecosystems so far performed laboratory through the nature of their car- vironment. “Shotgun libraries” of random has uncovered novel types of rRNA se- www.sciencemag.org ⅐ SCIENCE ⅐ VOL. 276 ⅐ 2 MAY 1997 737 quences, often representing major new lin- logenetic diversity of Archaea. All culti- explosive radiation of lineages, rather eages only distantly related to known ones. vated Crenarchaeota branch in the cluster than from the sequential divergence of The discovery of rRNA sequences in the bracketed by Pyrodictium and Thermofilum main lines seen, for instance, in the euca- environment that diverge more deeply in in Fig. 1. Discovery of a rich abundance of ryal domain (Fig. 1). Preliminary results phylogenetic trees than those of cultivated diverse crenarchaeal rRNA genes in Ob- from Obsidian Pool also call into question organisms is particularly noteworthy. It sidian Pool sediment (for example, pSL another supposition based on culture stud- means that the divergent organisms recog- sequences in Fig. 1), scores of new genera, ies, that Archaea dominate high-tempera- nized by rRNA sequence are potentially expanded the known phylogenetic diver- ture environments. Quantitative hybrid- more different from known organisms in the sity (estimated by specific line-segment ization of domain-specific oligonucleotide lineage than the known organisms are from lengths) of Crenarchaeota severalfold probes to rRNA genes obtained by PCR one another. The deepest divergences in (42). More surprising, other sequences indicates that bacterial genes outnumber both the Bacteria and Archaea were first from Obsidian Pool (pJP27 and pJP28 in archaeal genes by 50:1 in this environ- discovered in rRNA-based surveys of hot Fig. 1) seem to branch so deeply in the ment. Such conclusions, of course, are spring–associated communities in Yellow- overall archaeal tree that they constitute a compromised to an unknown extent by stone National Park. new kingdom-level branch of Archaea, considerations such as nonuniform ampli- The geothermal features of Yellow- recognized provisionally as “Korarcha- fication of different rRNA genes, but the stone National Park have been favorite eota” (43). It now will be interesting to trend seems to indicate that bacteria dom- haunts of high-temperature biologists for study other genes from these novel organ- inate this environment. decades (38). Currently, rRNA-based isms. These genes, as well as information It is not necessary to go to extreme methods are being used to survey phylo- on the physiology and other properties of environments to encounter exotic diversi- types present in a number of Yellowstone the organisms, will be obtained most ty; it is all around us. Phylotypes that, hot springs with disparate chemical set- readily if they can be cultured. Even with- because of their abundance, must be sig- tings. One of these, Octopus Spring (Fig. out cultivation, however, cloning large nificant contributors to the biosphere 2A)—a near-boiling, slightly alkaline, ex- fragments of environmental DNA and have escaped detection until the se- tremely low-nutrient flow near Old Faith- then “chromosome walking” to assemble quence-based methods developed. One ful geyser containing an abundant commu- contiguous clones offers access to the ge- example of an arena for research opened nity of pink filaments—yielded the first nomes of these or other uncultivated or- by the molecular methods involves the evidence for the lineage currently thought ganisms (44). recently discovered mesophilic (low-tem-

to be the most deeply divergent in the Continuing study of Obsidian Pool is perature) Crenarchaeota (represented in on September 6, 2007 Bacteria. When this lineage, represented expanding the known extent of bacterial, Fig. 1 by pGrf and marSBAR). On the by Aquifex and EM17 (pink filament as well as archaeal, diversity. Obsidian basis of culture-studies, crenarchaeotes clone) (39) in Fig. 1, was first encountered Pool, judged extremely inhospitable from had been thought to be restricted to high- by 5S rRNA sequence (40), little could be the human standpoint, contains a rich temperature environments. Cloned rRNA inferred about the physiology of the asso- diversity of sequence types representing gene analysis shows, however, that low- ciated organism because no cultivated spe- most of the known bacterial kingdoms, as temperature versions of Crenarchaeota are cific relative had yet been described. Sub- well as kingdom-level divergences never abundant globally in marine (19, 46) and sequent clues to the nature of the pink described by cultivation (45). Phyloge- terrestrial (47) environments, in typically

filaments came with the discovery of A. netic studies of cultured and environmen- 30 to 50% of planktonic rRNA genes in www.sciencemag.org pyrophilus, cultured from an Icelandic hot tal sequences have expanded substantially limited samplings of Atlantic, Pacific, and spring with a chemical character similar to our appreciation of the scope of bacterial Antarctic waters (48). The physiologies of that of Octopus Spring (41), and determi- diversity: In 1987, only about 12 phyloge- the low-temperature crenarchaeotes are nation of the 16S rRNA sequence for the netic kingdoms (main phyla) of Bacteria unknown; none has yet been cultivated. pink filaments (39). The A. pyrophilus and were recognized (Fig. 3, inset) (7), but The properties of their remote relatives— pink filament sequences are sufficiently now, at least 25 to 30 distinct, kingdom- the cultivated, high-temperature Crenar- closely related (Fig. 1) that many of their level phylogenetic divergences are re- chaeota—hint that the mesophilic types properties are likely to be shared. The solved (Fig. 3). The topology of the bac- might also engage in hydrogen metabo- Downloaded from mode of nutrition of the pink filaments, terial tree is remarkable. Bacterial diversi- lism, perhaps using some oxidation state of for instance, is predicted to be that of ty seems to have arisen mainly from an sulfur as an electron acceptor. Aquifex, consumption of hydrogen with low levels of oxygen and fixation of carbon dioxide. Many other representatives AB of the Aquifex-EM17 relatedness group (Aquificales) have now been cultured, mainly from high-temperature settings, and all are thermophilic hydrogen oxidizers (31). Hot springs on the northern flank of the Yellowstone caldera usually have high concentrations of iron (II), hydrogen sulfide, hydrogen, and carbon dioxide—a Fig. 2. Yellowstone National Park hot springs rich in microbial diversity. (A) Octopus Spring. The source wealth of foodstuffs for compatible physi- pool of this hot spring is 90° to 93°C and extremely low in nutrients but contains abundant biomass and ologies. Ongoing sequence-based studies the deepest known evolutionary divergences in the domain Bacteria. (B) Obsidian Pool. Molecular of the microbial inhabitants of one of studies find that the inhospitable environment of this hot spring, 75° to 95°C in temperature and these springs, Obsidian Pool (Fig. 2B), containing high concentrations of iron (II) and hydrogen sulfide, supports an extensive diversity of have radically revised our view of the phy- previously unknown microbial life, both archaeal and bacterial. 738 SCIENCE ⅐ VOL. 276 ⅐ 2 MAY 1997 ⅐ www.sciencemag.org ARTICLES

Fig. 3. Diagrammatic repre- sentation of the known phylogenetic span of Bacteria in 1987 (inset) and today. Phy- logenetic trees containing sequences from the indicated organisms or groups of organisms, chosen to represent the broadest diversity of Bacteria, were used as the basis for this diagram (compiled with P. Hugen- holtz). Filled sectors indicate that several representative sequences fall within the indicated depth of branching. Lines designated by OP represent one or more phylotypes that were identified in Obsidian Pool by means of molecular methods but have not been not cultivated. The inset is an outline of the bacterial tree compiled by Woese in 1987 (7).

Microbial Diversity and the Limits shot through with biomass, wherever the of Earth’s microbial biota is needless and, of of the Biosphere physical conditions permit. Metabolism of course, impossible. A representative survey, hydrogen is a dominant theme among or- however, is worthwhile. A representative Textbooks generally portray only a part of ganisms isolated from geothermal settings survey could be achieved with modest ef- the global distribution of life, the part that or deep aquifers (51). Hydrogen is gener- fort, with the use of automated sequencing

is immediately dependent on either the ated readily by abiotic mechanisms such as technology. Analysis of 1000 clones (to on September 6, 2007 harvesting of sunlight or the metabolism interaction of water with iron-bearing ba- detect the most abundant genome types) of the decay products of photosynthesis. salt, the main stuff of Earth’s crust; conse- from each of 100 chemically different envi- The molecular phylogenetic record shows, quently, a food source is unlikely to be ronments would be comparable to the effort however, that lithotrophic metabolism limiting in most subterranean environ- to sequence a single microbial genome. The preceded and is more widespread phyloge- ments. Rather, it is likely to be the oxi- questions are large and many: What kinds netically and geographically than is either dant, the terminal electron acceptor, that of organisms do we share this planet with phototrophy or organotrophy. The litho- limits growth. Nonetheless, it seems pos- and depend on? What model systems should trophic biosphere potentially extends ki- sible that much, perhaps most, of the bio- we choose for laboratory studies of environ-

lometers into the crust of Earth, an essen- mass on Earth is subterranean, a biological mental processes? How extensive is the www.sciencemag.org tially unknown realm (49). These consid- world based on lithotrophy. Although the fund of biodiversity from which we can erations may indicate that lithotrophy metabolic rate of this subterranean bio- draw useful lessons and resources? Can we contributes far more to the biomass of sphere is likely to be far slower than in the use the distribution of microbes as a biosen- Earth than currently thought. more dynamic, photic environment, life is sor array to map and monitor the chemis- Part of that lithotrophic biomass is in likely to be as pervasive in occurrence, and tries of Planet Earth? Are there deeper microhabitats all around us, usually away perhaps in cellular diversity, as we experi- branchings in the tree of life than the lin- from light and oxygen. It is not necessary ence on the surface. eages we know? to look far to find such environments: the The opportunities for the discovery of Downloaded from rumens of cattle and the guts of termites Opportunity for an Environmental new organisms and the development of re- and humans, for example, are significant Genome Survey sources based on microbial diversity are sources of methane, a signature of hydro- greater than ever before. Molecular se- gen metabolism. Most life that depends on It is clear from even the small number of quences have finally given microbial biolo- inorganic energy metabolism, however, environments so far studied with the mo- gists a way to define their subjects, through probably is in little-known environments, lecular methods that our understanding of molecular phylogeny. The sequences also based on poorly understood geochemis- the makeup of the natural microbial world are the basis of the tools that will allow tries. The oceans, for instance, cover 70% is rudimentary. The sequence-based meth- microbial biologists to explore the distribu- of Earth’s surface to an average depth of 4 ods, however, now provide a way to survey tion and roles of the organisms in the en- km. Most life in the ocean is microbial, biodiversity rapidly and comprehensively. vironment. Microbial biology can now be a and the metabolic patterns of such organ- Ribosomal RNA genes gathered from the whole science; the organism can be studied isms are not understood: Large standing environment are snapshots of organisms, in the ecosystem. crops of low-temperature crenarchaeotes, representatives of different types of ge- potential hydrogen oxidizers, may indicate nomes, targets for further characterization if REFERENCES an unsuspected, lithotrophy-based food they seem interesting or useful. If we want ______chain in the oceans. Another little-stud- to understand the biosphere, I think it im- 1. M. T. Madigan, J. M. Martinko, J. Parker, Brock Bi- ied environment with global significance portant, even essential, that we undertake a ology of Microorganisms (Prentice-Hall, Upper Sad- dle River, NJ, ed. 8, 1996). is the deep subsurface (50). There is in- representative survey of microbial diversity 2. A. T. Bull, M. Goodfellow, J. H. Slater, Annu. Rev. creasing evidence that the crust of Earth is in the environment. A complete cataloging Microbiol. 46, 219 (1992). www.sciencemag.org ⅐ SCIENCE ⅐ VOL. 276 ⅐ 2 MAY 1997 739 3. R. H. Whittaker, Science 163, 150 (1969). 21. O. Kandler, in Early Life on Earth, S. Bengtson, Ed. 49, 1379 (1985). 4. E. Chatton, Titres et Travoux Scientifiques (Sete, (Nobel Symp. 84, Columbia Univ. Press, New York, 41. R. Huber et al., Syst. Appl. Microbiol. 15, 340 (1992). Sottano, Italy, 1937). 1993), pp. 152–160. 42. S. M. Barns, R. E. Fundyga, M. W. Jeffries, N. R. 5. E. Zuckerkandl and L. Pauling, J. Theor. Biol. 8, 357 22. M. L. Sogin, ibid., pp. 181–192. Pace, Proc. Natl. Acad. Sci. U.S.A. 91, 1609 (1994). (1965); R. M. Schwartz and M. O. Dayhoff, Science 23. D. C. Smith and A. E. Douglas, The Biology of Sym- 43. S. M. Barns, C. F. Delwiche, J. D. Palmer, N. R. 199, 395 (1978). biosis (Edward Arnold, London, 1987). Pace, ibid. 93, 9188 (1996). 6. C. R. Woese and G. E. Fox, Proc. Natl. Acad. Sci. 24. V. Tunnicliffe, Am. Sci. 80, 336 (1992). 44. J. L. Stein, T. L. Marsh, K. Y. Wu, H. Shizuya, E. F. U.S.A. 74, 5088 (1977). 25. R. A. Lutz et al., Nature 371, 663 (1994). DeLong, J. Bacteriol. 178, 591 (1996). 7. C. R. Woese, Microbiol. Rev. 51, 221 (1987). 26. T. Fenchel and B. J. Finlay, Ecology and Evolution in 45. P. Hugenholtz, C. Pitulle, K. L. Hershberger, N. R. 8. ࿜࿜࿜࿜, O. Kandler, M. L. Wheelis, Proc. Natl. Acad. Anoxic Worlds, R. M. May and P. H. Harvey, Eds. Pace, in preparation. Sci. U.S.A. 87, 4576 (1990). (Oxford Ser. Ecol. Evol., Oxford Univ. Press, Oxford, 46. J. A. Fuhrman, K. McCallum, A. A. Davis, Nature 9. D. L. Swofford, G. J. Olsen, P. J. Waddell, D. M. 1995). 356, 148 (1992). Hillis, in Molecular Systematics, D. M. Hillis, C. 27. , Ophelia 30, 75 (1989). ࿜࿜࿜࿜ 47. T. Ueda, Y. Suga, T. Matsuguchi, Eur. J. Soil Sci. 46, Moritz, B. K. Mable, Eds. (Sinauer, Sunderland, MA, 28. T. M. Embley and B. J. Finlay, Microbiology 140, 225 415 (1995); K. L. Hershberger, S. M. Barns, A.-L. 1996), pp. 407–514. (1994). Reysenbach, S. C. Dawson, N. R. Pace, Nature 384, 10. W. F. Doolittle and J. R. Brown, Proc. Natl. Acad. 29. M. Kates, D. J. Kushner, A. T. Matheson, Eds., The 420 (1996); S. B. Bintrim, T. J. Donohue, J. Handels- Sci. U.S.A. 91, 6721 (1994). Biochemistry of Archaea (Archaebacteria) (Elsevier, man, G. P. Roberts, R. M. Goodman, Proc. Natl. 11. J. D. Palmer, Science 275, 790 (1997). Amsterdam, 1993). Acad. Sci. U.S.A. 94, 277 (1997); G. Jurgens, K. 12. C. J. Bult et al., ibid. 273, 1058 (1996). 30. K. O. Stetter, Am. Soc. Microbiol. News 61, 285 Lindstrom, A. Saano, Appl. Environ. Microbiol. 63, 13. The taxonomic term “kingdom” has no molecular (1995). 803 (1997). definition. I use the term to indicate main lines of 31. C. Pitulle et al., Int. J. Syst. Bacteriol. 44, 620 (1994). radiation in the particular domain; 14 such “king- 32. N. R. Pace, Cell 65, 531 (1991). 48. E. F. DeLong, K. Y. Wu, B. B. Prezelin, R. V. M. dom-level” lines are associated with the eucaryal line 33. B. K. Pierson, in Early Life on Earth (Nobel Symp. 84, Jovine, Nature 371, 695 (1994). of descent in Fig. 1 [see also (22)]. Columbia Univ. Press, New York, 1993), pp. 161– 49. W. C. Ghiorse, Science 275, 789 (1997). 14. J. Sapp, Evolution by Association: A History of Sym- 180. 50. T. Gold, Proc. Natl. Acad. Sci. U.S.A. 89, 6045 biosis (Oxford Univ. Press, New York, 1994). 34. R. I. Amann, W. Ludwig, K. H. Schleifer, Microbiol. (1992); J. K. Fredrickson and T. C. Onstott, Sci. Am. 15. T. Cavalier-Smith, Microbiol. Rev. 57, 953 (1993). Rev. 59, 143 (1995). 275, 68 (October 1996); D. Lovely and F. Chapelle, 16. E. T. N. Bui, P. J. Bradley, P. J. Johnson, Proc. Natl. 35. N. R. Pace, D. A. Stahl, D. J. Lane, G. J. Olsen, Am. Rev. Geophys. 33, 365 (1995). Acad. Sci. U.S.A. 93, 9651 (1996); A. Germot, H. Soc. Microbiol. News 51, 4 (1985); P. Hugenholtz 51. T. O. Stevens and J. P. McKinley, Science 270, 450 Philippe, H. le Guyader, ibid., p. 14614; A. J. Roger, and N. R. Pace, Trends Biotechnol. 14, 190 (1996). (1995); K. Pedersen, Earth Sci. Rev. 34, 243 (1993). C. G. Clark, W. F. Doolittle, ibid., p. 14618. 36. R. K. Saiki et al., Science 239, 487 (1988). 52. B. L. Maidak et al., Nucleic Acids Res. 25, 109 17. T. L. Marsh, C. I. Reich, R. B. Whitelock, G. J. Olsen, 37. D. L. Distel et al., J. Bacteriol. 170, 2506 (1988). (1997). ibid. 91, 4180 (1994). 38. T. D. Brock, Thermophilic Microorganisms and Life 53. I thank S. Kustu, G. Olsen, and C. Woese for helpful 18. T. Rowlands, P. Baumann, S. P. Jackson, Science at High Temperatures (Springer-Verlag, New York, comments on the manuscript, and S. Barns, S. 264, 1326 (1994). 1978). Dawson, C. Delwiche, and P. Hugenholtz for assist- 19. E. F. DeLong, Proc. Natl. Acad. Sci. U.S.A. 89, 5685 39. A.-L. Reysenbach, G. S. Wickham, N. R. Pace, Appl. ance with the figures. Research in my laboratory is (1992). Environ. Microbiol. 60, 2113 (1994). supported by grants from the U.S. Department of

20. L. I. Davis, Annu. Rev. Biochem. 64, 865 (1995). 40. D. A. Stahl, D. J. Lane, G. J. Olsen, N. R. Pace, ibid. Energy and NIH. on September 6, 2007 www.sciencemag.org Downloaded from

740 SCIENCE ⅐ VOL. 276 ⅐ 2 MAY 1997 ⅐ www.sciencemag.org