Methanogens:Reevaluation Unique Biological GroupMETHANOGENS: A NEWPERSPECTIVE 263 gate (50, 51). Theversatility ofthese anaerobic proceduresiswelldocumented(19,21,31,47,48, 51, 60,

MICROBIOLOGICAL REVIEWS, June 1979, p. 260-296 Vol. 43, No. 20146-0749/79/02-0260/37 $02.00/0

Methanogens: Reevaluation of a Unique Biological GroupW. E. BALCH,' G. E. FOX,2 L. J. MAGRUM,3 C. R. WOESE,3 AND R. S. WOLFE*

Departments ofMicrobiology' and Genetics and Development,3 University of Illinois, Urbana, Illinois 61801,and Department of Biophysical Sciences, University of Houston, Houston, Texas 770042

INTRODUCTION ........ ...... ........ 261GROWTH OF METHANOGENS 262Techniques for Growth of Methanogens ..... 262Principal Media......6..3................. .... 263Strain Histories.................2..6..4.............264Deoxyribonucleic Acid Base Composition Determination (Moles Percent Gua-

nine plusCytosine). 264GENERATION OF 16S RIBOSOMAL RIBONUCLEIC ACID OLIGONUCLEO-

TIDE CATALOGS ..............................2...................264Labeling and Ribosomal Ribonucleic Acid Isolation .......... 264Determination of Oligonucleotide Catalog .26........6..26Analysis ofData ............2......6..................6.........26

COMPARATIVE CATALOGING OF 17 METHANOGENS 266DEVELOPMENT AND OUTLINE OF A NEW TAXONOMIC TREATMENT

BASED ON 16S RIBOSOMAL RIBONUCLEIC ACID ................ 267Orders ................................................... 268Families .2......6.8......................6.......268Genera.......... 269Species ...2.6.......9.........269

EVIDENCE IN SUPPORT OF THE NEW TAXONOMIC TREATMENT ........ 276Cell Wall Structure and Composition .. .......... ........ 276Order I, Methanobacteriales.276Order II, Methanococcales.278Order m, Methanomicrobiales.278

LFimpid Composition.279Deoxyribonucleic Acid Base Composition (Moles Percent Guanine plus Cyto-

sine) ................................................................ 280Substrates for Growth and Methane Production 280Metabolic Pathways ................. .. 281Coenzyme M (2-mercaptoethanesulfonic acid) .. .. ....... .... 281Coenzyme F420 .......................................... 282Intermediary metabolism .... .... ...... .... .... 282

Ribosomal and Transfer Ribonucleic Acids of Typical Bacteria and Methan-ogens .............. 282

Genome Sizes of Typical Bacteria and Methanobacterium thermoautotrophi-cum 283

FORMAL DETAILED PROPOSAL: REVISION OF CLASSIFICATION ANDNOMENCLATURE 283

Order I,Methanobacteriales .................................... 283Family I, Methanobacteriaceae ............................................ 284Genus I, Methanobacterium............ 284Methanobacterium bryantii .................................... 284

Genus II,Methanobrevibacter ......................... ..... 284Methanobrevibacter smithii ................. 286

Order H,Methanococc s ....................................... 286Family I, Methanococcaceae ............................. 286Genus I, Methanococcus ....... ... 285Methanococcus voltae.............c. 286

Order Im , Methanomicrobiales .. .................................... 286Family I, Methanomicrobiaceae ................................. 286Genus I, Methanomicrobium .... ... 286Genus II, Methanogenium ..... .. 286Genus IH, Methanospirillum ...... 286

Family H, Methanosarcinaceae 288Genus I, Methanosarcina-. ........ 288

OV'ERVIEW ... 288

260

on Novem

ber 18, 2020 by guesthttp://m

mbr.asm

.org/D

ownloaded from

http://mmbr.asm.org/

METHANOGENS: A NEW PERSPECTIVE 261

Resolution of the Methanogen Group .......................... ... 288Phylogenetic Relationship of Methanogens to Other Organisms .. ...... 290

LITERATURE CITED ................2..........................292

INTRODUCTIONThe discovery by Italian physicist Alessandro

Volta (49) in 1776 that "combustible air" wasbeing formed in the sediments of streams, bogs,and lakes rich in decaying vegetation led tosubsequent discoveries by Bechamp, Popoff,Tappeneiner, Hoppe-Seyler, Sohngen, and Ome-lianski that defined a microbial basis for theorigin of methane gas (11). Within the pastdecade studies have established the widespreadand fundamental role of the methane-producingbacteria in anaerobic degradation processes innature (63, 118, 119, 124, 126). More elusive hasbeen an understanding of the systematic rela-tionships among the methane bacteria and theirrelationship to other procaryotes. On one hand,the methanogens are a morphologically diversegroup of organisms; on the other, they are aphysiologically coherent group of strict anaer-obes, sharing the common metabolic capacity toproduce methane. The uncertainty concerningtheir systematic relationships was reflected inearly taxonomic schemes which initially dis-persed the methanogenic bacteria among thebetter characterized bacterial groups accordingto their morphologies. Later, Barker (11) em-phasized the highly unique physiology of thegroup and thereby clustered the methanogensinto a single family, the Methanobacteriaceae.Although Barker's regrouping is presently rec-ognized (20), the scheme provides limited insightinto the relationships among the various meth-anogenic species. Methanogens metabolize onlya restricted range of substrates and are poorlycharacterized with respect to other metabolic,biochemical, and molecular properties. Of thethree genera proposed in the most recent editionof Bergey's Manual of Determinative Bacteri-ology (20), only one genus is moderately wellcharacterized and contains more than one spe-cies that is available in pure culture. Both gram-negative and gram-positive cells are representedin that genus; morphologies vary from short,lancet-shaped cocci to long, filamentous rods;and deoxyribonucleic acid (DNA) base compo-sition ranges widely from 28 to 51 moles percentguanine plus cytosine (mol% G + C).The dilemma posed by seemingly contradic-

tory characters (in this case an apparent unifyingphysiology versus diverse morphologies) and adearth of phenotypic characters to define rela-tionships are characteristic features of bacterial.classification for the last century. Our inabilityto formulate a constant basis for bacterial sys-

tematics stems from the conceptual view of mi-crobial relationships taken by early taxonomists.In contrast to a phylogenetic definition of rela-tionships found between the higher eucaryotes,bacterial systematics historically became fixedinto an empirical definition of relationships. Thisapproach reflected both the lack of genealogicalinformation and the need for a practicalscheme-one in which a number of simple phe-notypic characters were sufficient to identify anorganism. The problems of an empirical ap-proach were concisely summarized by R. Y.Stanier and C. B. van Niel in 1941 (93):

... an empirical system is largely unmodifiable becausethe differential characters employed are arbitrarilychosen and usually cannot be altered to any greatextent without disrupting the whole system. Its soleostensible advantage is its greater immediate practicalutility; but if the differential characters used are notmutually exclusive (and such mutual exclusivenessmay be difficult to attain when the criteria employedare purely arbitrary) even this advantage disappears.[P. 438-439]

The subsequent history of bacterial systematicshas confirmed the insight of these statements; aconstant reshuffling of taxa and the invention ofnew taxa have been the inevitable outcome ofthe introduction of new methods for classifyingbacteria.van Niel (106) proposed that the issues could

be simply resolved by shifting our emphasis from"indication of relationships" among the bacteriato "means of identification." In this light, anytaxonomic scheme is an effective tool for ascer-taining the identity of an organism. For thispurpose, van Niel encouraged the developmentof multiple-determinant schemes based on awide range of differential characters (106). Thisattitude is to some extent reflected in the mostrecent edition of Bergey's Manual, where themajor clusters of organisms are grouped intosections, or "parts." Whereas relationships arespecified on the basis of certain characteristicswithin the group, there is no indication of rela-tionships between the various parts. Such anapproach, although important as a means ofidentification, provides a limited basis for bac-terial systematics. The concept of genealogicalrelationships is a starting point for any trueunderstanding of microbial diversity. The ad-vantages of a phylogenetic or natural definitionof relatedness were summarized by Stanier andvan Niel (93):

VOL. 43, 1979

on Novem


mbr.asm

.org/D

ownloaded from


262 BALCH ET AL.

Even granting that the true course of evolution can,never be known and that any phylogenetic system hasto be based to some extent on hypothesis, there isgood reason to prefer an admittedly imperfect naturalsystem to a purely empirical one. A phylogenetic sys-tem has at least a rational basis, and can be alteredand improved as new facts come to light; its veryweaknesses will suggest the type of experimental worknecessary for improvement. [P. 438]

Indeed, certain groups, such as the methano-gens, can only be approached by developing ameasure of their relationships.The capacity to sequence macromolecules

provides the key to a natural basis for bacterialclassification. Macromolecular sequences are ineffect historical records. These "semantides" (in-formational macromolecules) (136) can be readthrough comparative analysis to reveal geneal-ogical relationships (111). A number of ap-proaches have compared amino acid sequencesof proteins, such as cytochrome c and ferredoxin,to elucidate phylogenetic groupings-e.g., thephylogeny constructed for the purple nonsulfurbacteria on the basis of cytochrome c (3). Theseapproaches were limited, however, by the distri-bution among organisms of the protein in ques-tion and by the fact that in many cases one hadto compare functionally related but not equiva-lent molecules, making interpretation of the ex-tent of sequence difference uncertain. Schwartzand Dayhoff (84) have combined published se-quence data from 5S ribosomal ribonucleic acid(rRNA), cytochrome c, and ferredoxin to presenta composite phylogeny for a limited number ofeucaryotes and procaryotes. Although a numberof assumptions are required to integrate such adiverse range of sequence information, the ap-proach represents a significant contribution tothe beginnings of a bacterial phylogeny.A more direct phylogenetic approach is based

on the comparative analysis of 16S rRNA. Theribosome is of ancient origin, is universally dis-tributed, and appears to be functionally equiva-lent over the broad range of bacteria (12, 75,120). The 16S rRNA primary structure is suffi-ciently constrained that it has not changed ex-tensively with time (114). It has regions both ofextreme conservation and of variability, makingit useful in establishing distant as well as rela-tively close relationships. Moreover, the mole-cule is large enough (1,540 nucleotides) to be a"statistical ensemble," which makes it a morecertain indicator of phylogenetic relationshipsthan are smaller molecules (e.g., 5S rRNA)."Comparative cataloging" of 16S rRNA has

been used to detect the phylogenetic relation-ships among the chloroplasts of Lemna (C. R.Woese, unpublished data), Euglena (122), Por-

phyridium (15, 16), and the cyanobacteria (workdone in the laboratory of W. F. Doolittle; 16, 17,29) clearly delineating the procaryotic nature ofthe eucaryotic chloroplast. The technique hasbeen used (36) to confirm and extend the class-ically defined genus Bacillus, based originallyon spore morphology and related structures (41).To date, well over 100 species of procaryoteshave been characterized in terms of their 16SrRNA oligonucleotide catalogs, including repre-sentatives from the genera Acetobacterium,Acholeplasma, Acinetobacter, Actinomyces,Aeromonas, Alcaligenes, Aphanocapsa, Ar-throbacter, Azotobacter, Bacillus, Bdellovibrio,Bifidobacterium, Brevibacterium, Cellulo-monas, Chlorobium, Chloroflexus, Chroma-tium, Clostridium, Corynebacterium, Dactylo-sporangium, Escherichia, Eubacterium, Geo-dermatophilus, Halobacterium, Halococcus,Lactobacillus, Leptospira, Leuconostoc, Micro-bacterium, Micrococcus, Mycobacterium, My-coplasma, Myxococcus, Nocardia, Paracoccus,Pasteurella, Pediococcus, Peptococcus, Photo-bacterium, Planococcus, Propionibacteriumr,Pseudomonas, Rhodomicrobium, Rhodopseu-domonas, Rhodospirillum, Ruminococcus,Sphaerotilus, Spirillum, Spirochaeta, Spiro-plasma, Staphylococcus, Streptococcus, Strep-tomyces, Sulfolobus, Synechococcus, Thermoac-tinomyces, Thermoplasma, Treponema, and Vi-brio (36, 61, 80, 81, 115, 117, 121, 123; Woese,unpublished data).

Recently, Fox et al. (35) applied comparativecataloging of 16S rRNA to a variety of methan-ogenic bacteria. The methanogens were shownto be a unique, coherent biological groupingphylogenetically distant from the typical bacte-ria. Subsequent studies on DNA structure (69),intermediary metabolism (32, 38, 39, 97, 109,108), lipid composition (64, 101, 101a, 102), andcell wall composition (52, 54-56) have confirmedand amplified these results. Herein, we formallypresent a new taxonomic scheme for the meth-anogens based on their phylogenetic relatednessas revealed by 16S rRNA comparisons. In thescheme the species are arranged into a simplepattern consistent with their now known phe-notypic properties. The present phylogeneticanalysis makes it apparent that the generallyaccepted taxonomic classification of bacteria atthe higher levels (40) is not structured to includethe methanogens.

GROWTH OF METHANOGENSTechniques for Growth of Methanogens

Recent advances in the culture of methano-gens are based on the pioneering efforts of Hun-

MICROBIOL. REV.

on Novem


mbr.asm

.org/D

ownloaded from



gate (50, 51). The versatility of these anaerobicprocedures is well documented (19, 21, 31, 47, 48,51, 60, 67). The technique used in this study forroutine growth of methanogens in a pressurizedatmosphere of 80% H2 and 20% C02 has beendescribed by Balch and Wolfe (7). The systemconsists of three components: a serum bottle orglass tube modified to accommodate a thickrubber septum (Fig. 1), a Freter-type anaerobicchamber (5), and a gassing manifold (Fig. 2).Medium without reducing agent (Table 1) isbrought to a boil under an atmosphere of 80%N2 and 20% C02 by use ofthe Hungate technique(51). Reducing agent is added, and the flask isstoppered and transferred into the anaerobicchamber, where the medium is dispensed intotubes. Tubes are sealed in the chamber. Whentransferred outside the chamber, the gas phasein each tube is exchanged for an 80% H2-20%C02 gas mixture by the use of a gassing manifold(Fig. 2).This procedure offers several advantages over

the classical techniques. Medium which is dis-pensed into tubes in an anaerobic chamber isexposed to a uniform, low-oxygen (2 Ad/liter)environment. The thick rubber stopper (Fig. 1),when crimped into place, provides a seal at bothpositive and negative pressures and withstandsnumerous needle penetrations. The use of sy-ringes for all transfers offers a maximal degreeof protection against oxygen and bacterial con-taminants, since the tube is never opened. Theuse of elevated pressure decreases the necessityfor frequent addition of substrate as the methanebacteria consume the gas mixture. When neces-sary, tubes can be repressurized aseptically in afew seconds (Fig. 3) (7). In general, the techniquerequires a minimal investment of time to yieldconsistent, quantitative results.The system, when applied to cultivation of

methanogens in 100- to 200-ml amounts of me-dium (Fig. 4), renders the use of shake flasks(21) obsolete. The bottle offers all the advan-tages of the tube technique; it eliminates therequirement for continuous flushing with a gasmixture and subsequent loss of sulfide from themedium, which may produce erratic growth re-sponses for certain methanogens.Methanogens are easily cultivated on agar

plates by use of the anaerobic chamber. Re-duced, sterile, 2% agar medium is poured intoplastic petri plates maintained in the chamberfor at least 12 h to remove dissolved oxygen.After inoculation, plates are transferred into acylinder (Fig. 5). Outside the chamber the cyl-inder is pressurized to 2 atmospheres with an80% H2-20% C02 gas mixture in a manner similarto that described above for tubes (7). Growth

I I .-Uncrimped Aluminum

n. Cap.*--Block Rubber Stopper

I.*-Crimped Aluminum Cap---Sealed Stopper

18X 150mm Tubewith Serum Viol Lip

--- 23 ml Gas Phase

--o- ~~ 5 ml Assay Medium

FIG. 1. Serum tube (67) and specially manufac-tured black rubber stopper (Belico Glass, Inc., Vine-land, N.J.; no. 2048-00150 serum tube and no. 2048-11800 stopper) used in thepressurized tube technique(7). The stopper is crimped intoplace with a one-piecealuminun seal (no. 224183; Wheaton Scientific Div.,Wheaton Industries, Millville, N.J.).

can be followed by methane production or bythe decrease in cylinder pressure as observed onthe pressure gauge (Fig. 5). Edwards and Mc-Bride (31) describe an alternative ultralow-oxy-gen chamber built inside the anaerobic chamberfor cultivation.

Principal MediaThree basis media (Table 1) were used to

culture the methanogens. Methanospirillumhungatii, Methanosarcina barkeri (strains MS,227, and W), Methanobacterium formicicum,Methanobacterium arbophilicum, Methano-bacterium ruminantium strain PS, and Meth-anobacterium strains M.o.H., M.o.H.G., AZ, andDC were grown in medium 1. M. ruminantiumstrain Ml and Methanobacterium mobile werecultured in medium 1 containing only 2.5 g ofNaHCO3, the medium being additionally supple-mented with (in grams per liter): valeric, isoval-eric, a-methylbutyric, and isobutyric acids, 0.5each; 2-mercaptoethanesulfonic acid, 0.0001. Forculture of M. mobile, an unidentified partiallypurified factor found in rumen fluid was added(W. E. Balch, unpublished data). Methanobac-terium thermoautotrophicum was cultured inmedium 2. Black Sea isolate JR1, Cariaco isolate

VOL. 43, 1979

on Novem


mbr.asm

.org/D

ownloaded from


264 BALCH ET AL.

GASSING MANIFOLD

14 ~~~35cm

A

FIG. 2. Gassing manifold with apparatus for supply of oxygen-free gas (7). (A) Gas mixture tank; (B)reduced copper column (oxygen scrubber) with heater; (C) three-way valve with Swagelok fittings (CrawfordFitting Co., Solon, Ohio); (D) vacuum-pressure gauge; (E) Nupro-Fine metering valves (Whitney Co.,Oakland, Calif.); (F) alternate tubing connectors for gassing probes which may be used in the normalHungate procedure (51); (G) thin-bore polyethylene tubing (Bel-Art Products, Pequannock, N.J.); (H)Vacutainer-Holder Needle (Becton, Dickenson & Co., Rutherford, N.J.). Squares indicate Swagelok brassfittings. Sections of copper tubing 6.35 and 3.18 mm thick are indicated by the shaded and solid regions,respectively.

JR1, Methanococcus strain PS, and Methano-coccus vannielii were cultured in medium 3.The substrate for growth as well as the gas phasefor all media was a mixture of 80% H2 and 20%C02 (Linde Specialty Gas Div., Union CarbideCorp., New York, N.Y.) at 2 atmospheres ofpressure. Media could be stored for up to 2weeks before use.

Strain Histories

The sources and histories of methanogensused in this study are summarized in Table 2.Where available, the culture collection numberof the Gernan collection of microorganisms(Deustche Sammlung von Mikroorganismen[DSM], D-3400, Gottingen, Federal Republic ofGermany) is given to facilitate requests for cul-tures that are maintained by H. Hippe.

Deoxyribonucleic Acid Base CompositionDetermination (Moles Percent Guanine

plus Cytosine)Methanobacterium strains AZ and DC, M.

ruminantium strains MI and PS, Methanobac-

terium strains M.o.H. and M.o.H.G., M. formi-cicum, M. thermoautotrophicum, and M. bar-keri strains MS and 227 were incubated for 30min at 600C in 2% sodium dodecyl sulfate, fol-lowed by disruption in a French pressure cell at3,000 lb/in2. M. mobile, M. vannielii, Methano-coccus strain PS, Cariaco isolate JR1, and BlackSea isolate JR1 were disrupted by freezing andthawing. DNA was purified from the lysed cellsby the method of Marmur (65); the buoyantdensity of the DNA was determined by cesiumchloride density gradient centrifugation in aBeckman model E ultracentrifuge. The base ra-tio was calculated by the method of Schildkrautet al. (83). DNA from Micrococcus lysodeikticus(a gift from C. L. Hershberger) was used as astandard.

GENERATION OF 16S RIBOSOMALRIBONUCLEIC ACID

OLIGONUCLEOTIDE CATALOGSLabeling and Ribosomal Ribonucleic

Acid IsolationThe procedures for determination of the oli-

MICROBIOL. REV.

on Novem


mbr.asm

.org/D

ownloaded from


VOL43,1979~~METHANOGENS:A NEW PERSPECTIVE 265

60 gonucleotide catalogs of rRNA have been de-Z If6Q5 00 scribed extensively (36, 82, 105, 116). In essence,

00 the technique involves labeling cells with* ~ [2PIorthohoshate at a final concentration ofc ito.o 6.3 0.5 to 1.0 mCi/mi. For this purpose, media (Ta-000 C~~~2~..~ ble 1) depleted in phosphate were prepared.

rX4 Yeast extract and Trypticase were dephospho-4 0 05 rylate-d by addition of magnesium acetate (0.35

eq§ £ g) toloo0ml containing2geachofyeast extractoz and Trypticase. The solution was titrated to pH

u M 'k 8.5 with concentrated ammonium hydroxide,

.13 Ci 6' . heldforl1hon ice,andcentrifugedfor10min at

No 0 >~sq~~ rcptte el gonfrthe oforgnr4) a ~~~~~§ 48,000 x g to remove the magnesium phosphate2 0 pre~~~~~~cnitruation,e grwahdandfophened by passagenr

clito 4) ~~~ations in the presence of ['P]were collected by0 ~~~~~~through a French pressure cell at 10,000 lb/in.

eq Total rRNA was obtained by phenol extraction,. U and the 16S rRNA was separated by polyacryl-

0 0 ~ ~ amide gel electrophoresis followed by elution0~~~~1. and purification with Whatman CF-il celluloseco

L6t-4 c c 0 co 4 column chromatography as indicated previouslyz G O z~W e

~~~ ~~. 0 (6).C6. -1 ': Determination of Olligonucleotide Catalog

CD0 o ;., Purified 16S rRNA was digested with ribo-~~-s nuclease (guanyloribonuclease) T1, the resulting

0~~~~~~~~~~~'~a~ 9 ei 0 ~5 oligonucleotides being resolved by two-dimnen-

.~sional paper electrophoresis (82) with modifica-eq~~' ~ - ~~tions (105, 116). Briefly, the analysis involves an

0 $4~ b 0~ initial electrophoretic separation on cellulose0 L66o.4r o6- W, § acetate, followed by a replica transfer by blotting

o~~~ ~ ~ ~~~~~~i~ to diethylaminoethyl cellulose paper, with sub-

000Oeq2~~ -~ sequent orthogonal electrophoresis. The result-ing oligonucleotide "fingerprint" (or primaryEach isopleth is a coherent group defined by the

~~~oo~~~~~~.2 unurdine content per oligonucleotide. Within each~~.~ (n 9. isopleth, oligonucleotides are displayed by order

R O.U ~ of size, and within each size "isocline" they areeq ok separated to some extent on the basis of corn-

z ~ ~ ~ ~~a position. Figure 6 is a representative fingerprint'z.' _77 of this type. The sequences of the various oli-0.0a0 - gonclotiesare deemndb secondary and

0 o0 S. ~ Analysis ofData~~~ ~ ~~, ~ The resulting oligonucleotide catalogs are ex--~~~~ ~ amined by a standard taxonomic analysis (4).

7a- w~~ -" An association coefficient (SAB) for each binarycouple is defined as follows: SAB = 2NAB/ANA +

2 49.0 0 NB),whereNAandNBare the total numbersof.349r 9.9 . nucleotides (found in hexamers and larger) in

~~ ~ ~~ .~~ catalogsAandB. NABrepresentsthenumberof.2 ~~~~~~~nucleotides in oligomers common to the two5~~~~~ la

2 eq~~~~~~cataogs. SAB values range from 1.0 (complete- ~~0 0 cnOZ0sequence homology) to 0.02 (the random coin-

to LK~ cidence level). The values are an underestimate

VOL. 43, 1979

on Novem


mbr.asm

.org/D

ownloaded from


266 BALCH ET AL.

FIG. 3. (A) Use ofthe gassing manifold to aseptically repressurize cultures duringgrowth ( 7). (B) Detailedview of a sterile gassing syringe connected to a tube and gassing channel.

of the actual percent homology between twocatalogs, because related but nonidentical ohl-gomers are not considered. The matrix of SABvalues for each binary comparison over a givenset of organisms is used to generate a dendro-gram by average linkage (between the mergedgroups) clustering (4).SAB is a measure of the sequence homology

between two rRNA's. Branching patterns aregenerally independent of whether single, aver-age, or complete linkage clustering techniquesare used (4). Moreover, analysis of the data interms of families of related oligonucleotides (36;G. E. Fox, unpublished data) yields similarbranching patterns.

COMPARATIVE CATALOGING OF 17ME HANOGENS

The 17 organisms whose 168 rRNA oligonu-cleotide catalogs are listed in Table 3 includethe major types of methanogens now in pureculture. These lists include previously unpub-lished 16S rRNA oligonucleotide catalogs for 7methanogens: Methanobacterium strains AZand DC, Methanobacterium strain M.o.H.G., M.mobile, M. barkeri strain 227, M. vannielii, andMethanococcus strain PS. The extent of theirrelatedness to all methanogens thus far exam-ined (6,35) can be seen in Fig. 7; this dendrogramis constructed from the SAB values calculated forindividual binary comparisons given in Table 4.Values for three other groups (the enteric-Vi-brio, Bacillus, and cyanobacteria) are averagesof several representative species and are in-cluded in this dendrogram for perspective.

FIG. 4. Modified 1-liter storage bottle (no. 219760;Wheaton Scientific Co.) used forgrowing cells in 100-to 200-ml liquid volumes at 2 atmospheres ofpressure(9). The cap has a 21-mm-diameter hole to allow analuminum seal tube (Bellco Glass), which has beencut and flanged, to protrude from the bottle. The tubeis sealed in the bottle with a no. 4 black rubberstopper with a no. 9-size hole. The tube and stopperare held in place by screwing the bottle cap down.The bottle ispressurized after autoclaving and beforeuse. For safety, the bottle is maintained in a stainlesssteel canister after pressurization (9).

A dendrogram published when 10 methano-gens had been characterized showed the groupto possess two major divisions (35); we now seeat least three (Fig. 7). The first division containsthe Methanobacterium species with the excep-

MICROBIOL. REV.

on Novem


mbr.asm

.org/D

ownloaded from


METHANOGENS: A NEW PERSPECTIVE

77Z,4u~~~~- . - -. ..GE_,

7 ~~~~~~t*q w t 7*000ft0;yC:

--axf

M mbran

SWbrm

<: r@0. :-~- siw,inut' '*;. ff0

FIG. 5. Pressure cylinder used to incubate agar plates at an elevated pressure of 2 atmospheres of an 80%H2-20% CO2 gas mixture. The cylinder length is designed to fit into the air lock ofa Freter (5)-type anaerobicchamber (Coy Manufacturing Co., Ann Arbor, Mich.).

tion of M. mobile, the second contains Methan-ococcus species; and the third contains Meth-anosarcina, Methanospirillum, M. mobile, andtwo new marine isolates (Black Sea isolate JR1and Cariaco isolate JR1). Divisions I and IIIhave two subgroups each (see Fig. 7): IA com-

prises coccobacillus-like, gram-positive rods; IBcomprises long, gram-positive rods; IIIA com-

prises various gram-negative forms, and IIIBcontains the gram-positive sarcina-like methan-ogens.Although Fig. 7 was generated by average

linkage clustering analysis between the mergedgroups (4), application of complete linkage or

single linkage or both gives a dendrogram ofnear identical topology to that shown (G. E. Fox,unpublished analysis). It is apparent that thechoice of clustering procedure does not have a

significant effect on the branching orders re-

ported here. The standard deviations about themeans for all binary comparisons involved in a

cluster are presented as brackets about the SABvalues (Fig. 7).

DEVELOPMENT AND OUTLINE OF ANEW TAXONOMIC TREATMENTBASED ON 16S RIBOSOMAL

RIBONUCLEIC ACIDThe methanogenic bacteria are recognized as

a separate family in Part 13 of the 8th edition ofBergey'sManual ofDeterminative Bacteriology(20). Table 5 presents the current key to thegenera of the family Methanobacteriaceae. Themajor determinative feature is morphology. Theresult of such a qualitative approach is a treat-ment which provides little insight into the rela-tionships between the various species. A solutionto the present inadequate taxonomic status isprovided by the results obtained from compar-ative cataloging of 16S rRNA. The degree ofsequence homology found among the 16SrRNA's isolated from different organisms pro-vides a quantitative basis for establishing rela-tionships.The methanogens represent a coherent phy-

logenetic grouping quite distinct from other typ-ical bacteria. Just how distinct they may be isindicated in Fig. 7; even the enteric-Vibrio spe-cies and the cyanobacteria appear closely relatedby comparison. By use of these same criteria, itis apparent that the methanogens as a group arevery diverse. The deepest branches are as dis-tinct from one another as the gram-negativeenteric-Vibrio group is from the gram-positivebacilli (Fig. 7). We propose that the Methano-bacteriaceae Barker 1956 be redefined to con-tain only a limited range of methanogenic spe-cies. SAB values derived for all species provide

267VOL. 43, 1979

on Novem


mbr.asm

.org/D

ownloaded from


TABLE 2. Sources and histories of methanogenic bacteria used in this study

Organisma Strain Source and history of culture DSM Referenceno.

1. Methanobacteriumarbophilicum

la. Methanobacterium sp.

2. Methanobacterium sp.

3. Methanobacteriumruminantium

4. Methanobacterium sp.5. Methanobacterium sp.

6. Methanobacterium sp.

7. Methanobacterium for7nicicum

8. Methanobacteriumthermoautotrophicum

9. Methanococcus vannielii

10. Methanococcus sp.

11. Cariaco isolate

12. Black Sea isolate

13. Methanospirillum hungatii14. Methanobacterium mobile15. Methanosarcina barkeri

15a. Methanosarcina barkeri

15b. Methanosarcina barkeri

DH1 J. G. Zeikus, wet wood oftrees

DC Isolated from enrichmentculture provided by D.Castignetti

AZ A. J. B. Zehnder, digestedsewage sludge

PS M. P. Bryant, primarydigester, sewage

Ml M. P. Bryant, rumenM.o.H. M. P. Bryant, from syntrophic

culture of "M. omelianskii"M.o.H.G. S. Schoberth, Goittingen

isolate similar toMethanobacterium sp.strain M.o.H.

MF M. P. Bryant

A&H R. S. Wolfe, sewage sludge

SB H. Hippe, San Francisco Bayblack mud

PS P. H. Smith, WaccasassaEstuary, Fla.

JR1

JR1

JF1BPMS

227

W

J. A. Romesser, CariacoTrench sediment, Venezuela

J. A. Romesser, Black Seasediment

J. G. Ferry, sewage sludgeP. H. Smith, rumenM. P. Bryant, Urbana, Ill.,sewage sludge digester

R. A. Mah, Los Angeles, Calif.sewage sludge digester

R. A. Mah, gas-vacuolatedstrain

1125 Zeikus and Henning (130)

1536

744

861

1093863

862

D. Castignetti and D. A. Klein,Abstr. Annu. Meet. Am. Soc.Microbiol. 1977, 014, p. 247

Zehnder and Wuhrmann (125)

P. H. Smith, Abstr. Bacteriol.,1961

Smith and Hungate (88)Bryant et al. (23)

1535 C. G. T. P. Schnellen, Ph.D.,thesis, Technical, Universityof the Delft, Delft, theNetherlands, 1947

1053 Zeikus and Wolfe (131)

1224

1537

1497

1498

8641539800

1538

Stadtman and Barker (90)

J. M. Ward, Ph.D. thesis,University of Miami, Miami,Fla., 1970

Romesser et al. (81a)

Romesser et al. (81a)

Ferry et al. (33)Paynter and Hungate (79)C. G. T. P. Schnellen, Ph.D.

thesisMah et al. (62)

R. A. Mah, Abstr. Annu. Meet.Am. Soc. Microbiol., 1977,I32, p. 160

a Organism numbers correspond to those in Table 4.

the basis for description of three orders, fourfamilies, and seven genera. These categories re-flect taxa characteristic of the typical bacteria.

OrdersFigure 8 presents a new taxonomic treatment

of the methanogens based on comparative cat-aloging of 16S rRNA's. At the highest level themethanogens are divided into three new orders.Order I, Methanobacteriales (group I in Fig. 7);order II, Methanococcales (group II in Fig. 7);and order III, Methanomicrobiales (group III inFig. 7). Separate groupings at least at the rankof order are justified by the low SAB valuesobserved (0.22 to 0.28) between the proposedorders (Fig. 7 and 8). These SAB values arecomparable to those separating the typicalgram-negative and gram-positive bacteria.

FamiliesOrder I, Methanobacteriales, and order II,

Methanococcales, contain one family each, anemended description of the Methanobacteri-aceae Barker 1956 and the Methanococcaceae,respectively. Order III contains two new fami-lies, Methanomicrobiaceae and Methanosarci-naceae (Fig. 8). The two families in order III arejustified on the basis of the SAB value relatinggroups IIIA and IIIB (SAB = 0.35) in Fig. 7. TheSAB value is similar to that obtained in thebinary comparison of the phototrophic bacte-rium Chromatium sp. (family Chromatiaceae)with Rhodopseudomonas sphaeroides (familyRhodospirillaceae), 0.34, or that of Bifidobac-terium breve (family Actinomycetaceae) andStreptomyces griseus (family Streptomyceta-ceae), 0.35 (Woese, unpublished data).

GeneraThe family Methanobacteriaceae in order I is

proposed to contain two genera; an emendeddescription of the genus Methanobacterium

268 BALCH ET AL. MICROBIOL. REV.

on Novem


mbr.asm

.org/D

ownloaded from


METHANOGENS: A NEW PERSPECTIVE

Kluyver and van Niel 1936 and a new genus,Methanobrevibacter (Fig. 8). Comparative cat-aloging of 16S rRNA of the genus Bacillus sup-ports this separation (36). For example, in thebinary comparisons of Bacillus subtilis withBacillus cereus, Bacillus pasteurii, and Sporo-lactobacillus inulinus, one obtains SAB values of0.73, 0.65, and 0.56, respectively. The SAB valueof 0.5 obtained in the binary comparison ofgroups IA and IB (Fig. 7) is of sufficient depthto justify two genera.The Methanococcales pose a dilemma. An

SAB of 0.60 is observed in the binary comparisonof the two species cataloged (Table 4). Thisvalue is not substantially different than thatobserved for groups IA and IB (Fig. 7). Sinceonly two species are presently described, weconservatively propose that the family Meth-anococcales contain an emended description ofthe single genus Methanococcus (Kluyver andvan Niel) Barker 1936. This assessment of theirrelationship may have to be modified as moreisolates become available.The family Methanomicrobiaceae in order III

is proposed to contain three genera: Methano-genium Romesser and Wolfe 1979, Methanos-pirillum Ferry, Smith, and Wolfe 1974, and anew genus, Methanomicrobium. The relativepositions of the three genera with respect toeach other are presently not clear. For example,in Table 4 average SAB values of 0.51 and 0.52are observed in the binary comparisons of thetwo species in Methanogenium (Cariaco isolateJR1 and Black Sea isolate JR1) with the Meth-anospirillum sp. (M. hungatii) and the Meth-anomicrobium sp. (M. mobile), respectively. AnSAB value of 0.36 is observed in the binary com-parison of the Methanomicrobium sp. (M. mo-bile) and Methanospirillum sp. (Table 4). Thesedata suggest that the genera Methanomicro-bium and Methanospirillum are more distantfrom each other but equidistant from Methan-ogenium. The clustering technique used in Fig.7 defines a slightly different topology. Additionalisolates are needed to clarify the relationshipsamong the three genera.

SpeciesThe compositions of the proposed genera are

as follows (Fig. 8, Table 6). In the family Meth-anobacteriaceae the genus MethanobacteriumKluyver and van Niel 1936 is emended to containonly three species. These include M. formicicumSchnellen 1947 (C. T. G. P. Schnellen, Ph.D.thesis, Technical University of the Delft, Delft,the Netherlands, 1947) as the neotype species,M. thermoautotrophicum Zeikus and Wolfe1972, and a new species, M. bryantii, formerly

FIG. 6. Representative 16S rRNA oligonucleotidefingerprint (primary), using ribonuclease T,. Iso-pleths, with their characteristic uridine (U) contentare indicated by numbers. Size isoclines within eachisopleth are indicated by large letters. The oligonu-cleotide compositions of representative oligomers isso indicated, where G is guanosine, A is adenosine,and C is cytosine.

Methanobacterium strain M.o.H. Bryant 1971.The relationship of M. bryantii to M. formici-cum and M. thermoautotrophicum is defined bythe average SAB value of 0.6 (Fig. 7), a valuesimilar to that found in defining separate Bacil-lus species (36). M. bryantii is named for MarvinP. Bryant, who pioneered studies in the separa-tion and characterization of this species from theMethanobacillus omelianskii syntrophic asso-ciation (23). An SAB value of 0.93 was obtainedin the binary comparison of M. bryantii andMethanobacterium strain M.o.H.G. Such highSAB values are found for closely related membersof a genus or species. For example, SAB values of

269VOL. 43, 1979

on Novem


mbr.asm

.org/D

ownloaded from


270 BALCH ET AL.

TABLE 3. Oligonucleotide catalogs for 16S rRNA's of 17 methanogensa

Oligonucleotide sequence Present in organism Oligonucleotide sequence Present in organismno.: no.:

1-15; 1, 2, 6, 128,9159,10,15139,131-15; 12, 139-1410,11, 141-12, 15; 151-10; 1, 2, 5, 611-141-2, 8-10,14, 157, 11138-9, 11-12, 151-9,11-151-15; 9, 101-3, 5-71-10,125-7; 68135-141-10; 73-4, 12-13159-111534-8, 11-15; 4, 8, 11-

13; 117,992, 9-10; 1014141-8, 12; 75-6, 8, 10, 12, 141-10; 7, 91-2, 9-108, 10, 12115-6,9-14; 13611-15; 12-15; 1411-121-10, 154-7,11-123-4,151-15; 1-3, 5, 7-10,12, 151-3, 5-6, 9-10, 15; 34,7, 131-15; 1-3, 6, 8, 151-15; 1-8, 11, 13-14;

1114

UUUCGUUCUGUCUUGCUUUGUUUAGUUAUGUUUUG

6-mersCCCCAGCCCAAGCAACCGACCACGACACCGAAACCGCCACAGAAAACGCCCACGCCCUCGCCUCAGCUCCAGUCCCAGCCACUGACCUCGCCUAAGCUCAAGCCAUAGCAUACGACACUGAACCUGAAUCCGCUAAAGUAAACGACUAAGACAAUGAUAACGAAUACGAACAUGAAACUGAAAUCGAAUAAGCCCUAGUACACGCCUACGCCAUCGAUCCCGAACUAGCCCUUGCCUCUGUCCCUGCCUUAGCUCUAGCUUCAGUCCUAGUUCCAG

CCUAUGCUACUGUCACUGCUAUCG

7, 9,11, 13, 14412,131-6, 153, 10-11, 14; 147, 13; 133, 10,13

7-108, 10,1512, 1311-138,11-13, 1512,13, 1514149, 10

5-6, 12, 155-65-63, 11-155-7, 13131-6, 155-87, 141-25-7, 11-141-10, 15; 1, 29-11, 13-15; 1411-141-8, 12-13, 159-10, 131-15151-8, 1515; 151-15; 12, 13, 141-4, 111-3, 5-10, 142149-109-109-109-108, 1511-141-15; 97, 11, 121-4131-31-84, 101-4, 8-94, 10-1411-15

5-mersCCCCGCCCAGCCACGACCCGCCAAGCACAGCAACGACACGACCAGAACCGACAAGAAACGAAAAGCUCCGCCCUGUCCAGCUCAGCCAUGUCACGUACCGACCUGACUCGAUCCGUAACGCAAUGACUAGACAUGAUACGAAUCGUAAAGAUAAG

AAAUGCCUCGCUAAGUCCCGCACUGUUCCGCUUCGUCCUGCCUUGCUCUGUCUAGUUCAGCUAUGUACUGUAUCGACUUGAUCUGAUUCGUUAAG

UAAUGAUAUGAAUUGAUUAG

UAUAG

MICROBIOL. REV.

on Novem


mbr.asm

.org/D

ownloaded from


METHANOGENS:A NEW PERSPECTIVE 271

TABLE 3-ContinuedPresent in organism

cPresent in organismOligonucleotide sequence no.: Oligonucleotide sequence no.:

AACAUCGAAAUCCGUAAAAAGUAAACAGCUCACCGACCCUCGUCCCAAGCAACUCGCCCUUAGCAUCCUGUACUCCGAUCUCCGACCUUCGUCCUAAGUUACCAGCUAACUGUAACUCGAUUCCAGAUCAUCGAAUCUCGAACCUUGUCUAAAGCUUAAAGCAAUAUGAUACUAGAAUCUAGAAAUCUGUAAAAUGUAACCUGCCCUAUGUAAAUAGUCACCUGCUCCUUGUCCCUUGUUCUCCGCUCUUAGUACUUCGUACUCUGUCAUAUGUAAUCUGAAUUUAGUUAACUGAUCCUUGCAUUAUGUUCUUCGUCUCUUGCUUUAUGUUUAUCGUAUUUCGAUUAUUGCUUUUAGUAUUCUGUUCUUUGUAUUUUGUUUUUUG

8-mersCCACAACGACCCCAAGAAACCCCGUCCACCAG

1511-131-2, 4-89141414141-2, 4-811-15111213111-3, 5-8, 154, 71-4, 7, 9-12, 1511, 1484810, 1510-131511-3, 7, 11-12, 14151559-10, 141091-6, 8-151311, 14312151574149-10101511, 12, 14151-21151025-841-4

1-31-2,5-6,9-101313

UCAUCGCAUCUGACUCUGACCUUGAUCCUGUCUAAGUUACAGUAUCAGUAUACGUAAUCGAUACUGACAUUGAACUUGAAUCUGUAAAUGAUUAAGAAUAUGUUACCGCUCAUGCACUUGCAUAUGCCUUCGCCUUUGCUUUCGUCUCUGUUCCUGUCUUAGCUAUUGUUACUGUAUUCGAUUCUGACUUUGUAUAUGUCCUUGUCUAUGCUUUUGUCUUUGUUUUUG

7-mersACCCACGACCACCGAACCCCGCCAACAGCAACACGCAAACCGAAAAAAGCCCUACGCCCACUGUCCACCGCCACCUGCCCUAAGUCACACGCUACACGUAACCCGAUACCCGAACCUCGCCUAAAGUAACACGAUAACCGAAUCCAG

11, 131111-125-10, 141-8, 11-1511-12, 141213111-154, 9-12, 14-151-245-6, 8, 11-1471-8, 11-12, 14131499-10; 109-10101-3, 5-6, 9151-3, 5-85-65-61-4, 7-101543, 12-13, 1531210101-7; 1, 22, 711

1-8,11-1411, 1411, 1411, 12,141,3,6,812, 13101-15157,810,1511, 1242, 5, 7, 9-156,81-13, 15121-91-8, 11-151112-13, 15

VOL. 43, 1979

on Novem


mbr.asm

.org/D

ownloaded from


TABLE 3-ContinuedPresent in organism Oliltd Present in organism

Oligonucleotide sequence no:gonueone sequence no.:

CCCACAUGCUCAACCGACCCUCAGACCACCUGUAACACCGAUCCCAAGAAAUCCCGAUACCCCGAAACCUAGCCCUCAUGUACUCCCGAU(CCUC)CGCCUAUCAGCCUAACUGCUUAACCGUAAUCCCGCUACAAUGUACUACAGUAAUACCGAUUACCAGAUAACCUGACAAUCUGAAAUCCUGAUAAACUGAUAAAUAGUAACUACGCUACCCUGUACUCCAGAAUCUCCG(CU,CCUU)CGAUCCUUCGUCUAACUGCUUAACUGUAAUCCUGUCUAAAUGUUAAAUCGCAUAUAUGAAAUCUUGAAAUUCUGAUAAAUUGUACUUCAGAUAAUCUGCUUUUCAGUUCUCAUGUUUAAUCGUAUCAUUGUUUAAAUGAUAUUUAGUUUAAUUGUUUUUUCGUUUUAUUG

9-mersCCCACCAAGCACACACCGCACACAACGCCCAACAAGAACCCCAAGAAACCCAAGCCUCACCAG

11-12,1412111-2, 4-8, 12, 151-8,10,153-41-21491-2, 4, 775-6155-67, 11, 13131-15159,11-1445, 8,11-12, 15131-3, 8, 11-134-83149-1099771-23-6, 8,141-4, 81-21515153-4, 141-214148313133-451-12, 14-152, 341

5-7, 9-101-1514,1511-148712

CCUACCAAGCCUACAACGAUAACCCCGAAACCUCCGCACACUAAGAUAAACCCGUACUCCCAGUAAUCCCCGAAUCCCCUGCUUACCAAG(UC)ACACAUG(UC)ACAAUCGUCAUAACCGCUAAUACCGACCCUUAAGAUAACCCUGAUAACCCUGAUAAUACCGAUAUACAAGCUUAACCAGCACUCAUAGUACUCCAAGUCUUACCAGUCACUAUCGUAAUCCCUGUAAUCCUCGAAUUUCCCGAAUCCUCUGUCAUAAUCGCUAAUACUGCAUCAUAUGAUAAUUCCGAAUUAUCCGUUUAAAACGUAAACUAUGAUAAUACUGAAUCUCCUGUAAACUCUGCCUAUCUAGAUAACCUUGCUAUUACUGUUAAAUUCGUUUAAUAAGUAAUCCUUGUAAACUUUGUUAUAUUCGUAUUUCUAGUUUAUUAAGAUUUUUAAGCUUUUAUUGUUUUUCCUGUUUUUAUUGUUUUUUUCG

1O-mersCACAACCACGAAUAACCCCGACCACCUAUGAAUCUCACCGAAAUCUCACG

8158, 12, 14-151-81-881-8111-2, 4-81-443-47411131-75-7139-109-101015815121531,2,5,61,2151511-141111, 1439, 1091010131,22, 31410313114823, 71

4-6, 8, 13119, 10,13127, 14

H


on Novem


mbr.asm

.org/D

ownloaded from


METHANOGENS:A NEW PERSPECTIVE 273

TABLE 3-Continued

Oligonucleotide sequence no.:|n| Oligonucleotide sequence Presentin organismno.: ~~~~~~~~~~~~~~~~~~no.:

UAACUCAAAGAAACUUAAAGACCUUACCUGUUACCAUCAGUACCUACUAGAAUCACUUCGAACCCUUAUGUAAAUAACUGUAACUCAUAGCAUAACCUUGUUCUUCACCGACUCUACUUGCUUAACUAUGAUACUAUUAGAAUCAUUUCGUUCCCUAUUGUCUUCUUAAGAUUUUUUUCGAUUUCCUUUGUUUUCUUUUG

l1-mersCACCACAACAGACAACUCACCGAAAUCCCACAGCAUCUCACCAGUAACUCACCCGAAAUCUCACCGAAACACCUUCGAAAUCCCAUAGUCCCUCCCCUGCAUAUCCUCCGAAAUCCUAUAGCAUCUUACCAGUUUCAACAUAGA(UA,UCA,CUA)UGCUUUUAUCAAGUCUCCUAACUGUUUCAAAAUAGUUUCAAUAUAGCUUUUCUUAAGCUUUUCAUUAGUUCUUUAAUCGUUUCUUCUCCGAUUCUUUUCUGUUUUUUUCCUGUUUUUUUUAAG

12-mersCCACCCAAAAAG(AC,U,CCC)AAAAAGUCAAACCACCCGUCACACCACCCGUCAAACCAUCCGUCACACCAUCCGACAUCUCACCAGCCACUCUUAACGCUCAUUAACACGCCAUUCUUAACGCUCAACUAUUAGCUCAUUAUCACG

121-151541568131098131,22, 3, 5-7577125,6

121589-11, 131311, 1385,6151541411, 13891414151, 2, 431110913

1-3, 7-9412-13, 15911101-85, 6-891-41510

CCACUAUUAUUGCAAUUAUUCCUGUUCAUAAUACUGCCACUUUUAUUG(AA,UUUA,UA)UUCGCCAUUUUUAUUG(CUA, CUUUUA)UUG

13-mersUAAACUACACCUG(CAA, CCA)CAUUCUGUAAUACUCCAUAGUUUCAAAAUAACGACCACUCUUAUUGACCACUUUUAUUGAUAAUUUUUCCUG(UUU,CUU,CU)AAAUGUUUAAAUCCUCUG

14-mersAAAACUUUACCAUGUCAAACCAUCUUAGUUUAAAACACAUCGAAAACUUUACAAUGAACUACUUUCUCCGAAUAAUCUUUUCUGAUUUUU(CCU,CU)UUG

15-mersUCUAAAACACACCUGAUAACCUACCCUUAGAUAACCUAACCUUAGAUAAUACUCCUAUAGAAUAAUACCCUAUAGAAUAAUACUCCAUAGAUAAUCUACCCUUAGUCUUCUAUUUAAAUGUUAACUUAACCUCAGAUAAUUCUCCAUAAG

16-mersUAAUCCCCUAAACCAGAAAUCCUAUAAUCCCGAAUCUCCUAAACAUAGCAAUCUCUUAAACCUGUAAUCUCCUAAACCUGUUAACUUACCCUCAAGAAAUCCUAUAAUCCUG

17-mersUCAAUCUCCUAAAC-CUG

CAAUCUUUUAAA-CCUAG

UAAU(CCU,CU)AA-ACUUAG

AUAAU(CCU,CU)-AAACCUG

18-mersAACAAUCUCCUAAAC-CUG

CAAAUCUCCUAAACC-CUG

1139, 101222, 5-64

15813129, 1014469, 10

13141411-12, 14, 15993

121-471012116599

875,6117105,6

14

4

1-3

13

12

10

VOL. 43, 1979

on Novem


mbr.asm

.org/D

ownloaded from


TABLE 3-Continued

Oligonucleotide sequence Present in organism Oligonucleotide sequence no.:

24-mers Posttranscriptionally modified sequencesb(AAACA,UAAUCUCA)- 15 1. a. CCCG 1-8CCCAUCCUUAG b. ICCCG 9, 10

Termini, 5' end c. C(C,C)G 11-15pAG 7, 8 d. APUAG 9, 10pAAUCCG 5,6 2. CCGxCG 15pAAUCUG 1,2,4 3. a. AACCUG 1-10pAUUCUG 3,11-15 b. AAUCUG 11-15pAUUCCG 9,10 4. a. UAACAAG 1-10

Termini, 3' end b. UAACAAG 11-15AUCACCUCCUoH 1-8,11-15 5. a. AUNCAACG 1-10AUCACCUCCoH 9, 10 b. ACNCAACG 11-15

aThe fist column is the oligonucleotide sequence (C, cytosine; G, guanosine; A, adenosine; U, uridine); thesecond column shows organisms in which that sequence is found. Organisms are designated by number (seeTable 4) as follows: 1, M. arbophilicum and Methanobacterium sp. DC; 2, Methanobacterium sp. AZ; 3, M.ruminantium PS; 4, M. ruminantium Ml; 5, Methanobacterium sp. M.o.H.G.; 6, Methanobacterium sp. M.o.H.;7, M. fornicicum; 8, M. thermoautotrophicum; 9, M. vannielii; 10, Methanococcus sp. PS; 11, Cariaco isolateJR1; 12, Black Sea isolate JR1; 13, M. hungatii; 14, M. mobile; 15, M. barkeri strains MS and 227. Multipleoccurrences of a sequence in a given organism are denQted by repeating the organism's number in the secondcolumn, e.g., "1-4, 6-8; 3, 7; 3" signifies a double occurrence in organism 7 and a triple occurrence in organism3.bThe posttranscriptional modifications indicated by the superscript dots are identified as described in

footnote a of Table 13. ', Pseudouridine.

0.88, 0.87, and 0.87 are generated in the binary *#0MA wbwbacwncomparisons of Bacillus brevis strains 1028

M. strain DCe! #~~~M sftrin AZ

(NRS) and 8185 (ATCC), 1028 and 953 (NRS), A Mnmn stroin PSand 8185 and 953, respectively (Woese, unpub- Mn#*nIwm stroin Mllished data). Methanobacterium strain M.o.H.G.is designated M. bryantii strain M.o.H.G. I M. strain M.o.H.G.The genus Methanobrevibacter contains M. stroin M.o.H.

three species: M. ruminantium as type species, M ^M. arboriphilus, and a new species, M. smithii, M. I_fmodhcumformerly Methanobacterium ruminantiumstrain PS Smith 1961 (P. H. Smith, Abstr. Bac- E*#lV=Cs V0*nnie,iteriol., A40, p. 60, 1961). M. arboriphilus, for- M. stroin PSmerly Methanobacterium arbophilicum (sic)Zeikus and Henning 1975, and M. ruminantium, Cortoco solate JRIformerly Methanobacterium ruminantium A

strain Ml Smith and Hungate 1958, are justifiedas separate species based on the SAB valuespresented in Fig. 7. M. smithii is named for Paul B AM ,*kesobn kr27H. Smith, who isolated the organism in his pi- Entrtc-Vibaisp.esoneering work on the ecology and physiology ofthe methanogenic bacteria. Methanobacterium Cyanabactena spessp. strain AZ Zehnder and Wuhrmann 1977 Cyor_obocteri__sp___syields an SAB value of 0.84 in the binary com- o .1 .2 3 .4 .5 .6 .7 .8 .9 1.0parison with M. arboriphilus (Fig. 7, Table 4). SABThe 16S rRNA catalog of Methanobacteriumsp. strain DC was identical to the M. arboriphi- FIG. 7. Dendrogram of relationships of methano-hpstriCave tent.atived. ae bohi gens and typical bacteria. The figure was constructedlus catalog. We have tentatively designated both by average linkage clustering (between the mergedorganisms as strains of M. arboriphilus (Fig. 8, groups) from the SAB values given in Table 4. TheTable 6) until information on more features is standard deviation about the mean for all binaryavailable. comparisons involved in a given cluster is indicated

In the family Methanococcaceae the genus by the dotted lines.


on Novem


mbr.asm

.org/D

ownloaded from



TABLE 4. SAB values (lower sector) for each indicated binary comparison and total number of nucleotidesin sequence hexamers and larger that are common to both catalogs (upper sector) a

OrganismOrganism

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

1. M. arbophilicum, Meth- 540 389 372 307 314 308 316 174 155 151 159 123 129 156 58 58 52anobacterium sp. DC

2. Methanobacterium sp. AZ 0.84 426 386 342 335 325 309 181 162 158 166 130 136 163 60 59 493. M. ruminantium PS 0.62 0.67 371 283 290 288 293 166 147 163 169 141 157 166 58 59 494. M. ruminantium Ml 0.60 0.62 0.62 297 297 285 296 148 159 148 156 131 144 155 71 68 505. Methanobacterium sp. 0.51 0.56 0.48 0.51 533 354 340 165 164 150 170 156 155 176 61 60 52

M.o.H.G.6. Methanobacterium sp. 0.52 0.55 0.49 0.51 .93 347 346 158 1571 135 155 149 148 161 61 60 52

M.o.H.7. M. formicicum 0.51 0.53 0.49 0.49 0.62 0.61 323 175 168 157 163 135 157 150 62 59 538. M. thermoautotrophicum 0.510.50 0.49 0.50 .58 .60 0.56 165 158 151 174 151 151 184 63 58 509. M. vannielii 0.290.300.280.25 .29 .28 0.31 0.28 352 132 115 121 126 119 46 31 26

10. Methanococcus sp. PS 0.25 0.26 0.24 0.27 .28 .27 0.29 0.26 0.60 137 120 126 124 139 57 37 2911. Cariaco isolate JR 1 0.24 0.25 0.27 0.25 .26 .23 0.27 0.25 0.23 0.23 352 296 320 207 33 42 4812. Black Sea isolate JR 1 0.26 0.27 0.28 0.26 .29 .27 0.28 0.29 0.20 0.20 0.59 305 292 261 39 57 5413. M. hungatii 0.20 0.21 0.24 0.22 .27 .26 0.23 0.26 0.21 0.21 0.50 0.52 214 201 45 52 5114. M. mobike 0.21 0.22 0.26 0.24 .26 .25 0.27 0.25 0.22 0.21 0.54 0.490.361 197 40 48 5515. M. barkeri MS, 227 0.25 0.25 0.27 0.25 .29 .27 0.25 0.30 0.20 0.22 0.34 0.43 0.33 0.32 60 64 4616. Enteric- Vibrio species 0.10 0.08 0.08 0.10 .10 .09 0.09 0.10 .070.08 0.05 .06 0.07 0.10 0.10 157 14517. Bacillus species 0.100.100.100.14 .10 .11 0.11 0.12 .06 0.06 0.08 .10 .10j0.10 0.08 .27 15118. Cyanobacteria species 0.1010.10 0.10 0.10 .10 .10 .10 .11 .060.06 .08 .09 0.08 0.10 0.11 .24 .26

a Values given for the enteric-Vibrio species, Bacillus species, and cyanobacteria species represent averagesobtained from 11 (L. Zablen, Ph.D. thesis, University of Illinois, Urbana, 1975), 7 (36), and 4 (16) individualspecies, respectively.

Methanococcus (Kluyver and van Niel) Barker1936 is emended to contain two species: M. van-nielii Stadtman and Barker 1951 as the neotypespecies, and a new species, M. voltae, formerlyMethanococcus strain PS (J. M. Ward, M.S.thesis, University of Miami, Miami, Fla. 1970).M. voltae is named for the Italian physicistAlessandro Volta, who discovered the combus-tible nature of gas from anaerobic sediments.

In the family Methanomicrobiaceae, the ge-nus Methanomicrobium is proposed to containthe single member M. mobile, formerly Meth-anobacterium mobile Paynter and Hungate1968. Results of comparative cataloging suggestthat this organism is unrelated to the species ofgroup I Methanobacteriaceae (SAB = 0.24) inFig. 7. It is sufficiently distinct from other mem-bers of the Methanomicrobiaceae (Fig. 7) to beplaced in a new genus (Fig. 8).The genus Methanogenium contains the spe-

cies M. cariaci (formerly Cariaco isolate JR1) asthe type species and M. marisnigri (formerlyBlack Sea isolate JR1) (Fig. 7). These are twonew marine isolates recently described by Ro-messer et al. (81a).The genus Methanospirillum contains the

single species M. hungatei as the type species,formerly M. hungatii (sic) Ferry, Smith, andWolfe 1974. The organism was named in honorof R. E. Hungate (33). (Incorrect spelling neces-

sitates the correction of the species name fromhungatii to hungatei at this time.)

In the family Methanosarcinaceae, the genusMethanosarcina (Kluyver and van Niel) Barker1956 is emended to contain the single memberMethanosarcina barkeri Schnellen 1947(Schnellen, Ph.D. thesis) as the neotype species.Three other isolates of Methanosarcina wereobtained for 16S rRNA analysis (Table 2). M.barkeri strain 227 (62, 86) was indistinguishablefrom M. barkeri strain MS (Table 4, Fig. 7).Partial characterization of M. barkeri strain W(R. A. Mah, M. R. Smith, and L. Baresi, Abstr.Annu. Meet. Am. Soc. Microbiol. 1977, I32, p.160), a gas-vacuolated strain, indicated an SABvalue between 0.8 and 0.9 in the binary compar-isons with strains 227 and MS. We conserva-tively propose a strain designation for both iso-lates.The proposed taxonomic treatment for the

methanogens is summarized in Table 6 and Fig.8. All references for the remainder of this paperare to the new designations, unless specifiedotherwise.

EVIDENCE IN SUPPORT OF THENEW TAXONOMIC TREATMENT

Recent evidence provided by studies on cellwall structure and composition, lipid distribu-

VOL. 43, 1979

on Novem


mbr.asm

.org/D

ownloaded from


276 BALCH ET AL.

tion, intermediary metabolism, and nucleic acidcomposition substantiate the hierarchy of rela-tionships presented in Fig. 7 and 8.

Cell Wall Structure and CompositionAn examination of the chemical composition

of the cell wall, Gram reaction, cell morphology,and cell ultrastructure support the proposed tax-onomic treatment (Table 7). Electron micro-graphs of cell ultrastructure have been presentedin detail elsewhere (33, 52, 54, 56, 59, 81a, 126,128, 132, 133).Order I, Methanobacteriales. In the family

Methanobacteriaceae, species of the genusMethanobacterium possess fimbriae (27) andare straight to irregularly crooked gram-positiverods often forming filaments. The species of thegenus Methanobrevibacter are gram-positive,lancet-shaped cocci or short rods which formpairs, chains, or irregular clumps depending onthe growth conditions. M. arboriphilus and M.smithii each possess a single flagellum (27).A unifying characteristic of the family is a

gram-positive envelope (59, 125, 126, 128, 130,132). Members of the genus Methanobacteriumhave a sharply defined, smooth, gram-positive

TABLE 5. Key to the genera offamilyMethanobacteriaceae Barker 1956 (from the 8th

edition of Bergey's Manual ofDeterminativeBacteriology [20])

I.Rods or chain-forming, lancet-shaped coccoids.Genus I. Methanobacterium

II.Cocci other than chain-forming lancets.A. Large cocci in packets.

Genus II. MethanosarcinaB. Cocci occurring singly, in pairs, or in clumps.

Genus III. Methanococcus

cell wall 15 to 20 unm in width when viewed inultrathin sections of whole cells (59, 128). In thegenus Methanobrevibacter, ultrathin sections ofwhole cells of M. ruminantium reveal a distinc-tive triple layer in which a cell wall 30 to 40 nmin width consists of an inner electron-dense layeradjacent to a thick, more electron-transparentmiddle layer and a rough, irregular outer layer(59, 128). Ultrathin sections of isolated sacculifrom representative members of the family showonly one electron-dense layer 10 to 15 nm inwidth (56). This corresponds to the inner layerof M. ruminantium and the smooth layer ofMethanobacterium. The isolated sacculi are in-distinguishable morphologically from the cellwall sacculi of peptidoglycan-containing gram-positive bacteria (56).Kandler and Konig (56) have shown the chem-

ical composition of cell wall sacculi isolated fromrepresentative species of the Methanobacteri-aceae to consist of a polymer containing threeL-amino acids and N-acetylated sugars. In gen-eral, the sacculi contained L-lysine-L-alanine-L-glutamate-N-acetylglucosamine (or N-acetyl-galactosamine and/or N-acetyltalosaminuronicacid) in a molar ratio of 1:1.2:2:1 (56; H. Konigand 0. Kandler, Arch. Microbiol., in press). Nomuramic acid or D-amino acids were found. Fur-ther analysis of the cell wall structure of M.thermoautotrophicum (Konig and Kandler, inpress) defines a peptidoglycan polymer whichdiffers strikingly from that of typical bacteria inthe absence of muramic acid and the presenceof N-acetyltalosaminuronic acid and only L-amino acids (Fig. 9). Historically, the cell wallpolymer of typical bacteria was referred to asmurein, a muramic acid-containing peptidogly-can. This term has been replaced in the litera-ture by the more general term peptidoglycan.

ORDER FAMILY GENUS SPECIES STRAIN

M. formicicum MF

Mcthonobocterium -- bryontlii M oH G

eM thermoovlotrophicum AH-Alelhonobocleri/es - htethonoboter,oceoe lM rum;Oontium MI

DHIAlethonobrewbcter- A orboriphilus AZ

DCM smith,i PS

-Aethonococco/es Alethonococcoceoe M Aethonococcus M.vonnie/i SB

A.voltoe ~ Ps

Alethonomicrobium AM mobile- BP

Alethonomicrobioceoe Mebthoogenrum Aorisnioc i JRI

Me-ihonomicrobloles Alethonospirillum lM hungotei JFMs

Alehonosorcinoceoe Alethonosorcino AlM barkeri 227w

l 1022 0.28 0.34 0 36 046 0 51 0 55 065 084 10

Ronge of SAE volues for eoch foxonomic level,

FIG. 8. New taxonomic treatment for the methanogenic bacteria based on 16S rRNA comparative catalog-ing.

MICROBIOL. REV.

on Novem


mbr.asm

.org/D

ownloaded from



C C

eq~~~~~~~~~~~~ R~~~~~CCOD

~~~eaC6 eq _0

21 02 2_ 0 0°-0- eq -0 as as

ASAS S£ S{X e w 0 S Al~~~;-414 ; a eq203X O 8 N t 8 EB N X X O 08

0~~ ~ ~

Le.)ci 0

L LOL L00 0OXOX 00 0eg o 0 000X

- NCe_o oC_00 000 0

00~~~~~~~~~~C-

Csa CC- z E

0 ~ ~~~~~0

-0 0. . Q03Cst~~~~~~~~~~O 0C s-C 4 at

tsU00 0~~~~~~~A t 3

:cliC'S00 0CiS~~~~~~~~~~~~~~~~~~~~~C

VOL. 43, 1979

on Novem


mbr.asm

.org/D

ownloaded from


TABLE 7. Summary ofmajor features associated with cell wall structure and composition"Gram

Order Family Genus reac- Morphology Cell wall compositiontion

I__I_ Methanobacterium + Long rods PseudomureinMethanobrevibacter + Short rods

II- I Methanococcus - Regular to irregu- Protein subunits with tracelar coccus glucosamine

Methanomicrobium - Short, curved rod jI Methanogenium Highly irregular Protein subunits

coccusIII Methanospirillum - Long, curved rod Protein subunits with exter-

nal sheathIrregular coccus in Heteropolysaccharide

II Methanosarcina + packets

'Cell wall composition description from Kandler and Konig (56), Kandler and Hippe (55) and Jones et al.(52).

A B-G - M - G - -G - T - G -

Ala Glu

I YGlu Ala

Lys -DAla Lys-2YGlu

(DAla) Lys (Glu) Lys-Glu} Wtv te

(DAla) Glu (Ala) Ala

Ala Glut

-G - M - G -G - T - G-FIG. 9. Typical structure of murein (A) and pro-

posed structure ofpseudomurein (B), the peptidogly-can polymer found in cell walls ofM. thermoautotro-phicum (Konig and Kandler, Arch. Microbiol., inpress). G, N-Acetylglucosamine; M, N-acetylmuramicacid; T,N-acetyltalosaminuronic acid.

To distinguish between the two polymer types,the structure in methanogens is referred to as

pseudomurein (54).Individual differences in the cell wall compo-

sitions of the Methanobacteriaceae are mostlyrestricted to neutral sugar content (56). How-ever, M. ruminantium was shown to differ sub-stantially with respect to the polypeptide com-

ponent (56). L-Threonine completely replacedL-alanine, and the position of N-acetylglucosa-mine was taken by N-acetylgalactosamine. Inaddition, both M. ruminantium and M. arbori-philus cell walls were unique in their high phos-phate contents (56).Order II, Methanococcales. The two gram-

negative species in the genus Methanococcusare highly motile, irregular to regular coccoid

cells (0.3 to 5 am in diameter), depending ongrowth conditions. Electron micrographs of theouter surfaces by Jones et al. (52) showed aregular array of protein subunits. Ultrathin sec-tions of whole cells revealed a single layer ofwall material 18 unm thick (52). Only one freeze-fracture plane was observed in the plasma mem-brane-cell wall complex, presumably passingthrough the hydrophobic region of the plasmamembrane, as no fractures were observed tooccur in the wall itself (52). Treatment of freeze-dried cells with 2% sodium dodecyl sulfate at1000C for 30 min or disintegration of cells withglass beads followed by incubation with trypsinresulted in complete solubilization (O. Kandler,personal communication). Amino acid analysisof the total acid hydrolysate of whole cells didnot yield muramic acid or diaminopimelic acid(52). Trace amounts of glucosamine were foundin both M. vannielii and M. voltae (O. Kandler,unpublished data).Order m, Methanomicrobiale& Genera in

the Methanomicrobiales exhibit a more diverserange of morphological properties. In the gram-negative family Methanomicrobiaceae, themember of the genus Methanomicrobium is ahighly motile, short, slightly curved rod (0.7 by1.5 to 2.0,um) distinct from other methane bac-teria. No evidence for a rigid sacculus was foundby Kandler and Konig (56). Freeze-dried cellstreated with sodium dodecyl sulfate or disinte-grated with glass beads followed by incubationwith trypsin were completely solubilized (56).Muramic acid or amino sugars were not detectedin the total acid hydrolysate ofwhole cells (Kan-dler, personal communication).The genus Methanogenium includes poorly

motile, highly irregular coccoid cells (1 to 2,min diameter). Ultrathin sections of whole cells(81a) showed a cell wall 10 nm in width closely


on Novem


mbr.asm

.org/D

ownloaded from



opposed to the cytoplasmic membrane with aperiodic surface pattern similar to that of M.vannielii (52). In contrast to M. vannielii, saltis required for cell stability (81a). Treatment offreeze-dried cells with sodium dodecyl sulfate ordisintegration with glass beads followed by in-cubation with trypsin resulted in complete sol-ubilization of the cells (56). Muramic acid oramino sugars were not detected in total acidhydrolysates of whole cells (56).The genus Methanospirillum consists of mo-

tile, regularly curved, long rods that form contin-uous spiral filaments. Ultrathin sections ofwholecells show a well-defined, double-layered cellenvelope (33, 127, 128). It consists of an electron-dense outer sheath (composed of subunits ar-ranged in stacks perpendicular to filamentlength) adjacent to a more electron-dense innerlayer which envelopes individual cells that areseparated by a "spacer" element (127). Cell wallpreparations of M. hungatei yielded no roundedsacculi, only cylindrical sheaths of variablelength and of brittle structure (56). Total acidhydrolysates of the sheath revealed a com`plexspectrum of 18 amino acids, but not aminosugars, or sugars. The sheath was resistant tosodium dodecyl sulfate and trypsin treatments.It was concluded that the "cell wall" preparationrepresented only the outer layer (56), which isnot involved in septum formation (127, 128). Theinner layer was destroyed by the isolation pro-cedure used, a result similar to those obtainedfor other genera in the Methanomicrobiaeeae.No muramic acid, diaminopimelic acid, or aminosugars were detected in the total acid hydroly-sate of whole cells (56).Members of the family Methanosarcinaceae

are irregular, gram-positive cocci that fornpackets of varying size. These clumps are largeenough to present a grainy appearance to thenaked eye. Division planes of the cells in thepacket are not necessarily perpendicular. Ultra-thin sections showed a very thick (500 nm),amorphous outer layer which often appearedlaminated (128). Zhilina (133) reported a triple-layered appearance of the cell wall in a gas-vacuolated strain. Results of chemical analysesof isolated cell walls indicated that they consistof an acid heteropolysaccharide that containsgalactosamine, neutral sugars, and uronic acids(55). No muramic acid, glucosamine, glutamicacid, or other amino acids typical of peptidogly-can were found in the two strains examined (55).A summary of the major features associated

with cell wall structure and composition andtheir relationship to the proposed new taxo-nomic treatment is presented in Table 7. Thechemical differences from the muramic acid-con-

taining peptidoglycan of typical bacteria are re-flected in the lack of sensitivity of methanogensto antibiotics, such as penicillin, cycloserine, andvancomycin (52, 55, 56; Balch, unpublisheddata), as well as to the inability of lysozyme toeffect cell wall lysis (52, 56).Methanogens are represented by a diverse

range of both gram-positive and gram-negativecell wall types. A comparison of cell wall com-position and structure to reaction of the Gramstain reveals a strict correlation between thepresence of a thick, rigid sacculus and a gram-positive reaction, a negative Gram reaction beingcorrelated with the absence of a rigid sacculus(52, 54-56). These results correlate well with cellwall structure and Gram reaction in peptidogly-can-containing bacteria, supporting the assump-tion that the Gram reaction does not depend onthe chemical nature of the cell wall but ratheron its physical properties.

Lipid CompositionA survey of the lipid composition of the meth-

anogenic bacteria shows that ether-linked polyi-soprenoid (branched) chain lipids are the pre-dominant components (Table 8) (64, 101, lOla,102). Of the total lipids of whole cells, 20 to 30%are found to be neutral lipids, and 70 to 80% arepolar lipids. Tornabene and Langworthy (101)and Makula and Singer (64) have shown thatthe polar lipids are principally isopranyl ethersof glycerol and are found to consist primarily ofC20 phytanyl and C40 biphytanyl glycerol ethers(structures I and II, Fig. 10). The distributionbetween the major lipid ethers in nine repre-sentative methanogens is shown in Table 8. Boththe C20 phytanyl and the C40 biphytanyl glycerolethers were found in the Methanobactericeaeand in the one genus examined of the Methan-omicrobeaceae (Methanospirillum). Trace or nodetectable levels of the C40 biphytanyl glycerolethers were found in the Methanococcaceae orthe Methanosarcinaceae (Table 8) (101).The neutral lipids comprised a wide range (C14

to C30) of isoprenoid hydrocarbons (lOla). Theprinciple compounds were C30, C25, and C20acyclic isoprenoid hydrocarbons with a contin-uous range of hydroisoprenoid derivatives. NoC35 or C4o compounds were detected. All themethanogens with the exception of the Meth-anosarcinaceae contained the C30 bifarnesyl orsqualene components (Table 8; structure III, Fig.10). The C25 and C20 isoprenoid hydrocarboncomponents were found in five and three of thenine methanogens examined, respectively(lOla).Thus, methanogens lack saponifiable lipids

VOL. 43, 1979

on Novem


mbr.asm

.org/D

ownloaded from


TABLE 8. Distribution of the major neutral andpolar lipids in the methanogenic bacteria"

Major lipidsOr- Fam- GeuSpcsder fly Genus Species Neutral (isoprenoid hydrocar- Polar (isopranyl

bons) glycerol ethers)- Methanobacterium M. formicicum NDb C20+ C40 ethers

M. bryantii C30H5,o C0oH52 C20 + C40 ethersI- I- M. thermoautotrophicum C30H50, C.2oH52, C0HrA,, C25Hro C20 + C40 ethers

LMethanobrevibacter M. ruminantium C00Hwq, CGmH52, CO0HrA, C30Hr,6 C20 + C40 ethersM. arboriphilus (AZ) C30H50 C20 + C40 ethersM. smithii C20Hw,t C30H52 C2o + C4o ethers

II I-Methanococcus M. vannielii C30H50, C25H42 C20 ethersM. voltae C0H52, C.oHrA,, C:joHr,6 C20 ethers

-Methanomicrobium ND NDI- Methanogenium ND ND

III-- Methanospirillum M. hungatei C30Hr,og C0H52 C20+ C40 ethersLII UMethanosarcina M. barkeri C25H46, C25H48, C25H50, C25H52 C20ethers

"Data presented are from Makula and Singer (64), Tornabene and Langworthy (101), and Tornabene et al.(lOla, 102).'ND, Not determined.

0

II

E ~~J-OHn°OH~~~~~O

O 0

-OHAm

FIG. 10. Structures of C20 diphytanyl glyceroldiether (1 and C40 dibiphytanyl diglycerol tetraether(I), principal components of the polar lipid fractionof methanogenic bacteria. Structure III is CoHw0(squalene), a common isoprenoid hydrocarbon foundin the neutral lipid fraction ofmethanogens.

that are characteristic of typical bacteria. In-stead, the major lipid components are ether-linked polyisoprenoid (branched) chain lipids.The C30 (squalene) isoprenoid hydrocarbon com-ponents dominate the neutral lipid fraction,whereas C20 phytanyl and C4o biphytanyl glyc-erol ethers dominate the polar fraction. Thelatter are ether-linked analogs ofthe ester-linkedglycolipids and phospholipids found in typicalbacteria. These findings provide the first evi-dence for the microbial synthesis of multi-branched isoprenoid hydrocarbons in the carbonrange comparable to that found in sediments

and petroleum and may have major implicationsin the interpretation of biogeochemical evolu-tion (101).

Deoxyribonucleic Acid Base Composition(Moles Percent Guanine plus Cytosine)Table 9 provides comparative data on the

DNA base composition of 21 methanogenstrains. The DNA mol% G + C content variesfrom 27.5 to 61%. However, each of the proposedindividual groups shows a restricted range. Thespecies of the genus Methanobrevibacter definethe range of 27.5 to 32 mol% G + C, whereas thespecies of the genus Methanobacterium rangefrom 32.7 to 40.7 mol% G + C for the mesophilesand 49.7 mol% G + C for the thermophile. Thelatter value may reflect the unique habitat ofM.thermoautotrophicum. The two species in thegenus Methanococcus have mol% G + C con-tents of 30.7 and 31.1. This is in contrast to thetwo marine cocci in the genus Methanogenium,where the mol% G + C contents are 51.6 and61.2. The remaining genera in the Methanomi-crobiales exhibit a range ofmol% G + C contents(Table 9). The patterns observed reinforce theunique status of each genus.

Substrates for Growth and MethaneProduction

Table 9 summarizes the substrates for growthand methane production by methanogenic bac-teria. All methanogens thus far examined oxidizehydrogen and reduce carbon dioxide to methane.Some species metabolize formate. Poor growthon carbon monoxide has been reported for somestrains (25, 27). The members of the Methano-sarcinaceae will additionally metabolize meth-


on Novem


mbr.asm

.org/D

ownloaded from



TABLE 9. Summary ofDNA base composition (mol% G + C) and substrates that serve as sole electrondonors for methanogenesis and growth

OrderFamilySsmol%G + C Reference Substrates for growth and CH, pro-Order Family Species mol% G + C" Reference duction'

Methanobacterium formici- 40.7 (42.0) 126 H2, formatecum

Methanobacterium bryantii 32.7 (38.0) 131 H2Methanobacterium bryyantii 33.2c H2

strain M.o.H.G.Methanobacterium thermo- 49.7' (52.0) 131 H2

autotrophicumI__I- Methanobrevibacter rumi- 30.6' H2, formate

nantiumMethanobrevibacter arbori- 27.5 130 H2philus

Methanobrevibacter arbori- 31.6' H2philus strain AZ

Methanobrevibacter arbori- 27.7" H2philus strain DC

Methanobrevibacter smithii 31.0" (32.0) 126 H2, formater Methanococcus vannielii 31.1" H2, formateL Methanococcus voltae 30.7c H2, formate- Methanomicrobium mobile 48.8e H2, formate

I-I Methanogenium cariaci 51.6 81a H2, formateMethanogenium marisnigri 61.2 81a H2, formate

III- Methanospirillum hungatei 45.0 33 H2, formateMethanospirillum hungatei 46.5 78 H2, formate

strain GP1-II-Methanosarcina barkeri 38.8c (43.5) 108 H2, CH30H, CH.3NH2, acetate

Methanosarcina barkeri 38.8" H2, CH30H, CH3NH2, acetatestrain 227

Methanosarcina barkeri 40.5d H2, CH30H, CH3NH2, acetatestrain W

Methanosarcina barkeri 43.5 109 H2, CH30H, CH3NH2, acetatestrain UBS

Methanosarcina barkeri 51.0 135 H2, CH30H, CH3NH2, acetatestrain Z

Values not done in this laboratory are presented in parentheses.b C02 is reduced to CH4 when H2 is the substrate.'Unpublished data, methods used to determine mol% G + C were as described in Deoxyribonucleic Acid

Base Composition Determination (Moles Percent Guanine plus Cytosine).d Unpublished data provided by R. A. Mah, using bouyant density method of Schildkraut et al. (83).

anol, methylamine (dimethylamine, trimethyl-amine, and ethyldimethylamine), and acetate(46, 62, 86, 89, 107-109) as the sole electrondonor for growth and methane production.These results support the separation of theMethanosarcina sp. from other members of theMethanomicrobiales into at least a separatefamily.

Metabolic PathwaysCoenzyme M (2-mercaptoethanesulfonic

acid). The oxidation of hydrogen and the reduc-tion of C02 to produce methane by methanogensinvolves a highly unique biochemistry. Taylorand Wolfe (97) have identified a new coenzyme,coenzyme M (2-mercaptoethanesulfonic acid)involved in methyl transfer reactions in methanebacteria. It is the smallest of all known coen-

zymes and exceptional in its high sulfur contentand acidity. The cofactor is required by methyl-coenzyme M reductase, an enzyme present in allmethanogens (R. P. Gunsalus, Ph.D. thesis, Uni-versity of Mllinois, Urbana, 111., 1977) and activein the terminal steps of CO2 reduction to CH4(42-44, 66, 97).In a survey to determine the biological distri-

bution of coenzyme M among nonmethanogens(8), a broad range of procaryotic organisms andeucaryotic tissues were examined; coenzyme Mwas not detected. In contrast, coenzyme M wasfound in high levels in all methanogens availablein pure culture. An average intracellular concen-tration of0.2 to 2mM coenzymeM was observed(8). It was concluded that the coenzyme doesnot play a general role in other methyl transferreactions; its importance lies in its central role

VOL. 43, 1979

on Novem


mbr.asm

.org/D

ownloaded from


282 BALCH ET AL.

in methanogens, organisms which are pivotal fornormal biodegradation in the cecum, rumen,sludge digesters, and aquatic sediments.Coenzyme F420. The structure of F420, a low-

potential electron carrier, was recently eluci-dated by Eirich et al. (32). Evidence suggeststhat F420 is the flavin mononucleotide analog 7,8-didemethyl-8-hydroxy-5-deazariboflavin-5'-phosphate, which has an N-(N-L-lactyl-y-L-glu-tamyl)-L-glutamic acid side chain attached in aphosphodiester linkage (Fig. 11). F420 has beendetected in all methanogens but has not beendetected elsewhere (L. D. Eirich, Ph.D. thesis,University of Illinois, Urbana, Ill., 1978). F420participates as an electron carrier in the nicotin-amide adenine dinucleotide phosphate-linkedhydrogenase and formate dehydrogenase sys-tems in methanogens (34, 103, 104, 108). Re-cently, Zeikus et al. (129) have demonstratedthat F420 is reduced by coenzyme A-dependentpyruvate and a-ketoglutarate dehydrogenases inextracts ofM. thermoautotrophicum, a reactionnormally mediated by ferredoxin in most bac-teria (70). In addition to F420, chromophoric fac-tors F342 and F430 have been described in M.thermoautotrophicum (45). Their role is pres-ently unknown. However, both F420 and F342 areuseful for tentative identification of methano-gens by fluorescent microscopy (28, 68). Cyto-chromes, menaquinones, or ubiquinones havenot been detected in methanogenic bacteria todate (35, 100, 129).Intermediary metabolism. The pathway of

C02 fixation into cell carbon in methanogensremains to be elucidated. Taylor et al. (98),Weimer and Zeikus (108), and Zeikus et al. (129)have been unable to demonstrate key enzymesfor the serine pathway (hydroxypyruvate reduc-tase), the hexulose pathway (hexulosephosphatesynthetase), or the reductive pentose phosphatepathway (ribulose-1,5-bisphosphate carboxyl-ase) in cell-free extracts of M. thermoautotro-phicum or M. barkeri. Analysis of short-term14CO2 fixation products yielded similar results(26). Taylor et al. (98) found 14CO2 labelingpatterns inconsistent with either the reductivetricarboxylic acid pathway or the total synthesisof acetate from C02, a pathway typical of certainclostridia. Zeikus et al. (129) demonstrated thepresence ofkey oxidoreductases ofthe tricarbox-ylic acid cycle in M. thermoautotrophicum withthe exception of isocitrate dehydrogenase. M.barkeri was found to lack fumarate reductaseand a-ketoglutarate dehydrogenase (109). Theauthors concluded that either additional keyenzymes remain to be detected in each organismor that a new pathway of autotrophic CO2 fixa-tion is operative.

Acetate has been shown to be assimilated intocell carbon in a wide range of methanogens (22,39, 62, 86, 89, 98, 107-109), providing up to 60%of total cell carbon in cells grown in the presenceof acetate. Fuchs et al. (39) recently examinedacetate assimilation into alanine, aspartate, andglutamate in M. thermoautotrophicum. Cell car-bon synthesis from labeled acetate was consist-ent with a pathway of pyruvate (alanine) syn-thesis from 1 C2 compound and 1 C02, oxaloac-etate (aspartate) synthesis from 1 C2 compoundand 2 C02, and a-ketoglutarate (glutamate) syn-thesis from 1 C2 compound and 3 C02 via oxa-loacetate, malate, fumarate, and succinate (Fig.12). Weimer and Zeikus (109) found acetate as-similation into cell carbon in M. barkeri to differfrom that in M. thermoautotrophicum; synthesisof a-ketoglutarate occurred via oxaloacetate, cit-rate, and isocitrate from 2 C2 compounds and 1C02 (Fig. 12). The authors suggested that thedifferences observed in intermediary metabo-lism may reflect the phylogenetic divergencebetween the two organisms (109).Although acetate is an important precursor to

cell carbon, the net synthesis of acetate fromC02 in autotrophically grown cells remains anenigma. Fuchs and Stupperich (38) recentlydemonstrated the absence of a complete reduc-tive carboxylic acid cycle in M. thermoauto-trophicum grown on H2 and C02 in the presenceof ["4C]succinate. Only glutamate was found tobe significantly labeled, excluding the possibilitythat oxaloacetate and pyruvate can be synthe-sized from succinate via isocitrate and citrate orother intermediates of the tricarboxylic acid cy-cle. The results suggested that a cyclic mecha-nism for regeneration of C02 acceptor molecules(e.g., acetyl coenzyme A) from citric acid cycleintermediates is not operative.

Ribosomal and Transfer RibonucleicAcids of Typical Bacteria and

MethanogensrRNA sequences of methanogens are not

closely related to those of typical bacteria. Themethanogen 16S (23S) rRNA sequences bear nomore relationship to those of typical bacteriathan they do to the eucaryotic 18S (23S) rRNA's(30; D. Stahl, Ph.D. thesis, University of Illinois,Urbana, Ill., 1978). In the secondary structure oftheir 5S rRNA's (37) methanogens again seemno closer to the typical bacteria than to theeucaryotic examples (K. Luehrsen and C. R.Woese, unpublished data). The posttranscrip-tional modification patterns in 16S rRNA (seeTables 3 and 13) and 23S rRNA (Stahl, Ph.D.thesis) of methanogens differ strikingly fromthose of typical bacteria in spite of the fact that

MICROBIOL. REV.

on Novem


mbr.asm

.org/D

ownloaded from



O CH3O coo 0 coOeCH2-CH-CH-CH-CH2-O-P-0-CH -C-NH-CH-CH2-CH2-C-NH-CH. .. .

OH OH OH °e

HO<rOnY- r )+O?2t N > H~~N,

0

FIG. 11. Stri

C02

ACETATE

ALANINE .-----PYRUVATE

WCO2ASPARTATE2 OXALOACETATE

MALATE CITRATE

FUMARATE ISOCITRATE

\ /

SUCCINATE----,-a-KETOGLUTARATE---GLUTAMATE

CO2

FIG. 12. Proposed pathway of CO2 fixation andacetate assimilation into cell carbon in M. thermoau-totrophicum (dotted lines) and M. barkeri (solidlines). Data presented are from Fuchs et al. (39),Fuchs and Stupperich (38), and Weimer and Zeikus(109).

the modifications tend to occur in the samelocales in the two cases.Methanogen transfer RNAs are unique in that

they do not contain the universal common armsequence GT*CG; rather, they contain an ana-log, either G 'CG or GO*CG (the superscriptdot denotes a modification of the base, but U #

T). Best (14) reports no evidence in M. vannieliitransfer RNAs for either of the modified basesribothymidine and 7-methylguanosine, basestypical of Escherichia coli. R. Gupta and C. R.Woese (unpublished data) have confirmed theseresults and extended them to all the groups ofmethanogens. However, in disagreement withBest (14), they find no evidence for the presenceof dihydrouridine except in the case of M. bar-keri.

Genome Sizes of Typical Bacteria andMethanobacterium thermoautotrophicumMitchell et al. (69) have examined the DNA

organization of M. thermoautotrophicum. TheDNA of the methanogen was found to be as

homogeneous as that ofE. coli based on thernalmelting profiles. In addition, results of buoyant

CH2

CH2

-ucture of F420.

density centrifugation in CsCl indicated thatsatellites were absent, and results of renatur-ation studies indicated the DNA to be of oneclass. On the basis of three different DNA re-naturation kinetic experiments, the complexity(genome size) of the methanogen DNA wasfound to be two to three times smaller than thatof E. coli. Assuming the E. coli genome to be2.7 x 109 daltons, this yields an average value of1.1 x 109 daltons for the genome size of M.thermoautotrophicum. These results are impres-sive, considering that M. thermoautotrophicumcan be grown autotrophically with hydrogen asthe energy source and C02 as the sole carbonsource.

FORMAL DETAILED PROPOSAL:REVISION OF CLASSIFICATION AND

NOMENCLATUREIn the most recent taxonomic treatment of the

group (11, 20), the methanogens were placed ina single family, the Methanobacteriaceae. Thetreatment is no longer tenable in light of recentinformation. Below, we present revised defini-tions of the family Methanobacteriaceae, thegenus Methanobacterium, the genus Methano-coccus, and the genus Methanosarcina as wellas definitions of the proposed new orders, fami-lies, genera, and species outlined above. Wher-ever possible, current nomenclatural usage isretained to maintain the historical developmentof the group.

Order I, Methanobacteriales

Methanobacteriaks Balch and Wolfe (ord.nov.) Me.tha.no.bac.ter.i.al'es. M.L. fem. pl. n.Methanobacteriaceae type family of order; -alesending to denote order; M. L. fem. pl. n. Meth-anobacteriales the Methanobacteriaceae order.

Cells are short, lancet-shaped cocci to long,filamentous rods. Cells are typically gram posi-tive, although some strains are reported to begram variable. Cell wall structure appears typi-cally gram positive, when viewed in ultrathinsection, but does not contain muramic acid.Pseudomurein is the predominant peptidogly-

VOL. 43, 1979

on Novem


mbr.asm

.org/D

ownloaded from


284 BALCH ET AL.

can polymer. Ether-linked polyisoprenoid(branched) chain lipids are the predominantlipid components. The Methanobacteriales arevery strict anaerobes which obtain energy forgrowth by the oxidation of H2 and sometimesformate or CO or both with the reduction of C02to methane. Cells contain coenzyme M and co-enzyme F420. They form a highly specializedphysiological group which does not utilize car-bohydrates, proteinaceous material, or organiccompounds other than formate or CO as energysources. They are widely distributed in nature,being found in anaerobic habitats, such as sedi-ments of natural waters, soil, anaerobic sewagedigesters, the gastrointestinal tracts of animals,and ecosystems where geothermally producedH2 accumulates.Family I, Methanobacteriaceae Metha-

nobacteriaceae Barker 1956 (emend.).Me.tha.no.bac.ter.i.a'ce.ae. M. L. neut. n. Meth-anobacterium type genus of family; -aceae end-ing to denote family; M. L. fem. pl. n. Metha-nobacteriaceae the Methanobacterium family.Only one family, Methanobacteriaceae, is ac-

cepted in the order Methanobacteriales; thedescription of the family is the same as that ofthe order. The family Methanobacteriaceaecontains two genera.Genus I, Methanobacteriurn Methanobac-

terium Kluyver and van Niel 1936 (emend.).Me.tha.no.bac.ter'i.um. M. L. n. methanummethane; Gr. n. bakterion a small rod; M. L.neut. n. Methanobacterium the methane (-pro-ducing) rodlet.

Cells are straight to irregularly crooked longrods (0.5 to 1 ,um in width) often forming fila-ments. Cells are nonmotile, mesophiles to ther-mophiles. Cells possess fimbriae. DNA basecomposition range can be expected to include 33to 41 mol% G + C (determined for three strains)for the mesophiles. The one thermophilic strainis 50 mol% G + C. The type species, M. soehn-genii Barker 1936, was never obtained in pureculture (10) and is not included on the list ofapproved names (1). The neotype species of thegenus Methanobacterium is therefore desig-nated to be M. formicicum Schnellen 1947(Schnellen, Ph.D. thesis). The species M. for-micicum Schnelien 1947 and M. thermoauto-trophicum Zeikus and Wolfe 1972 are describedelsewhere (20, 22, 59, 74, 87, 99, 103, 131, 132).M. bryantii sp. nov., formerly Methanobacte-rium strain M.o.H. (23), is described below.Methanobacterium bryantiL Methanobac-

terium bryantii Balch and Wolfe (sp. nov.).bry.an.ti.i'. M. L. gen. n. bryantii of Bryant;named for M. P. Bryant for his pioneering workin the separation and characterization of this

MICROBIOL. REV.

organism from the M. omelianskii syntrophicculture.The original description by Bryant et al. (23)

is supplemented with material of Bryant et al.(22) and Langenberg et al. (59). Cells are slenderand cylindrical with blunt, rounded ends, oftenforming chains and filaments with unevenlycrooked cells. Cells are gram positive to gramvariable. Cell width is from 0.5 to 1.0 tum withchains up to 10 to 15,m in length. Cells arenonmotile and possess fimbriae. Surface colo-nies, which can reach 1 to 5 mm in diameter, areflat with diffuse to filamentous edges and havea characteristic gray to light gray-green appear-ance. Deep colonies are rounded and filamen-tous. Strains tend to clump in liquid culture.Cells grow optimally at 37 to 39°C at pH 6.9 to7.2. Cells utilize H2 as the sole energy source.Formate is not a substrate for growth and meth-anogenesis as found in M. formicicum. NH4' isessential as the main source of cell nitrogen.Acetate, cysteine, and B-vitamins are highlystimulatory for growth. The DNA base compo-sition of the type strain (M.o.H.) is 32.7 mol% G+ C.The type strains for the species in the genus

Methanobacterium are designated M. ther-moautotrophicum strain AH (DSM 1053) (131)and M. bryantii strain M.o.H. (DSM 863) (23).As the original strain of M. formicicum is notextant, M. formicicum strain MF (isolated fromsewage sludge by M. P. Bryant) is designatedthe neotype strain (DSM 1535).Genus H, Methanobrevibacter. Methano-

brevibacter Balch and Wolfe (gen. nov.).Me.tha.no.bre.vi.bac'ter. M. L. n. methanummethane; L. adj. brevis short; M. L. n. bactermasc. equivalent of Gr. neut. n. bakterion rod,staff; M. L. masc. n. Methanobrevibacter theshort methane (-producing) rod.

Cells are lancet-shaped cocci to short rodswhich often form pairs or chains 0.5 to 1.0 tm inwidth. Cells are nonmotile to poorly motile withan optimal growth range of 37 to 390C. DNAbase composition range can be expected to in-clude 27.5 to 32 mol% G + C (determined forfive strains). M. ruminantium (Smith and Hun-gate 1958) Balch and Wolfe (comb. nov.) is des-ignated the type species for the genus Methan-obrevibacter, as it was the first species charac-terized. The rumen strain Ml has been nutri-tionally characterized by Bryant et al. (18, 22).M. arboriphilus (Zeikus and Henning 1975)Balch and Wolfe (comb. nov.) is described else-where (112, 127, 132). M. smithii sp. nov., for-merly M. ruminantium strain PS Smith 1961 (P.H. Smith, Abstr. Bacteriol., A40, p. 60, 1961), isdescribed below.

on Novem


mbr.asm

.org/D

ownloaded from


VOL. 43, 1979

Methanobrevibacter smithiL Methanobrev-ibacter smithii (Smith 1961) Balch and Wolfe(sp. nov.). smith.i.i'. M. L. gen n. smithii ofSmith; named for P. H. Smith for his pioneeringwork on the ecology and physiology of the meth-anogenic bacteria.The original description by Smith (P. H.

Smith, Abstr. Bacteriol., A40, p. 60, 1961) issupplemented with material of Bryant et al. (22)on a sludge strain (PS) previously considered tobe M. ruminantium. Cells are short, lancet-shaped to oval cocci 0.5 to 1.0 ym in width and1.0 to 1.5 ,m in length. Cells frequently occur inpairs and long chains resembling streptococci.Cells are nonmotile and strongly gram positive.Surface colonies are translucent, convex, andcircular with entire margins and usually yellowwhite in appearance. Cells may use either H2 orformate as a substrate for energy and growth.Cells grow optimally at 37 to 390C. DNA basecomposition of the type strain is 30.6 mol% G +C. M. smithii requires ammonia as a nitrogensource and acetate as a major carbon source forgrowth. The species is morphologically identicalto M. ruminantium with the exception that M.smithii has a single polar flagellum but is non-motile when viewed in wet mount. The majordifferences are physiological and nutritional. M.smithii can be cultured on a simple chemicallydefined medium in the absence of a-methylbu-tyrate, amino acids, and coenzyme M, which areessential for the growth of M. ruminantium (7,22, 96). M. smithii grows optimally from pH 6.9to 7.4, whereasM. ruminantium grows optimallyfrom pH 6.3 to 6.8.The type strains for species in the genus Meth-

anobrevibacter are designated M. arboriphilusstrain DH1 (DSM 1125) (130) and M. smithiistrain PS (DSM 861) (20, 22, 87; Smith, Abstr.Bacteriol., A40, p. 60, 1961). M. ruminantiumstrain Ml (DSM 1093) (18, 20, 22, 59, 88, 128) isdesignated the neotype strain, because the orig-inal strain of Smith and Hungate (88) was notmaintained.

Order II, MethanocoalesMethanococcales Balch and Wolfe (ord. nov.).

Me.tha.no.coc.cal'es. M. L. fem. pl. n. Metha-nococcaceae type family oforder; -ales to denoteorder, M. L. fem. pl. n. Methanococcales theMethanococcaceae order.

Cells are regular to irregular cocci, 1.0 to 5.0um in diameter, and gram negative with a singlelayer of protein subunits external to the plasmamembrane. Muramic acid is not detected. Ether-linked polyisoprenoid (branched) chain lipidsare the predominant lipid components. TheMethanococcales are very strict anaerobes


which obtain energy for growth by the oxidationof H2 or formate with reduction of C02 to CH4.CelLs contain coenzyme M and coenzyme F420.They form a highly specialized physiologicalgroup which does not utilize carbohydrates, pro-teinaceous material, or organic compounds otherthan formate as energy sources. Widely distrib-uted in nature, being found in anaerobic habi-tats, such as sediments of natural waters.Family I, Meth tlcoccaceae Methanococ-

caceae Balch and Wolfe (fam. nov.).Me.tha.no.coc.ca'ce.ae. M. L. neut. n. Methano-coccus (Kluyver and van Niel) Barker 1936 typegenus of the family; -aceae ending to denote afamily; M. L. fem. pl. n. Methanococcaceae theMethanococcus family.Only the one family, Methanococcaceae, is

accepted in the order Methanococcales; the de-scription of the family is the same as that of theorder. The family Methanococcaceae containsone genus.Genus I, Methanococu& Methanococcus

(Kluyver and van Niel) Barker 1936 (emend.).Me.tha.no.coc'cus. M. L. n. methanum methane;M. L. n. coccus a spherical cell; M. L. masc. n.Methanococcus the methane coccus.

Cells are fragile, regular to irregular cocci (0.5to 5,um in diameter) that occur singly or in pairs.Cells are highly motile, with an optimal growthrange from 32 to 400C. DNA base compositionrange can be expected to include 30.7 to 31.1mol% G + C (determined for two strains). Thetype species, Methanococcus mazei Barker1936, was never obtained in pure culture (10)and is not included in the list of approved names(1). The neotype species of the genus Methano-coccus is therefore designated to be M. vannieliiStadtman and Barker 1951 and is described indetail elsewhere (20, 52, 53, 90). M. voltae sp.nov., formerly Methanococcus strain PS (Ward,M. S. thesis) is described below.Methanococcus voltae- Methanococcus vol-

tae Balch and Wolfe (sp. nov.). vol'tae. M. L.gen. n. voltae of Volta; named for the Italianphysicist Alessandro Volta for discovery of thecombustible nature of gas from anaerobic sedi-ments.The original description of J. M. Ward (M.S.

thesis) is supplemented with unpublished dataof W. E. Balch. Cells are highly motile regularto irregular cocci (0.5 to 3.0 /Am in diameter).Morphology is dependent on the physiologicalstate of the cell and on the ionic strength of themedium. Surface colonies can reach 1 to 3 mmin diameter and are clear and convex withsmooth edges. Subsurface colonies are creamcolored, with rough-textured interiors and irreg-ular edges. Cells grow optimally from 32 to 400C,

on Novem


mbr.asm

.org/D

ownloaded from


286 BALCH ET AL.

pH 6.7 to 7.4. H2 and formate serve as the soleenergy sources for growth and methane produc-tion. DNA base composition of the type strainis 30.7 mol% G + C. Cells are sensitive to lowosmotic environments; cells require 1.2 to 4.8%NaCl for optimal growth. A summary of themajor phenotypic properties which distinguishM. voltae and M. vannielii is presented in Table10.The type strains for species in the genus Meth-

anococcus are designated M. vannielii strain SB(DSM 1224) (52, 53, 90) and M. voltae strain PS(DSM 1537) (Ward, M.S. thesis).

Order m, MethanomicrobialesMethanomicrobiales Balch and Wolfe (ord.

nov.). Me.th.no.mi.cro.bi.al'es. M. L. fem. pl. n.Methanomicrobiaceae type family of order;-ales ending to denote order; M. L. fem. pl. n.Methanomicrobiales the Methanomicrobiaceaeorder.

Cells are cocci to rods, gram negative or grampositive, nonmotile to motile. In ultrastructure,a wide range of cell wall enclosures are observed,none of which contain muramic acid. Ether-linked polyisoprenoid (branched) chain lipidsare the predominent lipid components. Thesevery strict anaerobes obtain energy for growthby oxidation of H2 or formate with the reductionof C02 to CH4 or via fermentation of such com-pounds as methanol, methylamine (dimethyl-amine, trimethylamine, and ethyldimethyla-mine), and acetate with the formation of meth-ane and C02. Cells contain coenzyme M andF420. The Methanomicrobiales form a highlyspecialized physiological group which does notutilize carbohydrate, proteinaceous materials, ororganic compounds other than those listedabove as energy sources. They are widely dis-tributed in nature, being found in anaerobichabitats, such as sediments of natural waters,soil, anaerobic sewage digesters, and the gas-trointestinal tracts of animals.Family I, Methanomicrobweceae Metha-

nomicrobiaceae Balch and Wolfe (fam. nov.).Me.tha.no.mi.cro.bi.a'ce.ae. M. L. neut. n. Meth-anomicrobium type genus of the family; -aceaeending to denote a family; M. L. fem. pl. n.Methanomicrobiaceae the Methanomicrobiumfamily.

Cells are gram negative, cocci to straight,slightly curved rods. Cells oxidize H2 or formateas the sole energy source for growth and meth-ane production. Three genera are included inthe family Methanomicrobiaceae.Genus I, MethanomicrobiunL Methanomi-

crobium Balch and Wolfe (gen. nov.).Me.tha.no.mi.cro.bi'um. M. L. n. methanum

MICROBIOL. REV.

methane; Gr. adj. micros small; Gr. adj. bios life;M. L. neut. n. Methanomicrobium the methane(-producing) small life.

Cells are short, straight to slightly curved rods(0.7 by 1.5 to 2.0 ,um) with rounded ends. Cellsare highly motile, with a growth optimum of 38to 400C. Only H2 serves as the substrate forgrowth and methane production. DNA basecomposition of the lone species is 48.8 mol% G+ C. The type species is M. mobile (Paynter andHungate 1968) Balch and Wolfe (comb. nov.).Detailed descriptions of this genus are presentedelsewhere (20, 79). The type strain is strain BP(DSM 1539) (79).Genus II, Methanogeniuym Methano-

genium Romesser and Wolfe 1979.Me.tha.no.gen.i'um. M. L. n. methanum meth-ane; Gr. suff. v. -genes producing; M. L. neut. n.Methanogenium the methane producer.

Cells are highly irregular cocci (0.5 to 2.5,umin diameter) that occur singly. Cells are poorlymotile. The optimal growth temperature is 25 to300C. Both H2 and formate serve as substratesfor growth and methane production. Marineforms are sensitive to low osmotic environments;cells require 1.5 to 3.0% NaCl for growth. DNAbase composition range can be expected to in-clude 51.2 to 61 mol% G + C (determined fortwo strains). Only two species, M. cariaci Ro-messer and Wolfe 1979 and M. marisnigri Ro-messer and Wolfe 1979, are available, and theyare described in detail elsewhere (81a). The typestrains for the species are designated M. cariacistrain JR1 (DSM 1497) and M. marisnigri strainJR1 (DSM 1498) (81a).Genus m, MethanospirillunL Methano-

spirillum Ferry, Smith, and Wolfe 1974.Me.tha.no.spi.ril'lum. M. L. n. methanum meth-ane; Gr. n. speira a spiral; M. L. neut. n. Meth-anospirillum the methane(-producing) spiral.

Cells are slowly motile, regularly curved rods(0.5 by 7 ,um) often forming long spiral filaments.Cells have a distinctive double-wall appearancewhen viewed in thin section. Cells lack a rigidcell wall sacculus, the outer layer being a pro-teinaceous sheath. Both H2 and formate aresubstrates for growth and methane production.DNA base composition range can be expected toinclude 45 to 46.5 mol% G + C (determined fortwo strains). One species is included in the genusMethanospirillum, M. hungatei Ferry, Smith,and Wolfe 1974 (M. hungatii [sic] Ferry, Smith,and Wolfe 1974). A complete description of thisspecies is presented elsewhere (33, 34, 78, 127).M. hungatei Ferry, Smith, and Wolfe 1974 (33)is designated the type species of the genus Meth-anospirillum. The type strain is strain JF1(DSM 864) (33).

on Novem


mbr.asm

.org/D

ownloaded from



20t-. 1-

CO C

'4

_ ,14~~~~~~~~~3-Og + + a

j XCo @ + + 3

'4

. TZ + + 4

o coco + +

O 0 4

+ + ~~~~~~~~4.4~~~~~~~~~~~~4*

o~~~~~~~~c

o O o.

.0.

0~~~~~~~~~~~~- 0

0. . .2 4.

0 ~ ~~

.0 *0~~~~0 C'4c

+

O. IT lC's~~~~~~~~'

ITC) t

CO 0

00~~

._C)Cri

.b

t

0 t$

Q

U

U

Q,

C

U

.0

~C

3o

O.0

aa6a.a5

VOL. 43, 1979

C10

us

0

CC.

043C

L.0

4.

400

4..

0

0C.)

on Novem


mbr.asm

.org/D

ownloaded from


288 BALCH ET AL.

Family II, Methanosarcinaceae Methano-sarcinaceae Balch and Wolfe (fam. nov.).Me.tha.no.sar.cin.a'ce.ae. M. L. neut. n. Meth-anosarcina type genus of the family; -aceaeending to denote family; M. L. fem pl. n. Meth-anosarcinaceae the Methanosarcina family.The Methanosarcinaceae are large, spherical

to pleomorphic gram-positive cells (1.5 to 2.5 ,tmin diameter) often forming packets of varyingsize. Cell division planes are not necessarily per-pendicular. Cells are nonmotile mesophiles tothermophiles. Energy for growth is obtained bythe oxidation of H2 with the reduction of CO2 tomethane or by the metabolism of methanol,methylamine (dimethylamine, trimethylamine,and ethyldimethylamine), and acetate with theformation of methane and CO2 as end products;ammonia is an additional product from degra-dation of the amines. The cell wall is character-istically thick (500 nm), being composed of anacid heteropolysaccharide (55). Gas-vacuolatedforms have been reported (133; Mah et al. Abstr.Annu. Meet. Am. Soc. Microbiol., 1977, I32, p.160).Genus I, Methanosarcina. Methanosarcina

(Kluyver and van Niel) Barker 1956 (emend.).Me.tha.no.sar.ci'na. M. L. n. methanum meth-ane; L. n. Sarcina a generic name; M. L. fem. n.Methanosarcina the methane sarcina.Only one genus, Methanosarcina, is included

in the family Methanosarcinaceae; the descrip-tion of the genus is the same as that of thefamily. DNA base composition range can beexpected to include 39 to 51 mol% G + C (deter-mined for five strains). The type species, M.methanica (Smit) Kluyver and van Niel 1936,was never obtained in pure culture and is notincluded in the list of approved names (1). Wetherefore designate M. barkeri Schnellen 1947(Schnellen, Ph.D. thesis) the neotype species ofthe genus Methanosarcina. A complete descrip-tion of this genus is presented elsewhere (11, 20,46,62,86, 107-109, 128, 133-135). The type strainis strain MS (DSM 800) (a sewage sludge isolateof M. P. Bryant).A key showing the main differential charac-

teristics of the taxa under consideration is pre-sented in Table 11.

OVERVIEW

Resolution of the Methanogen GrouprRNA is a semantide of sufficiently broad

distribution, constant function, and conservedsequence to serve as a reliable indicator of phy-logenetic relationships over a wide spectrum oforganisms. The results of comparative catalog-ing of the 16S rRNA's from representative spe-

MICROBIOL. REV.

cies of methanogens defnes a comprehensiveand detailed taxonomy which reflects the hier-archy observed in Fig. 7, an arrangement con-sistent with other classical determinative char-acters, including morphology, Gram reaction,mol% G + C, cell wall composition, and lipiddistribution.

Cluster analysis demonstrates the methano-gens to be a major group. The most distantrelationships-having SAB values in the range0.2 to 0.3-are as distant as those observedamong the gram-positive bacteria, enteric bac-teria, and cyanobacteria, for example. The meth-anogens are also as phenotypically varied as arethe typical bacteria. It is reasonable, then, thattaxa at least at the level of class (or division) beassigned to the methanogen group. We considerthe present taxonomic assignments (based on alimited range of known species) to be conserva-tive estimates of the levels of genealogical andphenotypic relatedness within the entire group.Taxonomic categories within the proposed

methanogen group were chosen to reflect thosecharacteristic of typical bacteria. Present taxamay have to be revised upwards as more infor-mation becomes available. For example, the re-lationships of the Methanosarcinaceae to othermethanogens are defined by a single species. Aless conservative interpretation of the SAB valuein Fig. 7 would suggest that the Methanosarci-naceae alone deserve the rank of order. This issupported by a range of distinctive phenotypicdifferences circumscribing the group. A similarsituation applies to some of the lower taxa. M.arboriphilus strain AZ or M. barkeri strain Wmay deserve the rank of separate species on thebasis of several unique characters. Interspeciesand interstrain differences are poorly defined atthis time. Additional methanogenic isolates,when compared to known species on the basis ofphenotypic properties, immunological proper-ties, and relevant molecular markers (e.g., 16SrRNA sequence, protein sequence[s], and nu-cleic acid hybridization), should provide the nec-essary insight to refine the existing taxa anddefine new ones.

Systematics within the procaryotes is, in onesense, necessarily arbitrary; no consistent defi-nition has been available for the assignment ofthe rank of species or any of the higher taxa.The degree of subdivision of a group often re-flects the intensity with which it has been stud-ied, rather than its true depth. The subjectivityof bacterial taxonomic levels was recognized byBenecke in 1912 (13):

What, then are species, i.e., the lowest of the taxo-nomic units? The answer is: that which the investiga-tor who proposes the species wishes to include in

on Novem


mbr.asm

.org/D

ownloaded from


VOL. 43, 1979 METHANOGENS: A NEW PERSPECTIVE 289

TABLE 11. Determninative key to species of the methanogenic bacteriabased on simple phenotypic characters

I. Gram-positive to gram-variable rods or lancet-shaped cocci often forming chains and filaments.Order I. Methanobacteriales

Family I. MethanobacteriaceaeA. Slender, straight to irregularly crooked long rods often occurring in filaments.

Genus I. Methanobacterium1. Mesophilic.

a. Methane produced from formate.Methanobacterium formicicum

b. Methane not produced from formate.Methanobacterium bryantii

2. ThermophilicMethanobacterium thermoautotrophicum

B. Short rods or lancet-shaped cocci which often occur in pairs or chains.Genus II. Methanobrevibacter

1. Cells form short, nonmotile rods which do not utilize formate.Methanobrevibacter arboriphilus

2. Chain-forming, lancet-shaped cocci that produce methane from formate and require acetate as acarbon source.a. Growth requirement for 2-mercaptoethanesulfonic acid and D-a-methyl butyrate.

Methanobrevibacter ruminantiumb. Do not have an obligate growth requirement for 2-mercaptoethanesulfonic acid or D-a-methyl-

butyrate.Methanobrevibacter smithii

II. Gram-negative cells or gram-positive cocci occurring in packets.A. Gram-negative, regular to slightly irregular cocci often forming pairs.

Order II. MethanococcalesFamily I. MethanococcaceaeGenus I. Methanococcus

1. Cells inhibited by addition of 5% NaCl to medium.Methanococcus vannielii

2. Cells not inhibited by addition of 5% NaCl to medium.Methanococcus voltae

B. Gram-negative rods or highly irregular cocci occurring singly.Order III. MethanomicrobialesFamily I. Methanomicrobiaceae

1. Straight to slightly curved, motile, short rods.Genus I. MethanomicrobiumMethanomicrobium mobile

2. Irregular coccoid cells.Genus II. Methanogenium

a. Cells require acetate.Methanogenium cariaci

b. Cells do not require acetate.Methanogenium marisnigri

3. Regularly curved, slender, motile rods, often forming continuous spiral filaments.Genus III. MethanospirillumMethanospirillum hungatei

C. Gram-positive coccoid cells which usually occur in packets and ferment methanol, methylamine, andacetate.

Family II. MethanosarcinaceaeGenus I. MethanosarcinaMethanosarcina barkeri

accordance with his scientific tact. Obviously it is or how little he wants to make his packets-thatimpossible to answer the question in any other way depends upon his scientific attitude, which may varybecause Nature herself does not create species, but for different individuals. [P. 212-213]only individuals with their descendants, so-called"clones." And the systematist collects such clones into Although the present approach does not providegroups or bundles which he calls "species." How big an absolute definition of these "packets," it

on Novem


mbr.asm

.org/D

ownloaded from


290 BALCH ET AL.

promises to develop a consistent, relative meas-ure of taxonomic categories. Such an approachshould provide a basis for interpretation of mi-crobial diversity.

Phylogenetic Relationship of Methanogensto Other Organisms

It is customary to view an organism as beingeither procaryotic or eucaryotic. This distinctionwas originally formulated by Chatton in 1937(24) and refined by Stanier, van Niel, Murray,Allsopp, and others (2, 71-73, 91, 92, 94). Ini-tially, cytological features, such as organelles,formed the basis for the distinction, and theprocaryote was identified in an essentially neg-

ative way-by its not possessing the character-istically eucaryotic features. At this point theprocaryote-eucaryote distinction was organiza-tional, or taxonomic. Detailed molecular char-acterizations have subsequently provided a farmore extensive basis for making the distinction.However, in the process there has occurred asubtle and unrecognized shift in the concept.What was initially a purely organizational dis-tinction has now become a mutually exclusivegenealogical dichotomy: Procaryotae and Eu-caryotae are accepted today as the two primaryphylogenetic categories.

Ifone considers methanogens only in a generalorganizational sense, then they are clearly pro-caryotes. They represent a variety of sizes andshapes typical of bacteria. They lack the eucar-

yotic nuclear membrane and organelles. At thegross molecular level their ribosomes resemblebacterial ribosomes in overall size and size oftheir main component parts. Yet the methano-gens appear to challenge procaryote-eucaryote

as a phylogenetic dichotomy. In terms of rRNAsequence homologies, the methanogens are no

closer to ordinary bacteria than they are to the"cytoplasmic aspect" (18S rRNA) of the eucar-

yotic cell (112). This conclusion, based initiallyon the semantide rRNA, has been confirmedand extended in terms of phenotypic properties(8, 14, 32, 35, 38, 39, 45, 52, 54-56, 64, 69, 95, 96,101, 101a, 102, 108, 109, 113, 119). Such a findingdemands that we reexamine some basic assump-tions of bacterial systematics. Can we maintainthe present empirical definition of the subdivi-sions of the Procaryotae (40) and rationallyincorporate the methanogens? If not, how do weprovide a systematic hierarchy of the typicalbacteria, methanogens, and eucaryotes which isboth deterministically and genealogically con-sistent?The definition of the kingdom Procaryotae

(72) has recently been supplemented by Gibbonsand Murray (40); it constructs a hierarchy ofrelationships for the major bacterial groups on

the basis of cell wall type (Gram reaction). Theprocaryotes are formally subdivided into threenew divisions: division I, Gracilicutes, comprisesall the gram-negative murein cell wall types(inclusive of the cyanobacteria); division II, Fir-macutes, comprises the gram-positive mureincell wall types; and division III, Mollicutes (30),comprises the mycoplasmas, organisms withoutcell walls. In addition, Gibbons and Murray sug-gest that a fourth division, "mendocutes," maybe needed for organisms which "do not have a

clear-cut type of cell wall containing peptidogly-can" (40); this last would include the methano-gens.This proposal is commendable in that it is the

TABLE 12. Degree of similarity between 16S rRNA catalogs of methanogens and other archaebacteria(represented by Halobacterium), the typical bacteria, and the eucaryotic 18S rRNA"

GroupGroup

1 2 3 4 5 6 7 8 9

1. Halobacterium species 2912. Order I methanogens 170 3433. Order II methanogens 167 153 2654. Order III methanogens 119 163 126 3525. Cyanobacteria 51 50 49 26 2776. Rhodospirillaceae species 58 79 51 62 147 2637. Bacillus species 54 58 51 32 151 177 3338. Enteric-Vibrio species 43 61 43 52 145 194 157 3729. Eucaryotic 18S rRNA 56 60 48 51 38 37 35 33 339

" For every pair of catalogs the number of nucleotides in sequence hexamer and larger that are common toboth catalogs is determined. The numbers shown are averages for all such catalog pairs between any two groupsof organisms. (Posttranscriptionally modified nucleotides are for this purpose treated as unmodified sequences.)The values given for Halobacterium, the cyanobacteria, Rhodospirillaceae, Bacillus, and the enteric-Vibriogroups represent averages obtained from 2, 5, 5, 10, and 10 individual species, respectively. The value given forthe eucaryotic 18S rRNA represents an average obtained from three cell types (Lemna, yeast, and L-cells).Nucleotide numbers on the diagonal represent total numbers of nucleotides common to each group.

MICROBIOL. REV.

on Novem


mbr.asm

.org/D

ownloaded from



first attempt to establish formally an encom-passing taxonomy of the bacteria on the basis ofa unifying, molecularly defined characteristic.However, in the short time since its publicationour perception of the evolutionary significanceof the cell wall has changed dramatically. Duelargely to the work of Kandler and his associates(54-56), we now know that among organismsrecognized as bacteria, the murein wall (grampositive or gram negative) is but one of severalbasically different wall types. The pseudomureinwall, the polysaccharide wall, and the proteinsubunit wall also exist (54-56). All the latter walltypes are represented among the methanogens.Is it correct then, to consider the murein walltype to be somehow primary and thus the basisfor structuring a classification of all bacteria? Orshould all wall types be considered phylogenet-ically equivalent and a classification scheme bedevised that gives each type the same weight?These questions cannot be answered on the basisof cell wall type per se; cell wall type, thoughtaxonomically (deterministically) useful, ap-pears insufficient to define a bacterial phy-logeny.Use of semantide analysis (136) has proven to

be a reliable indicator of phylogenetic related-ness (3, 17, 36, 84, 111, 122). On the basis ofcomparative rRNA studies, Woese and Fox (112)have suggested that the categories Euccaryotaeand Procaryotae be viewed as different "levelsof organization"-as "domains" that are notphylogenetically comparable. The procaryotegenealogically can be directly compared to theeucaryote only when the latter is conceptuallydissected into the sum of its phylogeneticallyseparate parts (e.g., mitochrondrion, chloroplast,cytoplasm [18S rRNA]). The term "urkingdom"or "primary kingdom" was coined to designatethe highest phylogenetic category in the "pro-caryotic domain."Three urkingdoms have been proposed (112);

one, the "eubacteria," comprises all bacteria thatpossess murein cell walls (and includes the cy-

anobacteria and most of the wall-less mycoplas-mas); a second is defined by that aspect of theeucaryotic cell represented in the cytoplasmic18S rRNA; whereas a third, the "archaebac-teria," includes the methanogens. This lattercategory is also known to contain the extremehalophiles (halobacteria), and two thermoaci-dophiles, Sulfolobus and Thermoplasma (61,115; Table 12). Woese et al. (115) recently sum-marized their unique and unifying molecularproperties. None have murein cell walls, all havemembranes whose lipid components are thehighly unusual ether-linked polyisoprenoid(branched) chain lipids, and they possess char-acteristic transfer RNAs and rRNA's in their

translation machinery (Table 13). The membersof the archaebacteria are as ecologically andbiochemically diverse as the eubacteria (115).Indeed, existence of archaebacteria phenotypi-cally distinct from the methanogens arguesagainst an interpretation of the 16S rRNA datasuggesting that the rate of rRNA sequencechange in the methanogens is for some reasondifferent, e.g., more rapid, than that found in theeubacteria.Given the novelty of the procaryotic domain

concept, we do not feel justified in proposinghere a formal taxonomic rank at the highestlevel. There can no longer be any doubt that themethanogens and their relatives are a distinctgroup of organisms. In any taxonomic scheme,whether designed as a "means of identification"(72) or as an "indication of relationships" (72),

TABLE 13. Posttranscriptionally modifiedsequences in methanogens and their likely

counterparts in other bacteriaOccurrence in:

Sequence Methanogen treme Typicalhalo- bacteria

I II III philes

1. a. C-CCG + - - - Noneb. C-CCG - + - - Nonec. C(C,C)G - - + + Noned. C-CCG - - - - >95

2. CCGCG - - +h - >953. a. AACCUG + + - - -30

b. AAUCUG - - + + Nonec. AAG - -60

4. a. UAACAAG + + - - Noneb. UAACAAG - - + + Nonec. UAACAAG - - - - >95

5. a. AUNCAACG + + - - Noneb. ACNCAACG - - + + Nonec. AAG(CAACG - - - - >95d. AUGCAACG - - - - >95

'The modifications, to the extent they are known,are the following. C modifications: C- in la and lb istentatively identified as 2'OMeC; C- in Id is4Me2'OMeC; Cs in lb and c are unidentified, aresensitive to pancreatic nuclease, and are probably thesame, but they are distinguishable from C in 5c, whichis 5MeC. A modifications: all As in 3 are N6Me2A; asare the second As in 4a and b; the initial A in 4a isunique but unidentified. G modifications: G in 2 is7MeG, identified by its being quantitatively destroyedby concentrated ammonium hydroxide at room tem-perature (N. Leonard, personal communication). G in5c and d is either 2MeG (most cases) or another,unidentified modification (cleaved by T, nuclease); Nin 5a and b may be a G derivative in that it is slowlycleaved by T1 nuclease. U modifications: U in 4c isunidentified; it is cleaved by pancreatic nuclease at areduced rate.

b Found only in Methanosarcina sp.

VOL. 43, 1979

on Novem


mbr.asm

.org/D

ownloaded from


292 BALCH ET AL.

it is clear the methanogens and their relativesmust ultimately be assigned a rank equivalentto that covering all typical bacteria.

ACKNOWLEDGMENTSWe gratefully acknowledge the patient assistance of

M. P. Bryant, whose deep knowledge of the methan-ogenic bacteria provided the constant critical reviewnecessary in developing the new taxonomic treatmentand whose support and enthusiasm for the project wasunending. We are indebted for the scholarly counseland interest of H. Merxmuller of the UniversitiitMunchen and H. Triiper of the Universitiit Bonn, onthe correct forms of the generic and specific nameswhich we propose. We appreciate the help of ourcolleagues, L. Baresi, D. Boone, D. Castignetti, H.Hippe, R. A. Mah, J. A. Romesser, S. Schoberth, P.H. Smith, and A. J. B. Zehnder for providing culturesto make the study as complete as possible. We thank0. Kandler, L. C. Klotz, T. A. Langworthy, T. G.Tornabene, G. D. Vogels, and J. G. Zeikus for gener-ously providing unpublished data. We thank V. Ga-briel, J. McCrea, and D. Szurgot for excellent technicalassistance. We acknowledge the helpful criticisms andcomments of K. Luehrsen and E. Stackebrandt as wellas all others listed, whose generous cooperation madethis manuscript possible.

This work reported herein was performed underNational Aeronautics and Space Administration(NASA) grant NSG-7044 and National Science Foun-dation (NSF) grant DEB 77-27-492 to C.R.W., NASAgrant NSG-7440 to G.E.F., and Public Health Servicegrant Al 12277 from the National Institute of Allergyand Infectious Diseases and NSF grant PCM 76-02652to R.S.W.

LITERATURE CITED

1. Ad Hoc Committee of the Judicial Commis-sion of the International Committee onSystematic Bacteriology. 1976. First draft.Approved lists of bacterial names. Int. J. Syst.Bacteriol. 26:563-599.

2. Allsopp, A. 1969. Phylogenetic relationships ofthe Procaryota and the origin of the eucaryoticcell. New Phytol. 68:591-612.

3. Almassy, R. J., and R. E. Dickerson. 1978.Pseudomonas cytochrome c55, at 2.0 A resolu-tion: enlargement of the cytochrome c family.Proc. Natl. Acad. Sci. U.S.A. 75:2674-2678.

4. Anderberg, M. R. 1973. Cluster analysis forapplications. Academic Press Inc., New York.

5. Aranki, A., and R. Freter. 1972. Use of anaer-obic glove boxes for the cultivation of strictlyanaerobic bacteria. Am. J. Clin. Nutr. 25:1329-1334.

6. Balch, W. E., L. J. Magrum, G. E. Fox, R. S.Wolfe, and C. R. Woese. 1977. An ancientdivergence among the bacteria. J. Mol. Evol.9:305-311.

7. Balch, W. E., and R. S. Wolfe. 1976. Newapproach to the cultivation of methanogenicbacteria: 2-mercaptoethanesulfonic acid (HS-CoM)-dependent growth of Methanobacte-

MICROBIOL. REV.

rium ruminantium in a pressurized atmos-phere. Appl. Environ. Microbiol. 32:781-791.

8. Balch, W. E., and R. S. Wolfe. 1979. Specificityand biological distribution of coenzyme M (2-mercaptoethanesulfonic acid). J. Bacteriol.137:256-263.

9. Balch, W. E., and R. S. Wolfe. 1978. Transportof coenzyme M (2-mercaptoethanesulfonicacid) in Methanobacterium ruminantium. J.Bacteriol. 137:264-273.

10. Barker, H. A. 1936. Studies on methane produc-ing bacteria. Arch. Microbiol. 7:420-438.

11. Barker, H. A. 1956. Bacterial fermentations, p.1-95. John Wiley & Sons, Inc., New York.

12. Bellemare, G., R. Vigne, and B. Jordan. 1973.Interaction between Escherichia coli ribo-somal proteins and 5S RNA molecules: recog-nition of prokaryotic 5S RNAs and rejection ofeukaryotic 5S RNA. Biochimie 55:29-35.

13. Benecke, W. 1912. Bau und Leben der Bakter-ien. B. G. Teubner, Leipzig.

14. Best, A. N. 1978. Composition and characteri-zation oftRNA from Methanococcus vannielii.J. Bacteriol. 133:240-250.

15. Bonen, L., and W. F. Doolittle. 1972. On theprocaryotic nature of red algal chloroplasts.Proc. Natl. Acad. Sci. U.S.A. 72:2310-2314.

16. Bonen, L., and W. F. Doolittle. 1976. Partialsequences of 16S rRNA and the phylogeny ofblue-green algae and chloroplasts. Nature(London) 261:669-673.

17. Bonen, L., and W. F. Doolittle. 1978. Ribo-somal RNA homologies and the evolution ofthe filamentous blue-green bacteria. J. Mol.Evol. 10:283-291.

18. Bryant, M. P. 1965. Rumen methanogenic bac-teria, p. 411-418. In R. W. Dougherty, R. S.Allen, W. Burroughs, N. L. Jackson, and A. D.McGillard (ed.), Physiology of digestion of theruminant. Butterworths Publishing, Inc.,Washington, D.C.

19. Bryant, M. P. 1972. Commentary on Hungatetechnique for culture of anaerobic bacteria.Am. J. Clin. Nutr. 25:1324-1328.

20. Bryant, M. P. 1974. Part 13. Methane-producingbacteria, p. 472-477. In R. E. Buchanan, andN. E. Gibbons (ed.), Bergey's manual of deter-minative bacteriology, 8th ed. The Williams &Wilkins Co., Baltimore.

21. Bryant, M. P., B. C. McBride, and R. S. Wolfe.1968. Hydrogen-oxidizing methane bacteria. I.Cultivation and methanogenesis. J. Bacteriol.95:1118-1123.

22. Bryant, M. P., S. F. Tzeng, I. M. Robinson,and A. E. Joyner. 1971. Nutrient require-ments of methanogenic bacteria. Adv. Chem.Ser. 105:23-40.

23. Bryant, M. P., E. A. Wolin, M. J. Wolin, andR. S. Wolfe. 1967. Methanobacterium ome-lianskii, a symbiotic association of two speciesof bacteria. Arch. Mikrobiol. 59:20-31.

24. Chatton, E. 1937. Titres et travoux scientifiques.Sete, Sottano.

25. Daniels, L, G. Fuchs, R. K. Thauer, and J.

on Novem


mbr.asm

.org/D

ownloaded from



G. Zeikus. 1977. Carbon monoxide oxidationby methanogenic bacteria. J. Bacteriol. 132:118-126.

26. Daniels, L., and J. G. Zeikus. 1978. One-carbonmetabolism in methanogenic bacteria: analysisof short-term fixation products of "CO2 and'4CH30H incorporated into whole cells. J. Bac-teriol. 136:75-84.

27. Doddema, H. J., J. W. M. Derksen, and G. D.Vogels. 1979. Fimbriae and flagella of meth-anogenic bacteria. FEMS Microbiol. Lett. 5:135-138.

28. Doddema, H. J., and G. D. Vogels. 1978. Im-proved identification of methanogenic bacteriaby fluorescence microscopy. Appl. Environ.Microbiol. 36:752-754.

29. Doolittle, W. F. 1975. Sequence studies on 16Sribosomal RNA from a blue-green alga. J. Mol.Evol. 4:307-315.

30. Edward, D. G. ff., and E. A. Freundt 1967.Proposal for Mollicutes as name of the Classestablished for the Order Mycoplasmatales.Int. J. Syst. Bacteriol. 17:267-268.

31. Edwards, T., and B. C. McBride. 1975. Newmethod for the isolation and identification ofmethanogenic bacteria. Appl. Microbiol. 29:540-545.

32. Eirich, L. D., G. D. Vogels, and R. S. Wolfe.1978. The structure of coenzyme F420, a novelelectron carrier isolated from Methanobacte-rium strain M.o.H. Biocheniistry 17:4583-4593.

33. Ferry, J. G., P. H. Smith, and R. S. Wolfe.1974. Methanospirillum, a new genus of meth-anogenic bacteria, and characterization ofMethanospirillum hungatii sp. nov. Int. J.Syst. Bacteriol. 24:465-469.

34. Ferry, J. G., and R. S. Wolfe. 1977. Nutritionaland biochemical characterization of Methan-ospirillum hungatii. Appl. Environ. Microbiol.34:371-376.

35. Fox, G. E., L J. Magrum, W. E. Balch, R. S.Wolfe, and C. R. Woese. 1977. Classificationof methanogenic bacteria by 16S ribosomalRNA characterization. Proc. Natl. Acad. Sci.U.S.A. 74:4537-4541.

36. Fox, G. E., K. J. Pechman, and C. R. Woese.1977. Comparative cataloging of 16S ribosomalribonucleic acid: molecular approach to procar-yotic systematics. Int. J. Syst. Bacteriol. 27:44-57.

37. Fox, G. E., and C. R. Woese. 1975. 5S RNAsecondary structure. Nature (London) 256:505-507.

38. Fuchs, G., and E. Stupperich. 1978. Evidencefor an incomplete reductive carboxylic acidcycle in Methanobacterium thermoautotrophi-cum. Arch. Microbiol. 118:121-126.

39. Fuchs, G., E. Stupperich, and R. K. Thauer.1978. Acetate asimilation and the synthesis ofalanine, aspartate and glutamate in Methano-bacterium ther7noautotrophicum. Arch. Mi-crobiol. 117:61-66.

40. Gibbons, N. E., and R G. E. Murray. 1978.

Proposals concerning the higher taxa of bac-teria. Int. J. Syst. Bacteriol. 28:1-6.

41. Gordon, R. E., W. C. Haynes, and C. Hor-NayPang. 1973. The genus Bacillus. U.S. Dep.Agric. Agric. Handb. no. 427.

42. Gunsalus, R., D. Eirich, J. Romesser, W.Balch, S. Shapiro, and R. S. Wolfe. 1976.Methyl-transfer and methane formation, p.191-198. In H. Schlegel, G. Gottschalk, and N.Pfennig (ed.), Proceedings of Symposium onMicrobial Production and Utilization of Gases(H2, C02, CO). Akademie du Wissenschaftenzu Gottingen, Goltze Verlage, Gottingen.

43. Gunsalus, R. P., J. A. Romesser, and R. S.Wolfe. 1978. Preparation of coenzyme M ana-logues and their activity in the methyl coen-zyme M reductase system of Methanobacte-rium thermoautotrophicum. Biochemistry 17:2374-2377.

44. Gunsalus, R. P., and R. S. Wolfe. 1977. Stim-ulation of CO2 reduction to methane bymethyl-coenzyme M in extracts of Methano-bacterium. Biochem. Biophys. Res. Commun.76:790-795.

45. Gunsalus, R. P., and R. S. Wolfe. 1978. Chrom-ophoric factors F342 and F430 of Methanobac-terium thermoautotrophicum. FEMS Micro-biol. Lett. 3:191-193.

46. Hippe, H., D. Caspari, K. Fiebig, and G.Gottschalk. 1979. Utilization of trimethyl-amine and other N-methyl compounds forgrowth and methane formation by Methano-sarcina barkeri. Proc. Natl. Acad. Sci. U.S.A.76:494-498.

47. Holdeman, L. V., and W. E. C. Moore. 1972.Roll-tube techniques for anaerobic bacteria.Am. J. Clin. Nutr. 25:1314-1317.

48. Holdeman, L. V., and W. E. C. Moore (ed.).1972. Anaerobe laboratory manual. VirginiaPolytechnic Institute and State University,Blacksburg.

49. Hoppe-Seyler, F. 1886. Uber die Gahrung derCellulose mit Bildung von Methan und Koh-lensaure. I. Ueber das Vorkommen der En-twicklung von Methan und Kohlensaure inWasserhaltigen Erdboden. Hoppe-Seyler's Z.Physiol. Chem. 10:201-217.

50. Hungate, R. E. 1950. The anaerobic mesophiliccellulolytic bacteria. Bacteriol. Rev. 14:1-49.

51. Hungate, R. E. 1969. A roll tube method forcultivation of strict anaerobes, p. 117-132. InJ. R. Norris and D. W. Ribbons (ed.), Methodsin microbiology, vol. 3B. Academic Press Inc.,New York.

52. Jones, J. B., B. Bowers, and T. C. Stadtman.1977. Methanococcus vannielii: ultrastructureand sensitivity to detergents and antibiotics. J.Bacteriol. 130:1357-1363.

53. Jones, J. B., and T. C. Stadtman. 1977. Meth-anococcus vannielii: culture and effects of se-lenium and tungsten on growth. J. Bacteriol.130:1404-1406.

54. Kandler, 0. 1979. Zellwandstrukturen beiMethan-Bakterien. Zur Evolution der Prokar-

VOL. 43, 1979

on Novem


mbr.asm

.org/D

ownloaded from


294 BALCH ET AL.

yonten. Naturwissenschaften ":95-105.55. Kandler, O., and H. Hippe. 1977. Lack of pep-

tidoglycan in the cell walls of Methanosarcinabarkeri. Arch. Microbiol. 113:57-60.

56. Kandler, O., and H. Konig. 1978. Chemicalcomposition of the peptidoglycan-free cellwalls of methanogenic bacteria. Arch. Micro-biol. 118:141-152.

57. Kluyver, A. J., and C. G. Schnellen. 1947. Onthe fermentation of carbon monoxide by purecultures of methane bacteria. Arch. Biochem.14:57-70.

58. Kluyver, A. J., and C. B. van Niel. 1936. Pros-pects for a natural system of classification ofbacteria. Zentrabl. Bakteriol. Parasitenkd. In-fektionskr. Hyg. Abt. 2 94:369-403.

59. Langenberg, K. F., M. P. Bryant, and R. S.Wolfe. 1968. Hydrogen-oxidizing methane bac-teria. II. Electron microscopy. J. Bacteriol. 95:1124-1129.

60. Macy, J. M., J. E. Snellen, and R. E. Hungate.1972. Use of syringe methods for anaerobiosis.Am. J. Clin. Nutr. 25:1318-1323.

61. Magrum, L J., K. R. Luehr8en, and C. R.Woese. 1978. Are extreme halophiles actually"bacteria"? J. Mol. Evol. 11:1-8.

62. Mah, R. A., M. R. Smith, and L Baresi. 1978.Studies on an acetate-fernenting strain ofMethanosarcina. Appl. Environ. Microbiol.35:1174-1184.

63. Mah, R. A., D. M. Ward, L Baresi, and T. LGlass. 1977. Biogenesis of methane. Annu.Rev. Microbiol. 31:309-341.

64. Makula, R. A., and M. E. Singer. 1978. Ether-containing lipids of methanogenic bacteria.Biochem. Biophys. Res. Commun. 82:716-722.

65. Marmur, J. 1961. A procedure for the isolationofdeoxyribonucleic acid from micro-organisms.J. Mol. Biol. 3:208-218.

66. McBride, B. C., and R. S. Wolfe. 1971. A newcoenzyme of methyl-transfer, coenzyme M.Biochemistry 10:2317-2324.

67. Miller, T. L., and M. J. Wolin. 1974. A serumbottle modification of the Hungate techniquefor cultivating obligate anaerobes. Appl. Micro-biol. 27:985-987.

68. Mink, R. W., and P. R. Dugan. 1977. Tentativeidentification of methanogenic bacteria by flu-orescence microscopy. Appl. Environ. Micro-biol. 33:713-717.

69. Mitchell, R. M., L. A. Loeblich, L C. Klotz,and A. R. Loeblich ]I. 1978. DNA organiza-tion of Methanobacterium thermoautotrophi-cum. Science 204:1082-1084.

70. Mortenson, L E., and G. Nakos. 1973. Bacte-rial ferredoxins and/or iron-sulfur proteins aselectron carriers, p. 37-64. In W. Lovenberg(ed.), Iron-sulfur proteins. Academic Press Inc.,New York.

71. Murray, R. G. E. 1962. Fine structure and tax-onomy of bacteria. In Ainsworth and Sneath(ed.), Microbial classification. Cambridge Uni-versity Press, Cambridge.

72. Murray, R. G. E. 1968. Microbial structure as an

aid to microbial classification and taxonomy.Spisy Prirodoved Fak. Univ. J. E. PurkyneBrne 43:245-252.

73. Murray, R. G. E. 1974. A place for bacteria inthe living world, p. 4-9. In R. E. Buchanan andN. E. Gibbons (ed.), Bergey's manual of deter-minative bacteriology, 8th ed. The Williams &Wilkins Co., Baltimore.

74. Mylorie, R. L., and R. E. Hungate. 1954. Ex-periments on the methane bacteria in sludge.Can. J. Microbiol. 1:55-64.

75. Nomura, M., P. Traub, and H. Bechmann.1968. Hybrid 30S ribosomal particles reconsti-tuted from components of different bacterialorigins. Nature (London) 219:793-799.

76. Nottingham, P. M., and R. E. Hungate. 1968.Isolation of methanogenic bacteria from fecesof man. J. Bacteriol. 96:2178-2179.

77. Pantskhava, E. S., and V. V. Pchelkina. 1968.Corrmnoid compounds synthesized by a newlyseparated species of methane bacterium. Dokl.Akad. Nauk SSSR 182:457-458.

78. Patel, G. B., L A. Roth, LX van den Berg, andD. S. Clark. 1976. Characterization of a strainof Methanospirillum hungatii. Can. J. Micro-biol. 22:1404-1410.

79. Paynter, M. J. B., and R. E. Hungate. 1968.Characterization of Methanobacterium mobi-lis, sp. n., isolated from the bovine rumen. J.Bacteriol. 95:1943-1951.

80. Pechman, K. J., B. J. Lewis, and C. R. Woese.1976. Phylogenetic status of Sporosarcinaureae. Int. J. Syst. Bacteriol. 26:305-310.

81. Pechman, K. J., and C. R. Woese. 1972. Char-acterization of the primary structural homol-ogy between the 16S ribosomal RNAs of Esch-erichia coli and Bacillus megaterium by oli-gomer cataloging. J. Mol. Evol. 1:230-240.

81a.Romesser, J. A., R. S. Wolfe, F. Mayer, E.Spiess, and A. Walther-Mauruschat. 1979.Methanogenium, a new genus of marine meth-anogenic bacteria, and characterization ofMethanogenium cariaci sp. nov. and Meth-anogenium marisnigri sp. nov. Arch. Micro-biol. 121:147-153.

82. Sanger, F., G. G. Brownlee, and B. G. Barrell.1965. A two-dimensional fractionation proce-dure for radioactive nucleotides. J. Mol. Biol.13:373-398.

83. Schildkraut, C. L, J. Marmur, and P. Doty.1962. Determination of the base compositionof deoxyribonucleic acid from its buoyant den-sity in CsCl. J. Mol. Biol. 4:430-443.

84. Schwartz, R. M., and M. 0. Dayhoff. 1978.Origins of prokaryotes, eukaryotes, mitochon-dria, and chloroplasts. Science 199:395-403.

85. Smit, J. 1930. Die Garungssarcinen. Eine Mon-ographie. Pflanzen forschung Heft 14, p. 1-59.

86. Smith, M. R., and R. A. Mah. 1978. Growth andmethanogenesis by Methanosarcina strain 227on acetate and methanol. Appl. Environ. Mi-crobiol. 36:870-879.

87. Smith, P. H. 1966. The microbial ecology ofsludge methanogenesis. Dev. Ind. Microbiol. 7:

MICROBIOL. REV.

on Novem


mbr.asm

.org/D

ownloaded from



156-160.

88. Smith, P. H., and R. E. Hungate. 1958. Isola-tion and characterization of Methanobacte-rium ruminantium n. sp. J. Bacteriol. 76:713-718.

89. Stadtman, T. C., and H. A. Barker. 1951.Studies on the methane fermentation. IX. Theorigin of methane in the acetate and methanolfermentations by Methanosarcina. J. Bacte-riol. 61:81-86.

90. Stadtman, T. C., and H. A. Barker. 1951.Studies on the methane fermentation. X. Anew formate-decomposing bacterium, Meth-anococcus vannielii. J. Bacteriol. 62:269-280.

91. Stanier, R Y. 1961. La place des bacteries dansle monde vivant. Ann. Inst. Pasteur Paris 101:297-303.

92. Stanier, R. Y. 1970. Some aspects of the biologyof cells and their possible evolutionary signifi-cance. In Charles and Knight (ed.), Organiza-tion and control in prokaryotic and eukaryoticcells. Cambridge University Press, Cambridge.

93. Stanier, RK Y., and C. B. van Niel. 1941. Themain outlines of bacterial classification. J. Bac-teriol. 42:437-466.

94. Stanier, R. Y., and C. B. van Niel. 1962. Theconcept of a bacterium. Arch. Mikrobiol. 42:17-35.

95. Steitz, J. A. 1978. Methanogenic bacteria. Na-ture (London) 273:10.

96. Taylor, C. D., B. C. McBride, R S. Wolfe, andM. P. Bryant. 1974. Coenzyme M, essentialfor growth of a rumen strain of Methanobac-terium ruminantium. J. Bacteriol. 120:974-975.

97. Taylor, C. D., and R. S. Wolfe. 1974. Struc-ture and methylation of coenzyme M(HSCH2CH2SO3). J. Biol. Chem. 249:4879-4885.

98. Taylor, G. T., G. T. Kelly, and S. J. Pirt. 1976.Intermediary metabolism in methanogenicbacteria (Methanobacterium), p. 173-180. InH. G. Schlegel, G. Gottschalk, and N. Pfennig(ed.), Proceedings of Symposium on MicrobialProduction and Utilization of Gases (H2, C04,CO). Akademie du Wissenschaften zu Gottin-gen. Goltze Verlagen, Gottingen.

99. Taylor, G. T., and S. J. Pirt. 1977. Nutritionand factors limiting the growth of a methano-genic bacterium (Methanobacterium ther-moautotrophicum). Arch. Microbiol. 113:17-22.

100. Thauer, R K., K. Jungermann, and K.Decker. 1977. Energy conservation in chemo-trophic anaerobic bacteria. Bacteriol. Rev. 41:100-180.

101. Tornabene, T. G., and T. A. Langworthy.1978. Diphytanyl and dibiphytanyl glycerolether lipids of methanogenic archaebacteria.Science 203:51-53.

101a.Tornabene, T. G., T. A. Langworthy, G. Hol-zer, and J. Or6. 1979. Squalenes, phytanes,and other isoprenoids as major neutral lipidsof methanogenic and thermoacidophilic

"archaebacteria." J. Mol. Evol. 13:1-8.102. Tornabene, T. G., R. S. Wolfe, W. E. Balch,

G. Holzer, G. E. Fox, and J. Or6. 1978.Phytanyl-glycerol ethers and squalenes in thearchaebacterium Methanobacterium ther-moautotrophicum. J. Mol. Evol. 11:259-266.

103. Tzeng, S. F., M. P. Bryant, and R. S. Wolfe.1975. Factor 420-dependent pyridine nucleo-tide-linked formate metabolism of Methano-bacterium ruminantium. J. Bacteriol. 121:192-196.

104. Tzeng, S. F., R. S. Wolfe, and M. P. Bryant.1975. Factor 420-dependent pyridine nucleo-tide-linked hydrogenase system of Methano-bacterium ruminantium. J. Bacteriol. 121:184-191.

105. Uchida, T., L. Bonen, H. W. Schaup, B. J.Lewis, L Zablen, and C. R. Woese. 1974.The use of ribonuclease U2 in RNA sequencedetermination: some corrections in the catalogof oligomers produced by ribonuclease T,digestion of Escherichia coli 168 ribosomalRNA. J. Mol. Evol. 3:63-77.

106. van Niel, C. B. 1946. The classification andnatural relationships of bacteria. Cold SpringHarbor Symp. Quant. Biol. 11:285-301.

107. Weimer, P. J., and J. G. Zeikus. 1978. Acetatemetabolism in Methanosarcina barkeri. Arch.Microbiol. 119:175-182.

108. Weimer, P. J., and J. G. Zeikus. 1978. Onecarbon metabolism in methanogenic bacteria:cellular characterization and growth of Meth-anosarcina barkeri. Arch. Microbiol. 119:49-57.

109. Weimer, P. J., and J. G. Zeikus. 1979. Acetateassunilation pathway of Methanosarcina bar-keri. J. Bacteriol. 137:332-339.

110. Wellinger, A., and K. Wuhrmann. 1977. Influ-ence of sulfide compounds on the metabolismof Methanobacterium strain AZ. Arch. Micro-biol. 115:13-17.

111. Wilson, A. C., S. S. Carlson, and T. J. White.1977. Biochemical evolution. Annu. Rev. Bio-chem. 46:573-639.

112. Woese, C. R., and G. E. Fox. 1977. Phylogeneticstructure of the prokaryotic domain: the pri-mary kingdoms. Proc. Natl. Acad. Sci. U.S.A.74:5088-5090.

113. Woese, C. R., and G. E. Fox. 1978. Methano-genic bacteria. Nature (London) 273:101.

114. Woese, C. R., G. E. Fox, L Zablen, T. Uchida,L. Bonen, K. Pechman, B. J. Lewis, and D.Stahl. 1975. Conservation ofprimary structurein 16S ribosomal RNA. Nature (London) 254:8346.

115. Woese, C. R., L. J. Magrum, and G. E. Fox.1978. Archaebacteria. J. Mol. Evol. 11:245-252.

116. Woese, C. R., M. Sogin, D. Stahl, B. J. Lewis,and L. Bonen. 1976. A comparison of the 16Sribosomal RNAs from mesophilic and ther-mophilic Bacilli: some modifications in theSanger method for RNA sequencing. J. Mol.Evol. 7:197-213.

117. Woese, C. R., M. L. Sogin, and L. A. Sutton.

VOL. 43, 1979

on Novem


mbr.asm

.org/D

ownloaded from


296 BALCH ET AL.

1974. Procaryote phylogeny I: concerning therelatedness of Aerobacter aerogenes to Esch-erichia coli. J. Mol. Evol. 3:293-299.

118. Wolfe, R. S. 1971. Microbial formation of meth-ane, p. 107-146. In A. H. Rose and J. F. Wilk-inson (ed.), Advances in microbiological phys-iology, vol. 6. Academic Press Inc., New York.

119. Wolfe, R. S., and I. J. Higgins. 1979. Biochem-istry of methane-a study in contrasts, p. 267-283. In J. R. Quayl (ed.), Microbial biochem-istry. MTP Press Ltd., Lancaster, England.

120. Wrede, P., and U. A.Erdma 1973. ActivitiesofB. stearothermophilus 50S ribosomes recon-stituted with prokaryotic and eukaryotic 5SRNA. FEBS Lett. 33:315-319.

121. Zablen, L, L. Bonen, R. Meyer, and C. R.Woese. 1975. The phylogenetic status of Pas-teureUaapestis. J. Mol. Evol. 4:347-358.

122. Zablen, L, M. S. Kissil, C. R. Woese, and D.E. Buetow. 1975. Phylogenetic origin of thechloroplast and prokaryotic nature of its ribo-somal RNA. Proc. Natl. Acad. Sci. U.S.A. 72:2418-2422.

123. Zablen, L, and C. R. Woese. 1975. Procaryotephylogeny IV: concerning the phylogenetic sta-tus of a photosynthetic bacterium. J. Mol.Evol. 5:25-34.

124. Zehnder, A. J. B. 1978. Ecology of methaneformation, p. 349-376. In R. Mitchell (ed.),Water pollution microbiology, vol. 2. JohnWiley & Sons, Inc., New York.

125. Zehnder, A. J. B., and K. Wuhrmann. 1977.Physiology of a Methanobacterium strain AZ.Arch. Microbiol. 111:199-205.

126. Zeikus, J. G. 1977. The biology of methanogenicbacteria. Bacteriol. Rev. 41:514-541.

MICROBIOL. REV.

127. Zeikus, J. G., and V. G. Bowen. 1975. Finestructure of MethanospiriUnm hungatii. J.Bacteriol. 121:373-380.

128. Zeikus, J. G., and V. G. Bowen. 1975. Com-parative ultrastructure of methanogenic bac-teria. Can. J. Microbiol. 21:121-129.

129. Zeikus, J. G., G. Fuchs, W. Kenealy, and R.K. Thauer. 1977. Oxidoreductases involved incell carbon synthesis of Methanobacteriumthermoautotrophicum. J. Bacteriol. 132:604-613.

130. Zeikus, J. G., and D. L Henning. 1975. Meth-anobacterium arbophilicum sp. nov., an obli-gate anaerobe isolated from wetwood of livingtrees. Antonie van Leeuwenhoek J. Microbiol.Serol. 41:543-552.

131. Zeikus, J. G., and R. S. Wolfe. 1972. Methan-obacterium thermoautotrophicus sp. n., an an-aerobic, autotrophic, extreme thermophile. J.Bacteriol. 109:707-713.

132. Zeikus, J. G., and R. S. Wolfe. 1973. Finestructure of Methanobacterium thermoauto-trophicum: effect of growth temperature onmorphology and ultrastructure. J. Bacteriol113:461-467.

133. Zhilina, T. 1971. The fine structure of Methan-osarcina. Microbiology (USSR) 40:587-591.

134. Zhilina, T. N. 1976. Biotypes of Methanosar-cina. Microbiology (USSR) 45:414-421.

135. Zhilina, T. N., and N. I. Aleksandrushkina.1974. On the taxonomy of Methanosarcina.Microbiology (USSR) 43:620-621.

136. Zuckerkandl, E., and L. Paulng. 1965. Mole-cules as documents of evolutionary history. J.Theor. Biol. 8:357-366.

on Novem


mbr.asm

.org/D

ownloaded from


Documents

Methanogens:Reevaluation Unique Biological GroupMETHANOGENS: A NEWPERSPECTIVE 263 gate (50, 51). Theversatility ofthese anaerobic proceduresiswelldocumented(19,21,31,47,48, 51, 60,