APPLIED AND ENVIRONMENTAL MICROBIOLOGY,0099-2240/99/$04.0010

Aug. 1999, p. 3641–3650 Vol. 65, No. 8

Copyright © 1999, American Society for Microbiology. All Rights Reserved.

Isolation and Characterization of Methanomethylovorans hollandicagen. nov., sp. nov., Isolated from Freshwater Sediment, a

Methylotrophic Methanogen Able To Grow onDimethyl Sulfide and Methanethiol

BART P. LOMANS, RONALD MAAS, RIANNE LUDERER, HUUB J. M. OP DEN CAMP,*ARJAN POL, CHRIS VAN DER DRIFT, AND GODFRIED D. VOGELS

Department of Microbiology and Evolutionary Biology, Faculty of Science,University of Nijmegen, NL-6525 ED Nijmegen, The Netherlands

Received 28 January 1999/Accepted 20 May 1999

A newly isolated methanogen, strain DMS1T, is the first obligately anaerobic archaeon which was directlyenriched and isolated from a freshwater sediment in defined minimal medium containing dimethyl sulfide(DMS) as the sole carbon and energy source. The use of a chemostat with a continuous DMS-containing gasstream as a method of enrichment, followed by cultivation in deep agar tubes, resulted in a pure culture. Sincethe only substrates utilized by strain DMS1T are methanol, methylamines, methanethiol (MT), and DMS, thisorganism is considered an obligately methylotrophic methanogen like most other DMS-degrading methano-gens. Strain DMS1T differs from all other DMS-degrading methanogens, since it was isolated from a fresh-water pond and requires NaCl concentrations (0 to 0.04 M) typical of the NaCl concentrations required byfreshwater microorganisms for growth. DMS was degraded effectively only in a chemostat culture in thepresence of low hydrogen sulfide and MT concentrations. Addition of MT or sulfide to the chemostat signif-icantly decreased degradation of DMS. Transient accumulation of DMS in MT-amended cultures indicatedthat transfer of the first methyl group during DMS degradation is a reversible process. On the basis of its lowlevel of homology with the most closely related methanogen, Methanococcoides burtonii (94.5%), its position onthe phylogenetic tree, its morphology (which is different from that of members of the genera Methanolobus,Methanococcoides, and Methanohalophilus), and its salt tolerance and optimum (which are characteristic offreshwater bacteria), we propose that strain DMS1T is a representative of a novel genus. This isolate wasnamed Methanomethylovorans hollandica. Analysis of DMS-amended sediment slurries with a fluorescencemicroscope revealed the presence of methanogens which were morphologically identical to M. hollandica, asdescribed in this study. Considering its physiological properties, M. hollandica DMS1T is probably responsiblefor degradation of MT and DMS in freshwater sediments in situ. Due to the reversibility of the DMSconversion, methanogens like strain DMS1T can also be involved in the formation of DMS through methylationof MT. This phenomenon, which previously has been shown to occur in sediment slurries of freshwater origin,might affect the steady-state concentrations and, consequently, the total flux of DMS and MT in these systems.

Dimethyl sulfide (DMS) has an impact on global warmingand acid precipitation and on the global sulfur cycle because ofits oxidation products (e.g., methanesulfonic acid and SO2),which are released into the atmosphere. For this reason trans-formations of volatile organic sulfur compounds have beenintensively studied during the past few decades. Microbial for-mation and degradation of DMS and methanethiol (MT) havebeen shown to have a significant effect on the total flux ofsulfur compounds in the atmosphere (13, 14, 21).

In freshwater sediments, formation of MT and DMS is bal-anced by degradation of these compounds, which results in lowsteady-state concentrations (18–20). In contrast to marine andestuarine systems, in which DMS originates mainly from di-methylsulfoniumpropionate, volatile organic sulfur com-pounds in freshwater sediments are derived mainly from meth-ylation of sulfide and MT (7, 15, 18).

In systems with high salt contents, degradation of MT and

DMS has been attributed to members of various groups ofbacteria, including aerobes (e.g., thiobacilli and Methylophagaspp.) (4, 34–36) and anaerobes (anoxygenic phototrophs, sul-fate reducers, and methanogens) (8, 16, 17, 23, 25, 26, 33, 39).The activity of members of these trophic groups depends onthe light intensity and the availability of oxygen or alternativeelectron acceptors, such as sulfate and nitrate. Due to oxygenlimitation in freshwater sediments, DMS and MT are degradedmainly anaerobically by means of methanogenic activity (19).Methanogenic conversion of MT and DMS in sediment slur-ries was first demonstrated by Zinder and Brock (40, 41). Sincethen, various methanogens have been isolated with DMS orMT from marine, estuarine, salt marsh, and salt lake sediments(8, 12, 16, 17, 23, 25). These methanogens belong to the generaMethanosarcina, Methanolobus, and Methanosalsus. Althoughmethanogens have been identified as the dominant consumersof DMS and MT in sulfate-poor freshwater sediments (18–20,40, 41), previous attempts to isolate methanogens which areable to grow on DMS or MT from such sediments were un-successful (33, 41). Moreover, production of methane or car-bon dioxide (or [14C]methane and [14C]carbon dioxide) fromMT or DMS (or [14C]MT and [14C]DMS) was not detected inpure cultures of methanogens isolated from nonsaline systems

* Corresponding author. Mailing address: Department of Micro-biology and Evolutionary Biology, Faculty of Science, University ofNijmegen, Toernooiveld 1, NL-6525 ED Nijmegen, The Netherlands.Phone: 31 (0) 24 3652657. Fax: 31 (0) 24 3652830. E-mail: huubcamp@sci.kun.nl.

3641

on April 11, 2021 by guest

http://aem.asm

.org/D

ownloaded from

(e.g., Methanobacterium ruminantium, Methanobacterium ther-mautotrophicum, and Methanosarcina barkeri cultures) (28, 41).

In this paper we describe the isolation of a nonhalophilicmethylotrophic methanogen, strain DMS1T, from the sedi-ment of a eutrophic freshwater pond on the campus of theDekkerswald Institute, Nijmegen, The Netherlands. This strainhas the salt tolerance characteristic of freshwater bacteria andis able to use DMS, MT, methanol, and methylamines forgrowth and methanogenesis. Characteristics of strain DMS1T

are discussed in relation to its ecological niche. Phylogeneticanalysis revealed that this strain represents a novel genus in thefamily Methanosarcinaceae.

MATERIALS AND METHODS

Source of inoculum. Samples were collected from the top layer (5 to 10 cm) ofsediment (depth, 50 cm) in a eutrophic freshwater pond on the Campus ofDekkerswald Institute, Nijmegen, The Netherlands. The sediment samples wereobtained by suction and placed in anaerobic bottles as described by Lomans et al.(18).

Slurry incubation. Sediment slurries were prepared and incubated as de-scribed previously by Lomans et al. (18–20). Bromoethanesulfonic acid (BES),sodium molybdate, sodium tungstate, DMS, and MT were added from pH-neutral anaerobic stock solutions in distilled water.

Media and culture techniques. Cultivation was carried out in 60- and 120-mlserum bottles filled with 25 and 50 ml of medium, respectively. The definedsulfide-reduced and bicarbonate-buffered medium of Widdel and Bak (37) wasmodified slightly and used for isolation. This medium contained (per liter) 1.00 gof NaCl, 0.25 g of Na2SO4, 0.25 g of NH4Cl, 0.20 g of KH2PO4, 0.50 g of KCl,0.4 g of MgCl2 z 6H2O, 0.1 g of CaCl2 z 2H2O, 0.30 g of Na2S z 7H2O, and 2.50 gof NaHCO3. Sodium sulfide was sterilized separately and added to the basalmedium 1 to 2 h before it was used. One milliliter of a trace element solutioncontaining the following compounds was added per liter of medium: nitrilotri-acetic acid (NTA) (13.0 g/liter), FeSO4 (0.12 g/liter), H3BO4 (0.30 g/liter),H2SeO3 (0.32 g/liter), KAl(SO4)2 z 12H2O (0.32 g/liter), CuSO4 z 5H2O (0.32g/liter), CoCl2 z 6H2O (0.32 g/liter), NaMoO4 z 2H2O (0.032 g/liter), NiCl2 z 6H2O(0.31 g/liter), ZnSO4 z 7H2O (0.32 g/liter), and MnCl2 z 4H2O (0.32 g/liter). NTAand FeSO4 were dissolved in 600 ml of distilled water by adding NaOH. After theother components were added to the NTA-FeSO4-NaOH solution, the final pHwas adjusted to 6.5 and the volume was adjusted to 1,000 ml with distilled water.The bottles were sealed with black butyl rubber stoppers, gassed with an O2-freeN2-CO2 mixture (80/20, vol/vol), and sterilized (121°C, 20 min).

After sterilization (121°C, 20 min) the medium was supplemented with 1 ml ofan anaerobic sterile vitamin stock solution per liter of medium; this solutioncontained (per liter) 0.1 g of p-aminobenzoate, 0.1 g of riboflavin, 0.2 g ofthiamine, 0.2 g of nicotinate, 0.5 g of pyridoxin, 0.1 g of pantothenate, 0.1 g ofcobalamin, 0.02 g of biotin, 0.05 g of folate, and 0.05 g of lipoate.

Carbon sources were added from anaerobic sterile stock solutions that hadbeen prepared in distilled water 1 to 2 h before inoculation. Immediately beforeinoculation, the medium was supplemented with small amounts of sodium di-thionite obtained from a freshly prepared stock solution (final concentration,0.08 g/liter) and with Fe(II)Cl2 (final concentration, 10 mg/liter) since this stim-ulated the growth of strain DMS1T significantly (see below). The bottles wereincubated in the dark at 30°C.

The isolate was enriched in an anaerobic chemostat that was fed with lowconcentrations of DMS (20 to 250 nmol per ml of headspace) in the N2-CO2(80/20, vol/vol) gas stream that was passed through the system (Fig. 1). Afterenrichment, strain DMS1T was isolated by the deep agar method as described byPfennig (27). The absence of contaminants was confirmed by growing a culturein the medium described above supplemented with yeast extract (0.5%) andTrypticase peptone (0.5%) and also by fluorescence microscopy. Stock cultureswere transferred monthly into fresh medium, and cultures containing mediumsupplemented with glycerol (5%) were stored in glass ampoules under an N2-CO2 atmosphere (80/20, vol/vol) at 280°C.

Determining optimal growth conditions. Specific growth rates were deter-mined by measuring the amount of methane formed during growth on methanol.The specific growth rate during exponential growth was analyzed by linear re-gression of the logarithm of the total amount of methane that accumulatedversus time. When the effects of environmental parameters (pH, temperature,and salt concentration) were tested, growth rates were determined with culturesadapted to the conditions used. We transferred cultures under these conditionsat least two times sequentially. In particular, in order to obtain a culture at ahigher osmolarity, it was necessary to transfer a culture in several steps to mediahaving progressively higher osmotic values.

Cell suspension experiments. Conversion of DMS and MT was studied byusing samples from chemostat cultures. After a cell suspension was harvestedand placed in a 120-ml serum bottle (reduced with 0.5 ml of a solution containing4 g of sodium dithionite per liter; final concentration, 17 mg/liter), 7- to 10-mlaliquots were dispensed into 60-ml serum bottles that had been reduced with 0.1

ml of 100 mM Ti(III) citrate (final concentration, 1 to 1.4 mM) (38). Thesuspensions were flushed with N2 to remove the endogenous substrates (DMSand MT) and sulfide. The cell suspensions were incubated at 32°C with shaking(100 rpm).

Analytical techniques. Gas samples (0.5 to 1.0 ml) were removed from theincubation bottles with pressure lock syringes and were analyzed to determinetheir methane, MT, and DMS contents with a Hewlett-Packard model 5890 gaschromatograph equipped with a flame ionization detector and a Porapak Qcolumn (80/100 mesh) (10). Specific determinations of sulfur compounds (H2S,MT, and DMS) in gas samples from the incubation mixtures were performedwith a Packard model 438A gas chromatograph equipped with a flame photo-metric detector and a Carbopack B HT100 column (40/60 mesh) as describedpreviously (3, 18).

Microscopy and photography. Transmission and scanning electron micro-graphs were obtained with a Philips model 201 transmission electron microscopeand a JEOL model T300 scanning electron microscope by using cells from alate-exponential-phase culture that had been fixed with glutaraldehyde (2%,wt/vol) in 50 mM sodium chloride and osmium tetroxide (1%, wt/vol) anddehydrated in absolute ethanol.

Phylogenetic analysis. Strain DMS1T DNA was isolated by crushing a cellpellet (obtained from 20 ml of culture) in liquid N2 by using a mortar and pestle(11). The homogenate was suspended in 4 ml of TE extraction buffer (10 mMTris-HCl, 1 mM EDTA; pH 7.5). After sodium dodecyl sulfate (SDS) (1%,wt/vol) and proteinase K (50 mg/ml) were added, the suspension was incubatedfor 30 min at 50°C. Then the lysate was mixed with an equal volume of coldisopropanol and centrifuged (10 min, 10,000 3 g, 4°C). The resulting pellet wasdissolved in TE extraction buffer and subjected to phenol-chloroform-isoamylalcohol (25:24:1) extraction. The pellet obtained after ethanol precipitation ofthe water phase was dissolved in 100 ml of sterile demineralized water and storedat 220°C. The DNA was used as a template for PCR amplification of approxi-mately 1,350- and 1,000-base segments of the 16S rRNA gene. The PCR condi-tions were as follows: 2.5 mM MgCl2, annealing temperature of 55°C, and 30cycles. The PCR amplification primers used were REV007 (59-GTTGATCCTGCCAGAGGYYA-39), ARC1326 (59-TGTGTGCAAGGAGCAGGGAC-39),REV915 (59-GTGCTCCCCCGCCAATTCCT-39) (29), and 23S047 (59-CCCBGGGCTTATCGCAGCTT-39) (29). The amplification products were ligated in

FIG. 1. Schematic diagram of the continuous gas flow chemostat used forenrichment and isolation of Methanomethylovorans hollandica DMS1T. Mediumwas pumped into the culture vessel, resulting in a dilution rate of 0.005 to 0.02h21. The energy and carbon source for growth was added via the gas streambubbling through the chemostat. An oxygen-free N2-CO2 gas stream was passedthrough a regulation vessel. DMS from a stock solution was pumped into thisvessel. The DMS was then sparged out of the water phase, and the DMS-containing gas was then passed through the culture vessel. The concentration ofDMS in the incoming gas could be regulated by altering the N2-CO2 gas flow orthe pump flow rate of the DMS stock solution. In this way, the dilution rate andthe concentration of DMS to which the culture was exposed could be regulatedindependently. A, culture vessel; B, medium stock preparation; C, regulationvessel; D, waste vessel; E, gas trap bottle; F, tubing connected to a DMS stocksolution; G, tubing connected to an oxygen-free N2-CO2 gas stream (0.3 atm).Shaded tubing is tubing that contained liquid. Tubing with droplets is tubing thatcontained both gas and liquid, whereas tubing without shading or droplets istubing that contained gas. Rectangles represent sterile (gas) filters. Circles rep-resent peristaltic pumps; the triangles in the circles indicate the direction ofpumping. The arrows indicate the direction of the gas flow.

3642 LOMANS ET AL. APPL. ENVIRON. MICROBIOL.

on April 11, 2021 by guest

http://aem.asm

.org/D

ownloaded from

the pCR II vector and transformed into the Escherichia coli cells of a TA cloningkit (Invitrogen). Plasmid DNA of clones were isolated by using a FlexiPrep kit(Pharmacia P-L Biochemicals Inc.). The sequences of the cloned PCR products,which represented the original 16S rRNA sequence, were analyzed with a DNAsequencer (Applied Biosystems model 373A) by using the Taq DyeDeoxy ter-minator cycle sequencing method (1, 24). Besides the primers of the TA cloningkit (M13FOR and M13REV) and primers mentioned by Raskin et al. (29)(ARC915, REV344, and MC1109), the following primers were used for sequenc-ing: REV007 (59-GTTGATCCTGCCAGAGGYYA-39), ARC1326 (59-TGTGTGCAAGGAGCAGGGAC-39), REV954 (59-TCAAGCTAAAGACTTTACCA-39), and REV1299 (59-CGGTTTCCAACATAGCGCGG-39). These primerswere designed in our laboratory and were based on known sequences of othermethanogens. Primer REV954 was designed on the basis of a partial sequence ofthe 16S rRNA gene of strain DMS1T but did not appear to be highly strainspecific. The resulting sequences were assembled to produce a 1,449-base con-tinuous DNA sequence. The deduced 16S rRNA sequence of strain DMS1T wasaligned with homologous 16S rRNA sequences of closely related members of theArchaea domain by using the Pileup method (The Dutch CAOS/CAMM CenterFacility, Nijmegen, The Netherlands). These 16S rRNA sequences were ob-tained from the GenBank/EMBL and Ribosomal Database Project databases.Distance matrix trees were constructed by using the method of Fitch and Mar-goliash (9) and the neighbor-joining method of Saitou and Nei (30) in theFITCH and NEIGHBOR programs of the PHYLIP (version 3.4) program pack-age (6). Parsimony and bootstrap parsimony analyses were performed by usingthe DNAPARS and DNABOOT programs as implemented in the PHYLIPpackage.

Nucleotide sequence accession number. The deduced, almost complete se-quence (1,449 bases) of the 16S rRNA gene of strain DMS1T has been depositedin the GenBank database under accession no. AF120163.

RESULTS

Enrichment and isolation of strain DMS1T. We examinedvarious sediment samples and used the sediment slurry from aeutrophic freshwater pond (Campus of Dekkerswald Institute,Nijmegen, The Netherlands) which exhibited the highest levelof DMS consumption (67 nmol per ml of sediment slurry zh21) (19, 20) for isolation of anaerobic DMS-consuming mi-croorganisms. After unsuccessful enrichment on DMS in ster-ilized pore water amended with Trypticase peptone (final con-centration, 0.02%), yeast extract (final concentration, 0.02%),or rumen fluid (final concentration, 1% [vol/vol]), mineral an-aerobic medium (see above) was inoculated with 10% (vol/vol)sediment slurry. Enhancement of the DMS consumption ratein dilutions of the sediment slurries was found to be verydifficult (the maximal DMS degradation rate measured was 12nmol of DMS per ml of slurry z h21), and transfers did notresult in a DMS-degrading enrichment culture. One of thediluted sediment incubation preparations was used to inocu-late a chemostat that was fed with DMS via the N2-CO2 (80/20,vol/vol) gas stream at increasing concentrations (2 to 250 nmolper ml of headspace) (Fig. 1 and 2A). The advantage of thissystem was that unused DMS and the products (MT and H2S)did not accumulate to inhibiting levels in the system, since theywere flushed out of the system. In this way, a stable chemostatculture with a low H2S concentration could be maintained.Conversion of DMS in the chemostat was revealed by thedifference between the concentration of DMS in the incominggas and the concentration of DMS in the outgoing gas and bythe presence of methane and MT in the outgoing gas (Fig. 2).Use of the chemostat (dilution rate, 0.01 h21) resulted in anenrichment culture that had a much higher DMS-degradingcapacity (6800 nmol per ml of culture fluid z h21) and con-sisted of only one morphologically distinct methanogen. At-tempts to isolate the DMS-degrading methanogen by using aliquid dilution series with DMS were unsuccessful. Eventually,cultivation of a single colony obtained from a dilution series inwhich methanol and DMS in deep agar tubes were used led toa pure culture of a DMS-degrading methanogen, which wasdesignated strain DMS1T. A microscopic and F420 fluorescencemicroscopic analysis of cultures grown on medium supple-mented with methanol (20 mM), yeast extract (0.5%), and

Trypticase peptone (0.5%) or on the same medium withoutmethanol revealed that the culture was free of contaminants.

Morphology. Deep agar colonies of strain DMS1T werewhite circular disks which reached a diameter of 2 mm in 2weeks when they were grown on methanol. Phase-contrastmicroscopy and scanning and transmission electron micro-graphs revealed that the morphology of the cells could best bedescribed as intermediate between the morphology of typicalMethanosarcina cell clusters and the morphology of the coc-coid cells characteristic of Methanolobus and Methanococ-coides species (Fig. 3). The cells occurred mainly in clustersconsisting of two or four cells, and the clusters formed largeaggregates consisting of hundreds of cell clusters. The averagediameter of individual cells was 1 to 1.5 mm. Motility was notobserved. Individual cells and cell clusters or aggregates didnot lyse within 15 min after SDS was added to a final concen-tration of 1.0 g per liter. The isolate stained gram negative.

Optimal growth conditions. Initially, the logarithmic growthphase of cultures of strain DMS1T was short and was followedby a relatively long linear methane formation phase whichindicated that no growth or slow growth occurred. To deter-mine the optimal growth conditions, the effects of vitamins,complex nutrients, trace elements, pH, temperature, and saltconcentration were tested by using cultures of strain DMS1T

growing on methanol. Addition of FeCl2 to a culture signifi-cantly stimulated the growth of strain DMS1T and resulted ina short lag phase and a longer logarithmic growth phase. Sinceaddition of other metals (NiCl2, CoCl2, and NaMoO4) or ad-

FIG. 2. (A) Time courses of the DMS concentrations in the incoming gas (■)and outgoing gas (h) in the chemostat. The chemostat was inoculated (10%)with a 10-fold-diluted sediment slurry (see text) that had been preincubated withDMS. The slurry was prepared from a eutrophic pond sediment (Campus ofDekkerswald Institute). (B) Time courses (start-up phase) of the amount ofDMS consumed (Œ) and the amounts of MT (F) and methane (}) produced bya chemostat inoculated with a pure culture of strain DMS1T.

VOL. 65, 1999 METHANOGENIC CONVERSION OF DMS AND MT 3643

on April 11, 2021 by guest

http://aem.asm

.org/D

ownloaded from

FIG. 3. Electron micrographs of cell clusters in a late-logarithmic-phase culture of Methanomethylovorans hollandica DMS1T grown on methanol. (A) Scanningelectron micrograph clearly showing large aggregates of cell clusters consisting of two to four irregular coccoid cells. (B) Transmission electron micrograph of afreeze-etched sample. Bar, 1 mm.

3644

on April 11, 2021 by guest

http://aem.asm

.org/D

ownloaded from

dition of larger amounts of vitamin and trace element solutions(see above) did not result in additional stimulation of thegrowth of strain DMS1T, only FeCl2 was added in all subse-quent experiments. The FeCl2-amended cultures were charac-terized by large quantities of a black precipitate consisting ofFeS and FeS2. To avoid unnecessary large quantities of theprecipitates in cultures of strain DMS1T, we determined theminimal concentration of FeCl2 which resulted in maximalstimulation of growth (Fig. 4). Addition of FeCl2 to a finalconcentration of 10 mg/liter resulted in maximal stimulation ofstrain DMS1T growth.

Strain DMS1T grew optimally between pH 6.5 and 7.0. Thegrowth rate was low (,0.01 h21) or there was no growth incultures having pH values below 6.0 and above 8.0. The opti-mum temperature was about 34 to 37°C, and no growth wasobserved at temperatures above 40°C. Strain DMS1T exhibitedoptimal growth at NaCl concentrations of 0 to 40 mM. The salttolerance range of the strain was 0 to 300 mM. At salt concen-trations above 400 mM no growth was observed. The toleranceof strain DMS1T to high concentrations of sulfide was tested bygrowing the strain on methanol in the presence of varioussulfide concentrations, and these experiments revealed thatsulfide concentrations of 8 mM and higher affected the growthof strain DMS1T. Similarly, we tested the tolerance of strainDMS1T to high DMS concentrations by adding various con-centrations of DMS to cultures growing on methanol. Themaximal growth rate of strain DMS1T on methanol remainedunaffected at DMS concentrations up to 20 mM (the highestconcentration tested); however, cultures containing 10 and 20mM DMS had dramatically longer lag phases than culturescontaining 0, 2.5, and 5 mM DMS had.

Catabolic substrates. Strain DMS1T used methanol, DMS,MT, monomethylamine, dimethylamine, and trimethylaminefor growth and methanogenesis but did not use H2-CO2 oracetate. Strain DMS1T could not completely reduce MT orDMS with H2. In the presence of methanol, cultures had a lagphase of about 50 h (Fig. 5A). A prolonged lag phase that wasabout 250 h long was observed when methanol-grown cellswere transferred into monomethylamine-, dimethylamine-, ortrimethylamine-containing medium (Fig. 5A). However, afterrepeated transfers to media containing these substrates, the lagphase was shortened to 50 h (Fig. 5B). The maximum growthrates of strain DMS1T in batch cultures containing methanoland batch cultures containing methylamines were 0.04 to 0.06and 0.03 to 0.05 h21, respectively. Growth on MT and DMS(maximum growth rates, 0.005 to 0.02 h21) occurred only inbatch or chemostat cultures in which DMS was added via theN2-CO2 gas stream as described above. Degradation of DMS

was characterized by the presence of methane and MT in theoutgoing gas stream of the chemostat (Fig. 2B).

The concentration of H2S in a chemostat culture appearedto have a dramatic impact on the conversion of DMS and MTand thus on the formation of methane. Addition of sulfide toeither the gas stream or the reaction vessel itself (final con-centration, 1 mM; less than 50% of the sulfide concentrationin methanol-containing culture medium) instantaneously re-sulted in a dramatic increase in the MT concentration in theoutgoing gas, and this was followed by a complete collapse ofthe formation of both MT and methane (Fig. 6A and B).Addition of 1 ml of a similar medium that was anaerobic butnot sulfide reduced resulted in increases in both the MT con-centration and methane production (Fig. 6C).

Although accurate stoichiometric analysis of degradation ofDMS by strain DMS1T in the chemostat appeared to be diffi-cult, formation of methane, MT, and H2S (detected in theoutgoing gas) (Fig. 2B) indicated that DMS was probably de-graded in a manner similar to the manner described for othermethanogens (with MT as an intermediate), as described bythe following equation:

(CH3)2S 1 H2O3 0.5CO2 1 1.5CH4 1 H2S

We were not able to estimate the stoichiometric formation ofH2S because of the formation of black sulfide precipitates withiron and other metal ions present in the medium.

Cell suspension studies. To study the conversion of DMS bystrain DMS1T in more detail, samples of the culture fluid in thechemostat were frequently removed and used for cell suspen-sion experiments performed in serum bottles. In these incuba-tion experiments, DMS was converted to MT and methane

FIG. 4. Effect of adding iron (FeCl2) on the maximum growth rates of cul-tures of Methanomethylovorans hollandica growing on methanol.

FIG. 5. Growth curves for strain DMS1T grown on methanol (20 mM) (■),monomethylamine (20 mM) (F), dimethylamine (20 mM) (Œ), and trimethyl-amine (20 mM) (}). (A) Cultures inoculated with a methanol-grown culture. (B)Cultures after two transfers onto medium containing the appropriate substrate.

VOL. 65, 1999 METHANOGENIC CONVERSION OF DMS AND MT 3645

on April 11, 2021 by guest

http://aem.asm

.org/D

ownloaded from

(Fig. 7). Accumulation of MT was transient, and at lower DMSconcentrations MT was converted to methane. Addition of2-[bis(2-hydroxyethyl)amino]ethanesulfonic acid (BES) com-pletely inhibited MT and methane production from DMS. In-cubation of DMS-grown cells from the chemostat with eitherDMS or methanol resulted in immediate production of meth-ane, whereas addition of DMS to methanol-grown cells did not(data not shown).

In the chemostat culture, sulfide had a dramatic impact onMT and methane formation. To determine whether this wascaused by the dynamics of methyl transfer reactions or by toxiceffects, DMS-degrading and methanol-degrading cell suspen-sions were amended with MT or sulfide. A methanol-contain-ing cell suspension was included as a control to check for thetoxic effects of MT and sulfide. The effects of MT and sulfide

on DMS degradation differed dramatically from the effects onmethanol degradation. Addition of MT (1.5 to 3.6 mM) toDMS-degrading cell suspensions resulted in significant de-creases in DMS degradation (Fig. 8B). In contrast, addition ofMT (1.5 to 3.6 mM) to the same cell suspensions amended withmethanol revealed that MT at these concentrations was nottoxic (Fig. 8C). Surprisingly, addition of MT to methanol-containing cell suspensions stimulated both MT degradationand methane formation (Fig. 8B and C). The MT degradationwas accompanied by a strong accumulation of DMS (Fig. 8A).Apparent formation of DMS from MT and methanol has alsobeen observed in sediment slurry preparations (20).

Addition of sodium sulfide (1.2 to 2.9 mM) resulted in slightdecreases in DMS degradation and methane formation inDMS-containing preparations only 10 h after the addition.More MT accumulated in these preparations than in the con-trol preparations containing DMS. Incubation of DMS-growncells with methanol and various concentrations of Na2S re-vealed that like MT, sulfide was not toxic at concentrationsbelow 4 mM.

Phylogenetic and taxonomic analysis. Two DNA fragmentsof about 1,000 and 1,350 bases long that were homologous tothe rRNA gene of strain DMS1T were amplified in vitro,cloned in the pCR II vector and Escherichia coli, and partiallysequenced. Chimera Check analysis of the Ribosomal Data-base Project data (22) revealed that the strain DMS1T 16SrRNA gene sequence was derived from a single target DNAsequence. Database searches for homologous sequences of 16SrRNA genes of other organisms revealed that the highest sim-ilarity value (94.5%) was obtained with the sequences of strainDMS1T and Methanococcoides burtonii. A phylogenetic treebased on a matrix of binary phylogenetic distances whichwere calculated from the alignment of archaeal 16S rRNAsequences clearly showed that strain DMS1T clusters withinthe family Methanosarcinaceae (Fig. 9). In addition to the 16SrRNA sequences of closely related methanogens, all knownsequences of other DMS-degrading methanogens were incor-porated in this tree. The topology of the tree was reevaluatedby varying the positions included on the basis of the degrees ofconservation and by applying alternative treeing methods (par-simony and bootstrap parsimony analysis).

DISCUSSION

Isolation and physiology of strain DMS1T. The newly iso-lated methanogen strain DMS1T is the first obligately anaero-bic archaeon which has been directly enriched and isolated

FIG. 6. Effects of adding sulfide and adding anaerobic but not sulfide-re-duced medium on DMS degradation in chemostat cultures of strain DMS1T.Sulfide was added either in the gas stream (A) or to the culture vessel (1 mM)(B). Also, 1 ml of nonreduced medium was added to the culture vessel (C). Thearrows indicate when compounds were added. The concentrations of methane(h) and MT (■) in the effluent gas from the chemostat were determined.

FIG. 7. Conversion of DMS (}) and MT (Œ) and formation of methane (■)in 9-ml cell suspension samples from the chemostat incubated in 60-ml serumbottles under an N2 atmosphere after DMS was added.

3646 LOMANS ET AL. APPL. ENVIRON. MICROBIOL.

on April 11, 2021 by guest

http://aem.asm

.org/D

ownloaded from

from a freshwater sediment in a defined minimal mediumcontaining DMS as the sole carbon and energy source. Incontrast to most previously described isolation proceduresused for saline environments (5, 8, 12, 16, 17, 23, 32) strainDMS1T could not be obtained from dilution series of batchenrichment cultures. Using a chemostat with a continuousDMS-containing gas stream as a method of enrichment, fol-lowed by cultivation in deep agar tubes, resulted in a pureculture. The success of the continuous gas flow chemostat wasprobably due to maintenance of low sulfide and MT concen-trations in the culture, which enhanced DMS conversion (seebelow). Apparently, DMS degradation by methanogens in sa-line environments is affected less by MT accumulation andsulfide accumulation. This may be due to the higher in situconcentrations of these compounds which the methanogensexperience in these habitats (14, 16). Strain DMS1T occurredas large aggregates of cell clusters that usually consisted of twoto four irregularly shaped cocci, which resembled the cocci ofMethanosarcina species. Cells of strain DMS1T do not have a

protein-containing cell wall, since the cells did not lyse afteraddition of SDS (1.0 g per liter). Like all other known DMS-degrading methanogens isolated from marine, estuarine, andsalt lake sediments, strain DMS1T disproportionated DMS tomethane, carbon dioxide, and sulfide with MT as an interme-diate (8, 16). Strain DMS1T differs from all other DMS-de-grading methanogens, since it was isolated from a freshwaterpond and required NaCl concentrations (0 to 0.04 M) that aretypical of the NaCl concentrations required by freshwater mi-croorganisms for growth. Since the only substrates utilized bystrain DMS1T are methanol, methylamines, MT, and DMS,this organism is considered an obligately methylotrophic meth-anogen like most other DMS-degrading methanogens.

Incubation experiments performed with cell suspensions ob-tained from chemostat cultures revealed that strain DMS1T

cannot completely reduce MT or DMS with H2, as has beenreported for methanol reduction by cultures of Methano-sphaera stadtmanae and Methanosphaera cuniculi (2). Com-pared to growth on methanol (growth rate, 0.04 to 0.06 h21)

FIG. 8. Conversion of DMS and MT and formation of methane by 9-ml cell suspension samples taken from the chemostat amended with methanol (A, C, and E)or DMS (B, D, and F) and incubated in 60-ml serum bottles. Symbols: ■, controls containing methanol or DMS; Œ, samples containing methanol plus MT or DMSplus MT; }, samples containing methanol plus Na2S or DMS plus Na2S.

VOL. 65, 1999 METHANOGENIC CONVERSION OF DMS AND MT 3647

on April 11, 2021 by guest

http://aem.asm

.org/D

ownloaded from

and methylamines (growth rate, 0.03 to 0.05 h21), growth onDMS and MT was slow (growth rate, 0.005 to 0.02 h21). Thesegrowth rates were obtained only in batch or chemostat culturesin which DMS was supplied in a continuous gas stream. Chemo-stat experiments revealed that accumulation of MT or sulfidesignificantly affected DMS degradation. This was confirmed bythe inhibition of DMS degradation by MT observed in cellsuspension experiments (Fig. 8). Transient accumulation ofDMS in MT-containing cultures also indicated that transfer ofthe first methyl group during DMS degradation is a reversibleprocess. This was confirmed by the formation of DMS in sus-pensions of strain DMS1T cells containing methanol plus MT.Transient accumulation of DMS in MT-containing cultures hasalso been reported by other authors (8). Accumulation of MTand sulfide in batch cultures makes the methyl transfer ener-getically less favorable. This may explain the lack of success inprevious attempts to enrich DMS-degrading methanogensfrom freshwater sediments.

Production of the enzymes needed for degradation of meth-ylamines has to be induced, since inoculation of methanol-grown cells of strain DMS1T into mono-, di-, or trimethyl-amine-containing medium resulted in long initial lag phases(Fig. 5). These lag phases were shorter after repeated transferson these substrates. Addition of methanol to DMS-grown cellsresulted in immediate formation of methane, indicating thatproduction of the enzymes needed for methanol conversion isconstitutive.

Phylogenetic and taxonomic analysis. In its physiologicalproperties, strain DMS1T resembles previously described meth-anogenic archaea belonging to the genera Methanolobus, Meth-anohalophilus, and Methanococcoides, the genera to which mostof the other DMS- and MT-degrading methanogens belong.Like the methanogens belonging to these genera, strain DMS1T

is an obligate methylotroph. In contrast to all of the membersof these genera, however, strain DMS1T has the low salttolerance typical of nonhalophilic microorganisms. Since themorphology of strain DMS1T resembles the morphology ofmethanogens belonging to the genera Methanolobus and Meth-anococcoides, as well as the genus Methanosarcina, a precisephylogenetic analysis of the 16S rRNA gene of strain DMS1T

had to be performed in order to classify strain DMS1T. On thebasis of its position on the phylogenetic tree, we concluded thatstrain DMS1T is clearly not related to the genus Methanosar-cina. The results of the bootstrap analysis, as indicated by thebootstrap values on the phylogenetic tree, revealed, however,that the branching of the genera Methanolobus, Methanococ-coides, and Methanohalophilus and strain DMS1T is not welldefined. A comparison of the distance between two speciesbelonging to one genus (e.g., Methanococcoides burtonii andMethanococcoides methylutens) with the distance between aspecies belonging to one of the three genera and strain DMS1T

clearly showed that strain DMS1T does not belong to thesegenera. According to a combined analysis of both the 16SrRNA gene and the MCRI gene (31) of methanogens, levels ofsimilarity of 98% or higher represent interspecies relation-ships, whereas levels of similarity of 90 to 95% represent in-tergeneric relationships.

On the basis of its low level of similarity with the most close-ly related methanogen, Methanococcoides burtonii (94.5%), itsposition on the phylogenetic tree, its morphology (which isdifferent from the morphology of members of the genera Meth-anolobus, Methanococcoides, and Methanohalophilus), and itssalt tolerance and optimum (which are characteristic of fresh-water bacteria), we propose that strain DMS1T is a represen-tative of a novel genus and species. This organism was namedMethanomethylovorans hollandica.

Ecological niche of Methanomethylovorans hollandica. Likethe concentrations of MT and DMS in marine, estuarine, andsalt lake sediments, the concentrations of MT and DMS infreshwater sediments are low due to the balance between theformation and degradation of these compounds (18). Degra-dation of MT and DMS occurs anaerobically due to the steepoxygen gradient at the water column-sediment interface, whichresults in anaerobic conditions in the sediment (19). Variousinhibition studies performed with the specific inhibitors BESand sodium tungstate revealed that most (about 95%) of theendogenously produced MT in freshwater sediments is con-verted by methanogenic archaea (20, 40, 41). Microscopic anal-ysis of DMS-containing sediment slurries with a fluorescencemicroscope revealed the presence of methanogens which weremorphologically identical to Methanomethylovorans hollandica,the organism described in this study. The enrichment condi-tions used resembled the in situ conditions and therefore madeit plausible that Methanomethylovorans hollandica DMS1T isthe most important DMS and MT consumer in its naturalfreshwater environment. This conclusion is supported by thephysiological properties of the organism (substrate use, lowgrowth rate, and iron requirement). Whether Methanomethy-lovorans hollandica is also an important utilizer of methanoland methylamines in situ remains to be investigated. The ob-ligately methylotrophic archaea are able to form a stable com-munity along with the acetoclastic and hydrogenotrophic meth-anogens.

FIG. 9. Phylogenetic tree based on a distance matrix prepared from an align-ment of partial 16S rRNA sequences (1,404 bases) of Methanomethylovoranshollandica and closely related methanogens. Reference sequences were obtainedfrom the GenBank, EMBL, and Ribosomal Database Project databases. Meth-anospirillum hungateii was used as the outgroup. Scale bar, 10 base substitutionsper 100 bases. The names of methanogens which are able to utilize DMS as acarbon source are underlined.

3648 LOMANS ET AL. APPL. ENVIRON. MICROBIOL.

on April 11, 2021 by guest

http://aem.asm

.org/D

ownloaded from

Due to the reversibility of the DMS conversion, methano-gens like strain DMS1T can also be involved in the formationof DMS through methylation of MT. This phenomenon, whichpreviously has been shown to occur in sediment slurries with afreshwater origin (19), might affect the steady-state concentra-tion and consequently the total flux of DMS (and MT) in thesesystems. Moreover, stimulation of the production of MT andDMS by high concentrations of sulfide (18) also appears toinhibit degradation of these compounds. Therefore, highersteady-state concentrations and fluxes of MT and DMS insulfide-rich freshwater systems are conceivable.

Description of Methanomethylovorans hollandica gen. nov.,sp. nov. Methanomethylovorans hollandica (Me.tha.no.me.thy.lo9vo.rans. M. L. n. methanum, methane; M. L. n. methylum,methyl; L. adj. vorans, devouring; M. L. n. Methanomethylovo-rans, methane producing, methyl group consuming; hol.lan9di.ca. L. adj. hollandica, from The Netherlands [Holland], refer-ring to the origin of the type strain). Cells are irregular, non-motile, coccoid (diameter, 1 to 1.5 mm), and gram negative.Cells normally occur in clusters consisting of two to four cellswhich form large aggregates. Cells are not sensitive to lysis by1.0 g of SDS per liter. Trimethylamine, dimethylamine, mono-methylamine, methanol, DMS, and MT are catabolic sub-strates, but H2-CO2 and acetate are not. Growth is most rapidin the presence of 0 to 0.04 M NaCl, and no growth occurs atNaCl concentrations higher than 0.4 M. Optimal growth occursat pH 6.5 to 7.0, and no growth occurs at pH values lower than6.0 and higher than 8.0. Growth is most rapid at 34 to 37°C, andvery slow or no growth occurs at temperatures below 12°C andabove 40°C. Type strain DMS1 was isolated from a slurryprepared from a eutrophic pond sediment (Campus of Dek-kerswald Institute, Nijmegen, The Netherlands). Strain DMS1T

has been deposited in the culture collection of the DeutscheSammlung von Mikroorganismen (Braunschweig, Germany).

ACKNOWLEDGMENTS

We thank Wim Willems of the Technical Service Department, Uni-versity of Nijmegen, for constructing the chemostat headplate, RonHochstenbach and Wander Sprenger for discussions related to the 16SrRNA sequence analysis, and Erik van Wesel for assistance with elec-tron microscopy.

This work was supported by The Netherlands Organization for theAdvancement of Pure Research (NWO) as part of the program “Ver-storing van Aardsystemen.”

REFERENCES

1. Applied Biosystems, Inc. 1992. Taq DyeDeoxy terminator cycle sequencingkit user bulletin no. 901497, revision E. Applied Biosystems, Inc., FosterCity, Calif.

2. Biavati, B., M. Vasta, and J. G. Ferry. 1988. Isolation and characterization of“Methanosphaera cuniculi” sp. nov. Appl. Environ. Microbiol. 54:768–771.

3. Derikx, P. J. L., H. J. M. Op den Camp, C. van der Drift, L. J. L. D. vanGriensven, and G. D. Vogels. 1990. Odorous sulfur compounds emittedduring production of compost used as a substrate in mushroom cultivation.Appl. Environ. Microbiol. 56:176–180.

4. De Zwart, J. M. M., P. N. Nelisse, and J. G. Kuenen. 1996. Isolation andcharacterization of Methylophaga sulfidovorans sp. nov.: an obligate methyl-otrophic aerobic, DMS oxidizing bacterium from a microbial mat. FEMSMicrobiol. Ecol. 20:261–271.

5. Elberson, M. A., and K. R. Sowers. 1997. Isolation of an aceticlastic strain ofMethanosarcina siciliae from marine canyon sediments and emendation ofthe species description for Methanosarcina siciliae. Int. J. Syst. Bacteriol. 47:1258–1261.

6. Felsenstein, J. 1982. Numerical methods of inferring evolutionary trees. Q.Rev. Biol. 57:379–404.

7. Finster, K., G. M. King, and F. Bak. 1990. Formation of methyl mercaptanand dimethyl sulfide from methoxylated aromatic compounds in anoxic ma-rine and freshwater sediments. FEMS Microbiol. Ecol. 74:295–302.

8. Finster, K., Y. Tanimoto, and F. Bak. 1992. Fermentation of methanethioland dimethylsulfide by a newly isolated methanogenic bacterium. Arch.Microbiol. 157:425–430.

9. Fitch, W. M., and E. Margoliash. 1967. Construction of phylogenetic trees:a method based on mutation distances as estimated by cytochrome c se-quences is of general applicability. Science 155:279–284.

10. Hutten, T. J., M. H. de Jong, B. P. H. Peeters, C. van der Drift, and G. D.Vogels. 1981. Coenzyme M derivatives and their effects on methane forma-tion from carbon dioxide and methanol by cell extracts of Methanosarcinabarkeri. J. Bacteriol. 145:27–34.

11. Jarrell, K. F., D. Faguy, A. M. Hebert, and M. L. Kalmokoff. 1992. A generalmethod of isolating high molecular weight DNA from methanogenic archaea(archaebacteria). Can. J. Microbiol. 38:65–68.

12. Kadam, P. C., D. R. Ranade, L. Mandelco, and D. R. Boone. 1994. Isolationand characterization of Methanolobus bombayenis sp. nov., a methylotrophicmethanogen that requires high concentrations of divalent cations. Int. J. Syst.Bacteriol. 44:603–607.

13. Kiene, R. P., and T. S. Bates. 1990. Biological removal of dimethyl sulfidefrom sea water. Nature 345:702–705.

14. Kiene, R. P., and S. K. Service. 1991. Decomposition of dissolved DMSP andDMS in estuarine waters: dependence on temperature and substrate con-centration. Mar. Ecol. Prog. Ser. 76:1–11.

15. Kiene, R. P., and M. E. Hines. 1995. Microbial formation of dimethyl sulfidein anoxic Sphagnum peat. Appl. Environ. Microbiol. 61:2720–2726.

16. Kiene, R. P., R. S. Oremland, A. Catena, L. G. Miller, and D. G. Capone.1986. Metabolism of reduced methylated sulfur compounds in anaerobicsediments and by a pure culture of an estuarine methanogen. Appl. Environ.Microbiol. 52:1037–1045.

17. Liu, Y., D. R. Boone, and C. Choy. 1990. Methanohalophilus oregonense sp.nov., a methylotrophic methanogen from an alkaline, saline aquifer. Int. J.Syst. Bacteriol. 40:111–116.

18. Lomans, B. P., A. Smolders, L. Intven, A. Pol, H. J. M. Op den Camp,and C. van der Drift. 1997. Formation of dimethyl sulfide and methane-thiol in anoxic freshwater sediments. Appl. Environ. Microbiol. 63:4741–4747.

19. Lomans, B. P., H. J. M. Op den Camp, A. Pol, and G. D. Vogels. 1999.Anaerobic versus aerobic degradation of dimethyl sulfide and methanethiolin anoxic freshwater sediments. Appl. Environ. Microbiol. 65:438–443.

20. Lomans, B. P., H. J. M. Op den Camp, A. Pol, and G. D. Vogels. 1999. Therole of methanogens and other bacteria in the degradation of dimethylsulfide and methanethiol in anoxic freshwater sediments. Appl. Environ.Microbiol. 65:2116–2136.

21. Lovelock, J. E., R. J. Maggs, and R. A. Rasmussen. 1972. Atmosphericdimethyl sulfide and the natural sulfur cycle. Nature 237:452–453.

22. Maidak, B. L., J. R. Cole, C. T. Parker, Jr., G. M. Garrity, N. Larsen, B. Li,T. G. Lilburn, M. J. McCaughey, G. J. Olsen, R. Overbeek, S. Pramanik,T. M. Schmidt, J. M. Tiedje, and C. R. Woese. 1999. A new version of theRDP (Ribosomal Database Project). Nucleic Acids Res. 27:171–173.

23. Mathrani, I. M., D. R. Boone, R. A. Mah, G. E. Fox, and P. P. Lau. 1988.Methanohalophilus zhilinae sp. nov., an alkaliphilic, halophilic, methylotro-phic methanogen. Int. J. Syst. Bacteriol. 38:139–142.

24. McBride, L. J., S. M. Koepf, R. A. Gibbs, W. Salser, P. E. Mayrand, M. W.Hunkapiller, and M. N. Kronick. 1989. Automated DNA sequencing meth-ods involving polymerase chain reaction. Clin. Chem. 35:2196–2201.

25. Ni, S., and D. R. Boone. 1991. Isolation and characterization of a dimethylsulfide-degrading methanogen, Methanolobus siciliae HI350, from an oil well,characterization of M. siciliae T4/MT, and emendation of M. siciliae. Int. J.Syst. Bacteriol. 41:410–416.

26. Oremland, R. S., and J. P. Zehr. 1986. Formation of methane and carbondioxide from dimethylselenide in anoxic sediments and by a methanogenicbacterium. Appl. Environ. Microbiol. 52:1031–1036.

27. Pfennig, N. 1978. Rhodocyclus purpureus gen. nov. and sp. nov., a ring-shaped, vitamin B12-requiring member of the family Rhodospirillaceae. Int. J.Syst. Bacteriol. 28:283–288.

28. Rajagopal, B. S., and L. Daniels. 1986. Investigations of mercaptans, organicsulfides, and inorganic sulfur sources for the growth of methanogenic bac-teria. Curr. Microbiol. 14:137–144.

29. Raskin, L., J. M. Stromley, B. E. Rittmann, and D. A. Stahl. 1994. Group-specific 16S rRNA hybridization probes to describe natural communities ofmethanogens. Appl. Environ. Microbiol. 60:1232–1240.

30. Saitou, N., and M. Nei. 1987. The neighbor-joining method: a new methodfor reconstructing phylogenetic trees. Mol. Biol. Evol. 4:406–425.

31. Springer, E., M. S. Sachs, C. R. Woese, and D. R. Boone. 1995. Partial genesequences for the A subunit of methyl-coenzyme M reductase (mcrI) as aphylogenetic tool for the family Methanosarcinaceae. Int. J. Syst. Bacteriol.45:554–559.

32. Stetter, K. O. 1989. Genus II. Methanolobus Konig and Stetter 1983, 439,p. 2205–2207. In J. T. Staley, M. P. Bryant, N. Pfennig, and J. G. Holt (ed.),Bergey’s manual of systematic bacteriology, vol. 3. The Williams & WilkinsCo., Baltimore, Md.

33. Tanimoto, Y., and F. Bak. 1994. Anaerobic degradation of methyl mercaptanand dimethyl sulfide by newly isolated sulfate-reducing bacteria. Appl. En-viron. Microbiol. 60:2450–2455.

34. Visscher, P. T., P. Quist, and H. van Gemerden. 1991. Methylated sulfurcompounds in microbial mats: in situ concentrations and metabolism by a

VOL. 65, 1999 METHANOGENIC CONVERSION OF DMS AND MT 3649

on April 11, 2021 by guest

http://aem.asm

.org/D

ownloaded from

colorless sulfur bacterium. Appl. Environ. Microbiol. 57:1758–1763.35. Visscher, P. T., and B. F. Taylor. 1993. A new mechanism for the aerobic

catabolism of dimethyl sulfide. Appl. Environ. Microbiol. 59:3784–3789.36. Visscher, P. T., B. F. Taylor, and R. P. Kiene. 1995. Microbial consumption

of dimethyl sulfide and methanethiol in coastal marine sediments. FEMSMicrobiol. Ecol. 18:145–154.

37. Widdel, F., and F. Bak. 1992. Gram-negative mesophilic sulfate-reducingbacteria, p. 3352–3378. In A. Balows, H. G. Truper, M. Dworkin, W. Harder,and K. H. Schleifer (ed.), The prokaryotes, 2nd ed., vol. IV. Springer-Verlag,New York, N.Y.

38. Zehnder, A. J. B., and K. Wuhrmann. 1976. Titanium(III) citrate as a

non-toxic oxidation-reduction buffering system for the culture of obligateanaerobes. Science 194:1165–1166.

39. Zeyer, J., P. Eicher, S. G. Wakeham, and R. P. Schwarzenbach. 1987. Oxi-dation of dimethyl sulfide to dimethylsulfoxide by phototrophic purple bac-teria. Appl. Environ. Microbiol. 53:2026–2032.

40. Zinder, S. H., and T. D. Brock. 1978. Methane, carbon dioxide, and hydrogensulfide production from the terminal methiol group of methionine by anaer-obic lake sediments. Appl. Environ. Microbiol. 35:344–352.

41. Zinder, S. H., and T. D. Brock. 1978. Production of methane and carbondioxide from methane thiol and dimethyl sulfide by anaerobic lake sedi-ments. Nature 273:226–228.

3650 LOMANS ET AL. APPL. ENVIRON. MICROBIOL.

on April 11, 2021 by guest

http://aem.asm

.org/D

ownloaded from