carbon cycling by cellulose-fermenting nitrogen-fixing bacteria

4
Adv. Space Res. VoL 9, No. 8. pp (8)149—(8)152. 1989 0273—1177/89 S0.0O + .50 Printed in Great Britain. All rights reserved. Copyright © 1989 COSPAR CARBON CYCLING BY CELLULOSE- FERMENTING NITROGEN-FIXING BACTERIA S. B. Leschine and E. Canale-Parola Department of Microbiology, University of Massachusetts, Amherst, MA 01003, U.S.A. ABSTRACF The most abundant organic materials on Earth are plant polysaccharides such as cellulose and hemicelluloses. Inasmuch as vast quantities of these polymers are present in anaerobic environments (e.g., in soils and sediments), anaerobic microorganisms that ferment plant polysaccharides play a central role in carbon cycling on the planet as a source of CO 2 and, indirectly, of CH4. Cellulose-fermenting bacteria from soil and pond sediment were isolated in a medium (incubated in N2) which lacked a source of combined nitrogen. The isolates had the ability to utilize atmospheric N2 as the nitrogen source for cell growth. Nitrogenase (the enzyme which catalyzes the reduction of N2 to ammonia) was demonstrated by means of the acetylene reduction test in these isolates and in several previous- ly described anaerobic cellulolytic bacteria isolated from various natural environments. Thus, cellulose-fermenting bacteria that fix N2 may be widespread and may play a role in nitrogen cycling as well as in carbon cycling on a global scale. Knowledge of the physiology and ecology of these organisms is crucial to detailing the mechanisms producing local sources and sinks of atmospheric gases, interpreting data obtained using space-based sensors, and understanding the effects of atmospheric warming on fermentations as major sources of CO2 and CH4. INTRODUCTION Key processes of biogeochemical cycles are mediated by microorganisms. For example, an important step in the global carbon cycle is the microbial degradation of cellulose, the most abundant organic material on Earth /1/. Photosynthesis yields annually up to 1.5 x 1011 tons of dry plant material worldwide, almost halfof which consists of cellulose /2/. For life to continue on Earth, the carbon present in this polymer must be reintroduced into the atmosphere in the form of CO2. A substantial amount of cellulose (5-10%) is degraded in anaerobic environments /2/. Anaerobic activity occurs in proximity to the surface in soils, composts, and freshwater, marine, and estuarine sediments, indicating that aerobic conditions normally prevail only in a thin crust /2/. In anaerobic environments, cellulose is initially decomposed by cellulose-fermenting microorganisms yielding CO2. H2, organic acids, and ethanol. Some products of cellulose fermentation serve as growth substrates for other bacteria which produce acetate, C02, and H2. These products are then converted to CH4 by methanogenic bacteria. Thus, as a source of CO2 and, indirectly, of CH4, anaerobic cellu- lolytic microorganisms play a major role in carbon cycling on the planet (Fig. 1). Although the microbiota involved directly and indirectly in cellulose degradation in the rumen has been studied extensively /3,4/, relatively little is known about the complex interactions among free-living microorganisms involved in the anaerobic degradation of cellulose in other natural habitats /2/. The distribution of anaerobic cellulolytic bacteria in nature is largely unknown. Due primarily to recent interest in the potential use of cellulolytic bacteria for alcohol fuel production from biomass, several species of anaerobic cellulolytic bacteria from sediments, compost, and sewage sludge have been de- scribed /5-1 1/. The long-term goal of our research is to advance understanding of the physiology and ecology of these microorganisms. Mostly, we have studied strains of a Clostridium species (referred to as “C strains”) that we isolated from the sediment of a freshwater swamp. These strains actively ferment not only cellulose but also components of the hemicellulosic portion of biomass (e.g., xylan, pentoses), forming primarily ethanol, acetic acid, H2, and CO2 /5/. Information obtained from these studies may help us predict the response of cellulose-fermenting bacteria to current global warming and probable altered precipitation patterns. For example, if anaerobic cellulose (8)149

Upload: sb-leschine

Post on 21-Jun-2016

214 views

Category:

Documents


1 download

TRANSCRIPT

Adv. Space Res. VoL 9, No. 8. pp (8)149—(8)152. 1989 0273—1177/89 S0.0O + .50Printedin Great Britain. All rights reserved. Copyright© 1989 COSPAR

CARBON CYCLING BY CELLULOSE-FERMENTINGNITROGEN-FIXINGBACTERIA

S. B. LeschineandE. Canale-Parola

DepartmentofMicrobiology, UniversityofMassachusetts,Amherst,MA 01003,U.S.A.

ABSTRACF

Themost abundantorganicmaterialson Earth areplant polysaccharidessuchas celluloseandhemicelluloses.Inasmuchasvastquantitiesof thesepolymersarepresentin anaerobicenvironments(e.g.,in soilsandsediments),anaerobicmicroorganismsthat fermentplant polysaccharidesplay acentralrole in carboncycling on theplanetasasourceof CO2 and,indirectly,of CH4. Cellulose-fermentingbacteriafromsoil andpondsedimentwereisolatedinamedium(incubatedin N2) whichlackedasourceof combinednitrogen. The isolateshadtheability to utilizeatmosphericN2 asthenitrogensourcefor cell growth. Nitrogenase(theenzymewhichcatalyzesthereductionofN2to ammonia)wasdemonstratedby meansoftheacetylenereductiontestin theseisolatesandin severalprevious-ly describedanaerobiccellulolytic bacteriaisolatedfromvariousnaturalenvironments.Thus,cellulose-fermentingbacteriathat fix N2 maybewidespreadandmayplay arole in nitrogencycling aswell asin carboncycling on aglobalscale. Knowledgeof thephysiologyandecologyof theseorganismsis crucial to detailingthemechanismsproducinglocal sourcesandsinksof atmosphericgases,interpretingdataobtainedusingspace-basedsensors,andunderstandingtheeffectsof atmosphericwarmingonfermentationsasmajorsourcesof CO2andCH4.

INTRODUCTION

Key processesof biogeochemicalcyclesaremediatedby microorganisms.For example,animportantstepin theglobal carboncycleis the microbialdegradationof cellulose,themost abundantorganicmaterialon Earth/1/.Photosynthesisyieldsannuallyup to 1.5x 1011 tonsof dryplantmaterialworldwide,almosthalfof whichconsistsof cellulose/2/. For life to continueon Earth,thecarbonpresentin thispolymermust bereintroducedinto theatmospherein theformof CO2.

A substantialamountof cellulose(5-10%)is degradedin anaerobicenvironments/2/.Anaerobicactivity occursinproximity to thesurfacein soils,composts,andfreshwater,marine,andestuarinesediments,indicating thataerobicconditionsnormallyprevailonly in athin crust/2/. In anaerobicenvironments,celluloseis initially decomposedbycellulose-fermentingmicroorganismsyielding CO2.H2, organicacids,andethanol. Someproductsof cellulosefermentationserveasgrowthsubstratesfor otherbacteriawhichproduceacetate,C02,andH2. Theseproductsarethenconvertedto CH4by methanogenicbacteria.Thus,asasourceofCO2 and,indirectly,of CH4, anaerobiccellu-lolytic microorganismsplayamajorrolein carboncyclingon theplanet(Fig. 1). Althoughthemicrobiotainvolveddirectly andindirectly in cellulosedegradationin therumenhasbeenstudiedextensively/3,4/, relativelylittle isknownaboutthecomplexinteractionsamongfree-livingmicroorganismsinvolvedin theanaerobicdegradationofcellulosein othernaturalhabitats/2/. Thedistributionof anaerobiccellulolyticbacteriain natureis largelyunknown.

Dueprimarily to recentinterestin thepotentialuseofcellulolytic bacteriafor alcoholfuelproductionfrombiomass,severalspeciesof anaerobiccellulolytic bacteriafrom sediments,compost,andsewagesludgehavebeende-scribed/5-11/. Thelong-termgoalof ourresearchis to advanceunderstandingof thephysiologyandecologyofthesemicroorganisms.Mostly, wehavestudiedstrainsof aClostridium species(referredto as“C strains”)thatweisolatedfrom the sedimentof afreshwaterswamp. Thesestrainsactively fermentnot only cellulosebut alsocomponentsofthehemicellulosicportion ofbiomass(e.g.,xylan,pentoses),formingprimarily ethanol,aceticacid,H2, andCO2/5/. Informationobtainedfrom thesestudiesmayhelpus predicttheresponseof cellulose-fermentingbacteriato currentglobalwarmingandprobablealteredprecipitationpatterns.For example,if anaerobiccellulose

(8)149

(8)150 S. B. LeschineandE. Canale-Parola

degradationis enhancedandtherateof CO2 andCH4 productionis increased,therewill be apositive feedbackontherateofclimatechangeby this microbialactivity /12,13/.

Environmentsrichin cellulosearefrequentlydeficientin nitrogen(e.g.,peatsoils, agriculturalwastes,composts).Thus,cellulose-fermentingbacteriathatsatisfytheir nitrogenrequirementsthroughthefixation of N2 would beex-pectedto haveastrongselectiveadvantageoverthosewhichrequireasourceof combinednitrogen.Moreover,suchmicroorganismsmight beexpectedto playamajorrolein nitrogencycling,aswell asin carboncycling,on aglobalscale. However,it hasnot beendeterminedwhethercelluloseis widely usedasan energysourceby nitrogen-fixingbacteria.This is surprisinggiventhe abundanceof celluloseand the suggestion/14/that nitrogen fixation by free-living heterotrophicbacteria in naturalecosystemsis limited by the availability of oxidizablegrowth substrates.Waterburyand coworkers/15/haveshownthat celluloseservesasagrowthsubstratefor anitrogen-fixingaerobicbacteriumwhichexistsin asymbioticrelationshipwith shipworms.Their findingsdemonstratethat thesetwocom-plex physiologicalprocesses,nitrogenfixation and cellulosedegradation,canbe performedby asinglebacterium.Amajor objectiveof our researchwasto determinewhether free-living anaerobiccellulolytic bacteria that are wide-spread in terrestrialenvironmentsfix nitrogenwhen they utilize celluloseas the fermentable substrate for growth.

~

~ H 0’‘~‘2 ~ ~1?. ~

c~ C~f~

] T C0~CH4

6 12 6 x

1. Cellulose 4 OrganIc Acids • Ethanol• CO2 • H2

2. organic Acids. Ethanol 4 Acetate + CO2 + H2

~ fAcetate 4 CO2 + CH4

1,,C02.H2 4 CH4

CELLULOSE 4 Co2 + Cl!4Fig. 1. Roleof cellulose-fermentingbacteria in carboncycling. Cellulose [(C6Hi2O6)~],amajor product ofthephotosynthetic fixation ofCO2. is fermented by anaerobiccellulolytic bacteria(step 1). Ethanol andtheacidicproducts of cellulosefermentation serveas growthsubstratesfor otherbacteriawhich produceacetate,CO2.andH2(step2). MethanogenicbacteriaproduceCH4 from acetateorby reductionof CO2 with H2 (step3).In anaerobicenvironments,thecompletedissimilationofcelluloseresultsin theformationofCO2 andCH4.

ISOLATIONOF NITROGEN-FIXING CELLULOSE-FERMENTINGBACTERIA

Four strains of anaerobiccellulolytic bacteriawere isolatedfrom forest soil andfreshwatermud usingaprocedurethat selectedfor nitrogen-fixing strains/16/. Enrichmentcultureswereprepared by seriallydiluting soil or mudsamplesinto anaerobicculturetubescontainingaliquid growthmedium,designated“MW-C”. Thismedium(whichwas similar to one describedby Daeschand Mortenson /17/) lacked a source of combined nitrogen, includedcellulose(ball-milled WhatmanNo. 1 filter paper;0.6%,thywt/vol) asthefermentablesubstrate,wasprereduced/18/, andwasmaintainedin anN2 atmosphere.After7-14daysof incubationat 30°C,enrichmentculturesshowedsignificantdisappearanceof cellulose. Spent mediumand remainingcellulosefibersfromenrichmentcultureswereseriallydilutedinto meltedcellulosesoft agarmediumin tubes.Thecontentsof thesetubeswerepouredOntoplatesof agar medium within an anaerobic chamber. After 2-4weeksof incubation, coloniessurroundedby zonesofclearingappearedin theotherwiseopaquemedium. Thesecolonieswere transferredby streakingonto platesofcellobioseagarmedium, and restreakedseveraltimes to obtain pure cultures.Colonieswere then transferredintoliquid cellulosemedium (MW-C) to confirm whether the isolatewascellulolytic. Three strains (B1A, B1B, andB lC) were isolatedfrom mud from thebottomof ashallowpond (Beaver’sPond,Shutesbury,Massachusetts)andone (strain B3B) from forest soil nearBeaver’sPond.

CHARACTERIZATION OF NITROGEN-FIXINGCELLULOLYTIC ISOLATES

The isolatesresembledone anothermorphologically (Fig.2). All weremotilecurved rodsmeasuring0.6 to 0.8 jimx 3 to 6 jim. Under thegrowthconditions used,sporeswere never observedeither within cellsor free in culturesupematant fluids. Cellsof all four strainsstainedgram-negative.Electronmicroscopyof thin sectionsof strainBIA cells /16/showedthatthe cytoplasmicmembranewassurrounded by amultilayered cell wall thatdifferedfromtypicalcell wadsof mostgram-negativeanaerobicbacteria(eg.,thoseof mostBacreroidaceae/19/),but resembledthoseof other gram-negativemesophiic cellulolytic bacteria /5,20/. All isolateswereobligatelyanaerobicandfermentedpolysaccharides,hexoses,andpentosesthatarecommonly present in plantmaterials.Noneof thestrainsutilized maltose, glycerol, or amino acids as fermentablesubstrates. Further phenotypicand genotypiccharacterizationoftheisolatesis requiredto determinetheirtaxonomicposition.

CarbonCycling (S)151

i’

‘S

/

‘: ~:‘~‘~k~ ~

/ ~ l~t.• ~ ~ —

/ ~‘r~,’ ~j•;_” ~

‘.•, ~ ..~. ~

FIG. 2 A-C. Phase-contrastphotomicrographsof nitrogen-fixingcellulolytic isolates (wet-mountpre-parations),strains(A) BIA, (B)BIB, (C) B 1C,and(D) B3B. Cellswereculturedto late-exponentialphaseina cellobiose-containingmedium. Cells of strainsBIC (C) and B3B (D) areentangledin cellulosefibersintroducedinto theculturealongwith theinoculum(a7-dayculture in cellulosemedium).Scalebar,10 p.M.

Thepresenceof nitrogenasein the isolateswasdemonstratedby meansof the acetylenereductiontest/21/. Cellsreducedacetyleneto ethylene(e.g., 1310nmol/h/mgcell protein)when growingin mediumMW-C but not whengrowingin thesamemediumsupplementedwith 0.2%NH4C1. Inasmuchastheisolatesgrewin amediumlackingcombinednitrogenandcontainingcelluloseasthefermentablesubstrate,it wasconcluded/16/ thattheyutilized thispolysaccharideastheenergysourcefor N2fixation.

N1TROGENASEACTIVITY IN OTHERCELLULOSE-FERMENTINGBACTERIA

Severalpreviouslydescribedanaerobiccellulolytic species[Clostridium strainC7, from mud of afreshwaterswamp in Woods Hole, Massachusetts,U.S.A. /5/;Clostridium papyrosolvens,from estuarinesedimentsof theRiver Don in Aberdeenshire,Scotland/6/; strainJW-2, from wetwoodof an elm in Amherst,Massachusetts,U.S.A. 120/] were culturedin a defmed,cellulose-containingmedium/16/. Thepreviouslydescribedcellulolyticspecies,aswell asthenewisolates,reducedacetyleneto ethylenewhengrownin this medium,butnot whengrownin thesamemediumsupplementedwith NH4CI, aresultthat indicatedthepresenceof nitrogenasein thesebacteria.Moreover,growth in a cellobiose-containingmediumlacking combinednitrogenwasdependenton N2 /16/. Nogrowthoccurredwhencultureswereincubatedin anargonatmosphere.TheseobservationsindicatedthatN2 servedasthenitrogensourcefor thegrowingcells. Thus,cellulolytic bacteriafrom avarietyof environmentssynthesizedactive nitrogenaseandapparentlyincorporatedN2 into cell materialduringgrowth. Although othercellulolyticbacteriamaypossessnitrogenase/15,22/,this is the first demonstrationof nitrogenfixation duringthe anaerobicdegradationofcellulose. Ourfinding of nitrogenaseactivity in cellulose-fermentingbacteriaisolatedfrom avarietyof environments is consistentwith the view /16/thatnitrogen-fixingcellulolytic bacteriaarewidespreadin nature.

DISCUSSION

In the study describedabove,we did not attempt to enumeratenitrogen-fixing cellulolytic bacteriain specificterrestrialenvironments.Theresultsof suchexperiments,if performedusingclassicalmethodsinvolving viablecounts,probablywould not bemeaningfulbecausethe~bacteriaoftenadhereto or areentangledin cellulosefibers(e.g., Fig. 2C,D). Furthermore,the growthrequirementsof diversenitrogen-fixingcellulolytic bacteriaare notknown. Futurestudiesaimedat identifying cellulolytic speciesin their naturalenvironmentsandestimatingtheiractivitieswill beconductedwith theuseof oligonucleotideprobescomplementaryto ribosomalribonucleicacid(rRNA) sequencesuniqueto membersof groupsof cellulose-fermentingorganisms.Sincetheprobesarecomple-mentaryto rRNAs,andactively growingcellsmaycontainmore than l0~ribosomes/23/, eachapotentialprobetarget,singlecellsmaybelabeledandidentifiedby in situ hybridizationandmicroautoradiographicprocedures/24/.This methodpermitstheidentificationof singlecellsmicroscopically(cultivation is not required)andoffersaverypowerful tool to identifythepresenceofmembersof particularbacterialgroupsin naturalecosystems.Themetabolicactivity ofpopulationsof cellulolyticbacteriain naturalsamplesmaybeestimatedby measuringtheamountofprobeboundto bulk extractedrRNA. Sincethe rRNAcontentof cells is proportionalto growthrateoverawide rangeofgrowthrates/25/, theamountof group-specificprobehybridizedperunit of biomasswouldprovide anestimateofthe metabolicactivity of thatgroup/24,26/. Resultsof thesestudieswill contributeto ourunderstandingof thediversityof cellulose-fermentingnitrogen-fixingbacteria,their quantitativecontributionto cellulosebreakdownandnitrogenfixation, their potentialfor improvingsoil fertility, andtheeffectof atmosphericwarmingon their activity.

ACKNOWLEDGMENTS

We thankT.A. Warnickfor experttechnicalassistance.This work wassupportedby NationalScienceFoundationgrantBSR-8708469andby U. S. Departmentof EnergycontractDE-FGO2-88ER13898.

(3)152 S. B. Leschineand E. Canale-Parola

REFERENCES

1. E. B. CowlingandW. Brown,Structuralfeaturesofcellulosicmaterialsin relationto enzymatichydrolysis,in: AdvancesinChemistrySeries95.ed.0.3. Hajny andE. T.Reese,AmericanChemicalSociety,Washington1969,p. 152.

2. L. 0. LjungdahlandK.-E. Erilcsson,Ecologyofmicrobialcellulosedegradation.in: Advancesin MicrobialEcolo~v.Vol. 8.ed. K. C. Marshall,PlenumPress,New York 1985, p. 237.

3. R.a Hungate, TheRumenanditsMicrobes,AcademicPress,New York, 1966.

4. M. I. Wolin andT.L. Miller, Interactionsof microbialpopulationsin cellulosefermentation,Fed. Proc.42, 109(1983)

5. S. B. LeschineandE. Canale-Parola,Mesophiliccellulolytic clostridiafrom freshwaterenvironments,Aool. Environ

.

Microbiot 46, 728 (1983)

6. ft. H. Madden,M. 3. Bryder, andN.J.Poole,Isolationandcharacterizationofananaerobic,cellulolytic bacterium,CIos:ridiwnpapyrosolvens sp. nov.,mt. J. Syst. Bacteriol. 32, 87 (1982)

7. W. D. Murray,L. Hofmann,N.L. Campbell,andR.H. Madden,Clostridium lentocellum sp. nov., acellulotytic speciesfromriversedimentcontainingpaper-millwaste,System.Ail~LMicrobipl. 8, 181 (1986)

8. W. D. Murray,L. C. Sowden,andJ.R. Colvin,Bacteroides cellulosolvens sp. nov., acellulolytic speciesfrom sewagesludge,mt. 3. Svst. BacterioL34, 185 (1984)

9. G. B. Patel,A. W. Khan,B. J. Agnew, and3. R. Colvin, Isolationandcharacterizationof ananaerobic,cellulolyncmicro-organism,Acetivibrio celiulolyticus gen. nov., sp. nov., mt. J. Syst.Bacteriol.30, 179 (1980)

10. E. Petitdemange,F. Caillet, 3.Giallo, andC. Gaudin,Clostridium cellulolyzicum sp. nov., acellulolytic mesophilicspeciesfrom decayedgrass,Int. J. Svst.Bacteriol.34, 155 (1984)

11. 3. Sukhumavasi,K. Ohmiya,S. Shimizu,andK. Ueno,Clostridiwn josui sp. nov., acellulolytic, moderatethermophilicspeciesfrom Thai compost,hit. I. Syst. Bacteripl.38, 179 (1988)

12. J. E. HobbleandJ. M. Melillo, Roleof microbesin globalcarboncycling,in: CurrentPersoectivesin Microbial Ecolpsv,ed.M. 3. KIug andC. A. Reddy,AmericanSociety for Microbiology,Washington1984,p389.

13. A. E. Linkins, I. M. Meillo, andft. L. Sinsabaugh,Factorsaffectingcellulaseactivity in terrestrialandaquaticecosystems.in: CurrentPersoectivesin Microbial Ecology,ed.M~I. KIug andC. A. Reddy,AmericanSocietyfor Microbiology,Washington1984, pSl2.

14. ft. Brouzes,J. Lasik.andR.Knowles,Theeffect oforganicamendment,watercontent,andoxygenon theincorporationof15N

2by someagriculturalandforestsoils,Can. 3. Microbiol. 15, 899(1969)

15. 1. B. Waterbury,D.B. Calloway,andR. D.Turner,A cellulolytic nitrogen-fixingbacteriumculturedfrom theglandof Deshayes

inshipworms(Bivalvia: Teredinidae),Science221, 1401 (1983)

16. S.B. Leschine,K. HolweJl,andE. Canale.Parola,Nitrogen fixationby anaerobiccellulolytic bacteria,Science,in press.

17. G.DaeschandL. E. Mortenson,Effectof ammoniaon thesynthesisand functionoftheN2-fixing enzymesystemin Closrridiwn

pasteurianum,I. Bacteriol. 110,103(1972)

18. ft. E.Hungate,A roll tubemethodfor cultivation ofstrict anaerobes,in: Methodsin Microbiolosy. Vol. 3B, ed. 3. ft. NorrisandD.W, Ribbons,AcademicPress,New York 1969,p. 117.

19. L. V. Holdeman,R.W.Kelley.andW. E. C. Moore,Anaerobicgram-negativestraight.curvedandhelicalrods,in: Bergev’sManualof SystematicBacteriology.Vol. 1, ed.N. ft. Krieg and1.0. HoIt, Williams andWilkins, Baltimore 1984, p. 602.

20. 1. E. Warshaw,S. B. Leschine,andaCanale-Parola,Anaerobiccellulolytic bacteriafrom wetwoodof living treeS,AppLEnviron.Microbiol. 50,807 (1985)

21. 3. ft. Postgate,Theacetylenereductiontestfor nitrogenfixation, in: Methods in Microbiology. Vol. 6B, ed. I. R.Norris andD. W. Ribbons,AcademicPress,New York 1972, p. 343.

22. M. BogdahnandD. Kleiner, Inorganicnitrogenmetabolismin two cellulose-degradingclostridia,Arch. MicrobioL 145, 159(1986)

23. H. BremerandP.P.Dennis,Modulationof chemicalcompositionandotherparametersof thecell by growthrate,in:Escherjchiacpu andSalmonellazvvlthnuriwn. CellularandMolecularBiology.VoL 2, ad.F. C. Neidhasdt,AmericanSocietyfor Microbiology, Washington1987,p.

1527.

24. 5.1. Giovannoni,E. F.DeLong,G. I. Olsen,andN. R.Pace,Phylogeneticgroup-specificoligodeoxynucleotideprobesforidentificationofsingle microbialcells,I. BacterioL170, 720(1988)

25. L. 3. Ingraham,0. Maaloe,aridF. C. Neidhardt,Growthof the BactenalCell, SinauerAssociates,Sunderland,Mass.,USA.1983.

26. D. A. Stahl,B. Flesher,H.R. Mansfield,andL. Montgomery,Useofphylogeneticallybasedhybridizationprobesfor studiesofruminalmicrobialecology,App!. Environ. Microbiol. 54, 1079(1988)