Robust Mercury Methylation across Diverse Methanogenic Archaea

13 Pages·2017·1.7 MB·English

RESEARCH ARTICLE
crossm
Robust Mercury Methylation across Diverse Methanogenic
Archaea
CynthiaC.Gilmour,aAllysonL.Bullock,aAlyssaMcBurney,aMirceaPodar,b DwayneA.Eliasb
aSmithsonianEnvironmentalResearchCenter,Edgewater,Maryland,USA D
bBiosciencesDivision,OakRidgeNationalLaboratory,OakRidge,Tennessee,USA o
w
n
ABSTRACT Methylmercury (MeHg) production was compared among nine cultured lo
a
methanogenic archaea that contain hgcAB, a gene pair that codes for mercury (Hg) d
methylation. The methanogens tested produced MeHg at inherently different rates, e
d
even when normalized to growth rate and Hg availability. Eight of the nine tested f
r
were capable of MeHg production greater than that of spent- and uninoculated- o
m
medium controls during batch culture growth. Methanococcoides methylutens, an
h
hgcAB(cid:2) strain with a fused gene pair, was unable to produce more MeHg than con- t
t
p
trols. Maximal conversion of Hg to MeHg through a full batch culture growth cycle :
/
for each species (except M. methylutens) ranged from 2 to (cid:3)50% of the added Hg(II) /m
or between 0.2 and 17 pmol of MeHg/mg of protein. Three of the species produced b
io
(cid:3)10% MeHg. The ability to produce MeHg was confirmed in several hgcAB(cid:2) meth- .
a
anogens that had not previously been tested (Methanocella paludicola SANAE, s
m
Methanocorpusculum bavaricum, Methanofollis liminatans GKZPZ, and Methanosphae-
.
o
rula palustris E1-9c). Maximal methylation was observed at low sulfide concentra- r
g
tions ((cid:4)100 (cid:2)M) and in the presence of 0.5 to 5 mM cysteine. For M. hollandica, the /

o
addition of up to 5 mM cysteine enhanced MeHg production and cell growth in a n

concentration-dependent manner. As observed for bacterial Hg methylators, sulfide J
a
inhibited MeHg production. An initial evaluation of sulfide and thiol impacts on bio- n
u
availability showed methanogens responding to Hg complexation in the same way a
r
as do Deltaproteobacteria. The mercury methylation rates of several methanogens ri- y

val those of the better-studied Hg-methylating sulfate- and iron-reducing Deltapro- 6
,

teobacteria. 2
0
IMPORTANCE Archaea, specifically methanogenic organisms, play a role in mercury 1
9
methylationinnature,buttheirglobalimportancetoMeHgproductionandthesub- b
y
sequent risk to ecosystems are not known. Methanogenesis has been linked to Hg
g
methylation in several natural habitats where methylmercury production incurs risk u
e
to people and ecosystems, including rice paddies and permafrost. In this study, we s
t
confirm that most methanogens carrying the hgcAB gene pair are capable of Hg
methylation. We found that methylation rates vary inherently among hgcAB(cid:2) meth-
Received19January2018 Accepted13
anogens but that several species are capable of MeHg production at rates that rival March2018 Published10April2018
those of the better-know Hg-methylating sulfate- and iron-reducing bacteria. Meth- CitationGilmourCC,BullockAL,McBurneyA,
anogens may need to be considered equally with sulfate and iron reducers in evalu- PodarM,EliasDA.2018.Robustmercury
methylationacrossdiversemethanogenic
ationsofMeHgproductioninnature.
Archaea.mBio9:e02403-17.https://doi.org/10
.1128/mBio.02403-17.
KEYWORDS Archaea,methylmercury,bioavailability,complexation,cysteine,hgcAB, EditorDerekR.Lovley,Universityof
mercury,methyltransferase,methylation,thiols MassachusettsAmherst
Copyright©2018Gilmouretal.Thisisan
open-accessarticledistributedundertheterms
The discovery of the hgcAB gene pair led to the identification of several groups of oftheCreativeCommonsAttribution4.0
Internationallicense.
microorganisms not previously identified as capable of mercury methylation, in-
AddresscorrespondencetoCynthiaC.
cluding methanogens. With the exception of a few syntrophic and fermentative
Gilmour,[email protected].
Firmicutes(4),demonstratedmercury(Hg)methylatorsareallstrictlyanaerobicbacteria
®
March/April2018 Volume9 Issue2 e02403-17 mbio.asm.org 1
®
Gilmouretal.
that use sulfate, iron, and carbon dioxide as terminal electron acceptors. Several
hgcAB(cid:2) putative methylators have been identified in the phyla Bacteroidetes, Chloro-
flexi, and Nitrospirae (2) and other phyla, but none of these have been confirmed to
producemethylmercury(MeHg)inculture.Overall,~210microbialspeciescontaining
hgcAB have been identified among all of the bacteria and Archaea whose genomes
havebeensequenced,numbering(cid:3)10,000.
The first list of hgcAB(cid:2) organisms was published in 2013 (3), with subsequent
updates(2,4).Inallcases,thelisthasbeenbasedoninsilicoscreeningforbothhgcA
and hgcB by using the encoded amino acids to eliminate genomic variability such as
codon wobble and species-specific nucleotide bias. Screening for hgcA included the
presenceofthedistinctiveconservedmotifNVWCAAGKintheactivemethyltransferase
site and predicted C-terminal transmembrane helices. The ferredoxin-encoding gene
hgcBisgenerally,butnotalways,immediatelydownstreamfromhgcA.Allbuttwoof D
o
the hgcAB(cid:2) methanogens identified are in one class, Methanomicrobia. However, the w
overall importance of methanogens to MeHg production in nature is not well under- n
lo
stood.Hgmethylationhasbeenconfirmedincultureforonlyfourorganismstodate, a
d
Methanospirillum hungatei (5), Methanomethylovorans hollandica, Methanolobus tin- e
darius(4),andMethanomassiliicoccusluminyensis(2).Here,weexaminedtheHgmeth- d

f
ylationratesofseveralstrainstounderstandhowtheirMeHgproductioncomparesto ro
thatofthebetter-studiedmethylatorsintheDeltaproteobacteria. m

It is clear that there are several habitats, like rice paddies (1), some freshwater h
t
t
sediments (6), and periphyton (7, 8), where methanogens play an important role in p
:
MeHg production. However, their global importance to MeHg production and the //
m
subsequentrisktoecosystemsarenotknown.Asurveyofpublicallyavailablemetag- b
enomes showed that methanogens may also be dominant Hg methylators in termite io
.
hindguts,permafrostpeat,Arcticsediments,andsomebioreactors(2).Earlyworkwith a
s
clade-specific primers for hgcAB(cid:2) methanogens, Deltaproteobacteria, and Firmicutes m
.
suggested that archaeal Hg methylators are important in freshwater sediments from o
r
EastForkPoplarCreek,TN(9,10). g
/
Inthisstudy,severalculturedrepresentativesofhgcAB(cid:2)methanogenswereusedto o
n
understand the extent to which these organisms can produce MeHg and to begin to
J
evaluatethebiogeochemicalcontrolsonHgmethylationbymethanogens.Specifically, a
n
we asked whether hgcAB(cid:2) methanogens have the ability to methylate significant u
a
portionsoftheavailableHg,howmethylationratesvaryamongspeciesandmetabo- r
y
lisms, and how the methylation rates of methanogens compare to those of Deltapro- 6
teobacteriaandFirmicutes.WeexaminedhowtheabilitytomethylateHgisclustered ,
2
withintheArchaea,phylogeneticallyandmetabolically. 0
1
Last, we began to evaluate the Hg complexes that are available for uptake and 9

methylation by hgcAB(cid:2) methanogens. We wanted to know if the same types of Hg b
y
complexesthatareavailabletothebest-knowngroupofHgmethylators,sulfate-and g
u
iron-reducing Deltaproteobacteria, are also available to Hg-methylating methanogens.
e
Smallthiolsappeartobemostbioavailableformethylationbysulfate-reducingbacteria s
t
(SRB)andiron-reducingbacteria(FeRB)(11,12).SulfidereducesthebioavailabilityofHg
to Hg-methylating Deltaproteobacteria via precipitation of mercuric sulfides that are
lessavailableforuptake(13–15).Dissolvedorganicmatter(DOM)canslowtheprecip-
itationofnanoparticulateHgS(HgS )(16–18),enhancingbioavailabilitytoSRB(13,14).
np
ThecharacteristicsofDOM,includingitsaromaticity(19),thiolcontent,anddegreeof
sulfidization(20),allinfluencehowDOMimpactsHgandHgS bioavailabilitytoSRB.
np
UnderstandingtherangeofmethylationratesandthebiogeochemicalcontrolsonHg
methylation by methanogens is a critical part of understanding the role of methano-
gensinMeHgproductionandriskinnature.
RESULTS
Methylationresults.WequantifiedandcomparedtheMeHgproductionof9ofthe
now 19 identified hgcAB(cid:2) methanogens, all Methanomicrobia (see Table S1 in the
supplementalmaterial).ThespeciestestedincludedpreviouslyuntestedhgcAB(cid:2)meth-
March/April2018 Volume9 Issue2 e02403-17 mbio.asm.org 2
®
MercuryMethylationRatesamongMethanogens
D
o
w
n
lo
a
d
e
d

f
r
o
m

h
t
t
p
:
/
/
m
b
io
.
a
s
m
.
o
r
g
/

o
n

J
a
n
u
a
r
y

6
,

FIG1 MaximalconversionofinorganicHgtoMeHgduringbatchculturegrowthbyeachhgcAB(cid:2)methanogen 2
0
tested,expressedtwoways.(Top)MeHgasapercentageofthetotalmeasuredHgintheculturemedium(both 1
unfiltered) at the end of log growth. Green and blue bars show noncellular control values. (Bottom) MeHg 9

normalizedtoproteinmeasuredattheendoftheloggrowthphase.Notethelogscales.Barsareaveragesof b
triplicateseparateculturesorcontrols.Methylationwasmeasuredfroma1nM201Hg(II)spike.Errorbarsarebased y

onthestandarddeviationofMeHgonly.M.luminyensisdataarefromreference2. g
u
e
s
t
anogens (Methanocella paludicola SANAE, Methanocorpusculum bavaricum, Methano-
follisliminatansGKZPZ,Methanococcoidesmethylutens,andMethanosphaerulapalustris
E1-9c)pluspreviouslyreportedmethylatorsMethanospirillumhungateiJF-1(5),Metha-
nomethylovoranshollandica,Methanolobustindarius(4).Additionally,thebioavailability
ofHgsulfideandHgthiolcomplexestomethanogenswastestedforselectedspecies.
FullinformationisprovidedinDataSetS1.
Figure1(top)providesacomparisonofthemaximummeasuredMeHgproduction
levelsofallnineofthehgcAB(cid:2)methanogenstested.Methylationwasassayedinbatch
cultures.MeHgconcentrationsweremeasuredonceculturesenteredstationaryphase,
andpercentageofMeHgformationwascalculatedonthebasisofthetotalmeasured
Hg concentration in the culture medium at the same time. Controls included MeHg
productioninfreshandspentmediaforeachstrain.Severalorganismsweretestedwith
differentconcentrationsofsulfideorcysteineinculturemedium;thedatainFig.1are
March/April2018 Volume9 Issue2 e02403-17 mbio.asm.org 3
®
Gilmouretal.
methylationratesunderthemostfavorableobservedconditions,usuallybetween0.5
and4mMcysteineand10and100(cid:2)Msulfide.
Eight of the nine species tested were capable of more MeHg production than
controls during batch culture growth. Maximal conversion of Hg to MeHg for each
strainthroughafullbatchculturegrowthcyclerangedfrom2to~65%.Threeofthe
species produced (cid:3)10% MeHg, M. luminyensis, M. palustris, and M. hungatei. Using
lowersulfideconcentrations,M.tindariusandM.hollandicaproducedsomewhathigher
percentagesofMeHgthanpreviouslyreported.MeHgproductionbyM.hungatei(43%)
was similar to that reported by Yu et al. (5). Methanococcoides methylutens, a species
withafusedhgcABgenepair,didnotproducemoreMeHgthancontrolsintriplicate
assaysrepeatedthreetimes.
TherewassignificantlossofHgfromtheaqueousphaseinmanyofthemethylation
D
assays.Hglosseswerecomparableincontrols(spentandfreshmedia)andcultures(see
o
Data Set S1). Differences in Hg loss appear to be related to medium chemistry, with w
n
greaterlossesoftenobservedinhigher-sulfidemedia.However,nosystematictestsof lo
medium chemistry effects on Hg losses was done, nor were the mechanisms of loss a
d
(bottlesorptionversusreductiontoHg°)systematicallyevaluated. e
d
AlthoughthepercentageofMeHgisaneasilyunderstandablewaytoevaluatethe
f
r
magnitudeofMeHgproductionbycultures,abetterwaytoreportandcomparerates o
m
amongorganismsandconditionsismethylationnormalizedtogrowth(Fig.1,bottom).

h
Protein-normalized MeHg production by the Hg-methylating methanogens ranged t
t
p
from roughly 1 to 20 pmol of MeHg/mg of protein, similar to that in prior assays :
/
/
ofM.tindariusandM.hollandica.Yuetal.(5)reportedabout100pmolofMeHg/mgof m
proteinforM.hungateiduringlog-phasegrowth,whichwasreducedto(cid:4)10pmolof b
io
MeHg/mgofproteininthepresenceofasulfidereductant. .
a
Protein-normalized MeHg production by Deltaproteobacteria and Firmicutes, mea- s
m
sured by the same batch culture approach, ranged from about 1 to 600 and 1 to .
o
50 pmol of MeHg/mg of protein, respectively, with Geobacter bemidjiensis and Desul- r
g
fovibrio sp. strain ND132 the highest producers to date (4). A survey of eight hgcAB(cid:2) /
o
Desulfovibrio species produced between about 20 and 250 pmol of MeHg/mg of n

protein in short-term (3 h) washed cell assays with 500 (cid:2)M cysteine and no added J
a
sulfide(21). n
u
MethylationbymethanogenicArchaeaincomparisontoBacteria.Tocompare a
r
methylationabilitiesamongcladesandacrossstudieswithdifferentgrowthconditions y

6
andHglevels,wecomparedabsoluteconversionofHgtoMeHg(percentageofMeHg; ,

2
Fig.2,top)andalsonormalizedMeHgproductionforeachspeciestobothproteinand 0
1
Hg (Fig. 2, bottom). In a past study, we found that methylation rates in washed-cell
9
assayswereproportionaltotheinorganicHg(II)concentration(21).Forthiscomparison, b
y
wechosethehighestmeasuredmethylationrateacrossalloftheconditionstestedfor
g
each species measured, usually between 0.5 and 4 mM cysteine and 10 and 100 (cid:2)M u
e
sulfide.Figure2compilesdatafromthisstudy,ourpriorbatchcultureassaysofallthree s
t
clades(4),andwashed-cellassaysofDesulfovibriospecies(21).Wedidnotcorrectthe
methylation rates for time, although batch cultures were often incubated for several
days while washed cell assays ran 3 h. Our observation has been that most MeHg
productioninculturesoccursinthefirsthoursofincubationwithactivecells(12,21).
Nevertheless, the differences in timing, culture conditions, and Hg concentrations
amongassaysadduncertaintytothiscomparison.
On average, across all of the species tested, Deltaproteobacteria produced the
highestaveragepercentageofMeHgandthehighestpercentageofMeHgnormalized
to protein. There are few hgcAB(cid:2) Deltaproteobacteria that convert (cid:4)10% of Hg(II) to
MeHg under ideal conditions (moderate thiol levels, low sulfide levels) during a few
hoursoflog-phasegrowth.Of13speciestested,4areabletoconvertmorethanhalf
of the available Hg(II) to MeHg. However, normalized to protein and Hg, MeHg
production by methanogens rivals that of Deltaproteobacteria. While methanogens
tended to convert a smaller fraction of Hg(II) to MeHg during batch culture growth,
March/April2018 Volume9 Issue2 e02403-17 mbio.asm.org 4
®
MercuryMethylationRatesamongMethanogens
D
o
w
n
lo
a
d
e
d

f
r
o
m

h
t
t
FIG 2 Comparison of maximal MeHg production among the three major clades of Hg-methylating p
:
bacteria.(Top)MaximalconversionofinorganicHgtoMeHgby13Deltaproteobacteria,7Firmicutes,and //
m
10 Methanomicrobia species. (Bottom) MeHg production normalized to both the measured Hg and
b
protein concentrations at the end of methylation assays, for the same organisms and assays, as
(MeHg/Hg)/(mgofprotein/liter).Hgspikelevelsrangedfrom1to50nM.Datacombineassaysdoneover io
.
thecourseofbatchculturegrowthinthisstudy,inthestudydescribedinreference4,andduring3-h a
s
washed-cellassays(21).Inthesebar-and-whiskerplots,theaveragevalueforeachspeciesisrepresented m
byadot;thecenterline,top,andbottomofeachbararethemedianandupperandlowerquartiles, .
respectively,acrossallspecies;andthewhiskersaredrawntothefurthestdatapointwithin1.5(cid:5)the o
r
quartilevalue(thusexcludingoutliers). g
/

o
n

J
theircelldensitiesandgrowthratesincultureareusuallylowerthanthoseofsulfate a
n
and iron reducers. Expressed as the percentage of MeHg per milligram of protein, u
a
severalmethanogensexhibitedmethylationratesequaltothoseofthebestDeltapro- r
y
teobacteria. Including this study, the methylation rates of about the same number of
6
Deltaproteobacteria (13) and Methanomicrobia (9) have been tested quantitatively in ,
2
culture.SRBandFeRBhavebeentestedinseverallabs(4,12,15,21–37),butonlyone 0
1
hgcAB(cid:2)methanogenhasbeentestedincultureoutsideourgroup(5). 9

About 40 hgcAB(cid:2) Firmicutes have been identified to date, but only 7 have been b
y
testedformethylationinculturesofar(4).Almostallproduced(cid:4)10%MeHginculture g
andauniformlylowpercentageofMeHgpermilligramofprotein(Fig.2).Themajority u
e
ofthesearesulfiteandsulfatereducers. s
t
Influence of cysteine on methylation by methanogens. The presence of small
thiolsoftenenhancesthebioavailabilityofHgformethylationtoDeltaproteobacteria.
WeevaluatedHgmethylationbyM.hollandicaandM.tindariusacrosscysteinegradi-
ents to begin to assess the bioavailability of Hg thiols to hgcAB(cid:2) methanogens.
Methanogens often require a source of reduced S for growth. Addition of cysteine,
other thiols, or sulfide can stimulate growth, complicating the interpretation of Hg
methylationdata.Inthisstudyweusedmethionine,whichisnotastrongligandforHg,
to provide reduced sulfur for growth. The amino acid methionine contains S in a
thioethersidechainbutnotafreethiol.Byprovidinganalternativereduced-Ssource,
we hoped to test the influence of cysteine on Hg bioavailability, separately from its
influenceoncellgrowth.
Cysteine enhanced the production of MeHg by both cultures. For M. hollandica,
MeHg production was tested in batch culture across a cysteine gradient of 1 to
5,000 (cid:2)M, with 1 mM methionine added to the medium. However, even when
March/April2018 Volume9 Issue2 e02403-17 mbio.asm.org 5
®
Gilmouretal.
D
o
w
n
lo
a
d
e
d

f
r
o
m

h
t
t
p
:
/
/
m
b
io
.
a
s
m
FIG3 ImpactofcysteineonM.hollandicagrowthandMeHgproduction.(Top)GrowthassessedbyOD .o
andmethaneproduction.(Bottom)MeHgasapercentageofthetotalHgintheculturemediumand r
g
MeHgnormalizedtoprotein.Allmeasurementsweremadeonceallculturesreachedstationaryphase /
(312h).Methionineat1mMwasincludedinallculturemedia.Sulfidewasaddedtoallat10(cid:2)M,but o
theconcentrationwasto2to4(cid:2)Mattheendoflog-phasegrowth.Notethelogscales.Errorbarsare n

standarddeviationsoftriplicatecultures. J
a
n
u
a
methionine was provided, cysteine up to 500 (cid:2)M enhanced cell growth in a con- r
y
centration-dependentmanner(Fig.3).Growthwasinhibitedat5mMcysteine.Methi- 6
onine (1 mM) as the primary reduced-S source produced less growth than 500 (cid:2)M ,
2
cysteine (in comparison with a prior assay [4]). In prior experiments, M. hollandica 0
1
grown in batch culture methylation assays with 500 (cid:2)M cysteine, 50 (cid:2)M sulfide, and 9

10 nM Hg produced ~1% MeHg and ~5 pmol of MeHg/mg of protein. Under those b
y
conditions,culturesgrewwell,achievinganopticaldensity(OD)of~0.5and~22(cid:2)gof g
u
protein/ml.Cysteinemaybeamoreavailableformofreducedsulfurthanmethionine
e
andalsoservesasanadditionalreductant(althoughallmediawerereducedwithTiNTA s
t
(titanium nitrilotriacetic acid) and monitored visually with resazurin). Corrected for its
effects on growth, cysteine enhanced the bioavailability of Hg for methylation by
M. hollandica. The absolute amount of MeHg produced increased with the cysteine
concentrationinthemedium,asdidMeHgproductionnormalizedtocellprotein,with
themostMeHgproductioninthe5mMcysteinecultures,despitetheirrelativelyweak
growth.
CysteinealsoenhancedmethylationbyM.tindarius.M.tindariuswasgrownacross
thesamecysteinegradientinmediumwithorwithouttheadditionof100(cid:2)Msulfide
and 500 (cid:2)M methionine (Fig. 4). Cysteine enhanced growth more strongly in the
mediumwithoutaddedsulfideormethionine.Inmediumwithoutaddedsulfide,MeHg
asafractionoftheHginthemedium(attheendofincubation)andMeHgproduction
normalizedtocellproteinbothincreaseduptoabout100(cid:2)Mcysteinebutdeclinedas
thecysteineconcentrationincreasedfurther.Sulfidehadastronginhibitoryeffecton
methylation. MeHg production was lower in medium with added sulfide, and the
March/April2018 Volume9 Issue2 e02403-17 mbio.asm.org 6
®
MercuryMethylationRatesamongMethanogens
D
o
w
n
lo
a
d
e
d

f
r
o
m

h
t
t
p
:
/
/
m
b
io
.
a
s
m
.
o
r
g
FIG4 ImpactofcysteineonMeHgproductionbyM.tindariusinmediumwithout(red)andwith(black) /
100(cid:2)Msulfideand500(cid:2)Mmethionine.(Top)MeHgasapercentageofthetotalHgintheculture o
n
medium.(Bottom)MeHgproductionnormalizedtoboththemeasuredHgandproteinconcentrations
J
attheendofmethylationassays,as(MeHg/Hg)/(mgofprotein/liter).Allmeasurementsweremadeonce a
allculturesreachedstationaryphase.Errorbarsarestandarddeviationsoftriplicatecultures.Growthdata n
u
areinFig.S1.
a
r
y

additionofcysteinedidnotincreaseMeHgproduction,eventhoughgrowthwasmuch 6
,
stronger.Thesulfideconcentrationincultureswithoutaddedsulfideaveraged0.3(cid:2)M 2
0
attheendoflog-phasegrowth;cultureswith100(cid:2)Maddedsulfideaveraged1.4(cid:2)M. 1
9
Inmediumwithorwithoutsulfide,cysteineheldHginsolutionintheculturemedium
b
inaconcentration-dependentway,basedonmeasurementoftheHgremainingin y

the aqueous medium (including cells) at the end of growth (Fig. S1). Interestingly, g
u
in the absence of added sulfide, cysteine increased methylation even when nor- e
s
malizedtothefinalHgconcentration,suggestingthatthemechanismforcysteine t
enhancement is not just its ability to enhance Hg solubility, but that the Hg-
cysteinecomplexismoreavailablethanotherforms.Thisisconsistentwithfindings
on Deltaproteobacteria (11, 21).
Influence of sulfide on Hg methylation by methanogens. Sulfide reduces the
bioavailabilityofHgtoHg-methylatingDeltaproteobacteriaviaprecipitationofmercuric
sulfidesthatarelessavailableforuptake(13–15).SulfidealsoinhibitedMeHgproduc-
tionbythreehgcAB(cid:2)methanogensgrowninmediumwithsulfideadditionsbetween
10and1,000(cid:2)M(Fig.5).SulfideadditionsresultedinlossofHgfrommediumtoeither
bottlewallsorreduction(seeFig.S1foranexample).Evenso,sulfidereducedMeHg
productionnormalizedtothefinalmeasuredHgconcentrationintheculturemedium.
DISCUSSION
Newly identified Hg-methylating methanogens. This is the first confirmation of
methylation ability in culture of hgcAB(cid:2) Methanocella paludicola SANAE, Methanocor-
March/April2018 Volume9 Issue2 e02403-17 mbio.asm.org 7
®
Gilmouretal.
D
o
w
n
lo
a
FIG5 ImpactofsulfideonMeHgproductionbythreemethanogensbasedonvaluesmeasuredatthe d
end of log growth in triplicate batch cultures. Sulfide was added to all cultures (between 10 and e
1,000(cid:2)M);thesulfideconcentrationshownwasmeasuredattheendofthelogphaseofgrowth.MeHg d

productionisnormalizedtoboththemeasuredHgandproteinconcentrationsattheendofmethylation fr
o
assays,as(MeHg/Hg)/(mgofprotein/liter).Allculturesweregrownwith4mMcysteine.Errorbarsare m
standard deviations of triplicate cultures. All measurements were made once all cultures reached

stationaryphase. h
t
t
p
:
/
/
m
pusculum bavaricum, Methanofollis liminatans GKZPZ, and Methanosphaerula palustris
b
E1-9c. Methanocella paludicola SANAE is the first confirmed methylator in the order io
.
Methanocellales.WeidentifiedasecondorganismwithafusedhgcABgenepairthatis a
s
unabletoproduceMeHgundertheconditionstested,Methanococcoidesmethylutens; m
thefirstwasPyrococcusfuriosus(2). .o
r
RarityofhgcABamongmethanogens.Todate,hgcABorthologshavebeenidenti- g
/
fied in only 18 sequenced Euryarchaeota genomes, among several dozen sequenced o
n
genomes(Fig.6).Methylationinculturehasbeenconfirmedforeightofthose.Allbut
J
twooftheidentifiedHgmethylatorsareinoneclass,Methanomicrobia(Methanomassi- a
n
liicoccus has tentatively been placed in the class Thermoplasmata [38]). There are no u
a
identified hgcAB(cid:2) organisms in the other four classes of methanogens, including the r
y
Methanobacteria/Methanococci superclass, Methanopyri, or the newly identified WSA2
6
class (“Candidatus Methanofastidiosa”) (39). Among the members of the class Metha- ,
2
nomicrobia,mostofthehgcAB(cid:2)strainsareinthetwolargerordersMethanomicrobiales 0
1
andMethanosarcinaceae.HgmethylatorsaremostcommoninthegeneraMethanolo- 9

bus,Methanoregula,andMethanocellabut,asidefromthesesmallclusters,arescattered b
y
throughout a 16S rRNA gene-based phylogeny. In the two methanogenic orders that g
containhgcAB(cid:2)organisms,MethanomicrobiaandThermoplasmata,therearecurrently u
e
about85sequencedgenomes.About20%ofthosespeciesarehgcAB(cid:2). s
t
Incomparison,hgcABorthologsarefoundin5ofthe10ordersofDeltaproteobac-
teria(Desulfarculales,Desulfobacterales,Desulfovibrionales,Desulfuromonadales,and
Syntrophobacterales).Overall,thereare~110hgcAB(cid:2)speciesamongthe(cid:3)1,000iden-
tifiedspeciesinthesefiveorders,orroughly10%ofthespeciesidentified.KnownHg
methylatorsareconcentratedinthegenusDesulfovibriointheorderDesulfovibrionales
(26 species, or about 25% of the named species) and in the genus Geobacter in the
order Desulfuromonadales (19 species). Almost all of the members of the genus
Geobacter whose genomes have been sequenced contain hgcAB. In the large order
Desulfobacterales,thereare22identifiedspecieswithhgcABorthologs(Hgmethylation
has been demonstrated in 7 of them) spread among several genera and (cid:3)250
identifiedspecies.
Carbon metabolism of Hg-methylating methanogens. Substrates for methano-
genesis include H -CO , acetate, and methylated compounds. Among the methano-
2 2
gens whose genomes have been sequenced, Hg methylators were found among the
March/April2018 Volume9 Issue2 e02403-17 mbio.asm.org 8
®
MercuryMethylationRatesamongMethanogens
D
o
w
n
lo
a
d
e
d

f
r
o
m

h
t
t
p
:
/
/
m
b
io
.
a
s
m
.
o
r
g
/

o
n

J
a
n
u
a
r
y

6
,

2
0
1
9

b
y

g
u
e
s
t
FIG6 16SrRNAgenedistancephylogenyoftheMethanomicrobia/Thermoplasmatasuperclass(excludingtheHalobacteria)withinthemethanogenic
Euryarchaeota,showingalloftheavailablesequencedgenomes.RedcirclesshoworganismswithhgcABorthologs.Greencirclesareorganismsconfirmed
tomethylateHginculture.Branchpointssupportedwithbootstrapvaluesof(cid:3)50%areshown.Thescalebarindicates0.04substitutionpernucleotide
position.
methylotrophic and hydrogenotrophic methanogens but not among the small group
ofdescribedaceticlasticmethanogens.NoneofthehydrogenotrophicHgmethylators
are class I methanogens (40). This distribution was also observed in a broad query of
availablemicrobialmetagenomes(2).
Several methylotrophic hgcAB(cid:2) methanogens were identified in the order Metha-
nosarcinales, the only methanogens that contain cytochromes and can metabolize
methylatedcompoundsviaamembrane-boundelectrontransportchain.Mostofthe
Hg-methylating members of the order Methanosarcinales are Methanolobus species,
which are all methylotrophic and somewhat salt tolerant. Substrates for some of the
March/April2018 Volume9 Issue2 e02403-17 mbio.asm.org 9
®
Gilmouretal.
species include methanol (MeOH), methylamines, and methyl sulfides. We tested Hg
methylationbyMethanolobustindariusgrowingonMeOHunderaN -CO atmosphere.
2 2
TheotherhgcAB(cid:2)MethanosarcinalesmemberisM.hollandica,whichisalsoobligately
methylotrophic. We tested its methylation ability on trimethylamine-MeOH under a
N -CO atmosphere.
2 2
Other methylotrophic methanogens are obligately H dependent. Methanomassili-
2
icoccus is a recently identified genus of obligately methylotrophic and H -dependent
2
methanogens isolated from the intestinal tracts of humans, ruminants, and termites;
anaerobic digesters; and soils. Along with Methanoplasma, Methanogranum, and
Methanomethylophilus,theseorganismshavebeenclassifiedasanew,seventhorderof
methanogens, Methanomassiliicoccales, in the class Thermoplasmatales, phylogeneti-
callydistantfromothermethanogenicorders(41).TheorderlacksthepathwayforCO
2 D
reductiontomethylcoenzymeMandproducesmethanebyH -dependentreduction o
2 w
of MeOH or methylamines. The two known species of Methanomassiliicoccus both
n
contain hgcAB. Methanomassiliicoccus luminyensis B10 produced MeHg in cultures lo
a
growingonMeOHsupplementedwithrumenfluid(2). d
e
TheremaininghgcAB(cid:2)methanogensareobligatelyhydrogenotrophicorganisms d

intheordersMethanomicrobialesandMethanocellales.TheorderMethanomicrobia- fr
o
les is a highly polyphyletic group of hydrogenotrophic methanogens, but with a m
narrow substrate range, generally utilizing only CO2, with either formate or H2 as h
t
the reducing agent. We identified eight species with hgcAB orthologs in four tp
:
families. Most were in the family Methanoregulaceae, formerly the E1/E2 and R10 /
/
m
groups or “fen cluster” phylotypes (42). Five Methanoregulaceae species in three
b
genera,Methanoregula,Methanosphaerula,andMethanolinea,havebeensequenced io
.
(42). Neither of the two Methanolinea strains contain hgcAB, but both Methanoregula a
s
species(M.booneiandM.formicica)andthesingleMethanosphaerulaspecies(M.palus- m
.
tris)do.BothM.booneiandM.palustriswereisolatedfromoligotrophicpeatenviron- o
r
ments (43, 44). M. formicica was isolated from an anaerobic sludge reactor treating g
/

breweryeffluent(45).TheunclassifiedMethanoregulaceaearchaeonJGIM3C4D3-001- o
n
G22,whosegenomehasrecentlybeensequenced,isalsohgcAB(cid:2).Hgmethylationwas
J
a
tested in Methanosphaerula palustris E1-9c cultures growing in H -CO supplemented
2 2 n
withacetate. u
a
Two other methylators in the order Methanomicrobiales were identified. Methano- ry

spirillumhungateiistheonlyorganisminthefamilyMethanospirillaceaewhosegenome 6
,
has been sequenced. Various phylogenies place M. hungatei within or just outside 2
0
Methanoregula(46).Isolatedfromsewagesludge,itwasoneoftheearliestmethano- 1
9
gens characterized (47) and the first confirmed Hg-methylating methanogen (5). We
b
grew this organism in H -CO supplemented with acetate and formate for Hg meth- y
2 2
g
ylation testing. Methanocorpusculum bavaricum (family Methanocorpusculaceae) was
u
isolatedfromasugarwastewaterpond.InadditiontoH -CO methanogenesis,itcan e
2 2 s
utilizesecondaryalcoholsaselectrondonorstoreduceCO andrequiresrumenfluid t
2
inculture.TotestHgmethylation,thisorganismwasculturedunderH -CO inmedium
2 2
supplementedwithacetate,formate,rumenfluid,andfattyacids.
To summarize, the hgcAB gene pair appears to be confined to two classes of
methanogens, Methanomicrobia and Thermoplasmata, but is found in organisms
with a wide variety of pathways for methanogenesis. Methanosarcina may metab-
olize one-carbon compounds like MeOH, acetate, and methylamines via a
membrane-bound electron transport chain. Other methylotrophic methanogens con-
taininghgcABrequireH .OurmethylationassaysofseveralpreviouslyuntestedhgcAB(cid:2)
2
Methanomicrobia species lead us to believe that these methanogens should be con-
sidered equally with sulfate and iron reducers in evaluations of potential Hg-
methylating biogeochemical conditions in nature. Our understanding of the distribu-
tion and activity of hgcAB(cid:2) organisms in nature and of the controls on hgcAB
expressionremainsinitsinfancy.
March/April2018 Volume9 Issue2 e02403-17 mbio.asm.org 10