-- LETTERS TO NATURE

ments are able to use humic substancesas an electron accepto

Humic substances as electron for the anaerobic oxidation of organic compoundsand hydrogen
This electron transport yields energyto support growth. Micro
acceptors for microbial bial humic reduction also enhancesthe capacity for microorgan
isms to reduce other, less accessibleelectron acceptors,such a
respiration insoluble Fe(m) oxides, becausehumic substances can shutt!
electrons betweenthe humic-reducing microorganisms and th'
Derek R. Lovley*, John D. Coatest, Fe(m) oxide. The finding that microorganisms can donate elec
Elizabeth L Blunt-Harris*, Elizabeth J. P. Phillipst trons to humic acids has important implications for the mechan
& Joan C. Woodward*t isms by which microorganisms oxidize both natural anI
contaminant organics in anaerobic soils and sediments, anf
.Department of Microbiology, University of Massachusetts, Amherst,
suggestsa biological source of electrons for humics-mediate.
Massachusetts 01003, USA reduction of contaminant metals and organics.
t Water Resources Division, US Geological Survey, Reston, Virginia 22092, In a study on the effect of various Fe(m) chelatorson anaerobi
USA benzene degradationiri petroleum-contaminated aquifers, it wa
found that humic acids stimulated benzene degradation bette
HUMIC substances are heterogeneous high-molecular-weight than any of the synthetic chelatorsevaluated2.Synthetic chelator
organic materials which are ubiquitous in terrestrial and aquatic stimulate benzene degradation by solubilizing Fe(m) oxides anI
environments. They are resistant to microbial degradation! and thus making Fe(m) more available to benzene-oxidizingFe(m)
thus are not generally consideredto be dynamically involved in reducirig bacteria3.4.The superiority of the humic acids oV,e
microbial metabolism,especiallyin anoxic habitats. However,we synthetic chelatorssuchasnitrilotriacetic acid (NT A~ wassurpns
show here that some microorganisms found in soils and sedi- irig becausealthough humic acidscan chelate Fe(m) , they do no
NATURE. VOL 382 .1 AUGUST 1996 44!
 ~RS TO NATURE
/!
/I 2.75 I
~
.E. 1.75
B Fe(II>
~
e
f 'G;'
u. 0.75
0 1 2 3 4
Ti1Ie (hours)
AG. 1 Fe(m) reduction to Fe(lI) by cell suspensions of G. meta/lireducens.
Addition of soil humic acid from the IHSS or commercially available Aldrich
humic acids greatly stimulated Fe(lIl) reduction, as did the humic analogue
2,6-anthraquinone disulphonate (AQDS).
METHODS. Washed cells of G. meta/fireducens (0.4, 1.3 or 1.3 mg protein
per ml for IHSS soil humic acids, Aldrich humic acids and AQDS
respectively) were suspended in 10 ml sodium bicarbonate buffer
(30 mM; pH 6.8) under N2-CO2 (80:20) containing acetate (10 mM) as
the electron donor and synthetic, poorly crystalline Fe{lIl) oxide (~6 mM)22
as the electron acceptor. The incubation temperature was 30 °C. Sub-
sampleS were anaiysed for Fe(lI) with ferrozjne in HEPES buffef2. The
concentration of dissolved humic acids was 2 mgml-1 and of AQDS was
100jiM.
have the chelation capacityof synthetic chelators.
To determine whether humic acidsstimulated microbial Fe(m)
reduction as previously shown for synthetic chelators3.4,cell
suspensions of the dissimilatory Fe(m)-reducing bacterium
Geobactermetallireducenswere added to bicarbonate buffer con-
taining acetate as the electron donor and poorly crystalline Fe(m)
oxide as the electron acceptor. Cell suspensionsof G. metalli-
reducensonly slowlyreduced Fe(m) oxide4,but when humic acids
were also added to the buffer, Fe(m) reduction was greatly
stimulated (Fig. 1). The samedegreeof stimulation wasobserved
with highlypurified 'soil humic acids'suppliedby the International
Humic SubstancesSociety (IHSS) and with the commercially
-
Humics
CO2 reduced Fe(lII)
Humics
Acetat oxidized
AG. 2 Model for mechanism by which humic acids stimulate
reduction by G. metallireducens.
Fe(lIl)
available Aldrich humic acids that were previously found to
stimulate benzene metabolism2.Humic acids did not reduce
appreciable amounts of Fe(m) when G. metallireducenswas
omitted (Fig. 1).
Stimulation of microbial Fe(m) reduction by NTA is associated
with visible dissolution of Fe(m) oxide and yields millimolar
concentrationsof dissolvedFe(m)4.In contrast, the humic acids
did not visibly dissolve Fe(m) oxide and the concentration of
dissolvedFe( m) in the presenceof humic acidsbut without added
cells waslessthan 200~M. This suggestedthat the mechanismby
which humic acidsstimulated Fe(m) reductionwas different from
that for synthetic Fe(m) chelators.
An additional property of humic acidsis their ability to undergo
reduction-oxidation. For example, humic substancesand qui-
none-containing humic analogues can transfer electrons from
reduced inorganics (sulphide) and organics ~rbic acid) to
contaminantssuchas mercuf'/'7,nitroaromatics and cWorinated
solvents1o.Although direct reduction of humic acids by micro-
organisms has not been demonstrated', it is well known that
reduced humic acids can donate electronsto Fe(m)I1-13.
We speculated that if G. metallireducenscould transfer elec-
trons to humic substances,then humic substancesmight stimulate
Fe(m) reduction in a two-stageprocess in which (1) G. metalli-
reducensoxidizes acetate,with humic acidsacting as the electron
acceptor,and (2) reduced humic acidsdonate electronsto Fe(m)
(Fig. 2). Such electron shuttling would alleviate the need for
Fe(m) reducers to contact Fe(m) oxides directly in order to
reduce them and so would accelerateFe(m) reduction.
To evaluate step (1) in the model, cells of G. metallireducens
were suspendedin buffer containing acetateas the sole electron
donor and humic acids as a potential electron acceptor(Fig. 3a).
Acetate was readily oxidized to carbondioxide in the presenceof
 LETTERS TO
humics,but there waslittle oxidation in the absenceof humic acids. 3.00 2S
There wasno acetateoxidation in the presenceof humic acidswhen
G. metallireducenswas omitted, showingthat humic acidscanserve
2.50
asan electron acceptorfor acetateoxidationbyG. metallireducens. 20
To determine whetherthe electronstransferred from acetateto
humic acids by G. metallireducenscould then be transferred to I 2.00
~
Fe(m) (step (2) of the model), cell suspensionswere incubated ~
0'
IS
g
with humic substancesand the cells removed by filtration. When -< 1.50 ~.
-:<
Fe(m) wasaddedto the filtrates, the Fe(m) wasreduced (Fig. 3b).
This wastrue for a wide varietyof highlypurified humic substances
~
" to o..
~ \.00 ~
obtained from IHSS as well as for the commercially available ..:
Aldrich humic acids.As in Fig. 1,humic substancesnot incubated 0.50
with G. metallireducensdid not reduce Fe(m). Furthermore, in
accordancewith a previousdemonstrationthat G. metallireducens
does not release extracellular componentscapable of reducing 0.00 0
Fe(mY4,filtrates of cell suspensionsthat did not contain humic 2 6 8

acids did not reduce Fe(m). Another Fe(m)-reducing micro- Days

organism, Shewanellaalga, which can reduce Fe(m) with Hz or ,.00 $0

by incompletely oxidizing la!;tate to acetatelS.16, was found to
transfer electronsto soil huIriic acids in a comparable way, with
lactate or Hz serving as electron donor (data not shown). 4.00 40
To quantify the electron transfer from acetate to humics and
then from humics to Fe(m), the amount of acetateoxidized (Fig. i 3.00
~
3a) and the amountof Fe(m) reducedbyhumics to Fe(n) (Fig. 3b) ~
-<
30 R-
g.
were determined simultaneot;lslyon cell suspensionsincubated -<
with soil humic acidsfor 15 h. The ratio of the numberof molesof 1 2.00 ~ q,.
Fe(n) producedby the reducedhumics to the number of moles of ~ ~
acetate oxidized in the presence of humics was 8.3:i: 0.7 1.00 10
(mean :i: s.d.; n = 3). This compares favourably with the reduc-
tion of Fe(m) to Fe(n) by humics,which requiresone electron,and
the acceptanceby humics of eight electrons per mol of acetate 0.00

oxidized by G. metallireducens. This result indicates that electrons 20 30 ~ so 60 70 80

HOUB
transferred to humic acids from acetate by G. metallireducenscan
be quantitatively transferred to Fe(m). In comparable experi-
3.50
ments with S. alga, the amount of Fe(n) produced was ~
100:i: 10% (n = 3) of that expected for complete transfer of
3.00
electrons from lactate to humics and then from humics to Fe(m). 40
The ability of humic acidsto participate in abiological electron ~
transfer hassuggestedthat quinone moeties maybe the electron-
~ :g 2.00 30 ~
accepting groups, with the resultant hydroquinones donating ~. g.
electronsto the ultimate electron acceptor9,lO. As the complexity 6u
G.
~ e 1.50 20
q
of humic acid structure has precluded direct proof of this, the -'"
",'" 'q..
potential involvementof the quinone/hydroquinoneredox couple go
:::I 1.00 ~
in such electron transfer has been investigated with model com- ~ O.SO
10 -J-
pounds suchas 2,6-anthraquinonedisulphonate (AQDSYO,17.
Concentrationsof AQDS as low as 100 11Mgreatly stimulated 0.00
Fe(m) reductionby cell suspensions of G. metallireducens(Fig. 1).
There was no detectable solubilization of Fe(m) oxide by the IS 10 2S
AQDS. Once Fe(m) reduction was complete, the buffer turned Hours
orangeowingto the accumulationofreducedAQDS (namely,2,6-
anthrahydroquinone disulphonate (AHDS). H more Fe(m) was
added, then the orange colour immediately disappeared.When
Fe(m) wasaddedto known concentrationsof AHDS in anaerobic
buffer, the orange colour again disappeared immediately and AG. 4 Growth of G. metal/ireducens and S. alga with 2,6-anthraquinone
there was an accumulationof Fe(n) that was consistentwith the disulphonate (AQDS) (a, b) or humic acids (c) as electron acceptor.
oxidation of AHDS back to AQDS, with reduction of 2 mol of METHODS. Cells were inoculated into anaerobic freshwater medium14
containing acetate (G. metal/ireducens) or lactate (S. alga) as electron
Fe(m) per mol of AHDS oxidized. These results show that G. donor and AQDS or humic acids as potential electron acceptor. Reduction of
metallireducenscan reduce AQDS to AHDS, which in turn can AQDS was quantified by the absorbance of 2,6-anthrahydroquinone disul-
instantaneouslyreduce Fe(m) and regenerateAQDS. phonate (AHDS) produced at 450 nm. In separate incubations, the stoi-
Both G. metallireducensand S. alga could grow in medium chiometry of acetate or lactate consumption and AHDS formation were
containing AQDS as the sole electron acceptor (Fig. 4a, b). determined by measuring loss of electron donors using high-performance
Growth coincided with AQDS reduction. There was no liquid chromatography. Millequivalents of reduced humic acids produced
growth or AQDS reduction if the electron donor was omitted were estimated from the amount of Fe(lIl) reduced when a filtrate of a
(Fig. 4a, b), or if the electron donors were provided but subsample was exposed to Fe(III), as for Ag. 3. Cell growth on AQDS was
monitored by direct cell counting14. Monitoring cell growth by direct cell
the AQDS omitted (data not shown). The stoichiometry of
counting or by protein determination was not feasible when the organisms
acetate uptake and AQDS reduction during growth on G. were grown on humic acids as electron acceptor because of interference by
metallireducens (Fig. 4a) was consistent with oxidation of the humic acids. Therefore, cell numbers of S. alga were determined from
acetate and reduction of AQDS according to: acetate- + plate counts on aerobic heterotrophic medium. This method could not be
4HzO + 4AQDS -+ 4AHDS + 2HCO] + H+; in cultures of S. used with the strict anaerobe G. rnetal/ireducens, so growth of this organism
alga,lactate consumptionand AQDS reduction (Fig. 4b) was in on humic acids could not be quantified.

NATURE. VOL 382 .1 AUGUST1996 447

 , LETTERS TO NATURE

agreement with the reaction: lactate- + 2Hp + 2AQDS -+
.I 2AHDS + acetate- + HCO"i + H+.
Technical difficulties prevented cell growth of G. metalliredu-
celIS from being monitored with humic acids as the electron
acceptor (Fig. 4); however,S. alga grew in a medium in which
humic acidswere the sole electron acceptor(Fig. 4c). Growth was
associatedwith an accumulation of reducing potential in the
humic acids that could be transferred to Fe(m) when Fe(m) was
added to cell-free filtrates of the culture. Growth was not due to
degradationbyS. algaof humic acidsbecausethere wasno growth
in the presenceof humic acids if lactate,the electron donor, was
is not availableto recyclehumic substancesbackto their oxi
form, humic substancesmay still be important electron acce
for organic matter oxidation, given the abundance of t
substancesin many soils and sediments. The known abilj
reducedhumic acidsto donate electronsto a varietyof metal
organics6,9,Io,usuggeststhat microbial reduction of humic
may have an impact on the fate of other environmentalcon:
nantsaswell.
Receiwd 7 May; accepted 14 June 1996.

omitted (Fig. 4c). These results show that respiration with humic 1. Mcl<ni~t, D. M. et a/. In Organ/CAcids/nAquaticErosystems (edsPertlue, E. M.&~
acidsasthe terminal electron acceptorcanyield energyto support E. T.) 223-243 (Wiley, NewYor1I, 1990).
cell growth. 2. lJMey, D. R.. Woodward, J. C. & Chapelle, F. H.Appi. environ. MIcrob/o/.i2,
3. lJMey, D. R., Woodward, J. C. & Chapelle, F. H. Nature 170, 128-131
288-291
(1994).
(

Besidesrevealing a new form of microbial respiration, these 4. lJMey, D. R. & Woodward, J. C. 01en1. Geo/. (In the press).
findings mayhave important implications for the biogeochemistry 5. Jackson, K. 5., Jonasson,I. R. &SWppen, G. B. Earth Sc/. Rev. 14, 97-146 (1978).
6. Alberts, J. J., Sdlindler, J. E., Miller, R. W. & Nutter, D. E. ScIence 184, 895-897 (19-
of soils and aquatic sediments. For example, the reduction of 7. Sdlindler, J. E., Williams, D. J. & Zimmerman, A. P. in EnvIronmental ~
insoluble Fe(m) and Mn(IV) oxides is one of the mostgeochemi- (eds Nriagu, J. D.) 109-115 (Ann A/t)O( ScIence, Ann A/t)O(, Mid1igan, 1976).
8. Schwarzenbach, R. P., StieI1i, R., LBnz, K. & Z1:yer, J. EnvIron. Sc/. Techno/. 24, 1566
cally significantprocessesthat takes place in sedimentaryenvir- (1990).
onmentsl8.19. Previousinvestigationsinto the mechanismsfor the 9. DuMivant, F. M., Schwarzenbach, R. P. & Macalady, D. L EnvIron. Sc/. Techno/. H, :
2142 (1992).
reduction of Fe(m) and Mn(IV) have emphasized either the 10. Curtis, C. P. & Reinhard, M. EnvIron. Sc/. Techno/. 28, 2393-2401 (1994).
abiological reduction of these metals by organics suchas humic 11. Szilag;i, M. SoIl Sc/. U1, 233-235 (1971).
materials and related aromatic compoui1ds~,or the direct bio- 12. SkogeIboe, R. K. & Wilson, S. A. AnaJyt 01en1. U, 228-232
13. Kahn, T. R.,lBngfotd, C. H. & SWppen, G. B. Org. Geochem. 7. 261-266
(1981).
(1984).
logical reduction of the metals by specialized metal-reducing 14.lJMey, D. R. & Phillips, E. J. P.Appi. environ. MIcrob/o/. &4, 1472-1480 (1988).
microorganisms21.Our results suggest that, at least in some 15. Caccaw, F. Jr, Blakemore, R. P. & lJMey, D. R. Appi. environ. MIcrob/o/. 18, 3211-
(1992).
instances,the reduction of Fe(m) (and other metals such as 16. RosseIIo-Morn, R. A. et al. ~ appi. M/croblol. 1.7. 569-573 (1994).
Mn(IV) that can also accept electrons from humic substances) 17. T~ P. G.&Macalady, D. LJ..Agrku/. FoodO1en1.17, 248-254 (1989).
18. Ponnamperuma, F. N. Adv. Agron. 24, 29-96 (1972).
may actually be a combination of both processes,with micro- 19. lJMey, D. R. Adv. Agron. &4, 175-231 (1995).
organismsfirst donating electrons to humic substancesand the 20. LB~, J. S. & Stone, A. T. Ge<x:h/m. O)SfIK)C/1/m. Acta U, 961-971 (1989).
humic substancesthen reducing the metals. Such electron shut- 21.
22.
lJMey, D. R., Phillips, E. J. P. & L1Inergan, D. J. EnvIron. Sc/. Techno/. 21, 1062-1067
lJMey, D. R. & Phillips, E. J. P. Appl. EnvIron. MIcrob/o/. Ii., 683-689 (1986).
(1!

tling may greatlyfacilitate the 'ability of Fe(m)-reducingbacteria

to pass electrons to insoluble Fe(m) oxides and is the likely ACI(N(ml£DGEMENTS. WethankD. McKnight and P. T~forhelpfuldlscussions. This rest
explanation for the earlier obseIVation2that humic acids greatly was supported ~ the Office of Naval Research and ~ the American Petroleum Institute.

acceleratethe rate of benzenedegradationin aquifer sedimentsin CORRESPONDENCE and requests for materials shook! be addressed to D.R.L (e-mail: dkM
which Fe(m) is the terminal electron acceptor. Even when Fe(m) microbio.umass.edu).

VOL 382 1 AII~II~T 100&:

448 NATURE