.*5&12*38/'"2+"*.$/,. .*5&12*38/'"2+"*.$/,.
*(*3",/--/.2 .*5&12*38/'"2+"*.$/,.*(*3",/--/.2 .*5&12*38/'"2+"*.$/,.
"$4,384#,*$"3*/.2*.3)&*/,/(*$",$*&.$&2 "0&12*.3)&*/,/(*$",$*&.$&2
."&1/#*$1&%/7$8$,*.(/'*1/.#8'1&2)6"3&12&%*-&.3."&1/#*$1&%/7$8$,*.(/'*1/.#8'1&2)6"3&12&%*-&.3
-*$1//1(".*2-2-*$1//1(".*2-2
"11*&!&#&1
.*5&12*38/'"2+"*.$/,.
+6&#&14.,&%4
"3*,%& 1143*"
.*5&12*38/',"#"-"42$",//2"
&118)41$)*,,
.*5&12*38/',"#"-"42$",//2"
"5*4++"%"04
"$*:$/13)6&23"3*/.","#/1"3/18*$),".%!
1*$/%&.
.*5&12*38/',"#"-"42$",//2"
&1/%&.#2$"24"&%4
/,,/63)*2".%"%%*3*/.",6/1+2"3)3302%*(*3",$/--/.24.,&%4#*/2$*'"$04#
"13/'3)&*'&$*&.$&2/--/.2
!&#&1"11*& 1143*""3*,%&)41$)*,,&1184++"%"04"5*".%/%&.1*$."&1/#*$
1&%/7$8$,*.(/'*1/.#8'1&2)6"3&12&%*-&.3-*$1//1(".*2-2
"$4,384#,*$"3*/.2*.3)&
*/,/(*$",$*&.$&2
)3302%*(*3",$/--/.24.,&%4#*/2$*'"$04#
)*213*$,&*2#1/4()33/8/4'/1'1&&".%/0&."$$&22#83)&"0&12*.3)&*/,/(*$",$*&.$&2"3
*(*3",/--/.2 .*5&12*38/'"2+"*.$/,.3)"2#&&."$$&03&%'/1*.$,42*/.*."$4,384#,*$"3*/.2*.3)&
*/,/(*$",$*&.$&2#8"."43)/1*9&%"%-*.*231"3/1/'*(*3",/--/.2 .*5&12*38/'"2+"*.$/,.
Environmental Microbiology (2006)
8
(1), 100–113 doi:10.1111/j.1462-2920.2005.00873.x
© 2005 Society for Applied Microbiology and Blackwell Publishing Ltd
Blackwell Science, LtdOxford, UKEMIEnvironmental Microbiology 1462-2912Society for Applied Microbiology and Blackwell Publishing Ltd, 20058
1100113
Original Article
Anaerobic redox cycling of iron in sedimentsK. A. Weber
et al.
Received 16 December, 2004; accepted 2 May, 2005. *For correspon-
dence. E-mail eroden@bsc.as.ua.edu; Tel. (
+
1) 205 348 0556; Fax
(
+
1) 205 348 1403 or e-mail kweber@nature.berkeley.edu; Tel.
(
+
1) 510 642 4972; Fax (
+
1) 510 642 4995.
†
Present address: Uni-
versity of California, Department of Plant and Microbial Biology, Ber-
keley, CA 94720-3102, USA.
Anaerobic redox cycling of iron by freshwater sediment
microorganisms
Karrie A. Weber,
1
*
†
Matilde M. Urrutia,
1
Perry F.
Churchill,
1
Ravi K. Kukkadapu
2
and Eric E. Roden
1
*
1
The University of Alabama, Department of Biological
Sciences, Tuscaloosa, AL 35487-0206, USA.
2
Pacific Northwest National Laboratory, MSIN K8-96,
Richland, WA 99352, USA.
Summary
The potential for microbially mediated anaerobic
redox cycling of iron (Fe) was examined in a first-
generation enrichment culture of freshwater wetland
sediment microorganisms. Most probable number
enumerations revealed the presence of significant
populations of Fe(III)-reducing (approximately
10
8
cells ml
--
--
1
) and Fe(II)-oxidizing, nitrate-reducing
organisms (approximately 10
5
cells ml
--
--
1
) in the fresh-
water sediment used to inoculate the enrichment cul-
tures. Nitrate reduction commenced immediately
following inoculation of acetate-containing (approxi-
mately 1 mM) medium with a small quantity (1% v/v)
of wetland sediment, and resulted in the transient
accumulation of NO
2
–
and production of a mixture of
gaseous end-products (N
2
O and N
2
) and NH
4
++
++
. Fe(III)
oxide (high surface area goethite) reduction took
place after NO
3
–
was depleted and continued until all
the acetate was utilized. Addition of NO
3
–
after Fe(III)
reduction ceased resulted in the immediate oxidation
of Fe(II) coupled to reduction of NO
3
–
to NH
4
++
++
. No
significant NO
2
–
accumulation was observed during
nitrate-dependent Fe(II) oxidation. No Fe(II) oxidation
occurred in pasteurized controls. Microbial commu-
nity structure in the enrichment was monitored by
denaturing gradient gel electrophoresis analysis of
polymerase chain reaction-amplified 16S rDNA and
reverse transcription polymerase chain reaction-
amplified 16S rRNA, as well as by construction of 16S
rDNA clone libraries for four different time points dur-
ing the experiment. Strong similarities in dominant
members of the microbial community were observed
in the Fe(III) reduction and nitrate-dependent Fe(II)
oxidation phases of the experiment, specifically the
common presence of organisms closely related
(
≥≥
≥≥
95% sequence similarity) to the genera
Geobacter
and
Dechloromonas
. These results indicate that the
wetland sediments contained organisms such as
Geobacter
sp. which are capable of both dissimilatory
Fe(III) reduction and oxidation of Fe(II) with reduction
of NO
3
–
to NH
4
++
++
. Our findings suggest that microbially
catalysed nitrate-dependent Fe(II) oxidation has the
potential to contribute to a dynamic anaerobic Fe
redox cycle in freshwater sediments.
Introduction
Iron (Fe)-bearing minerals are abundant in soil and sedi-
mentary environments, where they exist predominantly as
solid-phase minerals containing Fe in the ferrous [Fe(II)]
and/or ferric [Fe(III)] oxidation state (Cornell and Schwert-
mann, 1996). Cycling between Fe(II) and Fe(III) (i.e. Fe
redox cycling) can significantly affect the biogeochemistry
of hydromorphic soils and sediments (VanBreemen, 1988;
Stumm and Sulzberger, 1992; Davison, 1993; Roden
et al
.,
2004). Direct microbial (enzymatic) reduction coupled to
oxidation of organic carbon and H
2
by dissimilatory iron-
reducing bacteria (DIRB) is recognized as the dominant
mechanism for Fe(III) oxide reduction in non-sulfidogenic
anaerobic soils and sediments [see Lovley (1991; 2000)
for review]. This process contributes to both natural and
contaminant (hydrocarbon) organic carbon oxidation in
sedimentary environments, and exerts a broad range of
impacts on the behaviour of trace and contaminant metals
and radionuclides (Lovley and Anderson, 2000).
When Fe(II) comes into contact with O
2
or other suitable
oxidants, Fe(II) can be re-oxidized to Fe(III). The dominant
role of microbial catalysis in Fe(II) oxidation in acidic envi-
ronments (e.g. acid mine drainage and acid hot springs)
is well-established (Brock and Gustafson, 1972; Singer
and Stumm, 1972; Johnson
et al
., 1993). In contrast,
Fe(II) is subject to spontaneous chemical oxidation by
dissolved O
2
at circumneutral pH (Davison and Seed,
1983; Millero
et al
., 1987), and the quantitative role of
microbial catalysis in Fe(II) oxidation by O
2
in circumneu-
tral aerobic environments is still a matter of debate (Emer-
son, 2000; Emerson and Weiss, 2004; Roden
et al
.,
2004). A previously unrecognized potential for microbial
This article is a U.S. government work, and is not subject to copyright in the United States.
Anaerobic redox cycling of iron in sediments
101
© 2005 Society for Applied Microbiology and Blackwell Publishing Ltd,
Environmental Microbiology
,
8
, 100–113
Fe redox cycling under anoxic conditions has been
revealed through the recent discovery of nitrate-reducing
microorganisms capable of enzymatic oxidation of Fe(II)
[Straub
et al
. (1996; 2004); see Fig. 1]. In contrast to abi-
otic Fe(II) oxidation by O
2
, the abiotic reaction of Fe(II)
with NO
3
–
is negligible under the temperature and aque-
ous geochemical conditions typical of natural soil and
sedimentary environments (Weber
et al
., 2001). Microor-
ganisms capable of oxidizing Fe(II) with reduction of NO
3
–
have been observed in several different freshwater sedi-
ments (Kluber and Conrad, 1998; Straub and Buchholz-
Cleven, 1998; Caldwell
et al
., 1999; Ratering and Schnell,
2000; Chaudhuri
et al
., 2001; Hauck
et al
., 2001; Finneran
et al
., 2002; Senn and Hemond, 2002; Shelobolina
et al
.,
2003) as well as sewage sludge systems (Nielsen and
Nielsen, 1998a,b).
The demonstrated potential for biological nitrate-
dependent Fe(II) oxidation in a wide variety of natural
systems suggests that this reaction may play a significant
role in the coupling of Fe and N redox cycles in sedimen-
tary environments. In addition, the recent demonstration
of biological nitrate-dependent Fe(II) oxidation by a pre-
dominant environmental Fe(III)-reducing bacterium,
Geo-
bacter metallireducens
(Finneran
et al
., 2002), suggests
that anaerobic Fe redox cycling could be catalysed by a
single group of microorganisms. A tight coupling between
Fe and N redox cycles in anaerobic sedimentary environ-
ments has significant implications for mechanisms of NO
3
–
removal and the regeneration of reactive Fe(III) oxides in
hydromorphic soils and sediments, as well as the trans-
formation of various natural and contaminant organic and
inorganic compounds.
Although the potential for enzymatic Fe(II) oxidation
coupled to NO
3
–
reduction has been well documented, the
microbial communities associated with Fe–N redox
cycling in natural environments are not yet well under-
stood. In this study, a first-generation enrichment culture
of freshwater wetland sediment was subjected to a
sequential shift in redox conditions [from organotrophic
NO
3
–
reduction, to organotrophic Fe(III) reduction, to
lithotrophic nitrate-dependent Fe(II) oxidation] in order to
explore the coupling between microbial N and Fe redox
cycling in sediments. Changes in microbial community
structure associated with redox shifts were monitored by
denaturing gradient gel electrophoresis (DGGE) analysis
of polymerase chain reaction (PCR)-amplified 16S rDNA
and reverse transcription polymerase chain reaction (RT-
PCR)-amplified 16S rRNA, and the phylogenetic associa-
tion of organisms predominant in the culture was
assessed through 16S rDNA clone libraries. A follow-up
study evaluated the potential for
G. metallireducens
to
catalyse anaerobic Fe redox cycling analogous to that
observed in the enrichment culture.
Results
Most probable number (MPN) enumerations
Approximately 10
5
cells (ml wet sediment)
-
1
of culturable
(MPN assay) nitrate-dependent Fe(II)-oxidizing microor-
ganisms were detected in Talladega Wetland surface
sediment (Table 1). The abundance of culturable acetate-
oxidizing [nitrate- and Fe(III)-reducing] microorganisms
was approximately three orders of magnitude higher.
Sequential nitrate reduction, Fe(III) reduction and
nitrate-dependent Fe(II) oxidation in the sediment
enrichment culture
Talladega Wetland sediment served as the inoculum (1%
vol:vol) to artificial groundwater (AGW) containing 1 mM
NO
3
–
, 2 mM acetate and 50 mmol l
-
1
of synthetic high
surface area goethite. Nitrate was consumed during the
initial 7 days of incubation, resulting in transient accumu-
lation of NO
2
–
and production of approximately 0.2 mM
NH
4
+
(Fig. 2A and B). The molar ratio of NH
4
+
produced to
NO
3
–
reduced (0.280,
r
2
=
0.940) was substantially lower
than 1.0, which indicates that gaseous end-products such
as NO, N
2
O and/or N
2
(not measured in this study) were
likely produced. A decrease in Fe(II) (0.75 mmol l
-
1
of
Fe(II) was introduced with the sediment inoculum) of
approximately 0.2 mmol l
-
1
occurred during the initial NO
3
–
Fig. 1.
Potential Fe–N redox pathways in anoxic sediments: Organ-
otrophic NO
3
–
reduction to N
2
(1) or to NH
4
+
(2); organotrophic dis-
similatory Fe(III) reduction (3); lithotrophic [Fe(II)-driven] NO
3
–
reduction to N
2
(4) or to NH
4
+
(5). Thick lines denote external loading
of NO
3
–
and organic carbon (CH
2
O). Temporal variations in NO
3
–
and
CH
2
O loading have the potential to cause temporal/spatial overlap of
organotrophic and lithotrophic pathways (see text).
N
2
,
NH
4
+
NO
3
-
Fe(III)
Fe(II)
1, 2
4, 5
3
CH
2
O
CH
2
O
+
CO
2
+
CO
2
N
2
,
NH
4
+
NO
3
–
Fe(III)
Fe(II)
1, 2
4, 5
3
CH
2
O
CH
2
O
+
CO
2
+
CO
2
+
CO
2
+
CO
2
Table 1.
MPN enumerations of nitrate-reducing, Fe(III)-reducing and
nitrate-dependent Fe(II)-oxidizing microorganisms in Talladega Wet-
land surface sediments.
Culture conditions MPN (cells ml
-
1
) 95% confidence interval
Acetate
+
NO
3
–
9.3
¥
10
7
2.1
¥
10
7
-
2.7
¥
10
8
Acetate
+
Fe(III) 9.3
¥
10
7
2.1
¥
10
7
-
2.7
¥
10
8
Fe(II)
+
NO
3
–
2.4
¥
10
5
4.8
¥
10
4
-
9.6
¥
10
5
102
K. A. Weber
et al.
© 2005 Society for Applied Microbiology and Blackwell Publishing Ltd,
Environmental Microbiology
,
8
, 100–113
0 5 10 15 20 25 30 35 40 45
0
1
2
3
4
5
6
7
8
9
10
Fe(
II)
A
NO
3
–
Reduction
Fe(III) Reduction
NO
3
–
Dependent
Fe(II) Oxidation
Total Fe
Fe(II), Total Fe (mmol l
–1
)
0 5 10 15 20 25 30 35 40 45
0
500
1000
1500
2000
2500
3000
NO
2
–
NO
3
–
0
50
100
150
200
250
300
NH
4
+
B
NO
3
–
Reduction
Fe(III) Reduction
NO
3
–
Dependent
Fe(II) Oxidation
Time (day)
NO
3
–
and NH
4
+
(
m
M)
NO
2
–
(
m
M)
NO
3
–
Addition
reduction phase. Because this loss of Fe(II) occurred dur-
ing the period of transient NO
2
–
accumulation, it is possible
that abiotic Fe(II) oxidation by NO
2
–
generated during
organotrophic NO
3
–
reduction was responsible for this
result. However, abiotic Fe(II) oxidation by NO
3
–
can be
ruled out based on the results of pasteurized control cul-
tures (see below).
Fe(III) reduction [Fe(II) accumulation] commenced
once NO
3
–
decreased to below approximately 0.5 mM
(Fig. 2A and B) and continued until acetate was depleted
(data not shown), yielding 7.6 mmol l
-
1
of 0.5 M HCl-
extractable Fe(II) [equivalent to approximately 15% of the
initial Fe(III) content of the slurry]. Approximately 35% of
the HCl-extractable Fe(II) was present as dissolved Fe(II)
at the end of the Fe(III) reduction phase. Reduction of the
synthetic goethite resulted in an obvious colour change in
the mineral from gold-yellow to dark greenish-brown.
Mixed Fe(II)–Fe(III) phases such as magnetite and/or
green rust were not detected by X-ray diffraction (XRD)
(Fig. 3A). However, comparison of a low-temperature
(77K) Mössbauer spectra for the reduced goethite with
that from a sample of microbially reduced (
Shewanella
putrefaciens
in AQDS and HCO
3
–
containing medium)
natural goethite (Kukkadapu
et al
., 2001) indicated the
presence of trace amounts of Fe(II) associated with
green rust (Fig. 3B). The formation of only minor
amounts of distinct Fe(II)-bearing mineral phases is con-
sistent with other recent studies of the end-products of
natural and synthetic goethite reduction by dissimilatory
Fe(III)-reducing bacteria (Kukkadapu
et al
., 2001;
Zachara
et al
., 2001). The vast majority of solid-
associated Fe(II) was presumably sorbed and/or
Fig. 2.
Change in 0.5 M HCl-extractable Fe(II)
and total Fe (A); and NO
3
–
, NO
2
–
and NH
4
+
(B)
over time in the wetland sediment enrichment
culture. Arrow denotes NO
3
–
amendment to
induce nitrate-dependent Fe(II) oxidation. Error
bars indicate standard error of triplicate cul-
tures; bars not visible are smaller than symbol.
Anaerobic redox cycling of iron in sediments 103
© 2005 Society for Applied Microbiology and Blackwell Publishing Ltd, Environmental Microbiology, 8, 100–113
surface-precipitated on residual Fe(III) oxide surfaces
(Zachara et al., 2001; Roden and Urrutia, 2002).
Addition of NO
3
–
following cessation of Fe(III) reduction
resulted in immediate oxidation of Fe(II) and consumption
of NO
3
–
(Fig. 2A). Subcultures removed from the primary
enrichment cultures and pasteurized prior to NO
3
–
re-
addition showed no Fe(II) oxidation or NO
3
–
consumption
(Fig. 4). Biological oxidation caused the microbially
reduced goethite to change from greenish-brown back to
its original goldish-yellow colour. Approximately 85% of
0.5 M HCl-extractable Fe(II) was oxidized within 15 days
in live cultures. Total 0.5 M HCl-extractable Fe
[Fe(II) + Fe(III)] decreased in parallel with HCl-extractable
Fe(II) during nitrate-dependent Fe(II) oxidation (Fig. 2A),
which suggests the production of crystalline Fe(III) oxide
phases not soluble in 0.5 M HCl. X-ray diffraction and
Mössbauer spectra of the nitrate-oxidized material were
virtually identical to those of the reduced mineral (Fig. 3),
suggesting that goethite was likely reformed.
In contrast to the initial organotrophic NO
3
–
reduction
phase of the experiment, nitrate-dependent Fe(II) oxida-
tion did not result in the transient accumulation of NO
2
–
(< 2 mM). Significant accumulation of NH
4
+
(approximately
0.9 mM) took place during nitrate-dependent Fe(II) oxida-
tion (Fig. 2B). The molar ratio of NO
3
–
reduced to Fe(II)
oxidized (0.191, r
2
= 0.983) was higher than the theoreti-
cal ratio for Fe(II) oxidation coupled to NO
3
–
reduction to
NH
4
+
(0.125), which indicates that small quantities of end-
products other than NH
4
+
(e.g. N
2
, NO, and/or N
2
O) were
likely produced.
10 20 30 40 50 60 70
2-Theta (°)
Intensity (%)
Reduced
NO
3
-
Oxidized
HSA Goethite
A
Reduced
NO
3
–
Oxidized
HSA Goethite
A
–10 –8 –6 –4 –2 0 2 4 6 8 10
94
95
96
97
98
99
100
B
Velocity (mm s
–1
)
% Transmission
–2 –1 0 1 2 3 4
98.8
99.2
99.6
100.0
Fig. 3. X-ray diffraction (A) and 77K Möss-
bauer (B) spectra of microbially reduced and
nitrate-dependent oxidized HSA goethite from
the sediment enrichment culture. The ‘HSA
goethite’ spectrum in panel A is from a mineral
preparation similar (but not identical) to the
material used in the enrichment culture experi-
ment; major peak lines for a reference goethite
phase are shown at bottom. Thick and thin solid
lines in panel B correspond to microbially
reduced and nitrate-dependent oxidized HSA
goethite, respectively, from the sediment
enrichment culture. The dashed line shows
results for microbially reduced natural goethite
from Kukkadapu and colleagues (2001). Arrows
in panel B point to an Fe(II) doublet (superim-
posed on the goethite sextet) that can be attrib-
uted to green rust (Kukkadapu et al., 2001).
