Bạn đang xem bản rút gọn của tài liệu. Xem và tải ngay bản đầy đủ của tài liệu tại đây (19.23 MB, 371 trang )
dealt with microorganisms that metabolize fermentable sugars and amino acids.
A detailed list of such organisms has been published (Lovley, 1987). None of
these organisms are able to conserve energy from Fe(II1) or Mn(IV) reduction and
all of them can grow in the absence of Fe(II1) or Mn(1V). Fe(II1) and Mn(IV)
merely serve as minor electron sinks for these organisms so that there is less
production of reduced organic compounds or H, when Fe(II1) and Mn(IV) are
Organisms that can completely oxidize fermentable sugars or amino acids to
carbon dioxide with Fe(II1) or Mn(IV) serving as the electron acceptor have never
been isolated and may not exist. Studies on the metabolism of glucose in several
types of sediments in which Fe(II1) reduction was the terminal electron-accepting
process indicated that glucose was first fermented to organic acids (primarily acetate) with the subsequent oxidation of the fatty acids to carbon dioxide. Thermodynamic considerations suggest that if organisms attempted to completely oxidize
glucose to carbon dioxide with Fe(II1) as the electron acceptor, they could be at a
competitive disadvantage with organisms that merely ferment glucose (Lovley
and Phillips, 1989). This is because the amount of energy released per mole of
electrons transferred is higher for fermentation than it is for glucose oxidation
coupled to Fe(II1) reduction.
Within the last decade, studies have indicated that there are many microorganisms which can conserve energy to support growth by coupling the oxidation of
organic compounds or H, to the reduction of Fe(II1) or Mn(1V). The first example
of this (Balashova and Zavarzin, 1980) was a Pseudomonas sp. that grew by oxidizing H2 with the reduction of Fe(II1) (reaction I , Table I). This organism did not
reduce Mn(IV) under the conditions evaluated. Shewanella (previously Alteromonas) putrefuciens, which has been isolated from a wide variety of environments (Obuekwe et al., 1981; Semple and Westlake, 1987; Myers and Nealson,
1988a; Nealson et al., 1991; DiChristina and DeLong, 1993), can use Mn(IV)
(reaction 2, Table I) as well as Fe(II1) as an electron acceptor for H, oxidation
(Lovley et al., 1989b). Another organism that can conserve energy to support
growth from H2 oxidation coupled to Fe(II1) reduction is the estuarine organism
known as strain BrY (Caccavo et al., 1992).
A number of microorganisms, most notably Desulfovibrio sp. which grow with
sulfate as the terminal electron acceptor, can also couple H, oxidation to Fe(II1)
reduction (Coleman et al., 1993; Lovley et al., 1993b). In contrast to the other
H2-oxidizing, Fe(II1) reducers, none of these organisms conserve energy to support growth from this metabolism. Despite this, Desulfovibrio may still be important catalysts for Fe(II1) reduction in some environments (see the following
Until recently, it was considered that HI-oxidizing, Fe(II1)-reducing microorganisms were limited in their ability to oxidize organic compounds with Fe(II1).
S. putrefaciens oxidizes formate with the reduction of Fe(II1) and Mn(IV) (Lovley
D. R. LOVLEY
Metal Reduction Reactions
+ 2Fe(III) + H,O
+ 2Fe(II) + 2H'
+ Mn(1V) + H,O
+ Mn(I1) + 2H'
+ 4Fe(III) + 2 H z 0
+ 2Mn(IV) + 2H,O
+ HC0,- + 4Fe(II) + 5H
+ H C 0 , - + 2Mn(II) + 5H
+ 2Fe(IlI) + 2H,O
+ HC0,- + 2Fe(II) + 3H '
+ Mn(1V) + 2H,O
+ HC0,- + Mn(I1) + 3H'
+ SFe(I1) + 9H '
+ I4Fe(II) + 16H'
+ 20Fe(II) + 23H'
+ 12Fe(II) + 14H'
Benzoate- + 3OFe(III) +
Phenol + 28Fe(III) + 17H,O
+ 30Fe(II) + 36H'
+ 28Fe(II) + 34H
+ 8Fe(III) + 4H,O
Propionate + I4Fe(III) +
Butyrate + 20Fe(III) + IOH,O
+ 34Fe(III) + 2 0 H z 0
+ 34Fe(II) + 41H'
+ 36Fe(11I) + 21H,O
+ 36Fe(lI) + 43H
+ 6Fe(III) + 4H,O
+ 6Fe(II) + 8H '
+ 3Mn0, + 4H
+ 3Mn(II) + 2 H 2 0
Mn(I1) + MnO,
2Mn(III) + 2H,O
+ IOFe(II1) + 7H,O
+ 9Fe(II) + lFeS + NH,' + 14H'
U ( I V ) + 2Fe(III)
2Fe(II) + Mn(1V)
2Fe(III) + Mn(I1)
+ 4%' + 4 H 2 0
+ 4Se0,-* + H '
3H2 + 2Cr(VI)
et al., 1989b) (reactions 3 and 4, Table I). However, the only multicarbon compounds it is known to oxidize with Fe(II1) as the electron acceptor are lactate and
pyruvate (Lovley et al., 1989b) and these are only incompletely oxidized to acetate and carbon dioxide (reactions 5-8, Table I). Other H1-oxidizing, Fe(II1)- and
Mn( 1V)-reducing microorganisms such as strain BrY and D. desulfuricans also
carry out these reactions. Clostridium pasteuranium can couple formate oxidation
to Fe(II1) reduction (Eisen, 1985; Lovley et al., 1991b), but does not conserve
energy to support growth from this metabolism (D. R. Lovley and E. J. P. Phillips,
Prior to 1987, no organisms that could completely oxidize organic compounds
to carbon dioxide with Fe(II1) or Mn(IV) serving as the electron acceptor had been
described. Now it is apparent that there is a large diversity of such organisms
(Coates et al., 1994). The first of these to be described was Geobacter metallireducens, a freshwater organism isolated from surficial river sediments (Lovley
et al., 1987, 1993a; Lovley and Phillips, 1988a). G. metallireducens oxidizes a
number of organic compounds completely to carbon dioxide with the reduction of
Fe( 111) or Mn(1V). Electron donors include acetate, propionate, butyrate, benzoate, ethanol, phenol, p-cresol, and toluene (reactions 9- 16, Table I).
To date, other organisms that can couple the complete oxidation of organic
compounds to the reduction of Fe(II1) or Mn(IV) have been found to be closely
related to G. metallireducens. Desulfuromonus is the genus in the delta proteobacteria that is phyologenetically closest to Geobacter, and the marine organism
Desulfuromonas acetoxidans as well as the freshwater strain, D. actexigens, can
grow by coupling the oxidation of acetate to the reduction of Fe(II1) (Roden and
Lovley, 1993a). Preliminary analysis of a partial sequence of the 16s rRNA of
strain PCA, an acetate-oxidizing organism isolated from ditch sediments, indi-
D. R. LOVLEY
H2, Acetate and
Other Short-Chain Fattv Acids)
Figure 1. Model for microbial oxidation of organic matter in soils and sediments with Fe(II1)
serving as the electron acceptor.
cates that it is closely related to Geobacter and Desulfuromonas (Caccavo et al.,
1994), as is strain H-2, an acetate-oxidizing organism isolated from a contaminated aquifer (Coates et al., 1994). An interesting finding with PCA and H-2 is
that they represent the first examples of organisms that use both H2 and acetate as
electron donors for Fe(II1) reduction.
From the known metabolism of the microorganisms that are available in pure
culture, as well as studies with purified enrichment cultures (Lonergan and Lovley,
199 1; Lovley, 199 1 ) and studies with natural communities of microorganisms living in sediments (Lovley and Phillips, 1989), it is possible to construct a model
for how complex organic matter can be oxidized to carbon dioxide with Fe(II1)
[and presumably by analogy, Mn(IV)] as the sole electron acceptor (Fig. I). In
this model, complex organic matter is first attacked by hydrolytic enzymes which
release monomeric compounds such as sugars and amino acids, long chain fatty
acids, and monoaromatics. Fermentative microorganisms metabolize the sugars
and amino acids, possibly reducing small amounts of Fe(III), but primarily fermenting the sugars and amino acids to short-chain fatty acids and H,. These
fermentation products are then oxidized with the reduction of Fe(1II). It was previously considered that the Fe(II1)-reducing microorganisms catalyzing the oxidation of the fermentation products conserved energy to support growth from this
metabolism, but organisms such as Desulfovibrio species which actively reduce
Fe(II1) but cannot grow by this reaction may be important in some environments
(Coleman et al., 1993). Furthermore, it was previously considered that the organisms which could completely oxidize the fermentation acids to carbon dioxide
were a distinct community from the H,-oxidizing organisms (Lovley, 199 1, I993),
but the discovery of Fe(II1) reducers that can metabolize both short-chain fatty
acids and H, indicates that this is not necessarily the case. Long-chain fatty acids
can be oxidized to carbon dioxide with the reduction of Fe(III), but there is no
organism in pure culture that can serve as a model for this metabolism. G. metallireducens is the only pure culture model for the direct oxidation of aromatic compounds to carbon dioxide with the reduction of Fe(II1). However, as with the other
types of metabolism involved with organic matter oxidation coupled to Fe(II1)
reduction, there are certain to be a wide diversity of as yet to be isolated organisms
that can carry out these reactions.
2. So-OxidizingFe(II1) and Mn(IV) Reducers
Under acidic conditions, Thiobucillus thiooxiduns, iT: ferrooxidans, and Sulfolobus acidocaldarius can oxidize S' to sulfate with Fe(II1) serving as the electron
acceptor (reaction 17, Table I). Early studies indicated that energy to support
growth was not conserved from this metabolism (Kino and Usami, 1982; Sugio
et al., 1988; Sand, 1989). However, an electron transport chain inhibitor, HQNO
(2-n-heptyl-4-hydroxyquinolineN-oxide), inhibited Fe(II1) reduction with S' by
7:ferrooxidans, suggesting that Fe(II1) reduction was linked to electron transport
(Corbett and Ingledew, 1987). More direct studies have demonstrated that S' oxidation coupled to Fe(II1) reduction can provide energy to support active transport
and growth (Pronk et al., 199 I , 1992).
Studies with aquatic sediments and soils have provided evidence suggesting
that Mn(IV) can also be reduced with S' as the electron donor (reaction 18,
Table I) (Tisdale and Bertranison, 1950; Vavra and Frederick, 1952; Aller and
Rude, 1988; King, 1990). In some instances this may be an indirect nonenzymatic
reaction in which Fe(II), sulfite, or other reductants produced during S' oxidation
chemically react with Mn(IV) (Vavra and Frederick, 1952; Sugio er al., 1988).
However, S oxidation coupled to Mn(IV) reduction in aquatic sediments appears
to be biologically mediated (Aller and Rude, 1988; King, 1990).
Sulfate is produced when Mn(IV) is added to anoxic, sulfide-containing sediments, but additions of Fe(II1) do not result in sulfate production (Aller and Rude,
1988; King, 1990). Both Mn(1V) and Fe(II1) nonenzymatically oxidize sulfide,
but only to the redox level of S' (Goldhaber and Kaplan, 1974; Burdige and Nealson, 1986). This suggests that the production of sulfate in the presence of Mn(IV)
is microbially catalyzed. This conclusion is supported by the finding that metabolic inhibitors inhibited the Mn(IV)-dependent sulfate production (Aller and
It has been suggested that the microorganisms responsible for S' oxidation coupled to Mn(IV) reduction might be sulfate reducers and related microorganisms
D. R. LOVLEY
(Lovley and Phillips, 1994b). Sulfate reducers catalyzing this reaction in cell suspensions include D. desulfuricans, Desulfomicrobium baculatum, and Desulfobacterium autotrophicum. The S *- and/or Fe(II1)-reducing microorganisms
D. acetoxidans and G. metallireducens also carry out this reaction. None of these
organisms could be grown with S' as the sole electron donor and Mn(IV) as the
electron acceptor. However, many of these organisms are likely to be already
growing in sediments via different mechanisms, and thus will often be present in
large enough numbers to catalyze S' oxidation as a side reaction (Lovley and
Phillips, 1994b). In concurrence with the previous results with marine sediments,
these organisms produced little or no sulfate from S o with Fe(II1) as a potential
It has been suggested that large rod-shaped magnetotactic bacteria that can be
observed in freshwater sediments might be capable of oxidizing sulfide with the
reduction of Fe(II1) (Spring et al., 1993). This speculation was based primarily on
the observation of the accumulation of sulfur and iron-containing magnetic particles in this organism. However, there is no direct proof for this metabolism and
the organism has yet to be grown under laboratory conditions or isolated in pure
3. Fe(II1) Reduction by Magnetotactic Bacteria
Magnetotactic microorganisms contain single domain magnetic particles which
permit the organisms to orient themselves in the Earth's geomagnetic field (Blakemore, 1982; Mann et al., 1990a). Magnetite is the magnetic mineral found in
many magnetotactic bacteria, but some, living in anoxic environments, may contain iron sulfides (Farina et ul., 1990; Mann etal., 1990b).An increasing diversity
of magnetotactic bacteria have been discovered (Moench, 1988; DeLong et al.,
1993; Sakaguchi et al., 1993; Spring et al., 1993). Furthermore, although magnetotactic bacteria were originally thought of as sediment organisms, they also
inhabit soils (Fassbinder et al., 1990) and water (Mann et al., 1987; Stolz, 1993).
The reduction of Fe(II1) that is part of the formation of magnetite inside magnetotactic bacteria (Mann et al., 1990a) may play an important role in the iron cycle
in some environments. This Fe(II1) reduction, which can take place either aerobically (Matsunaga et al., 1991) or under microaerophilic (Mann et al., 1984) or
anoxic conditions (Bazylinski et al., 1988), may have an important influence on
the iron geochemistry and magnetic properties of soils (Fassbinder et al., 1990),
sediments (Stolz et al., 1990), and water (Stolz, 1993).
It has been suggested that the most highly studied magnetotactic bacterium,
Aquaspirillum magnetotacticurn, can also act as a dissimilatory Fe(II1)-reducing
microorganism and reduce Fe(II1) with the release of extracellular Fe(I1) under
anaerobic conditions (Guerin and Blakemore, 1992). However, the importance of
this metabolism in natural environments is not clear. The rates of Fe(II1) reduction
by A . magnetoracricum were relatively slow in comparison with other dissimilatory Fe(II1) reducers and it could not be conclusively demonstrated whether energy to support growth was conserved from Fe(lI1) reduction (Guerin and Blakemore, 1992).
TO Fe(III) AND Mn(n3
As previously discussed in detail (Lovley, 1991), numerous early studies, most
notably those of Ottow and co-workers (Ottow, 1968, 1970b; Ottow and von Klopotek, 1969; Ottow and Munch, 1977; Munch and Ottow, 1983) (and references
therein), demonstrated that Fe(II1) and Mn(1V) reduction in the presence of dissimilatory Fe(II1)- and Mn(1V)-reducing microorganisms was the result of an enzymatic process. However, studies on enzymatic mechanisms for Fe(1II) and
Mn(1V) reduction in these organisms and many others (Bromfield, 1954a; Dailey
and Lascelles, 1977; Lascelles and Burke, 1978; Lundgren et al., 1983; De Vrind
er al., 1986; Short and Blakemore, 1986; Karavaiko er al., 1987; Ghiorse, 1988)
focused on organisms in which Fe(II1) or Mn(1V) reduction was not a primary
process for organic matter oxidation and growth. Thus, these studies may have
little relevance in explaining most of the Fe(II1) and Mn(1V) reduction in soils and
A major question has been the mechanisms by which electrons are transported
to the insoluble Fe(II1) and Mn(IV) oxides outside the cell. Numerous studies have
suggested that direct cell-oxide contact is required for Fe(II1) and Mn(1V) reduction (Munch and Ottow, 1983; Arnold er al., 1988; Lovley and Phillips, 1988a;
Lovley et al., 1991b; Caccavo et al., 1992). In G. metallireducens (Gorby and
Lovley, 199 1 ) and S. putrefaciens (Myers and Myers, 1993a) the Fe(II1)-citrate
reductase activity is localized in the membrane fraction. Growth under anoxic
conditions induces the production of c-type cytochromes in the outer membrane
of S. putrefuciens (Myers and Myers, 1992). This demonstrates a potential for
enhanced capacity for electron transport to metal oxides in contact with the outer
cell surface. However, it has yet to be demonstrated whether the c-type cytochromes themselves or another electron carrier which accepts electrons from the
c-type cytochrome actually donates electrons to the metal oxides.
The periplasmic c ? cytochrome in Desulfovibrio vulgaris can function as a
Fe(II1) oxide reductase (Lovley, 1993). Whether this is the physiological Fe(1II)
reductase has yet to be demonstrated. Two Fe(II1) reductases have been purified
from Thiobacillus ferrooxidans. A periplasmic enzyme oxidizes sulfide to sulfite
with the reduction of Fe(II1) (Sugio et al., 1987, 1989) and a membrane-bound
Fe(II1) reductase oxidizes sulfite to sulfate (Sugio et al., 1988). No Fe(II1) reductases have been purified from dissimilatory Fe(II1) reducers that conserve energy to support growth from Fe(II1) reduction at neutral pH. However, investiga-
D. R. LOVLEY
tions have indicated that the Fe(II1) reductases in such organisms are distinct from
the nitrate reductase (Myers and Nealson, 1990; Gorby and Lovley, 1991; DiChristina, 1992; Myers and Myers, 1993a).
It has been proposed that the electron carrier that transfers electrons from the
outer cell surface to the metal oxides may be metal adsorbed onto the outer surface of the cells (Ehrlich, 1993). In the Mn(1V)-reducing microorganism, strain
BIII 88, there is a strong correlation between the manganese content of the cell
envelopes and the capacity for the organism to reduce Mn(IV) (Ehrlich, 1993).
This has led to the suggestion that manganese in the outer membrane may be
directly involved in electron transport across the cell envelope/MnO, interface
(Ehrlich, 1993). In this model, Mn(I1) on the cell surface enters into a disproportionation reaction with the MnO,, generating Mn(II1) (reaction 19, Table I). The
Mn(II1) then accepts electrons from an as yet unspecified electron donor in the
electron transport chain and is reduced back to Mn(I1). An extension of this model
is the hypothesis that strain BIII 88 can reduce Mn(1V) aerobically as well as
anaerobically because Mn(I1) is not readily oxidized by 0, at neutral pH. It is
speculated that other organisms such as G. metallireducens and S. putrefuciens,
which can only reduce Mn(IV) in the absence of O,, use Fe(I1) rather than Mn(I1)
as an electron shuttle (Ehrlich, 1993). Fe(I1) cannot be effectively used as an electron shuttle in the presence of 0, because it is rapidly chemically oxidized. This
model is attractive but requires further verification.
For organisms that conserve energy to support growth from Fe(II1) reduction,
studies have implicated b- or c-type cytochromes (Obuekwe er al., 1981; Obuekwe and Westlake, 1982; Arnold et a/., 1986; DiChristina et al., 1988; Myers
and Myers, 1992; Lovley et al., 1993a; Naik et a/., 1993; Roden and Lovley,
1993a) and menaquinone (Lovley el al., 1993a; Myers and Myers, 1993b) as involved in electron transport to Fe(II1). However, detailed models for the electron
transport in these organisms have yet to be elucidated.
Although not providing direct information about rates of Fe(II1) and Mn( IV)
reduction, determinations of the numbers and types of Fe(II1)- and Mn(1V)reducing microorganisms can be helpful in evaluating this process. Many of the
techniques for culturing Fe(II1)- and Mn(1V)-reducing microorganisms in liquid
and solid media have already been reviewed (Lovley, 1991 ). The insoluble nature
of Fe(II1) and Mn(IV) oxides is often a limitation in these methods, especially for
culturing on solid medium where it is difficult to provide enough Fe(II1) or
Mn(1V) in contact with the organisms to permit colony development (Lovley,
1991). This problem can be overcome through the use of soluble Fe(II1) forms.
Fe(II1) citrate has been used as a soluble Fe(II1) form but is not always appropriate
as citrate degraders that are present in most soil and sediment samples consume
the citrate with the precipitation of the Fe(II1). Furthermore, citrate may inhibit
Fe(II1) reduction in some organisms (Lovley et al., 1993b; Roden and Lovley,
1993a). Fe(II1) chelated with nitrilotriacetic acid (NTA) is a much better source
of soluble Fe(II1) than Fe(II1) citrate because the NTA is resistant to anaerobic
degradation and does not inhibit Fe(II1) reduction in a wide range of Fe(II1)reducing microorganisms (Lovley et al., 1993b; Roden and Lovley, 1993a).
When the goal is merely isolation rather than enumeration, another potential
approach is to use an electron acceptor other than Fe(II1). For example, many of
the acetate-oxidizing Fe(II1)-reducing microorganisms can use fumarate or nitrate
as an electron acceptor. Thus, isolation of acetate-oxidizing fumarate or nitratereducing colonies may yield organisms that are also Fe(II1) reducers. This technique was used to obtain G. metallireducens in pure culture (Lovley and Phillips,
1988a) and several other acetate-oxidizing Fe(II1)-reducing microorganisms (J. D.
Coates, unpublished data). Fe(II1)-reducing microorganisms have even been purified from enrichments using standard aerobic culturing and isolation techniques
(Caccavo et al., 1992).
Techniques for enumerating Fe(II1)- and Mn(1V)-reducing microorganisms
which do not require isolation of the organisms from the environment are also
emerging. An oligonucleotide probe for the 16s rRNA of S. putrefaciens was
successfully employed for detecting S. putrefaciens in freshwater aquatic sediments (DiChristina and DeLong, 1993). The finding that acetate-oxidizing Fe(II1)and Mn(1V)-reducing microorganisms cluster in a tight phylogenetic group in the
delta proteobacteria (see previous discussion) suggests that the 16s rRNA probe
approach may prove useful in detecting organisms in this important metabolic
In a similar manner, the study of the ecology of magnetotactic bacteria has been
limited by the difficulty in culturing these organisms (Mann et al., 1990a). However, recent advancements in phylogenetic studies of the magnetotactic bacteria
have made it possible to develop 16s rRNA probes which might be used to study
populations of unculturable magnetotactic bacteria in soils and sediments (Eden
etal., 1991; Schleifer etal., 1991; Spring et al., 1992, 1993; Burgess et al., 1993;
DeLong et nl., 1993).
Fe(II1)-reducing microorganisms can also potentially be monitored through
analysis of lipids in soils and sediments. For example, G. metallireducens is the
only bacterium known to contain lipopolysaccharide hydroxy fatty acids of 16
carbons with the hydroxyl on the 9th, loth, or I I th carbon (Lovley e f al., 1993a).
However, these midchain hydroxy fatty acids have been observed in sediments
and may serve as a marker for G. metallireducens and related organisms.
The Fe(111)-reducing microorganism S. putrefaciens has a lipopolysaccharide
D. R. LOVLEY
structure that is different than that of most terrestrial Gram-negative bacteria
(Pickard et al., 1993). The overall lipid profile of this organism is also highly
unusual when compared with other Gram-negative bacteria (Moule and Wilkinson, 1987). These might be useful characteristics for screening for the potential
presence of S. putrefaciens.
Techniques for studying the activity of Fe(II1) and Mn(1V) reducers are similar
to those that are used for ecological investigations of other anaerobic organisms.
Although the following discussion focuses on Fe(II1) reduction, the study of
Mn(IV) reduction is similar. Environmental rates of Fe(II1) reduction can be estimated with laboratory incubations of sediments or soils under anaerobic conditions and following the production of Fe(I1) and/or the loss of Fe(II1) over time
(Kamura et al., 1963; Sorensen, 1982; Jones et al., 1984; Lovley and Phillips,
1986a,b, 1988a,b; Ellis-Evans and Lemon, 1989; Canfield et al., 1993a). As previously reviewed (Lovley, 1991), most of the Fe(I1) that is produced during Fe(II1)
reduction remains in solid phases. Therefore, for quantitative estimates it is important to use a technique such as HCl extraction (Lovley and Phillips, 1986a,b)
which accounts for this solid-phase Fe(I1).
The limitation of such anoxic laboratory incubations is that they may distort in
situ conditions (Lovley and Phillips, 1986a,b; Canfield et al., 1993a). For other
processes, most notably sulfate reduction (Jorgensen, 1978), this problem has
been significantly overcome by injecting radiolabeled tracers into undisturbed
sediment cores. However, this approach is not feasible for Fe(II1) reduction because there is rapid isotope exchange between the Fe(II1) and Fe(I1) pools (Roden
and Lovley, 1993b).
In some instances it may be possible through geochemical modeling of solid
and dissolved iron and manganese phases to estimates rates of Fe(II1) and Mn(IV)
reduction (Aller, 1980; Burdige and Gieskes, 1983; Aller, 1990; Baedecker et al.,
1993). However, this approach can require significant assumptions about the equilibrium solubility of the metals for which there is often little supporting data.
A particularly novel approach for estimating rates of Fe(II1) and Mn(IV) reduction was developed by Canfield and co-workers (1993a,b) for continental margin
sediments. At depths in the sediments where 0, and nitrate were depleted it could
be assumed that organic matter oxidation could be attributed to Mn(IV) reduction,
Fe(II1) reduction, or sulfate reduction. Overall rates of organic matter oxidation in
these zones were estimated from rates of ammonia production and in one instance
were further confirmed with estimates of inorganic carbon production. Subtraction of the contribution of sulfate reduction as measured with the 35SOd-tra~er
technique gave an estimate of the amount of organic matter oxidation coupled to
Fe(II1) and Mn(1V) reduction. Proportioning electron flow between Fe(II1) reduction and Mn(1V) reduction was based on results of sediment incubation studies
and geochemical considerations.
MECHANISMS FOR ENVIRONMENTAL
AND Mn(IV) REDUCTION
The fact that microorganisms can reduce Fe(II1) and Mn(1V) has been known
for a century (Adeney, 1894; Harder, 1919; Allison and Scarseth, 1942). However,
as previously discussed in detail (Lovley et al., 1991b), it has only become apparent within the last decade that enzymatic reduction of Fe(II1) is the major mechanism for Fe(II1) reduction in most soils and sediments. Although the relative importance of enzymatic and nonenzymatic mechanisms for Mn(1V) reduction have
not been studied in as much detail, there are instances in which enzymatic Mn(1V)
reduction must predominate (discussed next).
The relative lack of significance of nonenzymatic Fe(II1) reduction in sediments
can be clearly seen by the lack of Fe(II1) reduction in sterilized sediments and by
the fact that Fe(II1) reduction has temperature optima characteristic of an enzymatically catalyzed reaction (Bromfield, 1954b; Kamura et al., 1963; Aristovskaya and Zavarzin, I97 1; Sorensen, 1982; Jones et al., 1983; Lovley and Goodwin, 1988; Canfield, 1989). Furthermore, model organic compounds that were
oxidized to carbon dioxide in live Fe(lI1)-reducing sediments were not oxidized
when the microorganisms were killed with heat (Lovley et al., 1991b). These
findings indicate that the bulk of the organic matter in sediments cannot directly
reduce Fe(II1). In fact, testing of a wide variety of organics revealed that very few
can nonenzymatically reduce Fe(III), especially at the circumneutral pH typical of
most submerged soils and aquatic sediments (Lovley et al., 1991b). This includes
the common microbial metabolites that are produced during anaerobic metabolism. In contrast, all of these organics could be oxidized to carbon dioxide with
Fe(II1) serving as the electron acceptor when the appropriate Fe(II1)-reducing microorganisms were present.
Furthermore, even for those few organic compounds that can enzymatically
reduce Fe(III), the extent of Fe(II1) reduction and oxidation of the organic compound is trivial compared to oxidation of the same compounds coupled to microbial Fe(II1) reduction. For example, although most monoaromatic compounds do
not nonenzymatically reduce Fe(III), a few can (LaKind and Stone, 1989; Lovley
et al., 199 1b). However, the nonenzymatic reaction is typically only a two electron
transfer, it does not result in any carbon dioxide production, and it leaves an intact
ring structure. In contrast, in the presence of the appropriate Fe(I1I)-reducing mi-