Desulfovibrio vulgaris as a model microbe for the study of corrosion under sulfate‐reducing conditions

Abstract Corrosion of iron‐containing metals under sulfate‐reducing conditions is an economically important problem. Microbial strains now known as Desulfovibrio vulgaris served as the model microbes in many of the foundational studies that developed existing models for the corrosion of iron‐containing metals under sulfate‐reducing conditions. Proposed mechanisms for corrosion by D. vulgaris include: (1) H2 consumption to accelerate the oxidation of Fe0 coupled to the reduction of protons to H2; (2) production of sulfide that combines with ferrous iron to form iron sulfide coatings that promote H2 production; (3) moribund cells release hydrogenases that catalyze Fe0 oxidation with the production of H2; (4) direct electron transfer from Fe0 to cells; and (5) flavins serving as an electron shuttle for electron transfer between Fe0 and cells. The demonstrated possibility of conducting transcriptomic and proteomic analysis of cells growing on metal surfaces suggests that similar studies on D. vulgaris corrosion biofilms can aid in identifying proteins that play an important role in corrosion. Tools for making targeted gene deletions in D. vulgaris are available for functional genetic studies. These approaches, coupled with instrumentation for the detection of low concentrations of H2, and proven techniques for evaluating putative electron shuttle function, are expected to make it possible to determine which of the proposed mechanisms for D. vulgaris corrosion are most important.


INTRODUCTION
Understanding the mechanisms for the corrosion of metals is key to developing strategies for preventing this economically significant problem 1,2 .Following the first suggestion that microbes might be important catalysts for the corrosion of metals 3,4 , a thorough analysis of the available data led to the conclusion that sulfate-reducing microorganisms play a key role in iron corrosion 5 .However, in the 1930s, Spirillium desulfuricans was the only microbe known to be capable of sulfate reduction 5 .The genus and species names of this and similar strains of sulfatereducing bacteria investigated in corrosion studies have changed over time 6,7 , but most are now generally recognized as strains of Desulfovibrio vulgaris (Table 1).
There is substantial evidence that Desulfovibrio species are involved in the corrosion of iron-containing metals in anaerobic environments 69,70 .Desulfovibrio species were abundant within the microbial community on metal surfaces exposed to oil field production waters 71,72 , corroded oil pipelines 73 , corroding steel pipe carrying oily seawater 74 , rust layers on steel plates immersed in seawater 75 , and the inner rust layer on carbon steel 76 .Desulfovibrio species were recovered in culture from corrosion sites [77][78][79][80] , including a D. vulgaris strain isolated from an oil field separator in the Gulf of Mexico that was damaged by corrosion 41 .Microbial activity on the cathodes of bioelectrochemical systems is thought to be related to microbial corrosion 81 and Desulfovibrio species are often enriched on cathodes from diverse microbial communities [82][83][84] .
Several mechanisms for D. vulgaris corrosion of ironcontaining metals have been proposed (Figure 1).These mechanisms may not be mutually exclusive.As detailed in this review, each of these models still requires rigorous examination.However, with the increasing availability of molecular tools to probe microbial activity and tools for genetic manipulation of D. vulgaris 85,86 , it now may be the time to either eliminate or confirm some of the existing

IRON CORROSION VIA AN H 2 INTERMEDIATE
The corrosion of iron-containing metals results from the oxidation of metallic iron to ferrous iron:  As recognized in the early analysis of corrosion by sulfate reducers 5 , protons are a likely electron acceptor for the electrons derived from Fe 0 .In early studies, the product of proton reduction is often referred to as "metallic hydrogen," but in the absence of data demonstrating that this form of hydrogen exists on the surface of corroding iron or can serve as an electron donor for microbial respiration, we assume that proton reduction yields H 2 , a known electron donor for diverse microbes: H 2 is an electron donor for D. vulgaris: and growth on H 2 is possible if acetate is provided as a carbon source 87 .Substantial abiotic H 2 production from Fe 0 was discounted in the early version of the model in which H 2 serves as an electron carrier between Fe 0 and cells 5 .However, the mechanism by which cells promoted the oxidation of Fe 0 with the production of H 2 was not specified.It is now known that the extent of abiotic H 2 production depends upon the form of the iron-containing metal.For example, pure Fe 0 abiotically produces substantial H 2 when submerged in anoxic water at circumneutral pH whereas 316 stainless steel does not 88,89 .
This difference in H 2 production between Fe 0 and stainless steel could provide one method for evaluating whether D. vulgaris relies on H 2 production to consume electrons from iron-containing metals.The closely related sulfate reducer D. ferrophilus reduced sulfate when pure Fe 0 was the electron donor, but not in a medium with stainless steel 90 .In contrast, Geobacter species capable of direct electron uptake could use either iron form as an electron donor 88-90 .These results indicated that D. ferrophilus was incapable of direct electron uptake and required the production of H 2 to mediate electron transfer between Fe 0 and cells.
The most direct approach to evaluating whether H 2 is an important intermediate in electron uptake from extracellular electron donors may be to generate a mutant that is unable to consume H 2 91 .D. vulgaris Hildenborough has multiple hydrogenases that have different localizations and metal constituents: the periplasmic [NiFe] HynAB-1 and HynAB-2, the periplasmic [Fe] HydAB, the periplasmic [NiFeSe] HysAB, the cytoplasmic [Fe] HydC, and the cytoplasmic membranebound Coo and Ech hydrogenases 92 .Deletions of genes for HydAB, HynAB-1, or HysAB negatively impacted the growth of D. vulgaris Hildenborough with H 2 as the electron donor [93][94][95] .However, these single-gene deletion mutants and a double deletion mutant of HynAB-1 and HydAB 95 still grew on H 2 as the electron donor, indicating redundant, complementary functions of the multiple hydrogenases.Thus, the construction of a strain with multiple hydrogenase gene deletions may be required to rigorously evaluate the role of H 2 in corrosion.
Transcriptomic analysis comparing growth on H 2 supplied from proton reduction with an iron electrode poised at −1.1 V versus growth on H 2 simply bubbled into medium revealed that the genes for HynAB-1 and HydAB were more highly expressed during growth on the cathodic H 2 40 .Gene transcripts for HysAB were more abundant when H 2 was bubbled into the medium.Gene deletions that prevented the function of the HynAB-1 and HydAB hydrogenases inhibited electron uptake from the iron cathodes 40 , as might be expected for cathodes poised at a low potential to induce H 2 production.The impact of the hydrogenase gene deletions on corrosion of iron that was not artificially poised at a negative potential was not determined because wild-type cells could not be grown under these conditions 40 .Lack of growth on unpoised iron suggests an inability to use Fe 0 as an electron donor.This difference between artificially poised iron cathodes and unpoised iron metal is an important consideration when evaluating other studies 19 that have concluded that H 2 is an important intermediate in iron corrosion by D. vulgaris based on experiments with electrochemically poised iron electrodes.
However, there is some indirect evidence for H 2 serving as an intermediary electron carrier between Fe 0 and D. vulgaris, especially when H 2 is not the sole electron donor.D. vulgaris did not reduce sulfate when steel wool was provided as the sole electron donor, but when lactate was added as an additional electron donor, more sulfide was produced than was possible from lactate oxidation alone 20 .In contrast, D. sapovorans, which cannot utilize H 2 , did not produce substantially more sulfide when grown in the presence of lactate and steel wool, than when grown with lactate alone.These results suggested that H 2 was an intermediary electron carrier for D. vulgaris electron uptake from the steel wool during growth with lactate 20 .The expression of one or more uptake hydrogenases is expected to be upregulated when H 2 is serving as an electron donor [96][97][98] .Thus, transcriptional and/ or proteomic studies may be useful in further assessing the role of H 2 as an electron donor during corrosion in the presence of lactate.Additional indirect evidence for the importance of H 2 as an electron carrier was the finding that D. vulgaris corroded "mild steel" faster than the grampositive D. orientis, which cannot consume H 2 12,13 .Other early studies suggested that H 2 production from iron was not a mechanism for corrosion 10,11 .However, the medium for investigating the possibility for H 2 serving as an electron donor did not include acetate, which is required as a carbon source for growth on H 2 .Therefore, no conclusion on the role of H 2 is possible from those studies.

FACTORS PROMOTING H 2 PRODUCTION
In a diversity of microbes, hydrogenases released from lysed cells, or specifically transported to the outer surface of living cells, facilitate the production of H 2 from Fe 0 99-102 .For example, methanogens highly effective in corrosion can produce an extracellular hydrogenase that enhances H 2 production from Fe 0 99 .Such extracellular hydrogenases have not been reported in Desulfovibrio species, but moribund cells of D. vulgaris release periplasmic hydrogenases that can retain activity for months 103 .Subjecting D. vulgaris to starvation, a condition likely to promote cell death and lysis, enhanced corrosion 45,58 .Therefore, studies to evaluate the role of extracellular hydrogenases in corrosion by D. vulgaris are warranted.
The iron sulfide that precipitates on iron-containing metals during corrosion coupled to sulfate reduction may also increase H 2 production.The addition of FeS reduced the overpotential necessary to produce H 2 from iron cathodes, suggesting a role for FeS in promoting the formation of H 2 16   .
In studies in which the culture was grown on fumarate rather than via sulfate reduction, the current was generated at more positive potentials when FeS was deposited on either mild steel or platinum cathodes 14 .These results further indicate that FeS may serve as a catalyst for H 2 generation.However, in studies with carbon steel coupons, it appeared that higher accumulations of sulfide inhibited corrosion 57 .The ability of D. vulgaris to corrode steel with either benzyl viologen as the electron acceptor 15 or when growing on fumarate 21 demonstrated that sulfide production was not essential for corrosion.Thus, a clear-cut concept for the role of FeS in corrosion has yet to be established.
Technology for measuring H 2 concentrations at extremely low concentrations during corrosion is available 89 .Thus, with the appropriate H 2 detector it should be possible to directly evaluate the role of FeS in facilitating H 2 production from iron-containing metals, simply by monitoring H 2 generation in the presence or absence of different quantities of FeS precipitate.

ELECTRON SHUTTLES OTHER THAN H 2
Soluble redox-active molecules promote extracellular electron exchange between microbes and minerals, electrodes, and other microbial species [104][105][106][107] .These electron shuttles typically accelerate extracellular electron exchange by alleviating the need for outer-surface electron transfer components to establish direct electrical contact with particulate extracellular donors and acceptors.The addition of riboflavin and flavin adenine dinucleotide enhanced D. vulgaris corrosion of carbon steel and stainless steel 49,50,63 .However, amendments of these cofactors, which are important for the function of numerous proteins, could influence D. vulgaris growth and metabolism in many ways.To determine whether flavins can serve as an electron shuttle for the corrosion of iron-containing metals coupled to sulfate reduction it is necessary to demonstrate that: (1) the metals are capable of reducing the flavins; and (2) that the reduced flavins can serve as electron donors for sulfate reduction.

DIRECT ELECTRON UPTAKE
Direct electron uptake from iron-containing metals has been demonstrated with Geobacter sulfurreducens and Geobacter metallireducens [88][89][90] .Strains unable to utilize H 2 readily reduced fumarate, nitrate, or Fe(III) with pure Fe 0 or stainless steel as the electron donor.Deletion of genes for outer-surface, multiheme c-type cytochromes previously shown to be involved in electron exchange with other extracellular donors/acceptors inhibited the corrosion.
There are no examples of similar studies with sulfatereducing microorganisms.It was suggested that c-type cytochromes positioned in the outer membrane of D. vulgaris might be able to make an electrical connection with Fe 0108 .However, subsequent studies have indicated that D. vulgaris does not have outer-surface cytochromes 92 .Clear next steps in this line of investigation would be to rigorously verify whether cytochromes are exposed on the outer surface of D. vulgaris, and if so, evaluate their role in iron corrosion with the appropriate gene deletion studies.Genetic, biochemical, biophysical, and immunological approaches previously employed for investigating the location and function of the outer-surface cytochromes of Geobacter would be suitable [109][110][111][112] .
Although it has been suggested that several of the c-type cytochromes of D. ferrophilus may be localized in the outer membrane, no genetic studies have been conducted to determine whether they are involved in extracellular electron exchange 90,[113][114][115] .As noted above, experimental analysis of iron corrosion by D. ferrophilus has suggested that it relies on H 2 as an intermediary electron carrier rather than direct electron uptake 90 .

CONCLUSIONS AND FUTURE DIRECTIONS
Sulfate-reducing microorganisms are considered to be important agents for catalyzing the corrosion of ironcontaining metals and D. vulgaris has historically been the model microbe of choice for elucidating the mechanisms for corrosion by sulfate reducers.As summarized above (Figure 1), previous studies have suggested several mechanisms by which D. vulgaris may enhance corrosion, but each of the proposed mechanisms requires further experimental evaluation.The mechanisms for electron transfer between iron-containing metals and Geobacter species were elucidated with genome-scale transcriptomics coupled with phenotypic analysis of mutant strains in which the genes for proteins hypothesized to be involved in electron transfer were deleted 88,89 .Deletion of the genes for hydrogenases and hypothesized outer-surface electrical contacts made it possible to determine the role of H 2 as an intermediary electron carrier and to identify likely electrical contacts on the outer surface of the cell.A similar approach seems possible for the study of D. vulgaris corrosion mechanisms.Transcriptomics of D. vulgaris biofilms is possible 39,40 to aid in identifying components that may have increased expression during corrosion of iron-containing metals versus other growth modes.Other candidates for corrosion components may be identified from the known physiological roles of proteins or their cellular location.Methods for the targeted deletion of genes in D. vulgaris are available 85,86 making it possible to evaluate the function of proteins hypothesized to be of importance.In fact, an extensive library of D. vulgaris mutants is publicly available, potentially eliminating the substantial investment of time and resources required for mutant construction 86 .Comparison of corrosion capabilities with Fe 0 , which readily reduces protons to generate H 2 , versus stainless steel, which does not generate H 2 , provides an additional tool to evaluate the role of H 2 as an intermediary electron carrier for corrosion 89,90 .Mechanisms for the corrosion of carbon steel, the most common form of iron in structural materials, should also be investigated.The hypothesis that FeS deposits stimulate H 2 production from iron-containing metals should be readily addressable with highly sensitive H 2 detection systems 88,89 .Mechanistic and genetic [116][117][118][119][120] approaches for the study of the role of electron shuttles in electron transfer to minerals and electrodes should be applicable to the study of the role of flavins as electron shuttles for corrosion.Therefore, it is expected that D. vulgaris will continue to serve as an important model microbe for the further elucidation of mechanisms for corrosion under sulfate-reducing conditions.ORCID Derek R. Lovley http://orcid.org/0000-0001-7158-3555

Figure 1 .
Figure 1.Previously proposed mechanisms for Desulfovibrio vulgaris to promote the corrosion of iron-containing metals.These include the consumption of abiotically produced H 2 (1); consumption of H 2 generated via catalysis by FeS (2) or hydrogenase (3); direct electron transfer from Fe 0 to cells via outer-surface electron transport components on the cell surface (4); and Fe 0 oxidation via reduction of the oxidized form of soluble flavin electron shuttle (Flavin ox ) with reduced flavin (Flavin red ) serving as the electron donor for sulfate reduction (5).The studies proposing these mechanisms are cited in the main text.

Table 1 .
Iron corrosion studies with Desulfovibrio vulgaris.
6,7trains now considered to be Desulfovibrio vulgaris were previously designated as Spirillium desulfuricans, Vibrio desulfuricans, Desulfovibrio desulfuricans, and Desulfovibrio desulphuricans6,7.Therefore, microbes that were later renamed D. vulgaris are listed by the name designated in the original text.Only strain designations are listed for strains designated as D. vulgaris in the original text.Alternative designations for these strains are described in parentheses. b Pimary corrosion mechanism discussed.Ha, abiotic H 2 production from iron; Hs, H 2 production from iron-catalyzed by FeS mineral deposits; S, sulfide promoting iron corrosion; D, direct electron transfer; F, electron transfer with a flavin shuttle; N, not applicable (the studies focused on biofilm formation, growth inhibition, corrosion inhibition, etc.).+, lactate included; −, no lactate; − (+A), no lactate but acetate added; − (+F), cells grown on fumarate; CI, cast iron; CS, carbon steel; GS, galvanized steel; MS, mild steel; PP, pipeline steel; SS, stainless steel.