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

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 corrosion⁵. However, in the 1930s, Spirillium desulfuricans was the only microbe known to be capable of sulfate reduction⁵. The genus and species names of this and similar strains of sulfate-reducing 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 pipelines⁷³, corroding steel pipe carrying oily seawater⁷⁴, rust layers on steel plates immersed in seawater⁷⁵, and the inner rust layer on carbon steel⁷⁶. Desulfovibrio species were recovered in culture from corrosion sites^77-80, including a D. vulgaris strain isolated from an oil field separator in the Gulf of Mexico that was damaged by corrosion⁴¹. Microbial activity on the cathodes of bioelectrochemical systems is thought to be related to microbial corrosion⁸¹ and Desulfovibrio species are often enriched on cathodes from diverse microbial communities^82-84.

Several mechanisms for D. vulgaris corrosion of iron-containing 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 mechanistic models for D. vulgaris corrosion or to develop new models. The purpose of this review is to summarize the previously proposed routes for Desulfovibrio species iron corrosion and to suggest experimental approaches to further advance the understanding of corrosion by this popular model microbe.

IRON CORROSION VIA AN H₂ INTERMEDIATE

and growth on H₂ is possible if acetate is provided as a carbon source⁸⁷.

Substantial abiotic H₂ production from Fe⁰ was discounted in the early version of the model in which H₂ serves as an electron carrier between Fe⁰ and cells⁵. However, the mechanism by which cells promoted the oxidation of Fe⁰ with the production of H₂ was not specified. It is now known that the extent of abiotic H₂ production depends upon the form of the iron-containing metal. For example, pure Fe⁰ abiotically produces substantial H₂ when submerged in anoxic water at circumneutral pH whereas 316 stainless steel does not^{88, 89}.

This difference in H₂ production between Fe⁰ and stainless steel could provide one method for evaluating whether D. vulgaris relies on H₂ production to consume electrons from iron-containing metals. The closely related sulfate reducer D. ferrophilus reduced sulfate when pure Fe⁰ was the electron donor, but not in a medium with stainless steel⁹⁰. 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 H₂ to mediate electron transfer between Fe⁰ and cells.

The most direct approach to evaluating whether H₂ is an important intermediate in electron uptake from extracellular electron donors may be to generate a mutant that is unable to consume H₂⁹¹. 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 membrane-bound Coo and Ech hydrogenases⁹². Deletions of genes for HydAB, HynAB-1, or HysAB negatively impacted the growth of D. vulgaris Hildenborough with H₂ as the electron donor^93-95. However, these single-gene deletion mutants and a double deletion mutant of HynAB-1 and HydAB⁹⁵ still grew on H₂ 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 H₂ in corrosion.

Transcriptomic analysis comparing growth on H₂ supplied from proton reduction with an iron electrode poised at −1.1 V versus growth on H₂ simply bubbled into medium revealed that the genes for HynAB-1 and HydAB were more highly expressed during growth on the cathodic H₂⁴⁰. Gene transcripts for HysAB were more abundant when H₂ was bubbled into the medium. Gene deletions that prevented the function of the HynAB-1 and HydAB hydrogenases inhibited electron uptake from the iron cathodes⁴⁰, as might be expected for cathodes poised at a low potential to induce H₂ 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 conditions⁴⁰. Lack of growth on unpoised iron suggests an inability to use Fe⁰ as an electron donor. This difference between artificially poised iron cathodes and unpoised iron metal is an important consideration when evaluating other studies¹⁹ that have concluded that H₂ 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 H₂ serving as an intermediary electron carrier between Fe⁰ and D. vulgaris, especially when H₂ 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 alone²⁰. In contrast, D. sapovorans, which cannot utilize H₂, 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 H₂ was an intermediary electron carrier for D. vulgaris electron uptake from the steel wool during growth with lactate²⁰. The expression of one or more uptake hydrogenases is expected to be upregulated when H₂ is serving as an electron donor^96-98. Thus, transcriptional and/or proteomic studies may be useful in further assessing the role of H₂ as an electron donor during corrosion in the presence of lactate. Additional indirect evidence for the importance of H₂ as an electron carrier was the finding that D. vulgaris corroded “mild steel” faster than the gram-positive D. orientis, which cannot consume H₂^{12, 13}.

Other early studies suggested that H₂ production from iron was not a mechanism for corrosion^{10, 11}. However, the medium for investigating the possibility for H₂ serving as an electron donor did not include acetate, which is required as a carbon source for growth on H₂. Therefore, no conclusion on the role of H₂ is possible from those studies.

FACTORS PROMOTING H₂ 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 H₂ from Fe^099-102. For example, methanogens highly effective in corrosion can produce an extracellular hydrogenase that enhances H₂ production from Fe⁰⁹⁹. Such extracellular hydrogenases have not been reported in Desulfovibrio species, but moribund cells of D. vulgaris release periplasmic hydrogenases that can retain activity for months¹⁰³. 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 H₂ production. The addition of FeS reduced the overpotential necessary to produce H₂ from iron cathodes, suggesting a role for FeS in promoting the formation of H₂¹⁶. 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 cathodes¹⁴. These results further indicate that FeS may serve as a catalyst for H₂ generation. However, in studies with carbon steel coupons, it appeared that higher accumulations of sulfide inhibited corrosion⁵⁷. The ability of D. vulgaris to corrode steel with either benzyl viologen as the electron acceptor¹⁵ or when growing on fumarate²¹ 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 H₂ concentrations at extremely low concentrations during corrosion is available⁸⁹. Thus, with the appropriate H₂ detector it should be possible to directly evaluate the role of FeS in facilitating H₂ production from iron-containing metals, simply by monitoring H₂ generation in the presence or absence of different quantities of FeS precipitate.

ELECTRON SHUTTLES OTHER THAN H₂

Soluble redox-active molecules promote extracellular electron exchange between microbes and minerals, electrodes, and other microbial species^104-107. 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^88-90. Strains unable to utilize H₂ readily reduced fumarate, nitrate, or Fe(III) with pure Fe⁰ 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 sulfate-reducing 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 Fe⁰¹⁰⁸. However, subsequent studies have indicated that D. vulgaris does not have outer-surface cytochromes⁹². 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^109-112.

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, 113-115}. As noted above, experimental analysis of iron corrosion by D. ferrophilus has suggested that it relies on H₂ as an intermediary electron carrier rather than direct electron uptake⁹⁰.

CONCLUSIONS AND FUTURE DIRECTIONS

Sulfate-reducing microorganisms are considered to be important agents for catalyzing the corrosion of iron-containing 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 H₂ 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 construction⁸⁶. Comparison of corrosion capabilities with Fe⁰, which readily reduces protons to generate H₂, versus stainless steel, which does not generate H₂, provides an additional tool to evaluate the role of H₂ 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 H₂ production from iron-containing metals should be readily addressable with highly sensitive H₂ detection systems^{88, 89}. Mechanistic and genetic^116-120 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.