Metal Reduction and Protein Secretion Genes Required for Iodate Reduction by Shewanella oneidensis

Microbial iodate (IO₃⁻) reduction is a major component in the biogeochemical cycling of iodine and the bioremediation of iodine-contaminated environments; however, the molecular mechanism of microbial IO₃⁻ reduction is poorly understood. Results of the present study indicate that outer membrane (type II) protein secretion and metal reduction genes encoding the outer membrane MtrAB module of the extracellular electron conduit MtrCAB are required for IO₃⁻ reduction by S. oneidensis. On the other hand, the metal-reducing c-type cytochrome MtrC of the extracellular electron conduit is not required for IO₃⁻ reduction by S. oneidensis. These findings indicate that the IO₃⁻ electron transport pathway terminates with an as yet unidentified IO₃⁻ reductase that associates with the outer membrane MtrAB module to deliver electrons extracellularly to IO₃⁻.

ABSTRACT

The metal-reducing gammaproteobacterium Shewanella oneidensis reduces iodate (IO₃⁻) as an anaerobic terminal electron acceptor. Microbial IO₃⁻ electron transport pathways are postulated to terminate with nitrate (NO₃⁻) reductase, which reduces IO₃⁻ as an alternative electron acceptor. Recent studies with S. oneidensis, however, have demonstrated that NO₃⁻ reductase is not involved in IO₃⁻ reduction. The main objective of the present study was to determine the metal reduction and protein secretion genes required for IO₃⁻ reduction by Shewanella oneidensis with lactate, formate, or H₂ as the electron donor. With all electron donors, the type I and type V protein secretion mutants retained wild-type IO₃⁻ reduction activity, while the type II protein secretion mutant lacking the outer membrane secretin GspD was impaired in IO₃⁻ reduction. Deletion mutants lacking the cyclic AMP receptor protein (CRP), cytochrome maturation permease CcmB, and inner membrane-tethered c-type cytochrome CymA were impaired in IO₃⁻ reduction with all electron donors, while deletion mutants lacking c-type cytochrome MtrA and outer membrane β-barrel protein MtrB of the outer membrane MtrAB module were impaired in IO₃⁻ reduction with only lactate as an electron donor. With all electron donors, mutants lacking the c-type cytochromes OmcA and MtrC of the metal-reducing extracellular electron conduit MtrCAB retained wild-type IO₃⁻ reduction activity. These findings indicate that IO₃⁻ reduction by S. oneidensis involves electron donor-dependent metal reduction and protein secretion pathway components, including the outer membrane MtrAB module and type II protein secretion of an unidentified IO₃⁻ reductase to the S. oneidensis outer membrane.

IMPORTANCE Microbial iodate (IO₃⁻) reduction is a major component in the biogeochemical cycling of iodine and the bioremediation of iodine-contaminated environments; however, the molecular mechanism of microbial IO₃⁻ reduction is poorly understood. Results of the present study indicate that outer membrane (type II) protein secretion and metal reduction genes encoding the outer membrane MtrAB module of the extracellular electron conduit MtrCAB are required for IO₃⁻ reduction by S. oneidensis. On the other hand, the metal-reducing c-type cytochrome MtrC of the extracellular electron conduit is not required for IO₃⁻ reduction by S. oneidensis. These findings indicate that the IO₃⁻ electron transport pathway terminates with an as yet unidentified IO₃⁻ reductase that associates with the outer membrane MtrAB module to deliver electrons extracellularly to IO₃⁻.

INTRODUCTION

Iodine is a biologically active element commonly found in freshwater and marine environments in the forms of iodide (I⁻; −1 oxidation state) and iodate (IO₃⁻; +5 oxidation state) (1). IO₃⁻ is more thermodynamically stable than I⁻, yet I⁻ is the predominant form in the environment, potentially indicating that microbial IO₃⁻ reduction is a major component of the iodine biogeochemical reaction network (1–3). In marine environments, microbial IO₃⁻ reduction is the primary mechanism for IO₃⁻ reduction to I⁻ (4–8). Microbial IO₃⁻ reduction has also attracted interest as a component of alternative strategies for remediation of waters and sediments contaminated with radioactive iodine released to the environment (9). The nuclear waste product and radioactive isotope ¹²⁹I is produced during uranium and plutonium fission reactions and displays a half-life of 1.6 × 10⁷ years (10). Following the 2011 Fukushima nuclear reactor catastrophe, westerly winds deposited a large portion of the radioactive iodine in the Pacific Ocean, where radioactive IO₃⁻ and I⁻ were the predominant ¹²⁹I forms (11–13). Radioactive iodine is also found in contaminated groundwater at the U.S. Department of Energy Savannah River and Hanford sites (9, 13, 14). Despite the human health concerns surrounding the fate and transport of radioactive iodine in the environment, the molecular mechanism of microbial IO₃⁻ reduction remains poorly understood (15).

IO₃⁻-reducing microorganisms include the facultative anaerobe Shewanella oneidensis, which reduces a wide range of terminal electron acceptors, including oxidized forms of iron, manganese, nitrogen, sulfur, uranium, plutonium, technetium, and iodine (16–18). S. oneidensis also transfers electrons to a variety of extracellular electron acceptors, including Fe(III), Mn(III), and Mn(IV) oxides (19, 20). To transfer electrons to external Fe(III) oxides, S. oneidensis employs a variety of novel respiratory strategies, including (i) direct enzymatic reduction via decaheme c-type cytochromes associated with the extracellular electron conduit (EEC) located on the surface or surface extensions of the S. oneidensis outer membrane (21–23), (ii) extracellular electron transfer via endogenous or exogenous electron shuttling compounds (24–26), and (iii) nonreductive Fe(III) solubilization by organic ligands to produce more readily reducible soluble organic Fe(III) complexes (27–29).

Previous findings for other IO₃⁻-reducing microorganisms indicated that nitrate (NO₃⁻) reductase may catalyze the reduction of IO₃⁻ as an alternative electron acceptor (30–32). However, neither assimilatory nor dissimilatory NO₃⁻ reductases are required for IO₃⁻ reduction by S. oneidensis (33). The molecular mechanism of IO₃⁻ reduction by S. oneidensis has yet to be examined. The electron transport pathways of S. oneidensis consist of upstream dehydrogenases linked via the menaquinone pool and the inner membrane-tethered c-type cytochrome CymA to downstream terminal reductase complexes, including the metal-reducing EEC (19, 34, 35). The S. oneidensis EEC is comprised of outer membrane β-barrel protein MtrB (and essential cysteine residue C42) (36) and decaheme c-type cytochromes MtrA and MtrC (34, 37–40). MtrC is translocated to the outside face of the outer membrane through GspD, the outer membrane secretin of the type II protein secretion system (21, 41, 42). Other proteins essential for electron transport to external metal oxides include the c-type cytochrome maturation permease CcmB (43) and the cAMP receptor protein (CRP), required for anaerobic respiratory gene expression in S. oneidensis (44).

Although NO₃⁻ reductase is not required for IO₃⁻ reduction by S. oneidensis, identification of metal reduction and protein secretion genes involved in this process will aid in development of biomarkers to examine the potential for microbial IO₃⁻ reduction, a prominent process in iodine cycling in natural and contaminated environments such as the ¹²⁹I-contaminated Hanford and Savannah River sites. Likewise, such biomarkers could be used to track ¹²⁹I cycling in a pump-and-treat system currently treating contaminated groundwater at Hanford. Formation of I⁻ could lead to increased adsorption onto organic material, such as granular activated carbon (GAC) in a fluidized bed reactor (FBR) that is part of the pump-and-treat process, or by organic matter in environmental systems such as those found at Savannah River. The main objective of the present study was to test the hypothesis that the S. oneidensis metal reduction and protein secretion pathways required for Fe(III), Mn(III), and Mn(IV) oxide reduction are also involved in IO₃⁻ reduction. The experimental strategy to test the hypothesis included (i) construction of additional S. oneidensis gene deletion mutants lacking metal reduction and protein secretion pathway components and (ii) tests of the battery of metal reduction and protein secretion pathway mutants for IO₃⁻ reduction activity.

RESULTS

Effect of electron donor on IO₃⁻ reduction activity by the S. oneidensis wild-type strain.

A set of anaerobic incubations with batch cultures of the S. oneidensis wild-type strain was carried out to determine the optimum IO₃⁻ concentration that avoided IO₃⁻ or produced I⁻ toxicity and maximized the IO₃⁻ reduction activity of the S. oneidensis wild-type strain at cell densities of 10⁸ ml⁻¹. IO₃⁻ concentrations of >500 μM inhibited IO₃⁻ reduction activity, while 250 μM was the optimum initial IO₃⁻ concentration (Fig. 1). The IO₃⁻ reduction activities of the S. oneidensis wild-type strain with lactate and formate as electron donors were similar (512 and 455 nmol h⁻¹mg of protein⁻¹, respectively), while the IO₃⁻ reduction activity with H₂ as the electron donor was approximately 4-fold lower (120 nmol h⁻¹mg of protein⁻¹) (Table 1). The extents of reaction (of the initial 250 μM IO₃⁻ starting concentration) for IO₃⁻ reduction by the S. oneidensis wild-type strain with lactate and formate as electron donors (60% and 55%, respectively) were approximately 4-fold greater than the extent of reaction with H₂ as the electron donor (16%) (Table 1).

FIG 1.

Effect of IO₃⁻ concentration on IO₃⁻ reduction activity of S. oneidensis MR-1. IO₃⁻ reduction was performed in M1 medium amended with 20 mM lactate and 250 μM IO₃⁻, ranging from 0.1 to 2 mM iodate at room temperature and 300 rpm. Values are means of triplicate samples from anaerobic incubations. Error bars represent SDs. Some error bars cannot be seen due to small SDs. Symbols: ●, 0.10 mM; ■, 0.25 mM; ⧫, 0.50 mM; ▲, 1.00 mM; ▼, 1.50 mM; ○, 2.00 mM.

TABLE 1.

IO₃⁻ reduction activity of wild-type and mutant strains of S. oneidensis with lactate, formate, and H₂ as electron donors^a

Condition or strain	Lactate		Formate		H2
	IO3− reduction rateb (nmol h−1 mg of protein−1)d	Extent of reactionc (% of IO3− reduced to I−)d	IO3− reduction rateb (nmol h−1 mg of protein−1)d	Extent of reactionc (% of IO3− reduced to I−)d	IO3− reduction rateb (nmol h−1 mg of protein−1)d	Extent of reactionc (% of IO3− reduced to I−)d
Abiotic	0 ± 0 (0)	0 ± 0 (0)	0 ± 0 (0)	0 ± 0 (0)	0 ± 0 (0)	0 ± 0 (0)
MR-1	512.0 ± 19.9 (100)	59 ± 1 (100)	454.7 ± 25.7 (10)	55 ± 3 (100)	119.9 ± 48.3 (100)	16 ± 3 (100)
Δcrp mutant	77.5 ± 9.4 (15)	7 ± 0 (11)	122.1 ± 10.3 (27)	15 ± 1 (28)	45.2 ± 14.4 (38)	7 ± 4 (46)
Δcrp/crp mutant	274.9 (54)	50 (85)	ND	ND	ND	ND
ΔmtrA mutant	100.7 ± 14.1 (20)	9 ± 1 (16)	373 ± 9.2 (82)	43 ± 0 (79)	254.3 ± 41.1 (212)	32 ± 3 (197)
ΔmtrA/mtrA mutant	839.8 (164)	83 (141)	ND	ND	ND	ND
ΔmtrB mutant	14.6 ± 11 (3)	1 ± 0 (2)	488.1 ± 16.4 (107)	59 ± 3 (108)	174.4 ± 16.4 (145)	22 ± 0 (134)
ΔmtrB/mtrB mutant	916.2 (179)	85 (143)	ND	ND	ND	ND
ΔmtrC mutant	389.7 ± 29.3 (76)	49 ± 2 (82)	566.5 ± 20.5 (125)	62 ± 2 (113)	170.8 ± 5.1 (142)	24 ± 3 (147)
ΔomcA mutant	351.6 ± 19.9 (69)	44 ± 4 (75)	552.3 ± 0.5 (121)	63 ± 3 (116)	104.0 ± 7.2 (87)	16 ± 1 (98)
ΔmtrC ΔomcA mutant	565.3 ± 97.2 (110)	66 ± 15 (112)	644.9 ± 85.2 (142)	71 ± 3 (130)	171.1 ± 12.3 (139)	18 ± 3 (112)
ΔcymA mutant	52.7 ± 9.4 (10)	2 ± 7 (3)	114.1 ± 7.2 (25)	14 ± 1 (25)	67.7 ± 25.7 (56)	5 ± 0 (28)
ΔcymA/cymA mutant	407.1 (80)	58 (98)	ND	ND	ND	ND
ΔccmB mutant	51.9 ± 10.5 (10)	5 ± 2 (8)	102.2 ± 20 (22)	11 ± 2 (20)	65.5 ± 4.1 (55)	3 ± 3 (16)
ΔccmB/ccmB mutant	435.5 (85)	60 (101)	ND	ND	ND	ND
ΔtolC mutant	494.1 ± 22.3 (97)	62 ± 2 (104)	435.1 ± 21.6 (96)	50 ± 0 (92)	140.3 ± 27.7 (117)	16 ± 0 (101)
ΔgspD mutant	155.2 ± 68.5 (30)	16 ± 0 (27)	203.4 ± 4.1 (45)	22 ± 2 (40)	66.9 ± 37 (56)	9 ± 2 (55)
ΔSO3800 mutant	581.0 ± 9.4 (113)	64 ± 2 (109)	586.8 ± 26.7 (129)	65 ± 1 (118)	166.4 ± 42.1 (139)	20 ± 3 (124)
mtrB-C42A mutant	37 ± 71.4 (7)	4 ± 7 (7)	ND	ND	ND	ND
mtrB-C45A mutant	489.9 ± 23.4 (96)	60 ± 3 (102)	ND	ND	ND	ND

^aValues represent means of triplicate samples; errors represent 1 SD. ND, not determined.

^bReaction rate was calculated from the first 4-h anaerobic incubation (lactate and formate) or 8-h incubation (H₂) values.

^cExtent of reaction is reported as the percentage of IO₃⁻ reduced to I⁻ upon completion of the 24-h incubation period, after which further IO₃⁻ reduction was minimal.

^dThe values in parentheses are in comparison with the wild-type rates (percent) within each set of lactate, formate, or H₂ values.

IO₃⁻ reduction activity of S. oneidensis EEC mutant strains.

The IO₃⁻ reduction activity of the S. oneidensis EEC mutant strains was determined with either lactate, formate, or H₂ as the electron donor. The ΔmtrB and mtrB-C42A site-directed mutant strains were severely impaired in IO₃⁻ reduction activity with lactate as the electron donor (3% and 7% of the wild-type rate, respectively, and 2% and 7% of the wild-type extent of reaction, respectively). In contrast, the ΔmtrB mutant strain retained wild-type IO₃⁻ reduction activity with formate or H₂ as the electron donor (107% and 145% of the wild-type rate, respectively) (Fig. 2; Table 1).

FIG 2.

IO₃⁻ reduction activity of S. oneidensis wild-type (MR-1) and EEC mutant strains with IO₃⁻ as the electron acceptor and lactate, formate, or H₂ as the electron donor and mtrB-CxxC motif mutants and complemented strains of ΔmtrA and ΔmtrB with pBBRmtrA and pBBRmtrB, respectively (A), and with IO₃⁻ as the electron acceptor and lactate as the electron donor (mutant strains normalized to wild-type levels) (B). Values are means from triplicate samples from anaerobic incubations. Error bars represent SDs. Some error bars cannot be seen due to small SDs.

The IO₃⁻ reduction activity of the ΔmtrB mutant strain was restored to wild-type rates by providing a wild-type copy of mtrB in trans. With lactate as the electron donor, the ΔmtrB/mtrB transconjugant strain reduced IO₃⁻ at a rate almost 2-fold higher (179% of the wild-type rate) than that of the wild-type strain and displayed an extent of reaction approximately 143% of that of the wild-type strain (Fig. 2B; Table 1). In contrast, the mtrB-C45A site-directed mutant reduced IO₃⁻ at near wild-type rates (96% of the wild-type rate) and displayed a near wild-type extent of reaction (102% of the wild-type extent of reaction). The ΔmtrA mutant was also severely impaired in IO₃⁻ reduction activity with lactate as the electron donor (20% of the wild-type rate) and displayed a significantly lower extent of reaction (16% of the wild-type extent of reaction) (Fig. 2A; Table 1). However, with formate as the electron donor, the ΔmtrA mutant reduced IO₃⁻ at near wild-type rates (82% of the wild-type rate) and displayed a near wild-type extent of reaction (79% of the wild-type extent of reaction). With H₂ as the electron donor, the ΔmtrA mutant reduced IO₃⁻ at rates over 2-fold higher that of the wild-type strain (212% of the wild-type rate) and displayed an extent of reaction almost 2-fold higher than that of the wild-type strain (197% of the wild-type extent of reaction). With lactate as the electron donor, the ΔmtrA/mtrA transconjugant strain reduced IO₃⁻ at a rate nearly 2-fold greater than that of the wild-type strain (164% of the wild-type rate) and displayed a higher extent of reaction (141% of the wild-type extent of reaction) (Fig. 2B; Table 1).

Conversely, with lactate as the electron donor, the ΔmtrC, ΔomcA, and ΔmtrC ΔomcA EEC mutant strains reduced IO₃⁻ at near wild-type rates with lactate (76%, 69%, and 110% of the wild-type rate, respectively) and displayed near wild-type extents of reaction (82%, 75%, and 112% of the wild-type extent of reaction) (Fig. 2A; Table 1). In a similar fashion, with formate as the electron donor, the ΔmtrC, ΔomcA, and ΔmtrC ΔomcA mutant strains reduced IO₃⁻ at near wild-type rates (125%, 121%, and 142% of the wild-type rate, respectively) and displayed near wild-type extents of reaction (113%, 116%, and 130% of the wild-type extent of reaction, respectively). With H₂ as the electron donor, the ΔmtrC, ΔomcA, and ΔmtrC ΔomcA mutant strains also reduced IO₃⁻ at near wild-type rates (142%, 87%, and 139% of the wild-type rate, respectively) and displayed near wild-type extents of reaction (147%, 98%, and 112% of the wild-type extent of reaction, respectively) (Fig. 2A; Table 1).

IO₃⁻ reduction activity of additional S. oneidensis c-type cytochrome and crp mutants.

With lactate as the electron donor, ΔcymA and ΔccmB c-type cytochrome mutant strains were also severely impaired in IO₃⁻ reduction activity (both 10% of the wild-type rate) and displayed significantly lower extents of reaction (3% and 8% of the wild-type extent of reaction, respectively) (Fig. 3A; Table 1). With formate as the electron donor, the mutant strains were also impaired in IO₃⁻ reduction activity (25% and 22% of the wild-type rate, respectively) and displayed significantly lower extents of reaction (25% and 20% of the wild-type extent of reaction, respectively). With H₂ as the electron donor, the ΔcymA and ΔccmB mutant strains were also impaired in IO₃⁻ reduction activity (56% and 55% of the wild-type rate, respectively) and displayed significantly lower extents of reaction (28% and 16% of the wild-type extent of reaction, respectively). The ΔcymA/cymA and ΔccmB/ccmB transconjugant strains recovered near wild-type rates of IO₃⁻ reduction (80% and 85% of the wild-type rate, respectively), and displayed near wild-type extents of reaction (98% and 101% of the wild-type extent of reaction, respectively) with lactate as the electron donor (Fig. 3B; Table 1). With lactate, formate, and H₂ as electron donors, the Δcrp mutant strain was severely impaired in IO₃⁻ reduction activities (15%, 27%, and 38% of the wild-type rate, respectively) and displayed significantly lower extents of reaction (11%, 28%, and 46% of wild-type extent of reaction, respectively) (Fig. 3A; Table 1), while the IO₃⁻ reduction activity of the Δcrp/crp transconjugant strain was partially restored to wild-type rates with lactate as the electron donor (54% of the wild-type rate and 85% of wild-type extent of reaction) (Fig. 3B; Table 1). The IO₃⁻ reduction activities of the transconjugant strains with formate or H₂ as the electron donor were not determined.

FIG 3.

IO₃⁻ reduction activity of S. oneidensis wild-type (MR-1) and c-type cytochrome and crp mutants with IO₃⁻ as the electron acceptor and lactate, formate, or H₂ as the electron donor and their complemented strains with pBBRcymA, pBBRccmB, and pBBRcrp, respectively (A), and with IO₃⁻ as the electron acceptor and lactate as the electron donor (mutant strains normalized to wild-type levels) (B). Values are means of triplicate samples from anaerobic incubations. Error bars represent SDs. Some error bars cannot be seen due to small SDs.

IO₃⁻ reduction activity of S. oneidensis type I, II, and V protein secretion mutants.

With lactate as the electron donor, the type I (ΔtolC) and the type V (ΔSO3800) protein secretion mutants reduced IO₃⁻ at near wild-type rates (97% and 113% of the wild-type rate, respectively) and displayed near wild-type extents of reaction (104% and 109% of the wild-type extent of reaction, respectively) (Fig. 4; Table 1). With formate as the electron donor, the mutants similarly reduced IO₃⁻ at near wild-type rates (96% and 129% of the wild-type rate, respectively) and displayed near wild-type extents of reaction (118% and 92% of the wild-type extent of reaction, respectively). With H₂ as the electron donor, the type I and type V protein secretion mutants reduced IO₃⁻ at near wild-type rates (117% and 139% of the wild-type rate, respectively) (Fig. 4; Table 1) and displayed near wild-type extents of reaction (101% and 124% of the wild-type extent of reaction, respectively). However, with lactate, formate, and H₂ as electron donors, the ΔgspD type II protein secretion mutant was severely to partially impaired in IO₃⁻ reduction activity (30%, 45%, and 56% of the wild-type rate, respectively) and displayed significantly lower extents of reaction with all three electron donors (27%, 40%, and 55% of the wild-type extent of reaction, respectively).

FIG 4.

IO₃⁻ reduction activity of S. oneidensis wild-type (MR-1) and ΔtolC, ΔgspD, and ΔSO3800 protein secretion mutants with IO₃⁻ as the electron acceptor and lactate, formate, or H₂ as the electron donor (mutant strains normalized to wild-type levels). Values are means of triplicate samples from anaerobic incubations. Error bars represent SDs.

DISCUSSION

The molecular mechanism of microbial IO₃⁻ reduction is poorly understood. Under NO₃⁻-reducing anaerobic conditions, microorganisms, including the phytoplankton Navicula and the bacteria Pseudomonas sp. strain SCT, Agrobacterium-related strain DVZ35, and Escherichia coli, reduce IO₃⁻ to I⁻, which led to the hypothesis that NO₃⁻ reductase reduces IO₃⁻ as an alternative terminal electron acceptor (6, 7, 30, 32, 45, 46). The IO₃⁻-reducing NO₃⁻ reductase hypothesis was recently brought into question, however, by findings with S. oneidensis, which demonstrated that NO₃⁻ and IO₃⁻ reduction activities were not inhibited by the presence of saturating levels of the competing electron acceptor and that NO₃⁻ reductase-deficient deletion mutants retained wild-type IO₃⁻ reduction activity (33).

S. oneidensis also reduces external metal oxides via EEC-mediated electron transfer either at the outside face of the outer membrane or via outer membrane extensions (i.e., nanowires) (19, 35, 47). The S. oneidensis EEC (MtrCAB) is composed of outer membrane β-barrel protein MtrB, which forms a ternary complex with decaheme c-type cytochromes MtrC and MtrA (48, 49). S. oneidensis mutants lacking MtrCAB display Fe(III), Mn(III), and Mn(IV) reduction-deficient phenotypes (39, 47, 50, 51). In addition, methyl viologen-reduced proteoliposomes containing only the S. oneidensis MtrCAB complex transfer electrons to external soluble and solid Fe(III) substrates (20, 52, 53). In the present study, ΔmtrA and ΔmtrB deletion mutants were severely impaired in IO₃⁻ reduction activity with lactate as the electron donor, indicating that MtrA and MtrB also function as critical components of the lactate-dependent IO₃⁻ reduction pathway. However, IO₃⁻ was reduced at wild-type rates with formate or H₂ as the electron donor, indicating that the electron transport pathway components required for IO₃⁻ reduction are electron donor dependent.

In contrast to the lactate-dependent, IO₃⁻ reduction-deficient phenotypes displayed by the ΔmtrA and ΔmtrB mutant strains, the ΔmtrC mutant strain retained wild-type IO₃⁻ reduction activity regardless of electron donor, indicating that MtrC is not required for IO₃⁻ reduction by S. oneidensis. These findings differed from those of previous studies which demonstrated that MtrC is required for Fe(III), Mn(IV), Mn(III), Tc(VII), U(VI), and flavin reduction by S. oneidensis (28, 39, 54–57). In addition, previous studies demonstrated that OmcA was required for Fe(III), Mn(IV), and Mn(III) reduction (39). In the present study, the ΔomcA and ΔomcA ΔmtrC double mutant strains retained wild-type IO₃⁻ reduction activity with each electron donor, thus indicating that neither OmcA nor MtrC is required for electron transport to IO₃⁻.

Previous studies with the mtrB-C42A and mtrB-C45A site-directed mutants demonstrated that cysteine at MtrB amino acid position 42 (but not at position 45) was required for Fe(III), Mn(IV), and Mn(III) reduction by S. oneidensis (36, 39). In a similar fashion, results of the present study demonstrated that themtrB-C42A mutant was severely impaired in IO₃⁻ reduction activity, while the mtrB-C45A mutant reduced IO₃⁻ at wild-type rates. Residues C42 and C45 comprise a conserved CXXC motif in MtrB homologs of metal-reducing gammaproteobacteria (36). The biochemical function of the CXXC motif of S. oneidensis MtrB is currently unknown but may involve MtrB maturation via disulfide bond formation or metal cofactor binding (36, 58, 59). The detection of a CXXC motif in MtrB homologs of gammaproteobacteria is diagnostic for microbial Fe(III) reduction (36). Future IO₃⁻ reduction activity assays will be required to determine if the CXXC motif of MtrB homologs in gammaproteobacteria is also diagnostic for microbial IO₃⁻ reduction. Such information will guide interpretation of in situ meta(omic) signals indicative of microbial IO₃⁻ reduction in natural and contaminated environments such as the ¹²⁹I-contaminated Hanford and Savannah River sites. These types of molecular signatures will be important for monitoring the ¹²⁹I cycling in sites like Hanford, which may affect overall mobility of ¹²⁹I in the oligotrophic aquifer. Diagnostic markers may also be used to monitor conversion of ¹²⁹IO₃⁻ in FBRs currently treating contaminated groundwater at Hanford. Conversion of ¹²⁹IO₃⁻ to ¹²⁹I⁻ facilitates uptake of ¹²⁹I on biofilm or GAC in FBRs.

S. oneidensis CcmB functions as the integral membrane component of the cytochrome c maturation complex (43). The ΔccmB mutant strain was severely impaired in IO₃⁻ reduction activity regardless of electron donor, which indicates that the c-type cytochrome pool is involved in anaerobic electron transport to IO₃⁻. Correspondingly, the ΔcymA mutant strain was also severely impaired in IO₃⁻ reduction activity regardless of electron donor. Inner membrane-tethered tetraheme c-type cytochrome CymA functions as the central branch point in anaerobic electron transport by S. oneidensis (60, 61). CymA accepts electrons from the menaquinone pool for subsequent delivery to a variety of periplasmic or outer membrane localized terminal reductase complexes, including MtrA of the S. oneidensis EEC (61, 62). The IO₃⁻ reduction-deficient phenotype of the ΔcymA strain indicates that the electron transport chain to IO₃⁻ also includes CymA. The Δcrp mutant strain (lacking the cyclic AMP receptor protein) was also severely impaired in IO₃⁻ reduction activity. CRP regulates expression of genes required for anaerobic respiration by S. oneidensis (44). The IO₃⁻ reduction-deficient phenotype displayed by Δcrp indicates that IO₃⁻ reduction gene expression is also regulated by CRP/cAMP levels in S. oneidensis.

The S. oneidensis genome encodes type I, II, and V protein secretion systems (39, 41, 42, 63). ΔtolC and ΔSO3800 deletion mutants retained wild-type IO₃⁻ reduction activity regardless of electron donor, thus indicating that IO₃⁻ reduction requires neither TolC-mediated type I protein secretion (for efflux of antibiotics, heavy metals, or toxic proteins) (64) nor SO3800-mediated type V protein secretion (an autotransporter-like serine protease involved in S. oneidensis adhesion to Fe(III) oxide surfaces [63]). In contrast, the IO₃⁻ reduction-deficient phenotype displayed by the mutant with a deletion of gspD (encoding GspD, the outer membrane secretin of type II protein secretion) indicates that IO₃⁻ reduction is linked to type II protein secretion in a manner similar to that of Fe(III), Mn(IV), and Mn(III) reduction. The type II protein secretion system is required for outer membrane localization of MtrC and OmcA (21, 65), and impairment of type II protein secretion results in mislocalization of MtrC and OmcA, with corresponding Fe(III), Mn(IV), and Mn(III) reduction-deficient phenotypes. The ΔmtrC, ΔomcA, and ΔmtrC ΔomcA mutant strains retained wild-type IO₃⁻ reduction activity. These findings demonstrate that IO₃⁻ reduction by S. oneidensis does not require either of the EEC cytochromes MtrC and OmcA but does require type II protein secretion of an as-yet-unidentified IO₃⁻ reductase to the outside face of the outer membrane. In the current model of the lactate (MtrAB)-dependent S. oneidensis IO₃⁻ reduction system (Fig. 5), electrons originating from lactate dehydrogenase are transported via the menaquinone pool, CymA, and MtrAB to the terminal IO₃⁻ reductase that is translocated to the outside face of the outer membrane via type II protein secretion. Current work is focused on identification of the S. oneidensis IO₃⁻ reductase via comparison of the IO₃⁻-reducing protein fractions harvested from the outside face of the outer membrane of S. oneidensis wild-type and ΔgspD mutant strains.

FIG 5.

Working model of the lactate (MtrAB)-dependent IO₃⁻ reduction electron transport pathway in S. oneidensis, including a comparison with the MtrAB-dependent metal reduction pathway. In both the metal and IO₃⁻ reduction pathways, electrons originating from lactate dehydrogenase located at the head end of the electron transport chain are transferred to the inner membrane-localized menaquinone pool and subsequently to CymA, which facilitates electron transfer across the periplasmic space to decaheme cytochrome MtrA. At this location in the electron transport chain, the metal and IO₃⁻ reduction pathways diverge and terminate with either metal-reducing c-type cytochrome MtrC or an unknown terminal IO₃⁻ reductase, both of which associate with MtrA and β-barrel protein MtrB. MtrC and the unknown IO₃⁻ reductase are both secreted extracellularly by the type II protein secretion system to form a ternary complex with the MtrAB module on the outside face of the outer membrane. The formate- and H₂-dependent IO₃⁻ reduction pathways are MtrAB-independent and thus are not depicted in this working model.

MATERIALS AND METHODS

Growth and cultivation conditions.

S. oneidensis strains were routinely cultured aerobically at 30°C in lysogeny broth (LB) (10 g liter⁻¹ of NaCl, 10 g liter⁻¹ of tryptone, 5 g liter⁻¹ of yeast extract). IO₃⁻ reduction rate experiments were conducted under anaerobic conditions in M1 minimal medium (66) amended with 20 mM lactate, 10 mM formate, or 2% H₂ gas as the electron donor and 250 μM IO₃⁻ as the anaerobic electron acceptor. When required for selection, gentamicin (20 μg ml⁻¹) was amended to the appropriate growth medium.

In-frame deletion mutagenesis of S. oneidensis genes.

The genes crp and ccmB were deleted in frame from the S. oneidensis MR-1 genome following previously described procedures (67). Regions corresponding to ∼750 bp upstream and downstream of crp and ccmB were PCR amplified with iProof ultrahigh-fidelity polymerase (Bio-Rad, Hercules, CA) (primers D1/D2 and D3/D4 [Table 2]) and subsequently joined using overlap extension PCR (primers D1/D4 [Table 2]). The resulting fragment was cloned into suicide vector pKO2.0, which does not replicate in S. oneidensis, and mobilized into wild-type S. oneidensis MR-1 via conjugation with E. coli donor strain β2155 λ pir (Table 3) (68). S. oneidensis strains with the integrated plasmid were selected on LB agar containing gentamicin (15 μg ml⁻¹). Single-crossover integrations were verified using PCR with primers flanking the recombination region (TF/TR) and were resolved from the genomes by plating on LB agar lacking NaCl and containing sucrose (10% [wt/vol]). The in-frame deletion strains (Δcrp and ΔccmB) were verified by PCR with primers TF/TR (Table 2). Genetic complementation analysis of ΔmtrA, ΔmtrB, Δcrp, ΔcymA, and ΔccmB strains was carried out by cloning the wild-type gene into broad-host-range cloning vector pBBR1MCS (69) and conjugally transferring the recombinant vector into the respective mutant strains via biparental mating procedures (67).

TABLE 2.

Primers used in this study

Primer	Sequence (5ʹ–3ʹ)	Remark
Δcrp
D1	CTGATAGGATCCTCTTTATACCAACGTTCGGCC	BamHI (underlined)
D2	GGCTTAAATCAAGCTGAAGTCTAACTGTCGATGTTCCTCGATTGATTAA
D3	TTAATCAATCGAGGAACATCGACAGTTAGACTTCAGCTTGATTTAAGCC
D4	TCGATCGTCGACAGTGCCTGAATTCGCGCTA	SalI (underlined)
TF	GCGTAAATAAAACCTAAACGGAACT
TR	TAGCTAAGTTGCTTGTTGGGATT
ΔccmB
D1	CTGATAACTAGTACATCTAGTCCTAAGCAATTGTAAACC	SpeI (underlined)
D2	GGGTGTAACCATTTCCACATTT TATGCCAAACCTTATACGAAGC
D3	GCTTCGTATAAGGTTTGGCATA AAATGTGGAAATGGTTACACCC
D4	TCGATCGTCGACTTGCATTATTTGACCTCCTCAG	SalI (underlined)
TF	TCGATTTGTACGAGAAATATTGC
TR	TAAAAGGCATAGCCACCCAT

TABLE 3.

Strains used in this study

Strain or plasmid	Features	Source or reference
S. oneidensis
MR-1	Wild-type strain	ATCC
Δcrp mutant	In-frame crp deletion mutant	This study
ΔccmB mutant	In-frame ccmB deletion mutant	This study
ΔcymA mutant	In-frame cymA deletion mutant	39
ΔmtrA mutant	In-frame mtrA deletion mutant	39
ΔmtrB mutant	In-frame mtrB deletion mutant	36
ΔmtrC mutant	In-frame mtrC deletion mutant	39
ΔomcA mutant	In-frame omcA deletion mutant	39
ΔmtrC ΔomcA mutant	In-frame mtrC and omcA double deletion mutant	39
ΔtolC mutant	In-frame tolC deletion mutant	39
ΔgspD mutant	In-frame gspD deletion mutant	39
ΔSO3800 mutant	In-frame SO3800 deletion mutant	63
mtrB-C42A mutant	Site-directed mutant with cysteine at amino acid position 42 replaced with alanine	36
mtrB-C45A mutant	Site-directed mutant with cysteine at amino acid position 45 replaced with alanine	36
E. coli
E100D pir-116	F− mcrA Δ(mrr-hsdRMS-mcrBC) φ80dlacZΔM15 ΔlacX74 recA1 endA1 araD139 Δ(ara, leu)7697 galU galK λ− rpsL (Strr) nupG pir-116(DHFR)	Epicentre
β2155 λ pir	thrB1004 pro thi strA hsdS lacZΔM15(F9 lacZΔM15 laclq traD36 proA1 proB1) ΔdapA::erm pir::RP4 Kmr	71
Plasmids
pKO2.0	In-frame gene deletion vector; 4.5 kb γR6K, mobRP4 sacB Gmr lacZ	63
pBBR1MCS	Broad-host-range cloning vector; Cmr lacZ	69
pBBRmtrA	pBBR1MCS containing wild-type mtrA	39
pBBRmtrB	pBBR1MCS containing wild-type mtrB	39
pBBRcymA	pBBR1MCS containing wild-type cymA	This study
pBBRccmB	pBBR1MCS containing wild-type ccmB	This study
pBBRcrp	pBBR1MCS containing wild-type crp	This study

Anaerobic incubation conditions.

Mutant strains were initially inoculated in liquid LB growth medium and incubated at 30°C for 24 h. Ten-milliliter subcultures at an initial optical density at 600 nm (OD₆₀₀) of 0.02 were incubated at 30°C for 24 h. Subcultures were centrifuged at 4,000 rpm for 30 min, resuspended in 10 ml of M1 growth medium amended with 20 mM lactate, and incubated aerobically at room temperature for 8 h. The preconditioned cells were inoculated in 30-ml serum bottles at an initial OD₆₀₀ of 0.1 in M1 growth medium amended with 250 μM IO₃⁻ and either 20 mM lactate or 10 mM formate and incubated anaerobically via continuous sparging with 100% high-purity (hydrated) N₂ gas. For IO₃⁻ reduction activity assays with H₂ as the electron donor, the preconditioned cells were incubated anaerobically via continuous sparging with high-purity (hydrated) anaerobic gas mix consisting of 2% H₂ and 98% N₂. Cultures were incubated at room temperature with gentle stirring under anaerobic conditions maintained by continuous sparging with high-purity hydrated N₂ gas. At preselected time points, OD₆₀₀ was measured and IO₃⁻ concentrations were determined using the IO₃⁻-triiodide formation method described below.

Determination of IO₃⁻ concentrations via IO₃⁻-triiodide formation with I⁻ at acidic pH.

The extent of IO₃⁻ reduction was determined using the IO₃⁻-triiodide method (33, 70). Culture samples were added to 96-well 500-μl microtiter plates. Sodium citrate buffer (0.1 M; pH 3.3) and potassium iodide solution (75 mM) were added to each well to initiate triiodide formation (IO₃⁻ + 5I⁻ + 6H⁺ → 3H₂O + 3I₂). Absorbance at 352 nm was measured with a UV spectrophotometer (Multiskan Go; Thermo Scientific) after a 4-min reaction time. IO₃⁻ concentrations were determined from a previously generated calibration curve.