The Hyb Hydrogenase Permits Hydrogen-Dependent Respiratory Growth of Salmonella enterica Serovar Typhimurium

Abstract

Salmonella enterica serovar Typhimurium contains three distinct respiratory hydrogenases, all of which contribute to virulence. Addition of H₂ significantly enhanced the growth rate and yield of S. Typhimurium in an amino acid-containing medium; this occurred with three different terminal respiratory electron acceptors. Based on studies with site-specific double-hydrogenase mutant strains, most of this H₂-dependent growth increase was attributed to the Hyb hydrogenase, rather than to the Hya or Hyd respiratory H₂-oxidizing enzymes. The wild type strain with H₂ had 4.0-fold greater uptake of ¹⁴C-labeled amino acids over a period of minutes than did cells incubated without H₂. The double-uptake hydrogenase mutant containing only the Hyb hydrogenase transported amino acids H₂ dependently like the wild type. The Hyb-only-containing strain produced a membrane potential comparable to that of the wild type. The H₂-stimulated amino acid uptake of the wild type and the Hyb-only strain was inhibited by the protonophore carbonyl cyanide m-chlorophenylhydrazone but was less affected by the ATP synthase inhibitor sodium orthovanadate. In the wild type, proteins TonB and ExbD, which are known to couple proton motive force (PMF) to transport processes, were induced by H₂ exposure, as were the genes corresponding to these periplasmic PMF-coupling factors. However, studies on tonB and exbD single mutant strains could not confirm a major role for these proteins in amino acid transport. The results link H₂ oxidation via the Hyb enzyme to growth, amino acid transport, and expression of periplasmic proteins that facilitate PMF-mediated transport across the outer membrane.

IMPORTANCE

Complex carbohydrates consumed by animals are fermented by intestinal microflora, and this leads to molecular hydrogen production. Salmonella enterica serovar Typhimurium can utilize this gas via three distinct respiratory hydrogenases, all of which contribute to virulence. Since H₂ oxidation can be used to conserve energy, we predicted that its use may augment bacterial growth in nutrient-poor media or in competitive environments within H₂-containing host tissues. We thus investigated the effect of added H₂ on the growth of Salmonella Typhimurium in carbon-poor media with various terminal respiratory electron acceptors. The positive effects of H₂ on growth led to the realization that Salmonella has mechanisms to increase carbon acquisition when oxidizing H₂. We found that H₂ oxidation via one of the respiration-linked enzymes, the Hyb hydrogenase, led to increased growth, amino acid transport, and expression of periplasmic proteins that facilitate proton motive force-mediated transport across the outer membrane.

INTRODUCTION

When an animal consumes complex sugars that are not absorbed or are difficult to metabolize, these sugars reach the intestinal flora and are anaerobically fermented by resident microbes (1, 2). One result is production of molecular hydrogen (H₂), and it is well established that such H₂ production can vary with the animal’s diet (3), including that of humans (4–6). The colonically produced gas can be distributed to many tissues where pathogens reside (7, 8). Some pathogens capitalize on this, using the high-energy reductant as an energy source to facilitate their growth (8). One of these is Salmonella enterica serovar Typhimurium, in which H₂ has been shown to be an important energy source for virulence during host colonization. Study of Salmonella hydrogenase mutants has shown that each of the three uptake hydrogenases contributes to virulence, and a triple uptake mutant lacking all respiratory H₂-oxidizing ability was avirulent in a mouse model (9).

Electrons generated from H₂ splitting are passed along metabolically versatile bacterial electron transport chains to a variety of acceptors, including fumarate, nitrate, sulfate, CO₂, or O₂ (10), depending on the inherent terminal oxidase content of the particular microorganism. Such respiratory chains conserve energy in the form of ATP. The three H₂-consuming hydrogenases known as Hyb, Hyd, and Hya in Salmonella Typhimurium (11) are membrane bound and contain NiFe centers. The enterobacterial uptake-type (H₂-oxidizing) hydrogenases are viewed as auxiliary energy input providers contributing to the proton gradient across the cell membrane (10, 12).

Metabolically flexible H₂-utilizing bacteria (e.g., the facultative chemoautotrophs) turn to H₂ use when high-energy organic substrates become limiting (13). Since H₂ oxidation can be used to conserve energy, it may be predicted that use of H₂ may be especially important to augment bacterial growth in nutrient-poor media or in the competitive environments within H₂-containing host tissues. Still, the effect of exogenous H₂ on growth, including under carbon-limited conditions, has not been studied in Salmonella Typhimurium. We thus initially investigated the effect of added H₂ on Salmonella Typhimurium growth in carbon-poor media with various terminal respiratory electron acceptors. The positive effect of H₂ led to the realization that the cells have mechanisms to increase carbon acquisition when oxidizing H₂.

The culture conditions used herein were ones in which the H₂ uptake enzymes were produced but no H₂ was produced, so only the effect of (exogenously added) H₂ was addressed.

RESULTS AND DISCUSSION

Effects of exogenous H₂ on growth.

The ability to use hydrogen is important for S. enterica serovar Typhimurium survival within the animal host (9). To address possible growth-stimulating effects of H₂, the growth parameters of the parent strain and various double mutants (thus, each mutant strain contains only one of the three uptake-type enzymes) were compared in cultures with and without added H₂. This was done anaerobically with four different terminal electron acceptors: trimethylamine-N-oxide (TMAO); dimethyl sulfoxide (DMSO); sodium fumarate; and sodium nitrate. The strains used are shown in Table 1. We used CR-Hyd medium (14, 15) with the modification that no glucose was added, so the peptone and casein hydrolysate served as the carbon sources for the growth of the strains. The condition used was such that only the H₂ uptake enzymes were produced and no H₂ was produced, so only the effects of exogenously added H₂ were being addressed. Also, the growth of the strains was assigned to the ability of the strains to assimilate carbon by utilizing the amino acids and peptides in the medium as the sole source of carbon. In our growth experiments, cell yield increased by 3.5-fold for cells grown with fumarate and H₂, compared to the cell yield in fumarate alone. Growth yields with H₂ were increased 1.8-fold and about 3-fold for cells grown with TMAO and DMSO, respectively, compared to those under H₂-lacking conditions (Table 2). Growth was not significantly higher in cells grown with nitrate. When grown with fumarate, the triple mutant (strain ALZ43; Table 1), lacking all H₂ uptake ability, never responded to H₂ addition. The wild type (WT) growth rate (doubling time) with H₂ was 1.5 h, whereas without H₂ it was about 5 h (Fig. 1). Less pronounced but significant growth rate differences (in comparing bacteria with H₂ added versus those with no H₂ added) were observed for the WT on either TMAO or DMSO (data not shown).

TABLE 1 .

Strains and plasmids used in this study

Strain/plasmid	Genotype/description	Reference
S.  enterica serovar Typhimurium strains
JSG 210	14028s (WT)	ATCC
ALZ36	JSG210 with Δhyb::FRTa Δhyd::FRT (Hya only)	29
ALZ37	JSG210 with Δhyb::FRT Δhya::FRT (Hyd only)	29
ALZ42	JSG210 with Δhyd::FRT Δhya::FRT (Hyb only)	29
ALZ43	JSG210 with Δhyb::FRT Δhyd::FRT Δhya::FRT (triple mutant)	29
RLK1	JSG210 ΔexbD::FRT (ΔexbD)	This study
RLK2	JSG210 ΔtonB::FRT (ΔtonB)	This study
Plasmids
pCP20	Ampr; contains flippase gene for λ Red mutagenesis	30
pKD46	Ampr; contains λ Red genes γ, β, and exo	30
pKD4	Kanr; contains kan cassette	30

^a FRT, flippase recombinase recognition target.

TABLE 2 .

S. enterica serovar Typhimurium H₂-facilitated growth yield with various electron acceptors

Growth condition	No. of cells/ml culturea
	WT	ALZ36 (Hya only)	ALZ37 (Hyd only)	ALZ42 (Hyb only)	ALZ43 (triple mutant)
Fumarate
+H2	35 ± 1.8b	12 ± 1.5	11 ± 1.4	32 ± 2.5b	10 ± 1.0
−H2	10 ± 1.2	9 ± 1.4	9 ± 1.0	9 ± 1.6	9 ± 1.5
Nitrate
+H2	42 ± 1.9	36 ± 2.0	19 ± 1.2	22 ± 2.1b	21 ± 1.1
−H2	38 ± 2.7	35 ± 1.0	17 ± 2.0	12 ± 1.0	22 ± 1.6
TMAO
+H2	25 ± 1.0b	14 ± 1.8	16 ± 3.4	28 ± 1.0b	11 ± 1.1
−H2	15 ± 1.7	13 ± 1.2	9 ± 1.6	9 ± 1.0	9 ± 1.4
DMSO
+H2	23 ± 2.1b	8 ± 1.5	8 ± 2.5	13 ± 1.6b	8 ± 2.5
−H2	7 ± 1.0	7 ± 1.0	6 ± 1.4	7 ± 1.2	8 ± 1.0

^a Values represent the growth yield in 10⁷ cells/ml ± the standard deviation (n = 3) at 18 h of incubation.

^b Significantly higher growth yield than without H₂ (P < 0.005 [Student’s t test]).

FIG 1 .

Effect of hydrogen on growth of S. enterica serovar Typhimurium WT and ALZ42 (Hyb-only-containing strain) with fumarate as an electron acceptor.

To assign specific hydrogenase activity to growth effects, endpoint growth yields were determined for uptake-type hydrogenase double and triple mutant strains. The strain containing only the Hyb hydrogenase (ALZ42; Table 1) had increased growth yield in the presence of H₂ (more than 3.0-fold greater in fumarate-containing medium) compared to that of cells grown without added H₂. This H₂-facilitated growth occurred when ALZ42 was grown with fumarate, nitrate, TMAO, or DMSO (Table 2). The growth rate of ALZ42 was clearly also H₂ stimulated (Fig. 1). The strain that contained only Hya (ALZ36; Table 1) did not respond significantly to the presence of H₂ and had growth yields similar (in nitrate, TMAO, or DMSO) to those of the WT in those same media when grown without H₂ (Table 2). The strain that contained only Hyd (ALZ37; Table 1) had increased growth yield with H₂ when TMAO was provided as an electron acceptor; still, its final yield was much less than those of ALZ42 and the WT. Addition of H₂ had no effect on growth or amino acid uptake of the triple mutant strain ALZ43 that lacks all H₂-oxidizing ability, and this strain had the lowest growth yield among the strains used.

In the growth experiments, the effects of exogenously added H₂ were addressed. This is appropriate, as organs colonized by Salmonella were shown to contain significant levels of H₂ (8). Sawers et al. demonstrated that H₂ evolution is low (between 0.016 and 0.001 µmol of H₂ evolved per min) when S. enterica serovar Typhimurium cells are grown under anaerobic respiration with fumarate (16). We wanted to determine whether cells were producing H₂ (which likely would affect the growth yield) under the growth conditions used in our study. One milliliter of headspace gas from 8-h stationary-phase cultures of the triple uptake mutant (ALZ43) grown with fumarate, DMSO, nitrate, or TMAO was assayed for the presence of H₂ using an amperometric Clark-type electrode (17). There was no detectable H₂ (less than 10 nmol) present in headspace gas from these cultures. This result indicates that the cells were not producing appreciable H₂ in the medium and under the atmosphere conditions used in the study herein.

Collectively, our results indicate that addition of exogenous H₂ to the headspace greatly enhances the growth rate and yield on fumarate and some other anaerobic respiratory electron acceptors. The bulk of this growth rate increase can be attributed to the Hyb enzyme. Yamamoto and Ishimoto (18) reported that Escherichia coli cells continuously bubbled with hydrogen grew with nitrate, fumarate, or TMAO provided as an electron acceptor and that both H₂ uptake and H₂ evolution activities were the greatest with fumarate. Hydrogenase activity staining bands from gels indicated the expression of multiple forms of the E. coli enzymes in the fumarate-containing medium, and from growth yields, they suggested that 1 mol of ATP is produced per mol of H₂ oxidized.

Amino acid uptake.

The growth studies described above were performed in an amino acid-containing medium, so we measured the uptake of ¹⁴C-labeled amino acids when cells were in an atmosphere containing H₂ and in one without the gas. Both the WT strain and Hyb-only-containing strain ALZ42 demonstrated significantly increased amino acid uptake ability in the presence of H₂. Although the uptake by ALZ42 was initially one-half that of the WT, both strains had 4-fold greater uptake in the first 5 min with H₂ than without H₂ added (Table 3). After 5 min, the amino acid accumulation continued at a lower rate, but at all points, the uptake was greater for both strains when H₂ was provided. The results indicate that the energy for uptake/transport of amino acids in these strains is provided via oxidation of H₂, akin to what was observed in Helicobacter hepaticus by Mehta et al. (19). The similar results for ALZ42 and the WT, which contains three distinct H₂-oxidizing enzymes, indicate that the Hyb hydrogenase is important for amino acid accumulation. The activity of the Hyb hydrogenase probably suffices to allow the organism to glean energy from H₂ for significant solute transport and thus for growth and survival of S. Typhimurium under anoxic and nutrient-limiting conditions. To confirm that the Hyb enzyme plays the largest role in amino acid uptake in nutrient-limited medium, the other mutant strains, each containing a single uptake-type hydrogenase, were assessed for H₂-dependent amino acid uptake (Table 4). The strains containing Hya or Hyd as the only uptake hydrogenase were capable of much less uptake of the ¹⁴C-labeled amino acid pool than the WT, and H₂ had little effect on their amino acid accumulation abilities. Still, the accumulation by Hyd-only-containing strain ALZ37 was slightly stimulated by H₂. The triple mutant ALZ43, which lacks all three uptake-type hydrogenases, did not respond to H₂.

TABLE 3 .

¹⁴C-labeled amino acid uptake by the WT and ALZ42 strains

Strain and condition	14C-labeled amino acid uptake (cpm [103]/108 cells)a at:
	5 min	10 min	20 min
WT
+H2	31.0 ± 3.8b	37.8 ± 2.3b	35.0 ± 1.9b
−H2	7.8 ± 1.3	9.3 ± 0.9	9.9 ± 0.8
ALZ42 (Hyb only)
+H2	14.0 ± 1.6b	14.9 ± 1.4b	20.4 ± 3.5b
−H2	3.1 ± 0.7	3.4 ± 0.5	3.5 ± 0.5

^a Values represent ¹⁴C-labeled amino acid uptake by 10⁸ cells ± standard deviation (n = 4).

^b Significantly higher uptake level than without H₂ (P < 0.005 [Student’s t test]).

TABLE 4 .

¹⁴C-labeled amino acid uptake by strains ALZ36, ALZ37, and ALZ43

Strain and condition	14C-labeled amino acid uptake (cpm [103]/108 cells)a at:
	5 min	10 min
ALZ36 (Hya only)
+H2	2.3 ± 0.7	2.9 ± 1.6
−H2	1.9 ± 0.3	2.0 ± 0.2
ALZ37 (Hyd only)
+H2	3.0 ± 0.6	3.8 ± 1.6b
−H2	2.0 ± 0.2	2.5 ± 0.1
ALZ43 (triple mutant)
+H2	0.7 ± 0.2	1.1 ± 0.03
−H2	0.6 ± 0.1	1.1 ± 0.01

^a Values represent ¹⁴C-labeled amino acid uptake by 10⁸ cells ± standard deviation (n = 4).

^b Significantly higher uptake level than without H₂ (P < 0.005 [Student’s t test]).

The proton motive force (PMF) has been suggested to be the driving force of active transport of amino acids in several bacteria (20–22), yet the Enterobacteriaceae are known to contain numerous ATP-utilizing amino acid permeases as well. The uptake and/or transport of amino acids across bacterial cell membranes is facilitated either by carriers which utilize the electrochemical energy stored in the H⁺ and Na⁺ gradients or by the ABC-type uptake and efflux systems which utilize the chemical energy derived from ATP (23). In an attempt to identify the type of energy coupled to amino acid uptake/transport in the WT strain and Hyb-only-containing strain ALZ42, inhibitors of the different energy-coupling processes were used. Carbonyl cyanide m-chlorophenylhydrazone (CCCP) is a protonophore that inhibits PMF, and orthovanadate inhibits ATP synthesis by specifically inhibiting protein tyrosine phosphatases (24). The role of PMF versus that of ATP in H₂-mediated amino acid uptake was addressed by the use of these two inhibitors. The amino acid uptake activities of both the WT and ALZ42 strains markedly decreased upon treatment of the cells with CCCP before the addition of the ¹⁴C-labeled amino acid mixture (Table 5). Pretreatment of the cells with sodium orthovanadate also resulted in reduced amino acid uptake activity of both the WT and ALZ42 strains. However, considerable uptake activity remained; the uptake rate with inhibitor was still about 50% of the uninhibited rate for the WT at both 5 and 10 min. ALZ42 had 39% and 52% of the uninhibited rate at 5 min and 10 min, respectively. These results indicate that H₂-dependent amino acid transport in these strains is driven by both PMF and ATP but that the PMF likely plays the larger role in H₂-facilitated amino acid uptake.

TABLE 5 .

Effects of inhibitors on ¹⁴C-labeled amino acid uptake by the WT and ALZ42 strains

Strain and presence of added H2	14C-labeled amino acid uptake (cpm [103]/108 cells)a at:
	5 min	10 min
WT
None	29.0 ± 2.0	31.0 ± 1.1
CCCP	0.7 ± 0.5b	0.3 ± 0.2b
Orthovanadate	14.8 ± 1.2b	18.7 ± 1.0b
ALZ42 (Hyb only)
None	14.0 ± 1.6	18.2 ± 3.5
CCCP	0.3 ± 0.1b	0.04 ± 0.07b
Orthovanadate	5.5 ± 0.8b	9.5 ± 0.7b

^a Values represent ¹⁴C-labeled amino acid uptake by 10⁸ cells ± standard deviation (n = 4).

^b Significantly lower uptake level than without inhibitor (P < 0.005 [Student’s t test]).

Hydrogenase activity and membrane potential (ΔΨ).

Hyb-only-containing strain ALZ42 demonstrated H₂ uptake hydrogenase activity that was 65% of that of the WT (42.7 ± 8.1 nmol H₂ uptake/min/10⁹ cells), while ALZ36 (containing Hya only) and ALZ37 (containing Hyd only) showed 10.8% and 2.0% of the uptake hydrogenase activity of the WT, respectively. ALZ43 did not show any uptake hydrogenase activity. Therefore, under the conditions used in this study, the bulk of the H₂ uptake activity in Salmonella Typhimurium is accomplished by the activity of the Hyb hydrogenase.

We utilized the fluorescence ratio imaging technique to measure the membrane potential component (ΔΨ) of PMF in the WT and ALZ42 strains by using a cationic dye, JC-1 (5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethylbenzimidazolylcarbocyanine iodide; Molecular Probes, Invitrogen). A shift in the emission spectrum of the JC-1 dye from red (590 nm) to green (530 nm) indicates a decrease in membrane potential and hence an increased green (530 nm)/red (590 nm) ratio. ALZ42 showed a green/red fluorescence ratio comparable to that of the WT (Fig. 2), indicating similar PMF levels in the two strains. The membrane polarization of ALZ42 (Hyb-only-containing strain) was 75% of that of the WT (red/green ratios, 2.56 ± 0.41 and 1.91 ± 0.28 in the WT and ALZ42 strains, respectively), and the difference between the two strains was statistically insignificant at a 99% confidence level. ALZ43 (triple mutant lacking Hya, Hyb, and Hyd) showed significantly decreased membrane potential compared to that of the WT (Fig. 2). The WT treated with CCCP was included as a control to show the green/red ratio of a disintegrated membrane potential. As expected, CCCP-treated WT cells had the highest green/red ratio among the samples. These results support our hypothesis that the Hyb hydrogenase is involved in carrying out the bulk of the respiratory hydrogen oxidation, and therefore in maintaining the PMF of the cells, under H₂-added conditions.

FIG 2 .

Comparison of membrane potentials of the WT, ALZ42, and ALZ43 strains. A small ratio indicates a larger membrane potential (n = 6; P < 0.01).

Involvement of the TonB-ExbD system.

In E. coli and other Gram-negative bacteria, the cytoplasmic PMF is utilized by the TonB-ExbB-ExbD system for substrate transport by the TonB-dependent outer membrane transport proteins (TBDTs) (25, 26). Initially shown to be specifically the uptake of iron complexes and vitamin B₁₂, the role of the TonB-dependent transport has since been expanded to the transport of various other substrates, such as nickel, carbohydrates, cobalt, and copper (27). While the precise mechanism of transport remains unclear, it has been suggested that TonB transduces the PMF to the TBDTs via its periplasmic interaction with ExbD, forming a TonB-ExbB-ExbD complex, and that TonB requires PMF to form the complex with ExbD (28). In an effort to investigate the effect of added H₂ on the PMF-facilitated cross-linking between TonB and ExbD in our strain, we subjected the WT to formaldehyde-mediated cross-linking and visualization of the TonB-ExbD complex using TonB- and ExbD-specific antibodies. We were unable to visualize the ExbD-TonB complex in our strains, although a large ExbD-immunoreactive adduct was observed in the culture growing with H₂. Importantly, a marked increase in the production of the TonB and ExbD proteins under the condition with H₂ added was observed (Fig. 3). Based on densitometry, the increases in expression due to incubation with H₂ were 4.0-fold and 11.0-fold for TonB and ExbD, respectively. Quantitative real-time PCR showed elevated tonB and exbD transcript levels in the WT (about 2-fold- and 4-fold-higher expression of tonB and exbD, respectively) and a 1.8-fold increased exbD transcript level in ALZ42 when the strains were grown under exogenously added H₂. DNA gyrase B (gyrB) was used as an internal control to normalize the expression levels of tonB and exbD, since a microarray analysis (not discussed here) revealed the expression of gyrB to be unaltered under the conditions used in this study. As in other Gram-negative bacteria, the TonB-ExbD system of Salmonella Typhimurium could play a crucial role in fulfilling the increased demand for the delivery of substrates such as iron siderophores, vitamin B₁₂, nickel complexes, and carbohydrates in a nutrient-limited environment.

FIG 3 .

Immunoblot analyses of TonB-, ExbD-, and TonB/ExbD-linked complexes in WT S. enterica serovar Typhimurium. (A) TonB visualized with TonB-specific monoclonal antibodies. (B) ExbD visualized with ExbD-specific polyclonal antibodies. M, molecular mass standards.

To investigate whether the TonB-ExbD system is also involved in H₂-stimulated amino acid uptake, we made ΔexbD and ΔtonB single-deletion mutants (strains RLK1 and RLK2, respectively; Table 1) and subjected them to the amino acid uptake assays described previously. Clear phenotypes distinguishable from that of the WT (i.e., decreases in uptake by the mutants) were not observed (data not shown). Bacteria contain a wide variety of transmembrane amino acid transporters (23), and this includes transporters that are aided by energy-coupling proteins other than TonB-ExbD. Nevertheless, the ΔexbD mutant strain demonstrated 40% reduced nickel uptake compared to that of the WT (⁶³Ni uptake, 20.4 × 10² ± 2.3 × 10² cpm/10⁸ cells in RLK1 and 33.9 × 10² ± 6.7 × 10² cpm/10⁸ cells in the WT), indicating a role for ExbD in nickel uptake in Salmonella. It is possible that in the presence of hydrogen, the bacteria upregulate the expression of ExbD and TonB to transport more nickel into the cells for proper hydrogenase maturation.

Our study shows that Salmonella Typhimurium can grow in an H₂-dependent manner and that most of the H₂ oxidation by nutrient-limited anaerobically growing Salmonella Typhimurium is aided by the activity of the Hyb hydrogenase alone. Growth under anaerobic respiration with terminal electron acceptors was enhanced with H₂, and this H₂-facilitated growth ability was assigned to a specific H₂-using hydrogenase. The energy available from H₂ oxidation is coupled to the uptake of carbon, and the uptake is driven by both PMF and ATP. H₂ increases the expression of genes that encode specific proteins (ExbD and TonB) known to complex PMF to aid solute transport processes. Similarly, immunologically identified TonB and ExbD proteins were significantly induced by incubation with H₂. The increased expression of the PMF-dependent transport proteins (ExbD and TonB) under the H₂-added condition is likely a way for the bacteria to balance energy input with nutrient acquisition. The results herein link H₂ oxidation via the Hyb hydrogenase enzyme to H₂-dependent growth, solute transport, and expression of periplasmic proteins that facilitate PMF-mediated transport across the outer membrane.

MATERIALS AND METHODS

Strains, growth conditions, and reagents.

WT S. enterica serovar Typhimurium ATCC 14028s and hydrogenase mutant strains described by Zbell and Maier (29) were used in this study. All mutant strains were shown to be nonpolar (29). tonB and exbD gene single-deletion mutants were constructed using the lambda Red system as previously described (30). The deletions were confirmed by PCR using primers complementary to the regions flanking the deleted genes and by sequencing across the deletions (Georgia Genomics Facility, University of Georgia). The strains and plasmids used in this study are listed in Table 1, and the primers used are listed in Table 6.

TABLE 6 .

Primers used in this study

Primer	Sequence (5′→3′)	Application
exbD del-F	GTCATGGCAATGCGTCTTAACGAGAACCTGTGTGTAGGCTGGAGCTGCTTC	exbD deletion
exbD del-R	CTTACCCGGCCTACAGCGTCAGCAGAATACCATATGAATATCCTCCTTA	exbD deletion
exbD-check-F	TGCAAATTTCCGGCGGTCAAA	exbD deletion confirmation
exbD-check-R	TATTGCGCAAACGCAGACCA	exbD deletion confirmation
exbD-F	TATTTCGCTTTCGCGGTCTCTTCG	exbD real-time PCR
exbD-R	CGGTGAAAGCGGATAACACCATGT	exbD real-time PCR
tonB del-F	GGTTTTTCAACTGAAACGATTATGACTTCATGTGTAGGCTGGAGCTGCTTC	tonB deletion
tonB del-R	TGGTATATTCCTGGCTGGCGGCGCCAGAGACATATGAATATCCTCCTTA	tonB deletion
tonB -check-F	CCGCTATCGGCAATGCCTTATT	tonB deletion confirmation
tonB-check-R	TGGCGATGTCGTATGCTGCTAC	tonB deletion confirmation
tonB-F	TTCACCTTTACGCGGCCTTCAATA	tonB real-time PCR
tonB-R	AAAGGTTGAAGAGCAGCCGAAGC	tonB real-time PCR
gyrB-F	CGGGTTCATTTCACCCAGACCTTT	Real-time PCR internal control
gyrB-R	TACGCATGGCGTGGATACCGATTA	Real-time PCR internal control

Strains were maintained in Luria-Bertani (LB) broth or on LB plates. Experiments were performed in CR-Hyd medium (14, 15) containing bacteriological peptone (0.5%, wt/vol), casein hydrolysate (0.2%, wt/vol), thiamine (0.001%, wt/vol), MgCl₂ (1 mM), (NH₄)₆Mo₇O₂₄ (1 µM), and NaSeO₃ (1 µM). The medium was supplemented with sodium fumarate (0.5%), sodium nitrate (0.5%), TMAO (0.5%), and DMSO (0.25%) where indicated. No carbohydrate was added, but 5 µM NiCl₂ was included in the medium. Cells were grown at 37°C anaerobically with or without H₂. Anaerobic conditions with H₂ were established by sparging sealed 165-ml bottles with N₂ for 15 min and then with anaerobic mix (10% H₂, 5% CO₂, 85% N₂) for 20 min, and then 10% H₂ was injected to bring the volume of added H₂ to 20% partial pressure. Cells were grown anaerobically without H₂ in 165-ml bottles by sparging with N₂ for 15 min and then injecting the sealed bottles containing cells with CO₂ to 5% partial pressure.

Growth curves and endpoint growth yields.

To determine the effect of hydrogen on growth, growth curves and endpoint growth assays were performed. Sealed 165-ml bottles containing CR-Hyd medium with various electron acceptors (as described above) were inoculated with WT or hydrogenase deletion mutant S. enterica serovar Typhimurium cells at a ratio of 1.0 × 10⁷ CFU/ml. Cells were grown anaerobically with or without 20% H₂ for 18 h at 37°C with shaking at 200 rpm. A₆₀₀ (OD at 600 nm) was measured after growth to determine cell numbers. An A₆₀₀ of 1.0 corresponds to about 6.70 × 10⁸ CFU for the strains used. Standard curves of A₆₀₀ versus the number of CFU/ml (plate counts) confirmed that the A₆₀₀ was proportional to the viable cell number within the OD range used herein, including for final yield (i.e., saturation growth) numbers. All growth rate and yield studies were performed three times or more, with results similar to those shown (Table 2 and Fig. 1).

Amino acid and nickel uptake assays.

WT S. enterica serovar Typhimurium and mutant strains ALZ36, ALZ37, ALZ42, and ALZ43 were grown in CR-Hyd medium (without glucose, supplemented with 0.5% sodium fumarate and 5 µM NiCl₂). Cultures were grown in quadruplicate under anaerobic conditions as mentioned above, without added H₂. After the cultures reached an A₆₀₀ of 0.1, 20% H₂ (vol/vol of headspace) was injected into two of the bottles for each strain. After 60 min, uniformly ¹⁴C-labeled amino acids (specific activity, >50 mCi [1.85 GBq]/mmol; Perkin-Elmer, Boston, MA) were injected into the bottles to a final concentration of 0.5 µCi/ml of growth medium. The mixture contains 15 uniformly labeled amino acids (l-Ala, l-Arg, l-Asp, l-Glu, l-Gly, l-His, l-Ile, l-Leu, l-Lys, l-Phe, l-Pro, l-Ser, l-Thr, l-Tyr, and l-Val). ¹⁴C-labeled amino acid uptake by the cultures was measured at 5, 10, and 20 min by a previously described method (19). For the inhibitor effects, cells were grown as described above and CCCP or sodium orthovanadate (Sigma-Aldrich Co., St. Louis, MO) was added 1 and 10 min before the addition of the radiolabeled amino acid mixture, respectively. CCCP was added to a final concentration of 50 µM, and sodium orthovanadate was added to a final concentration of 10 mM. These concentrations have been used for studies of other enteric bacteria (28, 31). The experiments (Tables 1 to 3) were repeated a total of three times with similar results, and the data shown are from four replicate samples from one experiment.

For the nickel uptake assays, cultures were grown anaerobically without H₂ as described above. After the cultures reached an A₆₀₀ of 0.1, 20% (vol/vol of headspace) H₂ was injected into the bottles. After 60 min, ⁶³Ni (Amersham Biosciences, Sweden) was injected into the bottles to a final concentration of 0.5 µCi/ml of growth medium and the uptake activity was measured at 1- and 5-min intervals.

Real-time quantitative PCR.

RNA was isolated from the test (20% H₂ added to the medium) and control (no added H₂) cultures (A₆₀₀ = 0.4) of the WT and ALZ42 strains using the RNA extraction kit from Qiagen (Qiagen Inc., Valencia, CA) by following the manufacturer’s instructions. First-strand cDNA was synthesized from 200-ng purified RNA samples using random hexamers and Moloney murine leukemia virus SuperScript III reverse transcriptase (Invitrogen) at 42°C for 50 min. The iCycler iQ real-time PCR detection system (Bio-Rad Laboratories, Hercules, CA) and iQ Sybr Green Supermix (Bio-Rad) were utilized for real-time PCR of control and test cDNAs. The expression level (threshold cycle) of each sample was normalized using DNA gyrase B (gyrB) as an internal control. The relative n-fold change in gene expression for each sample was determined using the 2^−ΔΔCT method as previously described (32). The gene-specific primers used for real-time PCR are listed in Table 6.

Western blot assays.

To identify the effect of H₂ on the expression levels of the ExbD and TonB proteins under the condition provided, overnight cultures of the WT and ALZ42 strains grown anaerobically in the presence or absence of 20% H₂ (in CR-Hyd medium without glucose and supplemented with 0.5% sodium fumarate and 5 µM NiCl₂) were subjected to the in vivo formaldehyde cross-linking method previously described (28). The ExbD and TonB proteins and their cross-linked complexes were detected by immunoblotting using ExbD-specific polyclonal antibodies and TonB-specific monoclonal antibodies (33). The antibodies were kindly provided by Kathleen Postle, Pennsylvania State University, University Park. The entire cross-linking experiment was repeated three times with the same results, as shown in Fig. 3.

Hydrogenase assay.

The H₂ uptake hydrogenase activity of the WT and ALZ42 strains was assayed in whole cells by following the reduction of methylene blue spectrophotometrically by a method modified from Stults et al. and Peng et al. (34, 35). Cells were grown with 20% H₂ in CR-Hyd medium (without glucose and supplemented with 0.5% sodium fumarate and 5 µM NiCl₂) to mid-exponential phase (A₆₀₀ = 0.4). A 2-ml sample of the culture was centrifuged (8,000 × g, 10 min), and cells were resuspended in 1 ml of phosphate-buffered saline. Cells were permeabilized by adding 10 µl of 10% Triton X-100 and incubating them for 30 min at room temperature. A 500-µl aliquot of the suspension was transferred to a sealed glass cuvette previously flushed with H₂. Sodium dithionite was then injected to a final concentration of 200 µM, followed by the injection of H₂-flushed methylene blue to a final concentration of 400 µM. Hydrogen uptake activity was determined by measuring the reduction of methylene blue at 570 nm and is expressed as nmol H₂ taken up/min/10⁹ cells.

Measurement of membrane potential (ΔΨ).

The membrane potential of the WT, ALZ42, and ALZ43 strains was measured using confocal fluorescence microscopy as described by Jovanovic et al. (36), with modifications as described herein. Cells were grown to mid-exponential phase in the presence of 20% H₂ as described above. Cells were harvested by centrifuging 2 ml of culture (8,000 × g, 10 min) and resuspended in 1 ml of permeabilization buffer (10 mM Tris, pH 7.5, 1 mM EDTA, 10 mM glycerol) in sealed tubes previously sparged with N₂ and injected with 20% H₂. The cells were then incubated with 1 µg/ml JC-1 (Molecular Probes, Invitrogen) for 30 min at room temperature, followed by centrifugation. Cells were then resuspended in 500 µl permeabilization buffer, and 5 µl of the suspension was immediately transferred to an agarose-coated glass slide prepared as described by Glaser et al. (37). Cells were visualized using a Leica SP5 confocal microscope (Leica Microsystems) at an excitation wavelength of 485 nm, and fluorescence emission at 530 nm (green) and 590 nm (red) was observed. The cationic dye JC-1 forms red fluorescent aggregates at higher potential and remains as green fluorescent monomers at lower potential. A decrease in the membrane potential is hence indicated by a shift in the fluorescence emission from red (590 nm) to green (530 nm). The data shown (green/red ratio) for each strain are from 600 individual cells, with 100 cells taken from each of six different fields. The six fields are from two individual cultures of each strain, each assayed three times.