Hydrogen-Stimulated Carbon Acquisition and Conservation in Salmonella enterica Serovar Typhimurium

Abstract

Salmonella enterica serovar Typhimurium can utilize molecular hydrogen for growth and amino acid transport during anaerobic growth. Via microarray we identified H₂ gas-affected gene expression changes in Salmonella. The addition of H₂ caused altered expression of 597 genes, of which 176 genes were upregulated and 421 were downregulated. The significantly H₂-upregulated genes include those that encode proteins involved in the transport of iron, manganese, amino acids, nucleosides, and sugars. Genes encoding isocitrate lyase (aceA) and malate synthase (aceB), both involved in the carbon conserving glyoxylate pathway, and genes encoding the enzymes of the d-glucarate and d-glycerate pathways (gudT, gudD, garR, garL, garK) are significantly upregulated by H₂. Cells grown with H₂ showed markedly increased AceA enzyme activity compared to cells without H₂. Mutant strains with deletion of either aceA or aceB had reduced H₂-dependent growth rates. Genes encoding the glutamine-specific transporters (glnH, glnP, glnQ) were upregulated by H₂, and cells grown with H₂ showed increased [¹⁴C]glutamine uptake. Similarly, the mannose uptake system genes (manX, manY) were upregulated by H_2, and cells grown with H₂ showed about 2.0-fold-increased [¹⁴C]d-mannose uptake compared to the cells grown without H₂. Hydrogen stimulates the expression of genes involved in nutrient and carbon acquisition and carbon-conserving pathways, linking carbon and energy metabolism to sustain H₂-dependent growth.

INTRODUCTION

Salmonella enterica serovar Typhimurium is an important bacterial pathogen, and it is implicated in a majority of the documented food-borne gastroenteritis cases in humans (47). The metabolic flexibility of S. Typhimurium allows it to survive in diverse environmental conditions both outside and within the host (19, 30). An important factor augmenting the survival capability of this bacterium in macrophages and in animal hosts is its ability to utilize molecular hydrogen produced within the host as an energy source (36, 37, 57).

Uptake-type (H₂-oxidizing) hydrogenases are considered catalysts enabling auxiliary energy source use for the generation of a proton gradient across the cell membrane (53, 54). The electrons generated in the process are passed along bacterial electron transport chains to terminal electron acceptors such as fumarate, nitrate, sulfate, CO₂, or O₂ (54), while the “sidedness” of the H₂-splitting reaction generates protons in a way to promote a proton potential across the cytoplasmic membrane. The electrochemical potential thus generated (32) can be utilized by the cells for work, such as energizing transport of nutrients against a gradient and various other cellular processes. S. Typhimurium oxidizes molecular hydrogen by the H₂-oxidizing activity of three NiFe-containing respiratory hydrogenases—Hya, Hyb, and Hyd (58). The respiratory hydrogenases are important for the virulence of S. Typhimurium (37), as the host colonic flora produces the highly diffusible H₂ gas (36).

Hydrogen can be an important energy source for bacteria growing in an environment where high-energy organic substrates are limiting (53). The availability of H₂ in a nutrient-limited condition such as the competitive environment within the host intestinal tract could therefore be instrumental to the survival of salmonellae. We recently studied the effects of exogenous H₂ on the anaerobic growth of S. Typhimurium in a nutritionally challenging medium (32). Our study showed that addition of H₂ significantly augments the growth of S. Typhimurium in a culture medium containing amino acids as the only carbon source. This H₂-mediated growth augmentation is mainly due to the enhanced ability of the bacteria to acquire amino acids from the medium and is largely facilitated by the membrane proton motive force (PMF) generated by the Hyb hydrogenase (32). This caused us to examine whether the cells have mechanisms to increase carbon acquisition when oxidizing H₂.

Microarray studies of salmonellae have documented their metabolic flexibility when facing altered availability of a nutrient (23) or by mutation of a specific metabolic factor (17, 33). In our previous study, we observed that in addition to energizing the uptake and transport processes for increased carbon (amino acid) acquisition, H₂ stimulates the expression of the proteins TonB and ExbD. The TonB-ExbD system is involved in transducing the PMF to the outer membrane and thus energizes the transport of nutrients across the membrane (44). This led us to herein further investigate the transcriptional roles of H₂ availability of S. Typhimurium. In this study, we identified potential nutrient acquisition-associated genes that are linked to H₂ metabolism, and we then performed focused physiology studies on how H₂ stimulates the acquisition and conservation of carbon by S. Typhimurium growing under carbon limitation.

MATERIALS AND METHODS

Strains, growth conditions, and reagents.

Wild-type (WT) Salmonella enterica serovar Typhimurium ATCC 14028s strain JSG210 (57) was used for the microarrays and real-time PCR-based validation of the microarray data. In addition, aceA and aceB single-deletion strains were used for physiological experiments based on the microarray results. Single-deletion mutants were constructed using the lambda Red system as previously described (11, 57). 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 a supplementary table (see Table S1 in the supplemental material).

Table 1.

Strains and plasmids used in this study

Strain/plasmid	Genotype/descriptiona	Reference
S. enterica serovar Typhimurium strains
JSG210	ATCC 14028s (WT)	57
RLK3	JSG210ΔaceA::FRT (ΔaceA)	This study
RLK4	JSG210ΔaceB::FRT (ΔaceB)	This study
Plasmids
pCP20	Ampr; contains flippase gene for λ Red mutagenesis	11
pKD46	Ampr; contains λ Red genes γ, β, and exo	11
pKD4	Kanr; contains kan cassette	11

^aFRT, flippase recombinase recognition target.

Strains were maintained in Luria-Bertani (LB) broth or on LB agar (LBA) plates. Experiments were performed in CR-Hyd medium (2, 6) containing bacteriological peptone (0.5%, wt/vol), Casamino Acids (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%) as a terminal electron acceptor. No sugar 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, and 85% N₂) for 20 min; more H₂ was then 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 (32).

RNA isolation.

Total bacterial RNA for DNA microarrays and real-time quantitative PCR was isolated from the test (20% H₂ added to the medium) and control (no added H₂) cultures (A₆₀₀ = 0.4) of WT by following a previously described method (33) with some modifications described herein. The cultures were treated with 0.15 volumes of ethanol-phenol (95%:5%, vol/vol) to protect the mRNA in the samples. The cells were then pelleted (8,000 × g; 10 min; 4°C), suspended in lysis buffer (10 mM Tris, 1 mM EDTA, 0.5 mg/ml lysozyme) and treated with 1 ml of 10% SDS for 2 min at 64°C. The suspensions were treated with 11 ml of 1 M sodium acetate, pH 5.2, and then with an equal volume of phenol and incubated at 64°C for 6 min, followed by rapid chilling on ice. The suspensions were centrifuged (10,000 × g; 10 min; 4°C); the aqueous supernatant was treated with an equal volume of chloroform and again centrifuged (10,000 × g; 5 min; 4°C). Nucleic acid was then precipitated by treating the aqueous supernatant with 0.1 vol of 3 M sodium acetate and 1.0 vol of cold isopropanol, pelleted by centrifugation (10,000 × g; 25 min; 4°C), and washed with 80% ethanol. The pellet was resuspended in 1 ml of nuclease-free water and 500 U of RNasin Plus RNase inhibitor (Promega, Madison, WI), 250 U of RQ1 RNase-free DNase (Promega, Madison, WI), 20 μl of 1 M Tris (pH 8.3), and 10 μl of 1 M MgCl₂. The mixture was incubated at 37°C for 30 min. RNA was extracted by treating once with phenol, once with phenol-chloroform (50:50, vol/vol), and twice with chloroform. The RNA was precipitated with 1.0 vol of isopropanol and 0.1 vol of 3 M sodium acetate, washed once with 80% ethanol, dried, and resuspended in nuclease-free water. The RNA samples were checked for quantity and quality by agarose gel electrophoresis and UV spectrometry at A_260/280 and A_260/230.

DNA microarrays.

The multiserotype microarray contained 5,660 PCR products covering 95% of all genes of Salmonella enterica serovar Typhimurium strain LT2, S. Typhimurium strain SL1344, Salmonella enterica serovar Typhi strain CT18, S. Typhi strain Ty2, Salmonella enterica serovar Paratyphi A strain SARB42, and Salmonella enterica serovar Enteritidis strain PT4. Each gene probe was dissolved in 50% dimethyl sulfoxide (DMSO) and spotted onto Corning Ultra-GAPS glass slides (catalogue no. 40015; Corning). Each glass slide contained triplicate identical arrays. Methods were as previously described by Lawhon et al. (33). Briefly, cDNA was synthesized from a total of 50 μg RNA and labeled with Cy3- or Cy5-conjugated dUTP using Superscript II reverse transcriptase (Invitrogen, Carlsbad, CA) and random hexamers. The labeled probes were then incubated with 0.1 M NaOH at 65°C for 10 min, followed by the addition of HCl to a final concentration of 0.1 M. The labeled nucleotides were further purified using the PCR purification kit from Qiagen (Qiagen Inc., Valencia, CA), following the manufacturer's instructions. Equal volumes of labeled probes from the test (20% H₂ added to the medium) and control (no added H₂) samples were mixed with an equal volume of hybridization solution (50% formamide, 10× SSC [1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate], 0.2% SDS). Array slides were incubated in prehybridization buffer (25% formamide, 5× SSC, 0.2% SDS) at 42°C for 45 min. Labeled probes were added to the hybridization buffer simultaneously and incubated at 95°C for 5 min and then hybridized to the arrays at 42°C for 18 h. The hybridized array slides were washed in numerous steps using buffer I (2× SSC, 0.1% SDS at 42°C) and buffer II (0.1× SSC, 0.1% SDS at room temperature [RT]), followed by rapid rinses in distilled H₂O and 95% ethanol, respectively. Three independent microarray experiments were performed for the analysis of each sample. Each microarray experiment was performed in duplicate, with the second chip hybridized with dyes swapped in order to minimize dye bias for each sample analyzed. Array slides were scanned using a ScanArray 5000 laser scanner (GSI Lumonics). Signals were recorded with ScanArray 3.0 software, and signal intensities were quantified using the QuantArray 3.0 software package (Packard BioChip Technologies, Billerica, MA). Spots were analyzed by adaptive quantitation and subsequently statistically analyzed using WebArray (56). The following parameters were used: background subtraction was performed using the “half” method, print-tip Loess normalization was employed within arrays, and scale normalization was used between arrays. Genes showing expression ratios more than 2.0 were considered upregulated, and those showing ratios less than 0.5 were considered downregulated.

Real-time quantitative PCR.

The microarray data were validated by real-time quantitative PCR of six upregulated and six downregulated genes selected from the microarray-identified differentially altered genes. RNA obtained from the test (20% H₂ added to the medium) and control (no added H₂) samples for real-time PCR were the same as those used for DNA microarrays. 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 value [C_T]) of each sample was normalized using DNA gyrase B (gyrB) as an internal control. The relative fold change in gene expression for each sample was determined using the 2^−ΔΔCT method described previously (34). The gene-specific primers used for real-time PCR are listed in Table S1 in the supplemental material.

Growth curves and endpoint growth yields.

In order to determine the involvement of AceA and AceB on the H₂-facilitated growth of Salmonella, growth curves and endpoint growth assays were performed. Sealed 165-ml bottles containing 15 ml CR-Hyd medium with 0.5% sodium fumarate (as described above) were inoculated with 1.0 × 10⁷ wild-type or deletion mutant S. enterica serovar Typhimurium cells. Cells were grown anaerobically with or without 20% H₂ for 24 h at 37°C with shaking at 200 rpm. A₆₀₀ (optical density at 600 nm [OD₆₀₀]) was measured after growth in order to determine cell number. An A₆₀₀ of 1.0 corresponds to 6.74 × 10⁸ CFU/ml for the strains used. Standard curves of A₆₀₀ versus 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 (see Fig. 3).

Fig. 3.

Comparison of H₂-facilitated growth of S. enterica serovar Typhimurium strains JSG210 (WT), RLK3 (ΔaceA), and RLK4 (ΔaceB) in CR-Hyd medium.

Enzyme activity assays.

The AceA and AceB activities of the H₂-added and H₂-absent cultures were compared by the spectrophotometric enzymatic assay methods provided by the Sigma-Aldrich Co. with modifications described herein. Cells were grown in CR-Hyd medium (without glucose, supplemented with 0.5% sodium fumarate and 5 μM NiCl₂) with and without added H₂, in triplicates for each condition. After the cultures reached an A₆₀₀ of 0.4, cells were pelleted (8,000 × g; 10 min; 4°C) and suspended in 1 ml of 30 mM imidazole buffer (pH 6.8 for AceA assay; pH 8.0 for AceB assay). The suspensions were sonicated and then centrifuged (10,000 × g; 10 min; 4°C) to obtain cell extracts. Total protein in each sample was quantitated using a Pierce bicinchoninic acid (BCA) protein assay kit (Thermo Fisher Scientific Inc., Rockford, IL) following the manufacturer's instructions. The assays were performed at 25°C, and 4 μg of total protein from each sample was used for individual assays.

Preparation of [¹⁴C]glutamine.

[¹⁴C]glutamine was synthesized in vitro from ¹⁴C-uniformly labeled glutamic acid using purified glutamine synthetase. The reaction mixture contained 50 mM imidazole (pH 7.15), 100 mM NaCl, 40 mM MgCl₂, 30 mM NH₄Cl, 10 mM ATP, 18 μCi [¹⁴C]glutamic acid (specific activity, >50 mCi/mmol; Perkin Elmer, Boston, MA), and 40 μg/ml purified Helicobacter pylori glutamine synthetase. The reaction mixture was incubated for 3 h at 37°C and was quenched by raising the temperature to 85°C for 15 min. A 1-μl portion of the quenched sample was spotted onto a thin-layer chromatographic plate (0.25 mm silica gel; Analtech, Newark, DE) and was developed using tert-butanol-acetic acid-water (25:12:10) for 6 h. Spots corresponding to cold glutamate and glutamine controls were visualized using ninhydrin staining. Regions estimated to contain [¹⁴C]glutamic acid and [¹⁴C]glutamine were scraped from the plate and quantified using liquid scintillation spectrometry. The conversion of [¹⁴C]glutamic acid to [¹⁴C]glutamine was determined to be 97% ± 2.5% (n = 4).

[¹⁴C]glutamine uptake assay.

Cells were grown in CR-Hyd medium under conditions as previously described in this paper (without glucose, supplemented with 0.5% sodium fumarate and 5 μM NiCl₂). Cultures were grown with and without added H₂ and in triplicates for each condition. After the cultures reached an A₆₀₀ of 0.1, all bottles were sparged with N₂ for 8 to 10 min at RT; the vials were then injected with CO₂ to 5% partial pressure. The OD₆₀₀ of the cultures was unchanged during the sparging period, and the activities (see Table 3) are expressed based on the cell number determined at the time of the transport assay. [¹⁴C]glutamine prepared as described above was injected into the bottles to a final concentration of 0.2 μCi/ml of growth medium. [¹⁴C]glutamine uptake by the cultures was measured at 5, 10, 15, and 60 min following a previously described method (42).

Table 3.

Comparison of [¹⁴C]glutamine uptake activities of S. enterica serovar Typhimurium JSG210 grown with versus without H₂^a

Presence of H2	[14C]Glutamine uptake [(cpm × 103)/108 cells] atb:
	5 min	10 min	15 min	60 min
+	9.1 ± 0.7*	13.1 ± 1.3*	19.6 ± 2.7*	96.1 ± 2.3*
−	7.4 ± 0.3	10.4 ± 0.6	13.3 ± 1.6	66.6 ± 4.2

^aCells were grown with and without added H₂ to an A₆₀₀ of 0.1, and the H₂ present in the cultures was removed by sparging the culture bottles with N₂. ¹⁴C-labeled glutamine was added to a final concentration of 0.2 μCi/ml, and uptake was measured in the absence of H₂, so the uptake could be compared for the two types of cultures without the H₂-PMF influence.

^b*, significantly increased activity compared to the H₂-absent condition (P, <0.01, n = 3).

¹⁴C-labeled d-mannose uptake assay.

¹⁴C-labeled d-mannose was obtained from Moravek Biochemicals (Brea, CA; specific activity, 53 mCi/mmol). The procedure and the final concentration of ¹⁴C-d-mannose used were the same as described for the [¹⁴C]glutamine uptake assay. Uptake was measured at 1, 2, 5, and 10 min after the addition of ¹⁴C-labeled d-mannose. The uptake assays as well as other physiological experiments were repeated, with results obtained similar to those reported herein.

Microarray data accession number.

The complete microarray data set has been deposited in the NCBI Gene Expression Omnibus (GEO) database under accession number GSE29739.

RESULTS AND DISCUSSION

Microarrays revealed a significantly altered H₂-dependent expression of 597 S. enterica serovar Typhimurium genes when cells were cultured anaerobically in CR-Hyd medium. Of the significantly altered genes, 176 were upregulated and 421 were downregulated. Of the significantly H₂-upregulated genes, a considerable proportion is classified as being involved in transport and metabolism (Fig. 1). Interestingly, a significant percentage of the genes downregulated by the H₂-grown salmonellae are associated with virulence. Real-time quantitative PCR of selected microarray-identified genes with expression changes showed a significant correlation between the gene expression detected by the microarrays and separately by real-time quantitative PCR (correlation coefficient [R²] = 0.96), thus validating the microarray data (Fig. 2). DNA gyrase B (gyrB) was used as an internal control to normalize the expression levels of the genes used for real-time PCR validation of the microarray data. Another validation of the data is the strong correlation between previously described H₂-upregulated proteins TonB and ExbD (32) and their H₂-affected transcript levels reported herein.

Fig. 1.

H₂-affected differential expression of genes belonging to functional groups in S. enterica serovar Typhimurium. Bars represent percentage of differentially expressed genes belonging to each group. White bars indicate the proportion of downregulated genes, and gray bars indicate the proportion of upregulated genes.

Fig. 2.

Real-time quantitative PCR of selected differentially expressed genes for validation of microarray data. Bars represent relative fold change in expression of selected microarray-identified genes in S. enterica serovar Typhimurium cells grown with H₂ compared to that in cells grown without H_2. Gray bars indicate values obtained from real-time quantitative PCR, and white bars represent values obtained from microarrays. For a comparison of the expression ratios obtained by microarrays and real-time PCR, see Table S2 in the supplemental material.

The genes that are significantly upregulated by H₂ include those that encode proteins involved in the transport and metabolism of iron, amino acids, purine, pyrimidine and nucleosides, and sugars (Table 2). Additionally, genes encoding isocitrate lyase (aceA) and malate synthase (aceB), both involved in the carbon-conserving glyoxylate pathway, and the d-glycerate pathway-associated genes (garK, garL, garR), also involved in glyoxylate metabolism, are significantly upregulated in cells grown in the presence of H₂ (Table 2). Surprisingly, the expression of the hyd hydrogenase was downregulated (3.0-fold) by H₂ while the hyb hydrogenase gene was not altered by H₂ exposure (see Table S3 in the supplemental material). For the discussion relevant to our study, a representative table of selected microarray-identified upregulated genes is shown in this study. A full set of the 597 genes affected by H₂ is provided in Table S3.

Table 2.

Microarray-identified H₂-stimulated select genes in S. enterica serovar Typhimurium

Locus IDa	Genea	Functiona	Expression ratiob (fold change)
Carbon transport and metabolism
Carbohydrates
STM2962	gudT	d-Glucarate permease	5.7
STM4077/78	yneA/B	Putative sugar transport protein	3.6/3.0
STM2960	gudD	d-Glucarate dehydratase	3.5
STM3557	ugpB	Glycerol 3-phosphate transport protein	2.6
STM2190	mglB	Galactose transport protein	2.4
STM3884	rbsB	d-Ribose transport protein	2.4
STM1830/31	manX/Y	PTS, mannose-specific enzyme IIAB-IIC	2.3/2.4
STM4325	dcuA	Anaerobic dicarboxylate transport protein	2.2
STM0685	nagE	PTS, n-acetylglucosamine-specific enzyme IIABC	2.2
STM4074	ego	Putative aldose transport protein	2.2
STM4075/76	ydeY/Z	Putative sugar transport protein	2.1/2.1
Amino acids
STM0791	hutH	Histidine ammonia lyase	4.3
STM3106	ansB	l-Asparaginase II	3.8
STM3709	kbl	Glycine acetyltransferase	2.8
STM0890	artI	Arginine transport system	2.6
STM3708	tdh	Threonine 3-dehydrogenase	2.6
STM0828-30	glnQ/P/H	Glutamine high-affinity transporter	2.0–2.6
STM2071-78	hisA-I	Histidine metabolism	2.0–2.6
STM4007	glnA	Glutamine synthetase	2.3
STM2555	glyA	Serine hydroxymethyltransferase	2.3
STM4477	pepA	Aminopeptidase A	2.3
STM0002	thrA	Aspartokinase I, bifunctional enzyme	2.0
Purine/pyrimidine/nucleotides
STM0756	nadA	Quinolinate synthetase A	6.4
STM4452	nrdD	Anaerobic ribonucleoside-triphosphate reductase	3.4
STM0012/13	dnaK/J	DNA biosynthesis	3.2/2.2
STM0757	pnuC	NMN/NmR transporter	3.1
STM2277/78	nrdA/B	Ribonucleoside diphosphate reductase 1, α and β subunits	2.9/2.3
STM3511	yhgI	DNA uptake protein	2.6
STM0523	allB	Allantoinase	2.5
STM2807	nrdE	Ribonucleoside diphosphate reductase 2, α subunit	2.4
STM2641	nadB	Quinolinate synthetase B	2.3
Iron/manganese
STM0191/93	fhuA/D	OMP receptor for iron uptake	15.2/2.0
STM2861/62/63	sitA/B/C	Salmonella Fe(II)/Mn(II) transporter	13.0/8.2/3.6
STM4055	sodA	Manganese-dependent superoxide dismutase	8.2
STM1737	tonB	Energy transducer; uptake of substrates	2.4
STM3158/59	exbD/B	Uptake of enterochelin; tonB-dependent uptake	2.0/5.0
Carbon conservation and assimilation
STM4183	aceB	Malate synthase	5.3
STM3248	garR	Tartronate semialdehyde reductase	4.9
STM3249	garL	2-Dehydro-3-deoxy-galactarate aldolase	3.9
STM4184	aceA	Isocitrate lyase	3.6
STM3247	garK	Glycerate kinase	3.3
STM4187	iclR	Acetate operon transcriptional repressor	2.9
STM0772	gpmA	Phosphoglyceromutase 1	2.8
STM0210	cdaR	Carbohydrate diacid transcriptional activator	2.2
STM1173-79	flgA-G	Flagellar biosynthesis	2.2–3.0
STM2082-90	rfbJ-V	LPS biosynthesis	2.2–2.6
STM3326	mtgA	Biosynthetic peptidoglycan transglycosylase	2.1
STM0310	ghmA	Phosphoheptose isomerase/LPS biosynthesis	2.1
STM3326	mtgA	Peptidoglycan biosynthesis	2.1
STM1752	galU	Glucose-1-phosphate uridylyltransferase	2.0

^aBased on the published genome of S. enterica serovar Typhimurium strain LT2 using the NCBI GenBank database.

^bHyphens indicate the range of expression ratio corresponding to the genes indicated in column 1. Slashes indicate the expression ratios for the two or three genes indicated in colulmn 1.

H₂ stimulates expression of iron and/or manganese acquisition and transport-associated genes in Salmonella Typhimurium.

Genes encoding iron/manganese-acquisition and transport proteins (fhuA, sitA, sitB, sitC, fhuF) are the most upregulated in Salmonella Typhimurium grown with H₂. The CR-Hyd medium (2, 6) used in this experiment contains low iron and manganese, their sole source being the bacteriological peptone that reportedly contains 0.004% iron and less than 0.001% manganese (12). It is well established that bacteria adapt to iron scarcity by upregulating the expression of iron-scavenging and transport proteins; however, they can also downregulate the synthesis of proteins that require iron for enzymatic activity (for a review, see reference 39). Manganese can function as an important alternative to iron for the activation of some enzymes in bacterial cells; an example of this in enteric bacteria is the expression of a manganese-dependent isozyme of superoxide dismutase when the iron levels are low in the growth environment (7, 39). This could be an important adaptive mechanism pathogenic bacteria use to counter the sequestration of iron by the host upon bacterial invasion. It is noteworthy that H₂ causes upregulation of the gene sodA (8.2-fold upregulated) that encodes a manganese-binding superoxide dismutase. In addition, the sitABCD genes that encode an important transporter of Mn(II) and Fe(II) (5, 24) are also upregulated by H₂ (Table 2). SitABCD is also required for full virulence of S. Typhimurium (5, 22). It has also been suggested that Mn(II) plays an important role in central carbon metabolism, and several enzymes of the carbon metabolic pathway either require Mn(II) for activity or are highly induced by Mn(II) (for a review, see reference 25). The transcription of the genes gpmM and glpX that encode the Mn(II)-dependent enzymes phosphoglyceromutase (GpmM) and fructose-1,6-bisphosphate phosphatase (GlpX), respectively, in S. Typhimurium (see reference 25) were not affected by H₂ in our microarray experiments. However, some Mn(II)-dependent enzymes involved in carbon metabolism in S. Typhimurium have likely not been identified. The H₂-dependent upregulation of genes involved in Fe(II)/Mn(II) scavenging and transport may also promote maturation of iron-containing respiratory enzymes in the H₂ oxidizing pathway.

Several of the highly upregulated genes detected by the microarrays are iron-acquisition and transport-associated genes that encode TonB-dependent proteins in Escherichia coli and Salmonella (NCBI database; http://www.ncbi.nlm.nih.gov/). The TonB protein forms a complex with two other proteins, ExbD and ExbB, that function as a secondary transport system also known as the TonB-ExbD-ExbB system. This system utilizes the cytoplasmic proton motive force (PMF) for the transport of substrates by the TonB-dependent outer membrane transport proteins (TBDTs) in E. coli and other Gram-negative bacteria (45, 49). In addition to its previously suggested role in the uptake of iron complexes and vitamin B₁₂, the TonB-ExbD-ExbB system is now implicated in the transport of other substrates such as nickel, carbohydrates, cobalt, and copper (48). In our previous study, we showed that the expression of tonB and exbD genes as well as the TonB and ExbD proteins was increased by Salmonella Typhimurium when growing in an atmosphere containing H₂ (32). The microarray experiments in the current study substantiated upregulation of the tonB, exbD, and exbB genes in H₂-added cultures of S. Typhimurium. As we proposed previously, H₂-stimulated expression of the TonB-ExbD-ExbB system proteins of S. Typhimurium is likely a way for the bacteria to enhance the uptake of the much-required nutrients such as iron-siderophores, nickel, carbohydrates, vitamin B₁₂, etc.; in this way the increased energy input from the oxidation of H₂ can be coupled to macromolecule synthesis and growth.

Microarrays also detected significant H₂-dependent upregulation of the DNA synthesis-associated genes nrdD, nrdA, nrdE, and nrdB (3.4-, 2.9-, 2.4-, and 2.3-fold upregulated, respectively). NrdD is an anaerobic ribonucleotide reductase, and NrdAB is an aerobic isozyme. Another isozyme that shows ribonucleotide reductase activity in Escherichia coli is NrdEF (8). It has been shown that NrdEF is a Mn(II)-dependent protein that supports cell replication during iron limitation in E. coli (39), and therefore the possession of the alternative Mn(II)-dependent NrdE enzyme could aid the pathogen to survive iron-limited conditions. However, in our experiments we found that the transcription of both the anaerobic and aerobic reductase genes and the isozyme subunit gene nrdE is significantly stimulated by H₂.

H₂ stimulates carbon acquisition and transport in Salmonella Typhimurium.

As stated earlier, H₂ addition significantly upregulated the expression of several genes that encode proteins involved in the acquisition and transport of carbon sources. The genes encoding glutamine high-affinity transporters (glnH, glnP, and glnQ) were significantly upregulated by H₂ exposure, as were the arginine transport proteins encoded by genes artI and artP (Table 2). It was shown earlier that H₂ rapidly augments the uptake of amino acids by Salmonella in this same medium/condition; this augmentation in amino acid uptake was attributed to the energization of the transport processes by the proton motive force generated by the oxidation of H₂ by S. Typhimurium (32). The microarray data also showed that the mannose uptake protein-encoding genes manX and manY were upregulated (2.3- and 2.4-fold, respectively) by H₂ exposure (Table 2).

Physiological experiments were conducted to assess the significance of the carbon source uptake and conservation-related microarray data. In order to eliminate the direct influence of the H₂ oxidation-generated PMF on amino acid transport, cells were grown with and without added H₂ to an A₆₀₀ of 0.1 and then the H₂ present in the cultures was removed by sparging the culture bottles with N₂. The ¹⁴C-labeled glutamine and mannose uptake abilities were then tested in the absence of H₂, so the uptake could be compared for the two types of cultures without the influence of the H₂-generated PMF. By assaying the uptake in H₂-absent conditions for both cultures, we intended to detect any change in the uptake rates in the culture as a result only of the expression of the uptake/transport-associated proteins. We found that S. Typhimurium cells pregrown with added H₂ exhibited 1.5-fold-increased uptake of ¹⁴C-labeled glutamine compared to cells without added H₂ (Table 3). Though markedly lower than the 4.0-fold-increased uptake of amino acids when the uptake assay was conducted with H₂ in our previous study (32), the increase in glutamine uptake by cells pregrown under an H₂-added condition was statistically significant. This result is likely due to the H₂-stimulated expression of the glutamine transporters GlnH, GlnP, and GlnQ. The cells pregrown with H₂ also had 1.7-, 1.7-, 2.5-, and 2.0-fold-increased ¹⁴C-d-mannose uptake at 1, 2, 5, and 10 min, respectively, compared to the cells grown in the absence of H₂ (Table 4). Increased uptake of d-mannose in the H₂-added condition could be due to the H₂-stimulated expression of the mannose uptake proteins ManX and ManY.

Table 4.

Comparison of [¹⁴C]d-mannose uptake activities of S. enterica serovar Typhimurium JSG210 grown with versus without H₂

Presence of H2	[14C]d-Mannose uptake [(cpm × 103)/108 cells] ata:
	1 min	2 min	5 min	10 min
+	25.7 ± 1.2*	25 ± 1.6*	113 ± 6*	116 ± 5*
−	14.8 ± 2.9	15 ± 2.4	45 ± 1	59 ± 4

^a*, significantly increased activity compared to the H₂-absent condition (P, <0.01, n = 3).

H₂ stimulates the carbon-conserving glyoxylate pathway and the d-glucarate/d-glycerate pathway genes in Salmonella Typhimurium.

In addition to the genes encoding proteins involved in amino acid transport, several sugar acquisition and transport-protein-encoding genes (gudT, gudD, rbsB, mglB, ugpB, manY, nagE) are also upregulated by H₂ in S. Typhimurium (Table 2). The genes encoding the enzymes of the glyoxylate pathway, namely, isocitrate lyase (aceA) and malate synthase (aceB), were H₂ upregulated 3.6- and 5.3-fold, respectively, as detected by the microarrays and 5.2- and 4.5-fold, respectively, by real-time quantitative PCR. In E. coli, AceA cleaves isocitrate to glyoxylate and succinate (50), and AceB converts acetyl coenzyme A (CoA) and glyoxylate to malate (1). The glyoxylate bypass plays an important role in conserving carbon by avoiding decarboxylation steps; it is especially important when bacteria are growing in a carbon-limited environment and a carbon pool needs to be maintained via the tricarboxylic acid (TCA) cycle to sustain gluconeogenesis and other biosynthetic pathways (10, 28, 29). Under anaerobic conditions, the TCA cycle is not functional; nonetheless, the biosynthetic precursors are maintained by alternative pathways: the reductive pathway producing succinyl-CoA and the oxidative pathway producing 2-ketoglutarate (10). Additionally, carbon flow can be mediated between these pathways by the action of the glyoxylate cycle enzyme AceA, which converts isocitrate to succinate and glyoxylate (9).

In our microarray experiment, we compared the differential gene expression as a result of addition of H₂ to the cultures; the two cultures were otherwise grown under the same (anaerobic respiration) condition. Therefore, any change in gene expression observed indicates the response of the bacteria to the H₂ added to the growth medium. Gene expression analysis revealed that both aceA and aceB are significantly H₂ upregulated. More interestingly, the regulator of the ace operon, iclR, that encodes a repressor protein known to partly regulate the ace operon (51) is upregulated about 3.0-fold by H₂ (Table 2).

We thus investigated the physiological expression of the AceA and AceB proteins by enzyme activity assays. Equal amounts of total proteins obtained in crude cell extracts of S. Typhimurium incubated with versus without H₂ were compared for enzyme activities by spectrophotometric analyses. The enzymatic activities of AceA and AceB were higher in cells with H₂ than in cells without H₂ (Table 5). The activity of AceA was, however, markedly higher than that of AceB. We further confirmed the involvement of AceA and AceB in the H₂-facilitated anaerobic growth of S. Typhimurium by gene deletion experiments. The aceA and aceB genes were deleted individually using the lambda Red system to produce ΔaceA (RLK3) and ΔaceB (RLK4) mutants (Table 1). The growth of the mutants and the wild type was then compared in CR-Hyd medium (without glucose) in H₂-added versus H₂-absent conditions. Both mutants exhibited H₂-enhanced growth and achieved final yields similar to those for H₂-incubated wild-type cells. However, the growth rates of the mutants in the exponential phase were approximately 2.0- to 2.5-fold less than that of the wild type (Fig. 3 and data not shown). In repeated growth experiments, the wild type consistently reached maximum growth yield earlier than the mutants in a way similar to the results shown in Fig. 3. Without H₂, growth rates and yields of the two mutants were similar to those of the wild type. These findings suggest that the glyoxylate bypass enzymes aceA and aceB are important for the full-blown H₂-enhanced growth of S. Typhimurium. In a carbon-limited growth environment, H₂ likely stimulates carbon flow via glyoxylate in order to conserve carbon. Increased AceA activity indicates an increased carbon flow from isocitrate via glyoxylate, thus minimizing the loss of carbon through decarboxylation of isocitrate to 2-ketoglutarate.

Table 5.

Comparison of isocitrate lyase and malate synthase enzyme activities in S. enterica serovar Typhimurium strain JSG210 grown with versus without H₂

Enzyme	Activity (μmol/min/mg)a
AceA (isocitrate lyase)
+ H2	53.0 ± 7.8*
No H2	9.7 ± 6.9
AceB (malate synthase)
+ H2	18.9 ± 0.8
No H2	11.1 ± 3.7

^a*, significantly increased activity compared to the H₂-absent condition (P, <0.01, n = 3).

The carbon-conserving glyoxylate cycle enzymes play important roles in the pathogenesis of several fungal and bacterial pathogens (for a review, see reference 14). Transcriptional profiling of phagocytosed cells showed an upregulation of all the steps of the glyoxylate cycle in Candida albicans (35). In Mycobacterium tuberculosis, isocitrate lyase mRNA expression is increased in response to human macrophages (13, 20), and malate synthase enhances adherence of the bacteria to lung epithelial cells (26). The involvement of isocitrate lyase and malate synthase in the pathogenesis of Salmonella has also been studied, and it was shown that isocitrate lyase is required for persistence of Salmonella during chronic infection but not for acute lethal infection (18, 52). Our finding shows that in a carbon-poor medium the glyoxylate pathway is maximized by H₂ so that carbon is conserved. This may be useful for Salmonella survival under H₂-rich but carbon-poor conditions such as that expected to occur within the host intestinal tract.

The addition of H₂ also stimulates the expression of genes that encode proteins associated with the d-glucarate and d-glycerate utilization pathways (Table 2). The enzyme glycerate kinase (GarK) is central to both of these pathways; in E. coli, it catalyzes the synthesis of 2-phosphoglycerate by utilizing glyoxylate when the latter is available as the main carbon substrate (21, 27). When d-glucarate, d-glycerate, or d-galactarate is available as the carbon source, GarK utilizes the substrate through a catabolic pathway (3). The gene garK is 3.4-fold upregulated by H₂ (Table 2)_. The genes encoding glucarate permease (gudT), glucarate dehydrogenase (gudD), and tartronate semialdehyde reductase (garR) that are required for glucarate uptake and utilization in E. coli (46) are also significantly upregulated by H₂ in S. Typhimurium (Table 2). Furthermore, in E. coli, the operons that encode the enzymes of the d-glucarate, d-galactarate, and d-glycerate pathways are commonly regulated by an autogenous regulator, SdaR (43). S. Typhimurium also possesses protein CdaR (carbohydrate diacid transcriptional activator, encoded by cdaR) with a similar function and 97% homology with SdaR. The cdaR gene is 2.2-fold upregulated by H₂ (Table 2). Therefore, it is possible that the H₂-stimulated expression of the genes of the d-glucarate and d-glycerate pathways is linked to the increased transcription of cdaR by H₂.

These findings connect H₂ metabolism to increased carbon use via the glyoxylate and the d-glucarate/d-glycerate pathways. During anaerobic growth in CR-Hyd mediun with added H₂, Salmonella is faced with a condition whereby energy is adequate but carbon is limited. The increased expression of the carbon-acquisition genes as well as the glyoxylate and the d-glycerate pathway genes is likely a transcriptional response by the bacteria to acquire and conserve carbon while growing in a carbon-deficient condition.

The H₂-stimulated up-expression of the d-glucarate utilization pathway in salmonellae might have important implications for survival of the bacteria under carbon-limited conditions within the animal host. d-Glucarate is normally present in the tissues and body fluids of humans (15, 38, 40) and is a major serum organic acid in humans (4). d-Glucarate is also considered to have antitumor and chemopreventive properties (31, 55) and is taken as a dietary supplement. The possession of the glucarate uptake and utilization mechanism could be particularly beneficial to the pathogenic bacteria when growing in the competitive environment within the host intestine, where most of the readily utilizable sugar sources have been exhausted by the host microbial flora. In such a condition, activation of the glucarate utilization mechanism may be expected to increase the survival of the pathogen by permitting use of an alternative carbon source. The presence of a glucarate catabolic pathway in E. coli is considered to be an evolved mechanism for use of glucarate as an alternative carbon source (46).

H₂ downregulates the expression of virulence-associated genes in S. Typhimurium in vitro.

The microarrays revealed a 2- to 100-fold downregulation of virulence-associated genes by H₂. A majority of the downregulated genes belong to the functional group of prophages and Salmonella pathogenicity islands (SPI). This indicates a much compromised virulence of the strain under the condition tested by the microarrays. However, several genes involved in cell growth which also have important roles in virulence, such as the flagellar biosynthesis-, cell envelope-, and lipopolysaccharide (LPS) biosynthesis-associated genes, are upregulated (Fig. 1; Table 2). The cells are in an in vitro nutrient-limited condition, and it can be predicted that their primary foci under such condition would be survival and cell growth rather than invasion and proliferation. It is noteworthy that several genes associated with LPS and nucleotide biosynthesis were downregulated during intracellular infection (16, 41).

In this study, we observed that H₂ exposure causes an upregulation of expression of genes and respective proteins required for the transport and metabolism of necessary nutrients. One major change caused by H₂ is increased carbon use, metabolism, acquisition and conservation, linking carbon and energy metabolism to sustain H₂-enhanced growth. We propose the results have relevance to in vivo situations where carbon is limited and H₂ is abundant (36).