Impact of Hydrogen on the Transcriptome of Sinorhizobium meliloti 1021 Using RNA-sequencing Technology.

Hydrogen formed during nitrogen fixation in legumes can enter the surrounding soil and confer multiple benefits to crops. Here, we used Sinorhizobium meliloti 1021, whose genome was sequenced in 2001, as a model bacterium to study the relationship between the bacterium and legume. We investigated the effects of hydrogen on the gene expression in S. meliloti using RNA-sequencing technology. We identified 43 genes whose expression was altered by hydrogen treatment; among these, 39 were downregulated, and 4 were upregulated. These genes accounted for 1.5% of the total 2941 annotated genes of the S. meliloti genome. Gene ontology and pathway analyses revealed that the hydrogen-regulated genes were associated with catalytic activity and binding. Further, these genes were primarily involved in arginine, proline, and β-alanine metabolism. Real-time PCR revealed that the transcription levels of SMc02983, cyoB, cyoC, and cyoD were reduced after hydrogen treatment. These results provide a theoretical framework for exploring new metabolic pathways of S. meliloti. Hydrogen formed during nitrogen fixation in legumes can enter the surrounding soil and confer multiple benefits to crops. Here, we used Sinorhizobium meliloti 1021, whose genome was sequenced in 2001, as a model bacterium to study the relationship between the bacterium and legume. We investigated the effects of hydrogen on the gene expression in S. meliloti using RNA-sequencing technology. We identified 43 genes whose expression was altered by hydrogen treatment; among these, 39 were downregulated, and 4 were upregulated. These genes accounted for 1.5% of the total 2941 annotated genes of the S. meliloti genome. Gene ontology and pathway analyses revealed that the hydrogen-regulated genes were associated with catalytic activity and binding. Further, these genes were primarily involved in arginine, proline, and β-alanine metabolism. Real-time PCR revealed that the transcription levels of SMc02983, cyoB, cyoC, and cyoD were reduced after hydrogen treatment. These results provide a theoretical framework for exploring new metabolic pathways of S. meliloti.

Introduction

Many microorganisms associated with plants are active in the rhizosphere; those that promote plant growth, control diseases, and increase crop yield are collectively referred to as plant growth-promoting rhizobacteria (PGPR). Sinorhizobium meliloti is an important PGPR that carries out biological nitrogen fixation through symbiosis with its hosts. This symbiosis has several benefits in agriculture, such as reducing the need for adding nitrogen fertilizers to the soil and improving crop tolerance (Wang et al. 2016). In many legume nodules, H2 produced as a byproduct of N2 fixation diffuses out of the nodules and is absorbed by the soil (Dong et al. 2001).

Previous studies have demonstrated that nitrogen fertilizers cannot completely replace the beneficial rotation effect of legumes. The 75% increase in the yield of soybean cannot be explained by existing theories (Heaterman et al. 1986). The nodules on legumes release large amounts of H2 during nitrogen fixation. Approximately 1.5 mol of H2 is produced for every unit of N2 fixed. Soybeans with nitrogen-fixing capacity can produce approximately 5000 l of H2 per hectare per day at peak growth and 2.4 × 105 l H2 per hectare per the growing season. The production of H2 requires 5–6% of the net energy produced during photosynthesis (Dong et al. 2003). From a metabolic point of view, the release of H2 from the root is a waste of energy; however, neither long-term natural evolution nor manual screening strategies have successfully reduced this energy loss.

Most of the hydrogen released from legume nodules is absorbed by the soil (La Favre et al. 1983). A large amount of hydrogen is released during the nitrogen fixation process, but only a small amount is released into the atmosphere because of the actions of soil microorganisms that harvest the produced H2 for energy. Kärst et al. (1987) demonstrated that hydrogen-utilizing bacteria could promote plant growth because they possess hydrogenase, a metal-containing protein that catalyzes the reduction of protons to produce hydrogen as an end product of the electron transport chain (Kärst et al. 1987). The reversible reaction of catalytic protons to electron production is as follows: 2H+ + 2e– ↔ H2.

S. meliloti 1021 is a Gram-negative soil bacterium that can co-produce nitrogen in the nodules of legumes. S. meliloti 1021 fixes the gaseous-phase nitrogen into ammonia; consequently, the fixed nitrogen, in the form of ammonia or alanine, is supplied to the host plant, which in turn provides nutrients to the resident bacteria (Spaink 2000).

This study aimed to provide a new understanding of the role of hydrogen in the plant-rhizobacterium interaction and gain a better understanding of metabolic processes underlying these relationships using S. meliloti 1021 as a model bacterium. To this end, we conducted transcriptome analysis of S. meliloti 1021 using RNA-sequencing (RNA-Seq) technology to identify differentially expressed genes in the presence and absence of hydrogen.

Experimental

Materials and Methods

Bacteria. The strain is a model bacterial strain used to study the relationship between Rhizobium and legume (Barnett et al. 2001; Capela et al. 2001; Galibert et al. 2001). Single colonies were preserved in our laboratory and maintained on nutrient agar slants. Nutrient agar was prepared by dissolving 3 g beef extract, 5 g peptone, 5 g NaCl, and 15 g agar in 1000 ml distilled water (pH 7.2), followed by sterilization using an autoclave. The agar slants were stored at 4°C until use.

Hydrogen treatment. Ten milliliters of the prepared mineral salt medium was spread flat over the bottom of the glass culture flask and inoculated with S. meliloti 1021. This experiment was performed in triplicate. After the glass culture flask was sealed with a sterile rubber stopper, hydrogen (Wuhan Newradar Special Gas Company, Wuhan, Hubei Province, China) gas was injected into the flask from the sampler until uniformly distributed, and then the flask was placed horizontally on a shaker (80 rpm, 28°C). Hydrogen concentration was determined by gas chromatography at 13 000 ppm. In addition, the relative oxygen content was calculated to be 1.8% (Trace-1300; Thermo Fisher Scientific (Shanghai, China) Co., Ltd.) on a column (30 m × 0.32 mm × 30.0 μm) by setting the Thermal Conductivity Detector (TCD) temperature to 200°C and split ratio to 15:1. Air was provided to the control group. After incubation for 24, 48, and 72 h, the cells were washed with phosphate-buffered saline and centrifuged. The experiment was conducted with three biological replicates in each group.

Preparation and detection of RNA samples. Equal numbers of hydrogen-treated and control samples were combined to extract total RNA. The mRNA was fragmented, and the fragments were used as templates for reverse transcription to synthesize the first-strand cDNA using ProtoScript II Reverse Transcriptase (New England Biolabs, Ipswich, MA, USA), random primers, and actinomycin D. Second-strand cDNA was synthesized using Second Strand Synthesis Enzyme Mix (New England Biolabs). It was purified using AxyPrep Mag PCR Clean-up (Axygen) and treated with End Prep Enzyme Mix (Axygen) to repair both ends and add dA-tails in a single reaction. Following end-repair and adaptor ligation, RNA-seq data were obtained using a HiSeq instrument (Illumina, San Diego, CA, USA).

Data analysis. As the genome of S. meliloti 1021 has already been sequenced, DNA microarrays can be used to study the regulation of gene expression. Clean data from RNA-Seq were aligned to the S. meliloti 1021 reference genome using Bowtie2 (Langmesd et al. 2012). Significant differences between the hydrogen-treated and control groups were identified by setting the false discovery rate (FDR) p-value threshold at 0.05. The DESeq package based on the negative binomial distribution was used for this analysis. The open reading frames of the transcriptome data were annotated for taxonomic compositions and functions. We used the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (Kanehisa et al. 2000) for the analysis of pathways associated with the differentially expressed genes. Moreover, the differentially expressed genes were subjected to the Gene Ontology (GO) enrichment for functional annotation, an internationally standardized gene function classification system that provides a dynamically updated vocabulary to fully describe the properties of genes and gene products in an organism (Li et al. 2015).

RT-qPCR and physiological research verification. For the specific genes SMc02983, SMc01578, SMc01656, SMc02677, SM_b20752, cyoD, cyoC, and cyoB, real-time quantitative PCR was performed with the following primers. SMc02983:

5’-GGCACAATCGTGCGTGTCCTTCAAC-3’ and 5’-GCCATTCGCCCATCTCAT-3’;

SMc01578: 5’-GACTAGACTACGGTCCCAAC-3’ and 5’-TTATTACTCGCTCGTGATCTCATC-3’;

SMc01656: 5’-GCCTTGTGTCTGAGGTTATCATC-3’ and 5’- CGTGGGAGGATGGCTCGAATG-3’;

SMc02677: 5’- GCGTGCACAAGGGCCGCATGAG-3’ and 5’-CTCGGCGTCCTTGAGCACTCG-3’;

SM_b20752: 5’-GGACGAGCGCCTGAACCACG-3’ and 5’-GCTCAGGAGCCATGCCTGTC-3’;

cyoD: 5’-GCCTGGAGAGGCCGCATGAAG-3’ and 5’-GTCGCCGGTCCTTGAGACCTCG-3’;

cyoB: 5’- CACTACGATACGGCTCCAAC-3’ and 5’-TTATTTACTCGCTCGGCCCTC-3’;

cyoC: 5’-TGCAGGACGTTCGCGGAGCT- 3’ and 5’-GAGCGCAGCTCTGTCGG- 3’.

We used 16S rRNA as an internal reference gene to calculate normalized gene expression intensity using the Q-gene software (Muller et al. 2002).

The content of 17 amino acids was determined according to the method described by Sokolova et al. (2007) in an automatic amino acid analyzer (Hitachi 835–50 amino acid analyzer; Tokyo, Japan) with the measurement repeated three times for each sample. The S. meliloti respiratory complexes were measured using methods based on mitochondrial respiratory complex chain determination, as described by Vyatkina et al. (2004), and the enzyme activity was expressed as μmol min–1 mg–1 pro–1. For this, 20 μl of the bacterial suspension was added to the PBS buffer, followed by the addition of 80 μM benzoquinone, 2 mM NaN3, and 0.4 μM antimycin A (10 μl each). After mixing well, the mixture was incubated in a 28°C-water bath for 3 min, and then 10 μl of 200 μM NADH was added to start the reaction. Immediately after the reaction, a standard quartz cuvette was used to measure the OD value at 340 nm using a UV spectrophotometer. The CI activity was calculated using the standard curve.

Similarly, 50 μM 2,6-dichlorophenol indophenol, 2 mM NaN3, 2 μg/ml rotenone, 2 μg/ml antimycin A, and 25 μM benzoquinone (10 μl each) were added to the buffer to which 20 μl of the bacterial suspension was added, and after mixing well, the mixture was incubated in a 28°C-water bath for 3 min, followed by the addition of 10 μl 20 μM sodium succinate to start the reaction. Immediately afterward, the OD value was measured with a standard quartz cuvette at 600 nm in a UV spectrophotometer. The CII activity was calculated from the standard curve.

As indicated above, 10 μl of 2 mM NaN3 and 10 μl of 50 μM cytochrome c reductase were added sequentially to the buffer to which 20 μl of the bacterial suspension was added, and the mixture was incubated in a 28°C-water bath for 3 min, followed by the addition of 10 μl of 80 μM coenzyme Q to start the reaction. Immediately afterward, the OD value was measured with a standard quartz cuvette at 550 nm using a UV spectrophotometer. The CIII activity was determined from the standard curve.

Similarly, 10 μl of 0.025% n-dodecyl-p-d-maltoside was added to the buffer to which 20 μl of the bacterial suspension was added, and after mixing well, the mixture was incubated in a 28°C-water bath for 3 min, followed by the addition of 10 μl of 50 μM cytochrome c. The OD value was then measured immediately with a standard quartz cuvette at 550 nm using a UV spectrophotometer, and then the CIV activity was calculated from a standard curve. These measurements were repeated three times for each sample.

Results

Differentially expressed genes after hydrogen treatment. Analysis of the abundance of transcripts suggested that more than 99.58% of the sequenced data were meaningful. Relative to air-only control levels, 39 genes were downregulated and four genes were upregulated after hydrogen treatment for 72 h; the differential expression was defined as a fold change in expression level that was greater than 2. The fragments per kilobase of transcript per million fragments mapped (FPKM) values of differentially expressed genes before and after hydrogen treatment represented their expression levels, which were then used in the hierarchical clustering analysis. The clustering analysis resolved the three biological replicates into one class, indicating that the biological replicates were consistent and that the experimental results were reliable (Fig. 1). These genes are listed in Table I.

Fig. 1.

Differential gene clustering. Red dots indicate downregulated genes and blue dots indicate upregulated genes in comparison with the control.

Table I

Differentially expressed genes – change with hydrogen treatment.

Gene ID	Gene name	Change with hydrogen treatment	Gene description
Gene1418	sugE	Down	Quaternary ammonium compound efflux SMR transporter SugE
Gene1984	msbB	Down	Lipid A biosynthesis (KDO)2-(lauroyl)-lipid IVA acyltransferase
Gene2119	SMc04350	Down	Multidrug efflux system transmembrane protein
Gene2120	SMc04351	Down	Transmembrane ATP-binding ABC transporter protein
Gene2494	mtlK	Down	Oxygen-independent coproporphyrinogen III oxidase
Gene2940	SMc02983	Down	Arginine decarboxylase
Gene3019	SMc03139	Ups	Hypothetical protein
Gene3089	SMc02519	Down	ABC transporter ATP-binding protein
Gene3090	SMc02518	Down	ABC transporter ATP-binding protein
Gene3090	SMc02518	Down	ABC transporter ATP-binding protein
Gene3091	SMc02517	Down	ABC transporter permease
Gene3471	SMa0081	Ups	ABC transporter permease
Gene361	iolB	Down	5-deoxy-glucuronate isomerase
Gene362	iolE	Down	Myo-inosose-2 dehydratase
Gene4065	SMa1163	Down	Cation transport P-type ATPase
Gene4073	SMa1176	Down	Hypothetical protein
Gene4075	nosR	Down	NosR regulatory protein for N2O reductase
Gene4076	nosZ	Down	Nitrous-oxide reductase
Gene4078	nosF	Down	NosF ATPase
Gene4079	nosY	Down	NosY permease
Gene4082	fhp	Down	Nitric oxide dioxygenase
Gene4092	fixI1	Down	ATPase
Gene4093	fixH	Down	Nitrogen fixation protein FixH
Gene4094	fixG	Down	FixG iron sulfur membrane protein
Gene4095	fixP1	Down	FixP1 di-heme cytochrome c
Gene4097	fixO1	Down	Cbb3-type cytochrome c oxidase subunit II
Gene4123	hemN	Down	Oxygen-independent coproporphyrinogen III oxidase
Gene4550	SMa2051	Down	Desaturase
Gene4551	SMa2053	Down	MocE-like protein
Gene5192	SM_b20487	Down	Sugar ABC transporter permease
Gene5193	SM_b20488	Down	Hypothetical protein
Gene5194	SM_b20489	Down	Carbohydrate kinase
Gene5727	groEL	Down	Chaperonin GroEL
Gene5728	groES5	Down	Molecular chaperone GroES
Gene6038	cyoB	Down	Cytochrome O ubiquinol oxidase subunit I
Gene6039	cyoC	Down	Cytochrome O ubiquinol oxidase subunit III
Gene6040	cyoD	Down	Cytochrome O ubiquinol oxidase CyoD
Gene6083	SM_b20654	Ups	Hypothetical protein
Gene6166	SM_b20753	Ups	Acyl-CoA dehydrogenase
Gene6288	agaL1	Down	Alpha-galactosidase (melibiase) protein
Gene811	groES	Down	Co-chaperone GroES (Cpn10) binds to Cpn60 in the presence of Mg-ATP and suppresses the ATPase activity of the latte
Gene968	betA	Down	Choline dehydrogenase
Gene969	betB	Down	Betaine aldehyde dehydrogenase
Gene970	betC	Down	Choline-sulfatase

GO analysis. GO contains three parameters that describe the molecular function, cellular component, and biological processes associated with the gene of interest. The differentially expressed genes under hydrogen treatment were classified into 10 main functions. In the molecular function category, 13 genes were associated with catalytic activity, binding, transporter activity, and electron carrier activity. In the cellular component category, eight genes were enriched; these were mainly involved in membrane and membrane components (Fig. 2). The genes associated with biological processes were primarily related to localization and metabolic process. Overall, the GO results indicated that hydrogen treatment primarily influences catalytic activity and binding of S. meliloti 1021. Eight of the differentially expressed genes (29% of the total) were associated with these two processes, suggesting that catalytic activity plays a crucial role in the response of S. meliloti to hydrogen.

Fig. 2.

Histogram of GO terms associated with differentially expressed genes; the GO term is on the ordinate; the number of differentially expressed genes associated with each term is on the abscissa.

KEGG pathway analysis and verification of related genes. A total of 3525 genes were assigned to 19 metabolic pathways from the KEGG database. The pathways were then restructured to categorize the differentially expressed genes into multiple classes. This revealed that 12 genes were involved in metabolic pathways; notably, the ABC transporter and oxidative phosphorylation pathways were modulated by hydrogen treatment. In addition, two of the downregulated genes were associated with two-component systems, inositol phosphate metabolism and glycine/serine/threonine metabolism. The detailed results are shown in Fig. 3.

Fig. 3.

Differential gene bubble map in the KEGG metabolic pathway.

During the KEGG pathway analysis, we noted that the arginine and proline metabolism pathway (ko00330) was downregulated in the hydrogen treatment group. To clarify this pathway’s connection with hydrogen treatment, we checked if the pathway involves any hydrogenases. There are 3896 hydrogenases (IUBMB database, last accessed on January 2019). We shortlisted the hydrogenases (Table II) involved in the arginine and proline metabolism pathway to determine the effect of hydrogen treatment.

Table II

Hydrogenases involved in arginine and proline metabolism.

EC number	Name
1.2.1.19	Aminobutyraldehyde dehydrogenase
1.2.1.3	Aldehyde dehydrogenase (NAD+)
1.4.3.22	Diamine oxidase
1.2.1.71	Succinylglutamate-semialdehyde dehydrogenase
1.2.1.88	L-glutamate gamma-semialdehyde dehydrogenase
1.5.1.19	D-nopaline dehydrogenase
1.5.1.11	D-octopine dehydrogenase

Hydrogen treatment caused a decrease in the expression of the gene encoding aminobutyraldehyde dehydrogenase (EC: 1.2.1.19), which is an important hydrogenase in the synthesis of 4-amino-butanoate and in β-alanine metabolism. Moreover, butanoate, alanine, glutamate, and ADC activities were also decreased under hydrogen treatment, which is the same outcome observed during the synthesis of β-alanine (ko00410).

The results of the amino acid analysis showed that the content of arginine, glycine, proline, serine, and threonine decreased after hydrogen treatment, and the content of arginine and proline exhibited the most significant decrease (Table III). After 72 h, the arginine content decreased by 0.47 mg g–1, and the proline content decreased by 0.51 mg g–1. The content of some other amino acids was elevated, but the increase was small.

Table III

The content of amino acids after hydrogen treatment.

Amino acid	Concentration (mg/g)
	Control group	Test group
		24 h	48 h	72 h
Alanine	0.32 ± 0.012	0.292 ± 0.03	0.342 ± 0.011	0.372 ± 0.05
Arginine	0.0842 ± 0.006	0.0682 ± 0.002	0.0622 ± 0008	0.0372 ± 0.004
Aspartic acid	1.822 ± 0.015	1.862 ± 0.3	1.882 ± 0.29	1.772 ± 0.24
Cysteine	0.182 ± 0.07	0.152 ± 0.04	0.222 ± 0.06	0.142 ± 0.03
Glutamic acid	2.12 ± 0.25	2.252 ± 0.16	2.242 ± 0.19	2.22 ± 0.11
Glycine	0.0772 ± 0.004	0.0742 ± 0.006	0.0712 ± 0.003	0.0682 ± 0.009
Histidine	0.162 ± 0.07	0.182 ± 0.09	0.172 ± 0.05	0.152 ± 0.02
Isoleucine	0.72 ± 0.037	0.682 ± 0.02	0.772 ± 0.03	0.732 ± 0.08
Leucine	0.252 ± 0.022	0.222 ± 0.03	0.262 ± 0.04	0.292 ± 0.05
Lysine	3.62 ± 0.17	3.42 ± 0.11	3.32 ± 0.2	3.72 ± 0.19
Methionine	0.092 ± 0.003	0.062 ± 0.001	0.092 ± 0.004	0.0122 ± 0.002
Phenylalanine	1.52 ± 0.08	1.72 ± 0.05	1.622 ± 0.07	1.582 ± 0.03
Proline	0.0732 ± 0.002	0.0642 ± 0.008	0.0262 ± 0.002	0.0222 ± 0.004
Serine	0.0652 ± 0.007	0.0612 ± 0.002	0.062 ± 0.008	0.0552 ± 0.007
Threonine	0.0782 ± 0.005	0.0772 ± 0.005	0.0542 ± 0.007	0.0582 ± 0.003
Tyrosine	None	None	None	None
Valine	0.062 ± 0.003	0.0622 ± 0.007	0.0642 ± 0.004	0.0672 ± 0.008

Oxidative phosphorylation (ko00190) is an important aspect of metabolism and is closely related to nitrogen fixation (Allen and Arnon 1955). Hydrogen treatment increased the expression of cytochrome c oxidase (EC:1.9.3.1), inhibiting the H+ consumption process in complex IV to reduce the intensity of oxidative phosphorylation.

Nitrogenase can capture large amounts of H+ for nitrogen fixation, thereby reducing the consumption of H+ in oxidative phosphorylation and providing energy for nitrogen fixation by nitrogenase (Millar et al. 1995).

Cytochrome O ubiquinol oxidase subunit I (cyoB), cytochrome O ubiquinol oxidase subunit III (cyoC), and cytochrome O ubiquinol oxidase CyoD (cyoD) are located on plasmid pSymB and participate in oxidative phosphorylation. SMc02983 encodes arginine decarboxylase, SMc01656 encodes gamma-aminobutyraldehyde dehydrogenase, SM_b20752 encodes 3-hydroxyisobutyryl-CoA hydrolase, SMc01578 encodes aspartate aminotransferase, and SMc02677 (proC) encodes pyrroline-5-carboxylate reductase; these genes are involved in arginine and proline metabolism. The transcription levels of SMc02983, SMc01656, SMc01578, SMc02677, SM_b20752, cyoB, cyoC, and cyoD, before and after hydrogen treatment, were measured by real-time quantitative PCR (Fig. 4). SMc02983 and cyoB showed a difference after 24 h, and the differential expression levels were 0.72 and 0.59, respectively; this difference reached the maximum (1.52 and 1.29) after 72 h. However, the difference for cyoC was maximal (1.06) at 48 h, after which it decreased. The expression of cyoD gene showed the least difference after 24 h, but it increased and showed a maximum difference of 1.27 after 72 h. The genes SMc02677 and SM_b20752 were upregulated, and the fold difference was the highest at 48 h (0.55) and 72 h (0.85), respectively. The data are consistent with the RNA-seq data, indicating that the expression of the four genes was significantly downregulated. Among these, SMc02983 had the highest reduction in expression.

Fig. 4.

Fold change in the expression levels of SMc02983, SMc01578, SMc01656, SMc02677, SM_b20752, cyoD, cyoB, and cyoC after the hydrogen treatment as determined by real-time quantitative PCR.

The results of oxidative phosphorylation complex activity showed that hydrogen treatment reduced the activity of complex IV most significantly. There may also be other mechanisms underlying the role of CIV activity in reducing complex II activity by regulating the electron transfer between CIII and CIV (Fig. 5).

Fig. 5.

The oxidative phosphorylation complex activity map. Each experiment was performed three times.

Discussion

S. meliloti can live as a soil saprophyte and engage in nitrogen-fixing symbiosis with plant roots (Barnett et al. 2012). A large amount of hydrogen is produced during nitrogen fixation in symbioses (Osborne et al. 2010). S. meliloti 1021 is an important N2-fixing bacterium in the root nodule of legumes (Capela et al. 2001). It is a model strain for studying Rhizobium and has further research value. At present, there are many studies on drought and heavy metals, and few studies have focused on how hydrogen affects Rhizobium (Elboutahiri et al. 2010). In this study, the effect of hydrogen on Rhizobium was analyzed using RNA-Seq technology. The response to hydrogen treatment was visualized through up- and down-regulated genes. Hydrogen primarily affected amino acid metabolism. Hydrogen-metabolizing organisms use [NiFe]-hydrogenase to catalyze hydrogen oxidation and the NAD+-reducing soluble [NiFe]-hydrogenase (SH) couples reduction of NAD+ to the oxidation of hydrogen (Vinson 2017). Hydrogenases catalyze the oxidation of hydrogen into protons and electrons, to use H2 as an energy source or for the production of hydrogen through proton reduction (Albareda et al. 2012). This is consistent with the findings of several studies that indicated amino acid metabolism to be a major feature of the S. meliloti chromosome (Capela et al. 2001). Consistent with these reports, we found that aminobutyraldehyde dehydrogenase is primarily involved in arginine, proline, and β-alanine metabolism, and its expression is downregulated. The determination of amino acid content also revealed a decrease in the content of proline and arginine. A proline catabolic mutant had a negligible impact on nitrogenase activity as evaluated through acetylene reduction (Chien et al. 1991). However, S. meliloti exhibited positive chemotactic responses to all proteinogenic amino acids, except aspartate, citrulline, cysteine, gamma-aminobutyric acid, and ornithine (Webb et al. 2017). Hydrogen treatment provided microaerobic conditions to S. meliloti 1021. Expression of genes involved in nitrogen fixation and the associated respiration is governed by two intermediate regulators, NifA and FixK, respectively, which are controlled by a two-component regulatory system FixLJ in response to low-oxygen conditions. FixJ controls 74% of the genes induced in microaerobiosis (2% oxygen) (Bobik et al. 2006). According to Defez’s research, under microaerobic conditions, the overexpression of IAA in free-living S. meliloti cells induced many of the transcriptional changes that normally occur in nitrogen-fixing root nodules (Defez et al. 2016). Therefore, hydrogen treatment, in this study, restricted the two pathways containing hydrogenase, but provided a micro-oxygen environment for S. meliloti. Further, the fixN gene and cyoB and cyoD genes were downregulated. Real-time quantitative PCR results showed that some genes have no reaction time for hydrogen; nonetheless, the genes involved in the metabolism of arginine and proline were the most different, indicating that amino acid synthesis is an important biological process of S. meliloti 1021. The expression of SMc02677, SM_b20752, and other genes partially metabolized by arginine and proline was upregulated. These genes were not screened for the transcriptome differential genes. Therefore, after reducing the screening of multiple differential genes, the expression of genes related to nitrogen fixation and amino acid synthesis was upregulated.

The differential genes involved in the experiments were annotated in several aspects. Based on previous studies, we focused on hydrogenase pathways. These genes were verified by real-time quantitative PCR and were found to be consistent with RNA-Seq findings. Therefore, the research is somewhat reliable. Taken together, hydrogen affects S. meliloti 1021 by affecting amino acid synthesis and oxidative phosphorylation processes. It may be associated either with the competition for carbon compounds imported into the nodules between the energy production and nitrogen assimilation processes or with the competition for redox potentials between the oxidative phosphorylation and nitrogen fixation processes (Provorov et al. 2013). We speculate that hydrogen is a beneficial factor for the symbiotic system of S. meliloti 1021. Previous studies have shown that the release of hydrogen to the soil has a beneficial effect and that the development of PGPR as the seed inoculants is an important step toward sustainable agriculture (Golding 2009). As an important type of PGPR, S. meliloti 1021 has potential development significance for the symbiosis between legumes and rhizobia. This study explores these beneficial effects of the S. meliloti 1021 metabolic processes.

Conclusion

We determined the transcriptomic changes in S. meliloti 1021 under the hydrogen treatment. Overall, the results demonstrate that hydrogen alters the expression of genes related to aminobutyraldehyde dehydrogenase and cytochrome c oxidase, which in turn affect metabolic pathways, especially those involved in amino acid metabolism and oxidative phosphorylation. These results may provide a theoretical basis for more effective agricultural exploitation of S. meliloti 1021.