Analysis of [FeFe]-hydrogenase genes for the elucidation of a hydrogen-producing bacterial community in paddy field soil


  • Ryuko Baba,

    Corresponding author
    1. Laboratory of Soil Biology and Chemistry, Graduate School of Bioagricultural Sciences, Nagoya University, Chikusa, Nagoya, Japan
    • Correspondence: Ryuko Baba, Laboratory of Soil Biology and Chemistry, Graduate School of Bioagricultural Sciences, Nagoya University, Chikusa 464-8601, Nagoya, Japan. Tel.: +81 52 789 5323;

      fax: +81 52 789 4136;


    Search for more papers by this author
  • Makoto Kimura,

    1. Laboratory of Soil Biology and Chemistry, Graduate School of Bioagricultural Sciences, Nagoya University, Chikusa, Nagoya, Japan
    Current affiliation:
    1. Food and Agricultural Materials Inspection Center (FAMIC), Shintoshin, Chuo-ku, Saitama-shi, Saitama, 330-9731, Japan
    Search for more papers by this author
  • Susumu Asakawa,

    1. Laboratory of Soil Biology and Chemistry, Graduate School of Bioagricultural Sciences, Nagoya University, Chikusa, Nagoya, Japan
    Search for more papers by this author
  • Takeshi Watanabe

    1. Laboratory of Soil Biology and Chemistry, Graduate School of Bioagricultural Sciences, Nagoya University, Chikusa, Nagoya, Japan
    Search for more papers by this author


Hydrogen (H2) is one of the most important intermediates in the anaerobic decomposition of organic matter. Although the microorganisms consuming H2 in anaerobic environments have been well documented, those producing H2 are not well known. In this study, we elucidated potential members of H2-producing bacteria in a paddy field soil using clone library analysis of [FeFe]-hydrogenase genes. The [FeFe]-hydrogenase is an enzyme involved in H2 metabolism, especially in H2 production. A suitable primer set was selected based on the preliminary clone library analysis performed using three primer sets designed for the [FeFe]-hydrogenase genes. Soil collected in flooded and drained periods was used to examine the dominant [FeFe]-hydrogenase genes in the paddy soil bacteria. In total, 115 and 108 clones were analyzed from the flooded and drained paddy field soils, respectively. Homology and phylogenetic analysis of the clones showed the presence of diverse [FeFe]-hydrogenase genes mainly related to Firmicutes, Deltaproteobacteria, and Chloroflexi. Predominance of Deltaproteobacteria and Chloroflexi suggests that the distinct bacterial community possessed [FeFe]-hydrogenase genes in the paddy field soil. Our study revealed the potential members of H2-producing bacteria in the paddy field soil based on their genetic diversity and the distinctiveness of the [FeFe]-hydrogenase genes.


Anaerobic decomposition of organic matters is accomplished through complex pathways with diverse anaerobes, which utilize organic and/or inorganic substances other than molecular oxygen as electron acceptors (Schink, 1997). Molecular hydrogen (H2) is one of the most important intermediates in the anaerobic decomposition processes. As the production and competitive consumption of H2 regulate the decomposition pathways, for example syntrophy with interspecies electron transfer between H2 producers and consumers (Schink, 1997; McInerney et al., 2008) and competition among diverse anaerobes such as iron reducers, sulfate reducers, and methanogens (Robinson & Tiedje, 1984; Conrad, 1999), elucidation of H2-producing and consuming processes and the related microorganisms are essential to understand anaerobic decomposition of organic matter.

Hydrogenases are the enzymes that catalyze the production and consumption of H2 (Vignais & Billoud, 2007). They are classified into three groups depending on the metal composition in their active site: [FeFe]-, [NiFe]-, and [Fe]- (formerly called ‘metal-free') hydrogenases (Vignais & Billoud, 2007). [FeFe]-hydrogenases have been found in anaerobic bacteria and eukaryotes and are known to catalyze H2-forming reaction in anaerobic environments, although some types of [FeFe]-hydrogenases (i.e. periplasmic hydrogenases in Desulfovibrio) are considered to catalyze oxidation of H2 (Meyer, 2007; Vignais & Billoud, 2007). Recently, three sets of primers have been designed based on the sequences of the gene encoding the H-cluster domain of [FeFe]-hydrogenase ([FeFe]-hydrogenase gene) and have been used to elucidate H2-producing bacterial communities present in anaerobic environments, for example hydF1/hydR1 in bioreactors (Xing et al., 2008), FeFe-272F/FeFe-427R in saline microbial mat (Boyd et al., 2009) and Yellowstone National Park (Boyd et al., 2010), and HydH1f/HydH3r in a moderately acidic fen (Schmidt et al., 2010) and earthworm gut (Schmidt et al., 2011). In addition, there have been studies on [FeFe]-hydrogenase genes in termite gut, which have used metagenomic and genomic information (Ballor & Leadbetter, 2012; Ballor et al., 2012). These analyses have provided novel findings to elucidate the diversity and functions of H2-producing bacterial communities in various environments.

Irrigated paddy fields, which are flooded during most periods of rice cultivation, are one of the well-investigated methanogenic environments for methane emission (Le Mer & Roger, 2001), soil reduction processes (Takai & Kamura, 1966), carbon flow (Kimura et al., 2004), and microbial community (Liesack et al., 2000; Asakawa & Kimura, 2008). In previous studies, microbial communities that are involved in electron-accepting processes such as sulfate reducers (Liu et al., 2009) and methanogenic archaea (Watanabe et al., 2009) have been studied. However, information on H2-producing bacterial communities in paddy field soil is lacking.

We aimed to evaluate the potential members of H2-producing bacteria in paddy field soil using molecular biology techniques with specific emphasis on studying the [FeFe]-hydrogenase genes. We examined the applicability of three available primer sets that amplify [FeFe]-hydrogenase genes. Next, we studied the diversity and characteristics of [FeFe]-hydrogenase genes in the microorganisms present in the paddy field soil by clone library analysis using the short-listed primer set to estimate potential H2-producing bacterial community.

Materials and methods

Soil samples

Soil samples were collected from the paddy field located at the Aichi-ken Anjo Research and Extension Center, central Japan (Anjo field; latitude 34°8′N, longitude 137°5′E), on April 14, 2011 and were used for examining of three primer sets. To evaluate the diversity of [FeFe]-hydrogenase genes in the paddy field soil, two soil samples were taken from the same Anjo field on April 11, 2003 under drained condition and on July 28, 2003 under flooded condition, and these were used as a representative soil sample for clone library analysis with the selected primer set. Bacterial (Kikuchi et al., 2007) and methanogenic archaeal (Watanabe et al., 2006) communities in the same field soil were investigated with DGGE analysis for 16S rRNA gene (16S rDNA). Chemical properties of the Anjo soil were as follows: total C, 12.6 g kg−1; total N, 1.1 g kg−1; pH [H2O], 5.8; free iron content, 11.0 g kg−1. The soil was classified as Oxyaquic Dystrudept (Soil Survey Staff, 1999) with light clay texture. The field has been managed with double cropping of rice or soybean and wheat as summer and winter cultivation, respectively. Details of the field managements in 2003 were described by Watanabe et al. (2006). About 1 kg of the composite soil samples was taken from three or four spots from the plow layer (0–10 cm) using a trowel and was transferred into a polyethylene bag. The soil samples were then passed through a 2-mm mesh sieve, mixed thoroughly, and stored at 4 °C until use. The samples were normally subjected to DNA extraction on the same or next day of sampling.

DNA extraction

DNA extraction from the soil samples collected in 2003 was performed four times according to the beads-beating method, and the DNA samples were purified with Sephadex G-200 as described in the previous studies (Cahyani et al., 2003; Watanabe et al., 2004). ISOIL for beads beating (Nippon Gene, Tokyo, Japan) was used for DNA extraction, performed three times, from the soil sample collected in 2011 according to the manufacturer's instruction. The DNA samples were appropriately diluted with TE (10 mM Tris-HCl, 1 mM EDTA, pH8.0) buffer for the subsequent PCR assays depending on the DNA concentration of the samples.

PCR amplification of [FeFe]-hydrogenase gene

Three degenerate primer sets, hydF1 (5′-GCCGACCTKACMATMATGGA-3′)/hydR1 (5′-ATRCARCCRCCSGGRCAGGCCAT-3′) (Xing et al., 2008), FeFe-272F (5′-GCHGAYMTBACHATWATGGARGA-3′)/FeFe-427R (5′-GCNGCYTCCATDACDCCDCCNGT-3′) (Boyd et al., 2009), and HydH1f (5′-TTIACITSITGYWSYCCIGSHTGG-3′)/HydH3r (5′-CAICCIYMIGGRCAISNCAT-3′) (Schmidt et al., 2010), which target the [FeFe]-hydrogenase gene, were examined. The PCR conditions and primer concentrations were modified slightly from the original program (Supporting Information, Table S1). All reaction mixtures contained 5 μL of 10× Ex Taq™ buffer (20 mM Mg2+ plus, TaKaRa, Otsu, Japan), 5 μL of dNTPs (2.5 mM each, TaKaRa), 0.25 μL of Ex Taq™ polymerase (5 U μL−1, TaKaRa), and 4 μL of template DNA, and the reaction volume was made up to 50 μL with sterilized ultrapure water. The amplicons were checked by agarose gel electrophoresis followed by ethidium bromide staining. The PCR amplifications were performed with triplicate or quadruple DNA samples. The amplified product from one replicate was used for the comparison of the three primer sets, and the amplicons of the mixed four replications were used for the subsequent analysis with the selected primer set.

Cloning and sequencing analysis

Each PCR product was purified using NucleoSpin Extract II (Macherey-Nagel, Düren, Germany) and cloned into pT7-Blue-T-Vector (Novagen, Darmstadt, Germany) with Ligation Solution I (Takara). Escherichia coli XL1-Blue competent cells (Toyobo, Osaka, Japan) were transformed with ligated vectors. Colonies carrying positive clones were confirmed by blue/white selection and by colony PCR with the same primer set used for [FeFe]-hydrogenase genes. Plasmid DNA was extracted using the alkaline extraction method or using the Zyppy Plasmid Miniprep Kit (Zymo Research, CA). Sequencing analysis was carried out using the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, CA) and was carried out on ABI PRISM 3130 Genetic Analyzer (Applied Biosystems) or outsourced to Hitachi Solutions (Yokohama, Japan) with Applied Biosystems 3730 DNA Analyzer (Applied Biosystems).

Nucleotide sequences obtained from the clones were translated to amino acid sequences using the European Molecular Biology Open Software Suite (emboss) Transeq program (Rice et al., 2000). Closest relatives of the in silico-translated [FeFe]-hydrogenase amino acid sequences were searched using the Basic Local Alignment Search Tool (blast) program (Altschul et al., 1990) on the National Center for Biotechnology Information web site. The sequences were aligned using the clustalw 1.83 (Chenna et al., 2003) on the DNA Data Bank of Japan (ddbj) web site ( with default parameters, and then, phylogenetic trees were constructed by the neighbor-joining method on the ddbj web site. Unweighted unifrac analysis (Lozupone & Knight, 2005) was performed using the mothur 1.23.0 (Schloss et al., 2009). Criterion of operational taxonomic units (OTUs) and Chao1 indices (Chao et al., 2009) was determined in the threshold of 80% sequence similarity by the mothur according to the previous study (Schmidt et al., 2011). Coverage was calculated according to this formula OTU/Chao1 × 100.

Accession numbers

The nucleotide sequences of [FeFe]-hydrogenase genes determined in the clone libraries have been submitted to the ddbj database under accession numbers AB760556AB760947.

Results and discussion

Comparison between the three primer sets for analysis of [FeFe]-hydrogenase genes

Detection ranges of the three different primer sets (hydF1/hydR1, FeFe-272F/FeFe-427R, and HydH1f/HydH3r) designed in the previous studies were compared. Either blurred, nonspecific or no band was observed when the original PCR conditions were applied to the paddy soil sample, expected lengths of PCR amplicons (c. 700, 450, and 600 bp for hydF1/hydR1, FeFe-272F/FeFe-427R, and HydH1/HydH3r, respectively) were successfully obtained after modification of the PCR conditions (data not shown). We analyzed 52, 54, and 61 clones obtained from the hydF1/hydR1, FeFe-272F/FeFe-427R, and HydH1f/HydH3r libraries. Three sequences (two in the hydF1/hydR1 library and one in the HydH1f/HydH3r library) did not seem to associate with the [FeFe]-hydrogenases and therefore were omitted for the subsequent analyses. Almost all clones showed a preserved L2 sequence motif (PCxxKxxE; Vignais & Billoud, 2007; Meyer, 2007), although four clones showed a few substituted amino acids (hydF1_1, 17, and 37 showed CCTAKKYE, and HydH1f_10 showed PSTAKKFE; the substituted amino acids are underlined). In this study, the criterion of OTU was set at a threshold of 80% amino acid sequence similarity (Schmidt et al., 2011). The numbers and values of OTUs/Chao1/coverage (%) were 32/74/43 (hydF1/hydR), 20/25/80 (FeFe-272F/FeFe-427R), and 23/36/64 (HydH1f/HydH3r). The unifrac analysis showed significant difference (< 0.0010) between the HydH1f/HydH3r library and the other two libraries.

Protein blast analysis of the deduced amino acid sequences of the clones showed similar detection ranges of [FeFe]-hydrogenase genes among the primer sets (Table S2–S4). Most of the clones in the same OTU aligned with the same taxonomic group, suggesting 80% similarity level is an acceptable range in this study. The clones which were closely related to the [FeFe]-hydrogenase genes in Firmicutes, Chloroflexi, Proteobacteria, Spirochetes, Bacteroidetes, and Caldithrix were amplified using the same three primer sets. A few clones were closely related to the [FeFe]-hydrogenase genes in Acidobacteria, Elusimicrobia, and Verrucomicrobia. In all the primer sets, the clones related to the [FeFe]-hydrogenase genes in Firmicutes, Chloroflexi, Proteobacteria, and Bacteroidetes accounted for more than half of the clones.

All the primer sets showed similar detection ranges and, in this sense, were applicable for analyzing H2-producing bacterial communities in paddy field soil. However, the unifrac analysis showed a significant difference for the primer set HydH1f/HydH3r from the rest. In addition, this primer set was designed to detect the widest range of [FeFe]-hydrogenase genes from the largest number of [FeFe]-hydrogenase genes (Schmidt et al., 2010). Thus, for further analysis, we decided to use the primer set HydH1f/HydH3r.

Genetic diversity and characteristics of [FeFe]-hydrogenase genes in paddy field soil

Clone library analysis of [FeFe]-hydrogenase genes in the Anjo paddy field soil showed that diverse microorganisms harboring the [FeFe]-hydrogenase genes could be retrieved from the soil. In this study, two soil samples collected in both flooded and drained periods were used as the representative soil conditions to evaluate the diversity and characteristics of [FeFe]-hydrogenase genes in the paddy field soil. In total, 115 and 108 clones were obtained from the soil samples collected under flooded and drained conditions, respectively. All obtained sequences showed similarity with the [FeFe]-hydrogenase sequence. The L2 sequence motif (see above) of each clone was found, and only two clones had a single amino acid substitution (e20411052 had CCTCKKAE, and e20411057 had PCMAMKFE; substituted amino acids are underlined). The indices of OTUs/Chao1/coverage (%) were 41/72/67 under the flooded condition, 34/64/53 under the drained condition, and 57/99/58 in total. The unifrac analysis showed no significant difference between these two libraries. The number of unique OTUs, which consisted of single clone, was 24 and 21. The numbers of total OTUs, unique OTUs, and Chao1 richness suggest that diverse [FeFe]-hydrogenase-producing bacteria exist in both flooded and drained paddy soils. The protein blast analysis (Tables S5 and S6) and phylogenetic tree (Fig. 1) showed most clones obtained in the present study were closely related to Firmicutes, Chloroflexi, and Proteobacteria (Table 1 and Fig. 1). Most clones affiliated with Firmicutes and Proteobacteria were closely related to Clostridia and Deltaproteobacteria (Fig. 1 and Table 1). All clones affiliated with Chloroflexi were closely related to Dehalococcoides (Fig. 1 and Table 1). Therefore, these bacteria are speculated to be possible H2 producers in the paddy field soil, although some [FeFe]-hydrogenases catalyze H2 consumption rather than H2 production in some occasions (see below), and these results should be interpreted with caution as already mentioned in the previous study (Schmidt et al., 2010) as the phylogeny of [FeFe]-hydrogenase gene is slightly different from that of 16S rRNA gene. The top hits of the blast search of the 93 and 94 clones obtained from the flooded and drained paddy soils were uncultured environmental clones (Table S5 and Table S6), indicating the presence of diverse unknown [FeFe]-hydrogenase genes in the paddy field soil.

Table 1. Top three representative species in each class of closest relatives in clone libraries obtained by PCR with HydH1f/HydH3r (Schmidt et al., 2010) and the number of clones affiliated with them
PhylumClassSpecies with strain nameAccession no.Identity, %(similarity, %) rangeNumber of clones
Flooded paddy soilDrained paddy soil
  1. a

    NCBI reference sequence accession number.

Firmicutes     2316
Clostridia Moorella thermoacetica ATCC 39073 ABC20019 45–74 (67–89)61
Pelotomaculum thermopropionicum SI BAF60191 53–69 (71–84)41
Acetivibrio cellulolyticus CD2 ZP_09466277 a 50–78 (70–89)22
Proteobacteria     5133
Alphaproteobacteria Rhodopseudomonas palustris BisA53 ABJ07787 68 (81)10
Deltaproteobacteria Desulfovibrio fructosovorans JJ EFL52165 56–73 (74–84)2813
Desulfovibrio magneticus RS-1 BAH74274 55–74 (69–86)105
Pelobacter carbinolicus DSM 2380 ABA88877 82–83 (91–93)34
Gammaproteobacteria Thiorhodococcus drewsii AZ1 EGV33414 70–71 (86)21
Chloroflexi     3149
Dehalococcoidetes Dehalococcoides sp. BAV1 ABQ16813 67–71 (81–86)1726
Dehalococcoides sp. VS ACZ61328 65–67 (83–85)56
Dehalococcoides sp. CBDB1 CAI82422 68–69 (84–85)36
Bacteroidetes Bacteroidia Odoribacter splanchnicus DSM 20712 ADY31293 70–73 (87–88)31
Anaerophaga thermohalophila DSM 12881 ZP_08845393 a 70–74 (86–88)22
Elusimicrobia Elusimicrobia Elusimicrobium minutum Pei191 ACC98088 62–68 (81–85)33
Lentisphaerae Victivallis vadensis ATCC BAA-548 EFA99820 72 (85)01
Spirochetes Spirochaetia Spirochaeta smaragdinae DSM 11293 ADK79621 65–67 (79–80)02
Spirochaeta thermophila DSM 6578 AEJ60920 39 (57)01
Verrucomicrobia Opitutae Opitutus terrae PB90-1 ACB74828 81–83 (89)20
Figure 1.

Neighbor-joining tree of [FeFe]-hydrogenase gene sequences obtained from two clone libraries (drained and flooded paddy field soil) and reference sequences. The primer set HydH1f/HydH3r (Schmidt et al., 2010) was used to obtain two libraries. GenBank accession numbers are indicated in parentheses. The number of resampling is 1000 for the bootstrap analysis, and the number of bootstrap values above 500 (closed circles) is shown. Sequences were grouped into 57 different OTUs based on an amino acid sequence with the threshold sequence similarity of 80% (Schmidt et al., 2011). Representative sequences selected by mothur (Schloss et al., 2009) are shown for each OTU. The bar indicates a 0.1 change per amino acid. The bar graph on the right side displays the number of clones included in each OTU. Phylogenetic assignment represented besides brackets are based on the topology of the tree and the results of blast search.

Phylogenetic affiliation of [FeFe]-hydrogenase genes

Most of [FeFe]-hydrogenase clones in Firmicutes were related to the orders of Clostridiales. The Clostridiales group consisted of clones related to various genera such as Clostridium, Desulfotomaculum, and Pelotomaculum, members having an ability of growing syntrophically with methanogens (McInerney et al., 2008).

All clones that aligned with Deltaproteobacteria were related to sulfate reducers except one clone that aligned with Syntrophus. These clones were affiliated with the families of Pelobacteriaceae and Desulfovibrionaceae which include bacteria that are able to grow syntrophically with H2-consuming bacteria (McInerney et al., 2008). Only one clone in our study was closely related to Syntrophus aciditrophicus in Syntrophobacteriales, which exhibits obligate syntrophic growth by interspecies H2 transfer (Jackson et al., 1999). Although the periplasmic [FeFe]-hydrogenase of Desulfovibrio catalyzes H2 consumption during the sulfate reduction process (Vignais & Billoud, 2007), some of those [FeFe]-hydrogenases have a function to produce H2 under syntrophic conditions (Meyer et al., 2013), suggesting that [FeFe]-hydrogenases of those sulfate reducers may be bifunctional depending on the growth conditions. Deltaproteobacteria was not a dominant group in the other environments tested (Boyd et al., 2009, 2010; Schmidt et al., 2010, 2011). As sulfate reducers are one of the key players in the carbon cycles in paddy field soil, they may have important functions for H2 metabolism in paddy field soil.

Clones belonging to Chloroflexi, Deltaproteobacteria, and Firmicutes were also retrieved from the DGGE analysis of bacterial 16S rRNA gene (Kikuchi et al., 2007). Therefore, bacteria belonging to these three groups may be one of the major groups of the bacterial community and play an important role for H2 production in the paddy field soil. In addition, Firmicutes and Deltaproteobacteria, especially Firmicutes, were frequently detected in nutrient-rich spots, such as rice straw (Weber et al., 2001), rice straw compost (Tanahashi et al., 2005), plant residues (Matsuyama et al., 2007; Rui et al., 2009), rhizospheric soil and roots (Shrestha et al., 2011). These findings suggest that Firmicutes and Deltaproteobacteria actively produce H2 in nutrient-rich sites in paddy field soil.

Dehaloccoides, which was the sole group belonging to Chloroflexi detected in the present study, is able to catalyze reductive dehalogenation using H2 as electron donors (He et al., 2003), while syntrophic growth with hydrogenotrophic methanogens was reported for the species in Chloroflexi isolated from paddy field soil (Yamada et al., 2007). No [FeFe]-hydrogenase gene related to Chloroflexi was detected in the other environments tested (Boyd et al., 2009, 2010; Schmidt et al., 2010, 2011) except in a reductive dechlorinating soil column (Marshall et al., 2012). Therefore, Chloroflexi may be unique among the H2-producing bacteria in the paddy field soil, and elucidation of the roles of the members in Chloroflexi in H2 production and consumption in paddy field soil needs further investigations.

Small numbers of clones affiliated with Verrucomicrobiae were also detected in the present study and DGGE analysis of bacterial 16S rRNA gene not only in the bulk soil (Kikuchi et al., 2007) but also in rice straw (Sugano et al., 2005), rice straw compost (Tanahashi et al., 2005), and plant residue (Rui et al., 2009). Opitutus terrae of Verrucomicrobia was abundant in paddy field soil (Chin et al., 1999) and was found to produce H2 (Chin et al., 2001), indicating that this group may also participate in H2 production in paddy fields.

We examined [FeFe]-hydrogenase genes in the paddy field soil using molecular biology techniques. Members of Firmicutes, Deltaproteobacteria, and Chloroflexi are presumably predominant H2 producers in the paddy field soil, although some [FeFe]-hydrogenases, especially deltaproteobacterial ones, may also be involved in H2 consumption. Clones of Firmicutes were also detected in other anaerobic environments, and Chloroflexi and Deltaproteobacteria seem to be distinctive groups in the paddy field soil. These results indicate that paddy field soil has a unique H2-producing bacterial community. Further studies such as analyzing the temporal change and spatial distribution of the community, isolating the H2-producing bacteria, and determining the roles of the [FeFe]-hydrogenases in sulfate reducers and Chloroflexi will be needed to elucidate the dynamics of the microbial community and its role in the anaerobic decomposition of organic matters in paddy field soil.


The present study was supported by the Grant-in-Aid for Young Scientists from the Japan Society for the Promotion of Science. We thank H. Honjo and N. Saka of the Anjo Research and Extension Station, Aichi-ken Agricultural Research Center, Japan, for their help in collecting the soil sample.