2023 Volume 38 Issue 4 Article ID: ME23068
Nitrification is a key process in the biogeochemical nitrogen cycle and a major emission source of the greenhouse gas nitrous oxide (N2O). The periplasmic enzyme hydroxylamine oxidoreductase (HAO) is involved in the oxidation of hydroxylamine to nitric oxide in the second step of nitrification, producing N2O as a byproduct. Its three-dimensional structure demonstrates that slight differences in HAO active site residues have inhibitor effects. Therefore, a more detailed understanding of the diversity of HAO active site residues in soil microorganisms is important for the development of novel nitrification inhibitors using structure-guided drug design. However, this has not yet been examined. In the present study, we investigated hao gene diversity in beta-proteobacterial ammonia-oxidizing bacteria (β-AOB) and complete ammonia-oxidizing (comammox; Nitrospira spp.) bacteria in agricultural fields using a clone library analysis. A total of 1,949 hao gene sequences revealed that hao gene diversity in β-AOB and comammox bacteria was affected by the fertilizer treatment and field type, respectively. Moreover, hao sequences showed the almost complete conservation of the six HAO active site residues in both β-AOB and comammox bacteria. The diversity of nitrifying bacteria showed similarity between hao and amoA genes. The nxrB amplicon sequence revealed the dominance of Nitrospira cluster II in tea field soils. The present study is the first to reveal hao gene diversity in agricultural soils, which will accelerate the efficient screening of HAO inhibitors and evaluations of their suppressive effects on nitrification in agricultural soils.
Nitrification is a key process in the biogeochemical nitrogen cycle, wherein ammonia is aerobically oxidized to nitrate. In agricultural soils, this process is driven by four different microbial groups: ammonia-oxidizing archaea (AOA), ammonia-oxidizing bacteria (AOB), nitrite-oxidizing bacteria (NOB), and complete ammonia-oxidizing (comammox) bacteria (Hayatsu et al., 2021). These nitrifying microorganisms are closely associated with the application of nitrogen fertilizers because the chemoautotrophic bacteria proliferate by obtaining energy from the oxidization of ammonia and/or nitrite (Norton and Ouyang, 2019). Nitrification activity increases in response to nitrogen fertilization amendments, and often leads to environmental pollution via nitrate leaching as well as nitrogen loss from agricultural fields. Nitrogen use efficiency in crops is less than 50% of nitrogen fertilizer input due to nitrogen loss, including nitrification (Zhang et al., 2021).
Nitrification and denitrification in the nitrogen cycle are major emission pathways for the greenhouse gas N2O, and these pathways are highly dependent on environmental factors, such as the substrate type and its availability, soil oxygen concentrations, and water contents in soil (Khalil et al., 2004). Previous studies reported the existence of agricultural soils with high N2O emission derived from nitrification or denitrification pathways. For example, N2O emission via the nitrification pathway is dominant in soils with a water-filled pore space (WFPS) less than 60% (Bollmann and Conrad, 1998; Bateman and Baggs, 2005; Baggs et al., 2010; Huang et al., 2014). It has also been suggested to occur in lower pH soils (Liu et al., 2016). On the other hand, elevations in denitrification activity and denitrification-derived N2O emissions have been reported when soil moisture increases due to rainfall and O2 concentrations decrease (Butterbach-Bahl et al., 2013). High organic matter also increases N2O emissions derived from denitrification in soils (Thomson et al., 2012).
AOB generate N2O through two metabolic pathways, hydroxylamine oxidization and nitrifier denitrification (Ward et al., 2011). AOB play major roles in the nitrification of agricultural soils treated with inorganic nitrogen fertilizers and are responsible for N2O emissions (Prosser et al., 2020). In contrast, the N2O emission potential of comammox bacteria is considered to be very low based on in vitro pure culture experiments (Kits et al., 2019). Similar findings were confirmed in soil microcosm experiments (Tan et al., 2022). However, limited information is currently available on the N2O emission potential of comammox bacteria. Therefore, further studies are needed to demonstrate the contribution of N2O emissions by comammox bacteria.
Several nitrification control methods, such as polymer-coated fertilizers (PCF) and nitrification inhibitors (NIs), have been proposed to reduce N2O emissions. PCF reduce nitrification activity by controlling the release of ammonium into the soil via a physical barrier and consequently reduce N2O emissions. NIs decelerate nitrification activity for nitrogen from fertilizers applied to soil by temporarily inhibiting ammonia oxidation. NIs are considered to be the most effective N2O mitigation method, and three commercially available NIs are now widely used: 2-chloro-6-(trichloromethyl) pyridine (Nitrapyrin), dicyandiamide (DCD), and 3,4-dimethylpyrazole phosphate (DMPP). However, the reduction of N2O emissions by these NIs was previously reported to range between 31% and 44% (Akiyama et al., 2010) and 46% and 53% (Fan et al., 2022). Therefore, the development of more effective NIs is urgently desired.
In nitrification, ammonia oxidation occurs in multiple steps catalyzed by ammonia monooxygenase (AMO) and hydroxylamine oxidoreductase (HAO). AMO catalyzes the oxidation of NH3 to NH2OH, while HAO catalyzes the oxidation of NH2OH to NO. The enzyme responsible for the oxidation of NO to NO2– has not yet been identified (Ward et al., 2011; Caranto and Lancaster, 2017). The majority of commercially available NIs target AMO. However, difficulties are associated with the purification and crystallization of AMO, and the three-dimensional protein structure of this enzyme is unknown (Hayatsu et al., 2021). In contrast, HAO is a periplasmic enzyme and its steric structure has been elucidated (Igarashi et al., 1997; Cedervall et al., 2013; Maalcke et al., 2014; Nishigaya et al., 2019); therefore, the development of novel and more effective inhibitors is possible using structure-guided drug design.
A previous study revealed the three-dimensional protein structure of HAO in the gamma-proteobacterial AOB (γ-AOB) Nitrosococcus oceani, possessing six active site residues, similar to HAO in the beta-proteobacterial AOB (β-AOB) Nitrosomonas europaea (Nishigaya et al., 2019). However, since two of the six HAO active site residues are different between β-AOB and γ-AOB, the effects of the HAO inhibitor phenylhydrazine differ: 40% inhibition against β-AOB HAO and almost no inhibition against γ-AOB HAO (Nishigaya et al., 2016, 2019). The three-dimensional protein structure of comammox-HAO currently remains unclear. A genome analysis revealed that comammox bacteria have the β-AOB type of HAO (Palomo et al., 2018). Therefore, a more detailed understanding of the diversity of HAO active site residues in soil microorganisms involved in nitrification is important for the development of HAO inhibitors. However, hao gene diversity in agricultural soil has not yet been investigated.
To reveal hao gene diversity in agricultural soil, the abundance and community structure of AOB and comammox bacteria possessing HAO need to be examined. The abundance and diversity of AOB and AOA are generally investigated using the 16S rRNA gene and amoA encoding the alpha-subunit of ammonia monooxygenase as a phylogenetic marker gene (Purkhold et al., 2000; Hatzenpichler, 2012; Monteiro et al., 2014; Aigle et al., 2019). The 16S rRNA gene and nxrB encoding the beta-subunit of nitrite oxidoreductase are typically used for NOB and comammox bacteria (Pester et al., 2014).
In the present study, we designed PCR primer sets for hao genes containing HAO active sites and then investigated hao gene diversity in soils from agricultural fields in which different crops were cultivated under distinct nitrogen fertilization regimes. The hao gene sequences obtained in the present study provide insights into the diversity of HAO active sites in β-AOB and comammox bacteria. Furthermore, the community structure and abundance of AOB and comammox bacteria in soil samples were analyzed by amplicon sequencing and quantitative PCR (qPCR) using the amoA gene for AOB and comammox bacteria, and the nxrB gene for canonical-NOB and comammox bacteria. We also investigated the relationship between the AOB and comammox bacterial community structure and HAO diversity in soils under different nitrogen fertilization regimes.
Three agricultural fields (cabbage, soybean, and tea) were selected for this study. The cabbage field was located at the Institute of Fruit Tree and Tea Science, National Agriculture and Food Research Organization (NARO), Tsukuba, Ibaraki, Japan (36°02′N, 140°11′E). The soybean field was located at the Institute for Agro-Environmental Sciences, NARO Tsukuba, Ibaraki, Japan (36°01′N, 140°07′E). The tea field was located at Kagoshima Prefectural Institute For Agricultural Development, Minamisatsuma, Kagoshima, Japan (31°48′N, 130°34′E). The soil in each agricultural field was treated with chemical or organic nitrogen fertilizer. Details on fertilizer application are shown in Table S1. The soybean field to which the organic fertilizer was applied had been under non-tillage management for thirty years. The soil type in all three agricultural fields was Andosol. Surface soil at a depth of 0–1 cm was removed. Using a small sterile shovel, soil samples were collected at a depth of 1–10 cm in triplicate and sieved through a mesh with a pore size of 2 mm. All soil samples were stored at –80°C until DNA extraction or at 4°C for an ammonia oxidation potential analysis within 7 days. Air-dried soil was used to analyze soil properties.
Characterization of soil propertiesSoil properties were analyzed as previously described (Tago et al., 2015). Briefly, soil pH was assessed in 1:2.5 (w/v) soil/water suspensions. Total carbon and nitrogen contents were measured using an elemental analyzer (2400II CHNS/O, PerkinElmer). Available phosphate was extracted from soil suspensions to 2 mM sulfuric acid (w/v) for 30 min and the concentration was calorimetrically determined using the molybdenum blue method. To measure the contents of ammonium nitrogen and nitrate nitrogen, a 10-g soil sample was extracted with 100 mL of 2 M KCl, and the soil suspension was filtered through Whatman paper No. 10 filter paper. Ammonium nitrogen was measured by a continuous flow analyzer (TRRACS, Bran+Luebbe) using the indophenol blue method. Nitrate nitrogen was analyzed using the copper cadmium reduction method.
Ammonia oxidation potential analysisAmmonia oxidation potential was measured based on the nitrite accumulation rate in the presence of an inhibitor of nitrite oxidization, potassium chlorate (Belser and Mays, 1980). Briefly, 2.5 g soil was suspended in 10 mL of 1 mM phosphate buffer (pH 7.0) containing 1 mM (NH4)2SO4 and 10 mM KClO3 and incubated at 25°C with continuous rotation at 150 rpm for several hours. A 1-mL soil suspension incubated for several hours was centrifuged at 10,000×g at 4°C for 10 min. The concentration of accumulated NO2– in the supernatant was spectrophotometrically assessed using the Griess-Ilosvay method. Nitrification rates were calculated based on the accumulation of nitrite (nmole h–1 [g dry soil]–1) (Keeney and Nelson, 1982).
DNA extraction from soil samplesSoil DNA was extracted from 0.4 g soil with the Fast DNA SPIN Kit for Soil (Qbiogene) as described in previous studies (Takada-Hoshino and Matsumoto, 2004; Morimoto et al., 2011). The extracted DNA sample was purified by MicroSpin S-400 HR columns (GE Healthcare) and the DNA Clean and Concentrator-25 kit (Zymo Research).
Quantitative PCR (qPCR)Extracted DNA was subjected to qPCR using SYBR Premix Ex Taq (Takara Bio) and 200 nM of primers. The primers and PCR temperature profile used for AOA-amoA and AOB-amoA genes were described in previous studies (Rotthauwe et al., 1997; Nicolaisen and Ramsing, 2002; Tourna et al., 2008; Morimoto et al., 2011; Yang et al., 2017; Table S2). qPCR of the comammox-amoA and nxrB genes was conducted using two primer sets: comaA-amoA-F (5′–CBKCNTGGTGGTGGTGGTC–3′) and comaA-amoA-R (5′–AGCCCATRTAGTCNGCCC–3′), and nxrB-F (5′–GTGGTGGAACAAYGTSGARAC–3′) and nxrB-R (5′–GCATCGABGTNGSVGTRTC–3′), respectively. The PCR temperature profile was set to 95°C for 2 min, followed by 40 cycles of 95°C for 30 s, 65°C in comammox-amoA and 60°C in nxrB for 30 s, and 72°C for 1 min using the StepOnePlus Real-Time PCR System (Applied Biosystems). The copy numbers of the three target genes were calculated based on a standard curve generated using a dilution series of linearized pGEM-T Easy plasmids (Promega) containing clones of each PCR fragment.
Clone library analysis of β-AOB and comammox hao genesThe PCR primers covering the β-AOB and comammox HAO enzyme active sites were designed as follows: Hao-pira-F (5′–TGCCAYACCAACCAGAACAA–3′) and hao-pira-R (5′–ATCTTGGTGTTYTCGTCCATG–3′), and hao-AS-F2 (5′–TGCCAYRYCMABCARAAYAAG–3′) and hao-AS-R2 (5′–TCMTCRTCCATGATTTCSACA–3′), respectively (Fig. S1). The 770-bp and 758-bp fragments of the β-AOB and comammox hao genes, respectively, were amplified using soil DNA samples, the aforementioned primer sets (Table S2), and TaKaRa Ex Taq (Takara Bio). The PCR temperature profile was set to 94°C for 2 min, followed by 30 cycles at 94°C for 30 s, 50°C in β-AOB-hao or 55°C in comammox-hao for 30 s, and 72°C for 1 min using the Applied Biosystems Veriti thermal cycler (Applied Biosystems). PCR products were purified using a QIAquick PCR Purification Kit (Qiagen), ligated into pGEM-T Easy Vector Systems (Promega), and cloned into Escherichia coli JM109 competent cells (Takara Bio). Cloned PCR products were sequenced at the Takara Bio Biomedical Center (Takara Bio).
Amplicon sequencingEighteen soil DNA samples (6 soil samples in triplicate) were subjected to PCR amplification of the AOB-amoA, comammox-amoA, and nxrB genes for an amplicon sequence analysis. PCR amplification was performed using TaKaRa Ex Taq (Takara Bio) and fusion primers containing adaptor sequences, keys, multiplex identifiers, and gene-specific sequences (Table S2). The PCR temperature profile was set to initial denaturation at 94°C for 2 min, followed by 25 cycles at 94°C for 30 s, the annealing temperature of each gene (Table S2) for 30 s, and a final extension at 72°C for 1 min using the Applied Biosystems Veriti thermal cycler (Applied Biosystems). PCR products were purified using the QIAquick PCR purification kit (Qiagen), and gel extraction was performed with the QIAquick Gel Extraction kit (Qiagen). The quality and quantity of the purified PCR amplicons were analyzed using an Agilent 2100 Bioanalyzer (Agilent Technologies) and the Quant-iT PicoGreen dsDNA Assay kit (Thermo Fisher Scientific), respectively. The amplicon sequence was analyzed by the MiSeq sequencer using Miseq Reagent kit v2 (Illumina), according to the manufacturer’s instructions.
Sequence analysis and phylogenetic assignmentThe sequences of the β-AOB and comammox hao genes from the clone library analysis were assigned to operational taxonomical units (OTU) based on a 95% sequence identity threshold and >1% in all sequences using the cd-hit-est program on the server (Huang et al., 2010). The most similar bacterial species in OTUs were identified by a BLASTX search of the NCBI database. The representative nucleotide sequences of each OTU were used for a phylogenetic analysis. Amplicon sequence quality was examined using the DADA2 package (Callahan et al., 2016) in R software (R Core Team, 2021). After the removal of low quality and chimeric sequences, non-target sequences were identified by a BLAST search against a target gene at the FunGene database and removed using Mothur (Schloss et al., 2009). Filtered sequences were then clustered using Mothur into OTUs with cut-off values of 0.05 for AOB-amoA, 0.06 for comammox-amoA, and 0.04 for nxrB genes. The representative nucleotide sequences in each OTU were aligned with the MAFFT program on the EMBL-EBI server (Li et al., 2015). Phylogenetic trees were generated for the amoA, hao, and nxrB genes using the maximum likelihood (ML) method with the removal of gap-including and ambiguous sites and with a bootstrap analysis (1,000 replicates) in MEGA software version 10.1.8 (Kumar et al., 2018; Stecher et al., 2020). We selected the Tamura-Nei model of nucleotide substitutions with gamma distributed and invariant sites (G+I) (Tamura and Nei, 1993).
Confirmation of HAO enzyme active sitesThe 1,112 β-AOB-hao sequences derived from the clone library analysis and Nitrosospira multiformis ATCC25196 [CP000103] as a reference gene were aligned using the MAFFT program and translated into amino acid sequences using MEGA version 10.1.8. The six residues at the HAO active sites in the aligned sequences were manually confirmed using Jalview software ver. 2.11.2.4 (Waterhouse et al., 2009). A total of 837 comammox-hao sequences and Nitrospira inopinata ENR4 [LN885086] were aligned, and their HAO active site residues were verified as described above.
Statistical analysisCorrelations between qPCR data and soil parameters were estimated by calculating Spearman’s ρ values using R software.
Nucleotide sequence accession numbersThe nucleotide sequences of the hao genes obtained in the present study were deposited in the DDBJ/Genbank/EBI databases under the accession numbers LC724068–LC726016. The amplicon sequence reads reported in this study have been deposited in the DDBJ Sequence Read Archive (DRA) under the accession number: DRA014735 (See also Table S1).
Tea field soils had lower pH and higher total carbon, total nitrogen, and available phosphate than cabbage and soybean field soils (Table 1). The type of nitrogen fertilizer affected several soil properties. Organic fertilizer treatment increased NO3-N concentrations in the three agricultural fields. Total carbon and total nitrogen contents were higher in cabbage and soybean field soils treated with organic fertilizer. The content of NH4-N was higher in soybean and tea field soils treated with organic fertilizer (Table 1). In all fields, ammonia oxidation potential (AOP) levels were two- to six-fold higher with the organic fertilizer treatment than with the chemical fertilizer treatment (Table 1).
Summary of soil properties and ammonia oxidation potential in this study
| Field type | Fertilizer | pH | total C (%) |
total N (%) |
available P (μg [g dry soil]–1) |
NH4-N (μg N [g dry soil]–1) |
NO3-N (μg N [g dry soil]–1) |
AOPa
(nmole h–1 g–1) |
water content (%) |
|---|---|---|---|---|---|---|---|---|---|
| Cabbage field | Chemical | 5.23±0.10 | 3.11±0.10 | 0.32±0.02 | 51.81±1.47 | 13.84±2.58 | 54.78±3.07 | 18.17±1.69 | 11.77±0.58 |
| Organic | 6.95±0.02 | 5.16±0.34 | 0.63±0.03 | 47.96±2.98 | 12.92±1.62 | 206.89±28.08 | 32.88±4.81 | 12.00±0.11 | |
| Soybean field | Chemical | 6.27±0.06 | 5.00±0.14 | 0.38±0.01 | 17.28±2.04 | 56.55±9.08 | 36.32±1.35 | 24.36±3.90 | 28.34±0.42 |
| Organic | 6.19±0.20 | 17.79±0.69 | 1.10±0.04 | 18.29±3.57 | 86.98±34.28 | 343.23±49.82 | 142.24±7.96 | 44.00±1.62 | |
| Tea field | Chemical | 3.35±0.10 | 31.53±1.76 | 2.43±0.14 | 201.40±30.41 | 37.66±19.90b | 255.27±34.95 | 13.81±2.86 | 62.18±1.67 |
| Organic | 3.83±0.14 | 18.80±3.43 | 1.55±0.32 | 514.42±113.28 | 124.62±86.84b | 501.45±112.57 | 60.03±15.02 | 53.54±3.50 |
a AOP: Ammonia oxidation potential
b The values for the mean±SD are of two replicates because one replicate was not detectable.
The microbial population involved in nitrification in each agricultural field was investigated by qPCR. The abundances of AOA-amoA and AOB-amoA were similar, ranging from 9.7×106 to 3.9×107 and from 3.7×105 to 5.4×106 gene copies (g dry soil)–1, respectively, except for the tea field treated with chemical fertilizer, which showed the lowest abundances of 1.7×104 and 1.8×105 gene copies, respectively (Table 2). On the other hand, comammox-amoA in all three fields ranged from 1.5×107 to 1.3×108 gene copies (g dry soil)–1, i.e., markedly higher than AOA-amoA and AOB-amoA, except for chemical fertilizer-treated cabbage soil (9.1×105 gene copies). All three field soils had similar values ranging from 3.9×106 to 2.7×107 nxrB gene copies (g dry soil)–1 involved in nitrite oxidation (Table 2).
Abundances of amoA genes of AOA, AOB, and comammox and the nxrB gene of NOB in agricultural field soils
| Sample Name (gene copy [g dry soil]–1) | AOA-amoA | AOB-amoA | comammox-amoA | nxrB | |
|---|---|---|---|---|---|
| Cabbage field | Chemical | 1.4×107±1.1×106 | 3.3×106±5.5×105 | 9.1×105±2.6×105 | 5.6×106±1.1×106 |
| Organic | 1.2×107±6.9×106 | 3.7×105±1.6×105 | 2.7×107±1.3×107 | 1.6×107±5.0×106 | |
| Soybean field | Chemical | 9.7×106±1.8×106 | 4.6×106±3.3×105 | 1.5×107±2.9×106 | 8.2×106±1.5×106 |
| Organic | 2.1×107±6.6×106 | 4.9×106±4.4×105 | 5.9×107±6.7×106 | 2.7×107±3.4×105 | |
| Tea field | Chemical | 1.7×104±1.1×103 | 1.8×105±8.9×104 | 5.3×107±6.1×104 | 3.9×106±1.7×106 |
| Organic | 3.9×107±4.1×106 | 5.4×106±3.8×106 | 1.3×108±8.0×105 | 8.8×106±6.2×106 | |
We examined the relationships between eight parameters of soil properties and the abundance of four microbial gene copies in qPCR data (Table S3). AOP positively correlated with the abundances of AOB-amoA and nxrB, but not with the abundances of AOA-amoA or comammox-amoA (Table S3). In other soil parameters, the abundance of AOA-amoA positively correlated with pH, but negatively correlated with total nitrogen, available phosphate, and water content. The abundance of comammox-amoA positively correlated with total carbon, total nitrogen, nitrate-nitrogen, and water content (Table S3). The abundance of nxrB positively correlated with pH (Table S3).
The hao gene diversity of β-AOB and comammox bacteriaTo investigate the hao gene diversity of β-AOB and comammox bacteria in soil samples collected from cabbage, soybean, and tea fields, we designed two primer sets against the β-AOB and comammox hao genes, respectively, including six amino acid residues of the HAO enzyme active site (Fig. S1), and then conducted a clone library analysis of six soil samples. We obtained 1,949 hao sequences, derived from 1,112 and 837 clones in β-AOB and comammox, respectively (Tables S4 and S5). The 1,949 assembled sequences were assigned to 30 OTUs based on the 95% sequence identity threshold and >1% detection of all sequences. Sixteen β-AOB-hao OTUs were identified as Nitrosospira spp., whereas 14 comammox-hao OTUs were members of Nitrospira in the NCBI database using a BLASTX search.
We performed a molecular phylogenetic analysis of the 16 β-AOB-hao OTUs and other reference sequences in the β-AOB-hao gene. The 16 OTUs were classified into six subgroups in the genus Nitrosospira (Fig. 1). Since cluster 3b was more diverse than the other clusters, it was divided into three subgroups: cluster 3b -N. tenuis-, cluster 3b -N. briensis-, and cluster 3b -Nitrosospira sp. Nv4-. The relative abundance of the AOB-hao gene in the six soil samples at the AOB subgroup level is shown in Fig. 2A. It revealed that the β-AOB-hao gene diversity was altered more markedly by fertilizer type than field type. Nitrosospira cluster 3b -N. tenuis- was dominant in all three chemical fertilizer-treated soils, but was present at a low level in organic fertilizer-treated cabbage and tea field soils (Fig. 2A). In the soybean field treated with organic fertilizer, Nitrosospira cluster 3a -Nitrosospira sp. nsp2- accounted for more than 70% of the population, and was also detected in the cabbage field treated with organic fertilizer. Nitrosospira sp. 56-18 cluster accounted for 25% of the population present in the tea field treated with organic fertilizer (Fig. 2A). Furthermore, 25% of the population present in cabbage and tea fields treated with organic fertilizer was Nitrosospira cluster 3a -N. multiformis- (Fig. 2A). Nitrosospira cluster 3b -N. briensis- and cluster 0 accounted for less than 10% in all soil samples (Fig. 2A). These results suggest that β-AOB-hao gene diversity in agricultural fields was markedly affected by the fertilizer treatment.
Molecular phylogenetic analysis of β-AOB-hao gene diversity in agricultural field soils. A maximum-likelihood tree was generated based on 719 aligned nucleotide sites of the hao gene. Maximum-likelihood bootstrap values (%) were calculated with 1,000 replicates, and bootstrap values >50 are shown at the tree nodes. We used reference sequences reported in a previous study (Aigle et al., 2019). Accession numbers in the DNA database (DDBJ/EMBL/GenBank or JGI) are shown in square brackets. Representative β-AOB-hao OTUs are shown in blue with a bold case font. Symbols indicate >5% relative abundance in the six soil samples; open circle: cabbage field soil treated with chemical fertilizer; closed circle: cabbage field soil treated with organic fertilizer; open triangle: soybean field soil treated with chemical fertilizer; closed triangle: soybean field soil treated with organic fertilizer; open square: tea field soil treated with chemical fertilizer; closed square: tea field soil treated with organic fertilizer; dash: none of any field. Symbol colors indicate each field: blue, cabbage field; green, soybean field; red, tea field.
Relative abundance of hao gene diversity in agricultural fields in (A) β-AOB and (B) comammox bacteria. Subclades were identified based on each phylogenetic tree (Fig. 1 and 3). The total number of clones in each soil is shown in the graphs, and details are provided in Table S4 and Table S5.
A molecular phylogenetic analysis of the comammox-hao gene was performed based on 14 comammox-hao OTUs and other reference sequences in the comammox-hao gene. Fourteen OTUs were classified into three clusters in the genus Nitrospira (Fig. 3). Ten OTUs (OTU5–OTU14) belonged to comammox clade A1, two OTUs (OTU1 and OTU2) to comammox clade B, and two OTUs (OTU3 and OTU4) to comammox clade A2 (Fig. 3). The relative abundances of these comammox subgroups in agricultural fields are shown in Fig. 2B. In the four cabbage and soybean field soil samples, clade B accounted for 34%–79%, with small variations among soil samples (Fig. 2B). However, clade B was not detected in the tea field, in which clade A1 was dominant and accounted for more than 90% (Fig. 2B). Clade A2 was a minor population in all the field soil samples (Fig. 2B). Collectively, these results suggest that comammox-hao gene diversity was not significantly affected by the fertilizer treatment.
Molecular phylogenetic analysis of comammox-hao gene diversity in agricultural field soils. A maximum-likelihood tree was generated based on 713 aligned nucleotide sites of the hao gene. Maximum-likelihood bootstrap values (%) were calculated with 1,000 replicates, and bootstrap values >50 are shown at the tree nodes. Accession numbers in the DNA database (DDBJ/EMBL/GenBank) are shown in square brackets. We referred to the nucleotide sequence information of comammox bacteria and its criteria reported in a previous study (Palomo et al., 2022). Representative comammox hao OTUs are shown in blue with a bold case font. Symbols and colors are as shown in Fig. 1.
Based on a protein crystal structure analysis, the HAO enzyme in β-AOB possesses six residues at its active site (Nishigaya et al., 2016; Fig. 4A). Although the protein crystal structure in comammox HAO has not yet been identified, its protein sequence identity strongly suggests that it possesses six active site residues, similar to β-AOB HAO (Fig. 4A). To confirm the diversity of the HAO enzyme active site in β-AOB and comammox bacteria in agricultural field soils, we confirmed the 1,949 hao sequences identified using a clone library analysis. We found that 99.0% (1,101/1,112) of β-AOB-hao sequences corresponded to the six residues of the typical HAO active site in N. europaea (Fig. 4B). However, the remaining 1.0% (11/1,112) showed 1–2 residue substitutions at their HAO active sites (Fig. S3A). A diversity analysis of the comammox-hao gene demonstrated that 99.3% (831/837) of the comammox-hao sequences had the same active site residues as those of the reference strain Candidatus Nitrospira inopinata (Fig. 4C), and only 0.7% (6/837) showed 1 residue substitution at the HAO active sites (Fig. S3B). Collectively, these results demonstrated that the six residues of the substrate-binding active sites in HAO were highly conserved in β-AOB and comammox bacteria, whereas HAO active site residues differed between β-AOB and γ-AOB.
Confirmation of the hydroxylamine oxidoreductase (HAO) enzyme active site. (A) Multiple sequence alignment of HAO in (A) β-AOB, comammox, and γ-AOB, in (B) 17 β-AOB-hao OTUs, and in (C) 14 comammox-hao OTUs detected in agricultural field soils. Representatives of each OTU are shown. OTUs with different active site residues are summarized in Fig. S3. Colors indicate HAO active site residues conserved between β-AOB and comammox bacteria (red), with γ-AOB (blue), and among all three types of bacteria (yellow). Arrowheads indicate HAO active site residues. Accession numbers in the DNA database (DDBJ/EMBL/GenBank) are shown in square brackets.
Phylogenetic analyses of AOB have been performed using the marker gene, amoA (Purkhold et al., 2000). Therefore, an amoA amplicon sequence analysis was conducted to compare bacterial diversity between amoA and hao and confirm the diversity of nitrifying bacteria. High-quality AOB-amoA gene sequence reads were clustered into 16 OTUs based on a 95% sequence identity threshold and more than 1% coverage of the total reads (Table S6). Each soil sample contained 5–13 OTUs. Sixteen AOB-amoA OTUs were identified as members of Nitrosospira using a BLASTX search of the NCBI database. The nMDS plot showed that the AOB community structure was significantly affected by the field type and fertilizer treatment (Fig. S4). The AOB community in tea field soils markedly different from those in cabbage and soybean field soils, whereas the AOB community in cabbage and soybean fields was similar between chemical and organic fertilizer-treated soils (Fig. S4). We then performed a molecular phylogenetic analysis of 16 AOB-amoA OTUs and other reference sequences in β-AOB-amoA. Within the genus Nitrosospira, 16 AOB-amoA OTUs were classified into eight subgroups (Fig. 5). The relative abundances of these Nitrosospira subgroups in the agricultural fields are shown in Fig. 6. The majority of AOB in cabbage, soybean, and tea fields treated with chemical fertilizer consisted of cluster 3b -N. tenuis-, whereas Nitrosospira cluster 3a -Nitrosospira sp. nsp2- was dominant in the soybean field treated with organic fertilizer. These results were consistent with the pattern revealed by the hao clone library analysis. In cabbage and tea fields treated with organic fertilizer, no single cluster occupied more than 50% of the population, while the total of cluster 3a -N. multiformis- and cluster 3a -Nitrosospira sp. nsp2- was more than 50% in cabbage field soil treated with organic fertilizer, and the total of cluster 2 and cluster 3a -N. multiformis- was dominant in tea field soil treated with organic fertilizer (Fig. 6).
Molecular phylogenetic analysis of AOB-amoA gene diversity in agricultural field soils. A maximum-likelihood tree was generated based on 452 aligned nucleotide sites of the amoA gene. Maximum-likelihood bootstrap values (%) were calculated with 1,000 replicates, and bootstrap values >50 are shown at the tree nodes. Accession numbers in the DNA database (DDBJ/EMBL/GenBank or JGI) are shown in square brackets. Representative AOB-amoA OTUs are shown in blue with a bold case font. Symbols and these colors are as shown in Fig. 1.
The relative abundance of AOB-amoA gene diversity in agricultural field soils. Subclades were identified based on phylogenetic divergence (Fig. 5). Detailed sequence reads detected in each cluster are shown in Table S6.
We also evaluated the diversity of comammox bacteria in soil samples using an amoA amplicon analysis. High-quality comammox-amoA sequences clustered into 13 OTUs, and each soil sample contained 2–7 OTUs (Table S7). All of these bacteria were identified as members of Nitrospira spp. using a BLASTX search of the NCBI database. The nMDS analysis demonstrated that comammox-amoA gene diversity was affected by the field soil rather than the fertilizer treatment (Fig. S5). A phylogenetic analysis revealed that all 13 comammox-amoA OTUs belonged to comammox clade A; however, the node did not include well-known comammox strains including N. inopinata (Fig. S6). OTU2, OTU3, and OTU4 were dominant in cabbage field soil treated with chemical or organic fertilizers and soybean field soil treated with chemical fertilizer with minor variations in the OTU composition (Fig. S7). OTU6 occupied one fourth of the comammox population in cabbage field soil treated with organic fertilizer. The OTUs in soybean field soil treated with organic fertilizer were distinct from those in soybean field soil treated with chemical fertilizer. In the latter case, OTU8, OTU10, OTU11, and OTU12 were frequently detected (Fig. S7). On the other hand, OTU1 and OTU7 were dominant in tea field soil treated with chemical fertilizer. OTU1 accounted for more than 90%. OTU5 and OTU9 were detected in tea field soil treated with organic fertilizer (Fig. S7).
Amplicon sequence analysis of nxrB gene diversity in NOBThe genus Nitrospira was originally known as NOB and divided into six clusters. A part of Nitrospira cluster II is known as comammox bacteria (Daims et al., 2015; van Kessel et al., 2015). To confirm Nitrospira diversity, including canonical NOB and comammox bacteria, in soil samples, an amplicon sequence analysis of the nxrB gene was conducted. The sequence reads of the six soil samples revealed 19 nxrB OTUs (Table S8). The Nitrospira community structure markedly differed between the tea field and cabbage and soybean fields (Fig. S8). In cabbage and soybean fields, the Nitrospira community structure of soybean field soil treated with organic fertilizer differed from that of the other three soils (Fig. S8A). The relative abundances of the 19 nxrB OTUs in each soil sample revealed that the tea field had a smaller number of OTUs than cabbage and soybean fields (Fig. S9). Four OTUs (OTU2, OTU8, OTU10, and OTU13) comprised nxrB in the tea field. Of these, OTU2, OTU10, and OTU13 were specifically detected in the tea field. OTU2 accounted for more than 80% of the Nitrospira population in the tea field soil treated with chemical fertilizer (Fig. S9). In contrast, the nxrB diversity was large in cabbage and soybean fields, detected in more than 10 OTUs per sample. Despite variations among samples, the majority (>50%) comprised five OTUs (OTU1, OTU3, OTU4, OTU5, and OTU6) in cabbage and soybean field soils (Fig. S9). The relative abundances of Nitrospira clusters in agricultural fields are shown in Fig. 7, which showed that Nitrospira cluster II was exclusive to tea field soils. On the other hand, clusters I and V, in addition to cluster II, were confirmed in cabbage and soybean field soils (Fig. 7).
The relative abundance of nxrB gene diversity in agricultural field soils. Subclades were identified based on phylogenetic divergence (Fig. S10). Detailed sequence reads are shown in Table S8.
The present study is the first to reveal hao gene diversity in β-AOB and comammox bacteria inhabiting various agricultural field soils. In addition, the hao gene sequences obtained herein showed the almost complete conservation of the six residues at the substrate-binding active site of the HAO enzyme in β-AOB and comammox bacteria (Fig. 4). The genus Nitrospira, originally known as NOB, is involved in nitrite oxidization. Nitrospira has been classified into six lineages, in which comammox bacteria were found to be a part of the cluster II (Daims et al., 2015; van Kessel et al., 2015). The genes involved in ammonia oxidization in comammox bacteria are assumed to have been horizontally transferred from β-AOB (Palomo et al., 2018), which presumably affected the conservation of amino acid residues at the HAO active sites in β-AOB and comammox bacteria. HAO active site residues differ between β-AOB and γ-AOB (Fig. 4A), resulting in different HAO inhibitor effects, with phenylhydrazine inducing a 40% reduction in HAO activity in β-AOB, but having no effect on that in γ-AOB (Nishigaya et al., 2016). Nevertheless, the present study revealed the conservation of HAO enzyme active site residues across bacterial taxa in agricultural field soils, suggesting that an HAO inhibitor, such as phenylhydrazine, is capable of suppressing HAO activity in broad nitrifying bacteria inhabiting agricultural fields, including β-AOB and comammox bacteria.
The AOB community structure is affected by various environmental factors, such as soil pH, ammonia affinity, urease activity, and salt tolerance (Koops and Pommerening-Röser, 2001; Tago et al., 2015; Li et al., 2018; Aigle et al., 2019; Hayatsu et al., 2021). Organic fertilizer-treated soils contained higher comammox-amoA and nxrB copy numbers than chemical fertilizer-treated soils (Table 2). Some Nitrospira hydrolyze urea and/or harbor genes that regulate the assimilation of simple organic substrates, such as pyruvate, formate, and acetate (Lucker et al., 2010; Koch et al., 2015), which could be advantageous in an abundance of Nitrospira in the organic fertilizer-treated soils.
The comammox-amoA gene copy number was markedly higher than the nxrB gene copy number in all soil samples, except for those from the chemical fertilizer-treated cabbage field (Table 2), which may be affected by multiple copies of the amoCAB gene cluster in comammox genomes (Camejo et al., 2017) and/or differences in PCR amplification efficiency between the two primer sets. The abundances of AOA-amoA and AOB-amoA were markedly less in chemical fertilizer-treated tea field soil than in all other soils. The abundances of AOB and AOA were previously shown to be markedly affected by soil pH (Li et al., 2018; Aigle et al., 2019). The decreases observed in their abundances in the present study may have been due to the extremely acidic soil (pH: 3.35) of the chemical fertilizer-treated tea field. However, organic fertilizer-treated tea field soil (pH: 3.83) had similar AOA-amoA and AOB-amoA gene copy numbers to those of AOA-amoA and AOB-amoA in the other soil samples (Table 2). The relationship between AOA and AOB abundances and nitrogen fertilizer treatment is currently unclear. However, the abundance of AOA may have increased due to organic nitrogen mineralization (Levičnik-Höfferle et al., 2012), and AOB became more abundant at higher soil pH and with the release of ammonia during the mineralization of organic fertilizer (Dai et al., 2021), which could affect increases in AOA and AOB abundances in organic fertilizer-treated tea field soil.
Since AOA do not have canonical hao genes, nitrification inhibitors against HAO do not regulate AOA growth or N2O emissions from AOA. Moreover, the abundance of AOA-amoA ranged from 9.7×106 to 3.9×107 gene copies (g dry soil)–1, except for the tea field treated with chemical fertilizer (Table 2). However, AOB, not AOA, are responsible for N2O emissions in agricultural soils (Prosser et al., 2020), which indicates that nitrification inhibitors against HAO reduce major N2O emissions from agricultural soils.
The hao gene diversity of β-AOB was similar in three different fields treated with chemical fertilizer, in which Nitrosospira cluster 3b -N. tenuis- was dominant. Nitrosospira cluster 3b has high ammonia tolerance (Webster et al., 2005). The bacterial growth of N. tenuis is slower than that of other AOB, with an optimal temperature and pH of 25–30°C and pH 7.7–7.8, respectively. N. tenuis was originally named Nitrosovibrio tenuis because its cell morphology is curved rods (Harms et al., 1976). However, the physiological and genomic features of Nitrosospira cluster 3b -N. tenuis- remain largely unknown because there are few isolates in the cluster, except for the type strain N. tenuis Nv1. Further studies are needed to reveal the mechanisms by which Nitrosospira cluster 3b -N. tenuis- adapts to chemical fertilizer-treated soils. On the other hand, cluster 3a -Nitrosospira sp. nsp2- had higher relative abundance in the organic fertilizer-treated soybean field soil, but not in the cabbage and tea fields treated with organic fertilizer. Soil tillage practices affect the community structure of AOB (Haiming et al., 2022). The organic fertilizer-treated soybean field has been maintained under no-tillage management, which might contribute to the unique AOB community that formed in this soil.
An analysis of comammox-hao gene diversity revealed that comammox clade B was absent in tea field soils, but dominant in cabbage and soybean field soils (Fig. 2B). Comammox clade A1 was dominant in tea field soils (Fig. 2B), which was consistent with previous findings (Takahashi et al., 2020). Since the genomes of comammox clades A and B show high dissimilarity (Palomo et al., 2018), the absence of comammox clade B in tea field soils indicates ecological niche differentiation. Numerous PCR primer sets have been designed for comammox amoA (Pjevac et al., 2017; Fowler et al., 2018; Xia et al., 2018; Zhao et al., 2019; Jiang et al., 2020; Lin et al., 2020). However, there is no primer set that amplifies a wide range of amoA in the two clades. Therefore, the primer set designed in the present study may not have adequately amplified comammox clade B in the soils tested. The nxrB gene amplicon sequence analysis revealed that Nitrospira clusters I and V were present in cabbage and soybean field soils, but were absent in tea field soils (Fig. 7). Furthermore, within Nitrospira cluster II, the dominant OTUs in tea field soils markedly differed from those in cabbage and soybean field soils, suggesting that the dominant NOB in tea field soils adapted to the extremely low pH. However, the mechanisms underlying this adaptation warrant further investigation.
The present study revealed similar predominant bacterial clusters in analyses of hao and amoA gene diversities, suggesting that the hao gene is also applicable to a phylogenetic analysis of ammonia-oxidizing microorganisms. However, fewer hao gene sequences have been deposited in public databases than amoA gene sequences. For example, no hao gene sequences were available for Nitrosospira cluster 2 or 4. AOB diversity has so far been examined by amoA and 16S rRNA genes (Purkhold et al., 2000). To fill the gap in data available for amoA and hao genes, further studies are needed to 1) identify the hao gene by the genome sequencing of bacterial isolates and 2) examine hao gene diversity in various environments.
Ohbayashi, T., Wang, Y., Aoyagi, L. N., Hara, S., Tago, K., and Hayatsu, M. (2023) Diversity of the Hydroxylamine Oxidoreductase (HAO) Gene and Its Enzyme Active Site in Agricultural Field Soils. Microbes Environ 38: ME23068.
https://doi.org/10.1264/jsme2.ME23068
We thank M. Udagawa for technical assistance, N. Nakamura and S. Tokuda for soil sampling, and Y. Hirono, T. Yamazaki, and Y. Nishigaya for technical advice. This study was supported by MEXT KAKENHI (19H01156 to KT and MH) and by the Moonshot project JPNP18016, commissioned by the New Energy and Industrial Technology Development Organization (NEDO).
