Genomic and Physiological Characteristics of a Novel Nitrite-Oxidizing Nitrospira Strain Isolated From a Drinking Water Treatment Plant

Nitrite-oxidizing bacteria (NOB) catalyze the second step of nitrification, which is an important process of the biogeochemical nitrogen cycle and is exploited extensively as a biological nitrogen removal process. Members of the genus Nitrospira are often identified as the dominant NOB in a diverse range of natural and artificial environments. Additionally, a number of studies examining the distribution, abundance, and characterization of complete ammonia oxidation (comammox) Nitrospira support the ecological importance of the genus Nitrospira. However, niche differentiation between nitrite-oxidizing Nitrospira and comammox Nitrospira remains unknown due to a lack of pure cultures. In this study, we report the isolation, physiology, and genome of a novel nitrite-oxidizing Nitrospira strain isolated from a fixed-bed column at a drinking water treatment plant. Continuous feeding of ammonia led to the enrichment of Nitrospira-like cells, as well as members of ammonia-oxidizing genus Nitrosomonas. Subsequently, a microcolony sorting technique was used to isolate a novel nitrite-oxidizing Nitrospira strain. Sequences of strains showing the growth of microcolonies in microtiter plates were checked. Consequently, the most abundant operational taxonomic unit (OTU) exhibited high sequence similarity with Nitrospira japonica (98%) at the 16S rRNA gene level. The two other Nitrospira OTUs shared over 99% sequence similarities with N. japonica and Nitrospira sp. strain GC86. Only one strain identified as Nitrospira was successfully subcultivated and designated as Nitrospira sp. strain KM1 with high sequence similarity with N. japonica (98%). The half saturation constant for nitrite and the maximum nitrite oxidation rate of strain KM1 were orders of magnitude lower than the published data of other known Nitrospira strains; moreover, strain KM1 was more sensitive to free ammonia compared with previously isolated Nitrospira strains. Therefore, the new Nitrospira strain appears to be better adapted to oligotrophic environments compared with other known non-marine nitrite oxidizers. The complete genome of strain KM1 was 4,509,223 bp in length and contained 4,318 predicted coding sequences. Average nucleotide identities between strain KM1 and known cultured Nitrospira genome sequences are 76.7–78.4%, suggesting at least species-level novelty of the strain in the Nitrospira lineage II. These findings broaden knowledge of the ecophysiological diversity of nitrite-oxidizing Nitrospira.

Nitrite-oxidizing bacteria (NOB) catalyze the second step of nitrification, which is an important process of the biogeochemical nitrogen cycle and is exploited extensively as a biological nitrogen removal process. Members of the genus Nitrospira are often identified as the dominant NOB in a diverse range of natural and artificial environments. Additionally, a number of studies examining the distribution, abundance, and characterization of complete ammonia oxidation (comammox) Nitrospira support the ecological importance of the genus Nitrospira. However, niche differentiation between nitrite-oxidizing Nitrospira and comammox Nitrospira remains unknown due to a lack of pure cultures. In this study, we report the isolation, physiology, and genome of a novel nitrite-oxidizing Nitrospira strain isolated from a fixed-bed column at a drinking water treatment plant. Continuous feeding of ammonia led to the enrichment of Nitrospira-like cells, as well as members of ammonia-oxidizing genus Nitrosomonas. Subsequently, a microcolony sorting technique was used to isolate a novel nitriteoxidizing Nitrospira strain. Sequences of strains showing the growth of microcolonies in microtiter plates were checked. Consequently, the most abundant operational taxonomic unit (OTU) exhibited high sequence similarity with Nitrospira japonica (98%) at the 16S rRNA gene level. The two other Nitrospira OTUs shared over 99% sequence similarities with N. japonica and Nitrospira sp. strain GC86. Only one strain identified as Nitrospira was successfully subcultivated and designated as Nitrospira sp. strain KM1 with high sequence similarity with N. japonica (98%). The half saturation constant for nitrite and the maximum nitrite oxidation rate of strain KM1 were orders of magnitude lower than the published data of other known Nitrospira strains; moreover, strain KM1 was more sensitive to free ammonia compared with previously isolated Nitrospira strains. Therefore, the new Nitrospira strain appears to be better adapted to oligotrophic environments compared with other known non-marine nitrite oxidizers. The complete

INTRODUCTION
Nitrification is a key process in the biogeochemical nitrogen cycle. Conventionally, this reaction takes place in two-steps: the first step is ammonia oxidation, which is carried out by ammoniaoxidizing archaea (AOA) and ammonia-oxidizing bacteria (AOB), and the second step is nitrite oxidation, which is carried out by the nitrite-oxidizing bacteria (NOB). Recently, it has been reported that a group of Nitrospira species that are conventionally recognized as NOB, carry out complete ammonia oxidation (comammox), transforming ammonia to nitrate (Daims et al., 2015;van Kessel et al., 2015). This surprising finding suggests the ecological importance of Nitrospira and indicates that our understanding about nitrification is still insufficient.
In systems producing drinking water, nitrification plays a key role in decreasing monochloramine and producing nitrate in DWTPs (Wang et al., 2014). Culture-independent approaches revealed that nitrite-oxidizing Nitrospira were more abundant than Nitrobacter in DWTPs (Regan et al., 2003;Tatari et al., 2017). As nitrite concentration is suggested as an important environmental factor differentiates niches among NOB (Daims et al., 2016) and between lineages I and II Nitrospira (Maixner et al., 2006;Fujitani et al., 2013), an increase of nitrite concentration selected for Nitrotoga rather than Nitrospira on biofilm systems fed with tap water (Kinnunen et al., 2017). To examine niche differentiation within NOB, it is important to characterize pure strains kinetically (Nowka et al., 2015a;Ushiki et al., 2017;Kitzinger et al., 2018). In rapid sand filters flowing into DWTPs in Denmark, Nitrospira were found to be present in abundance as a core taxon based on metagenomic approaches Palomo et al., 2016). Further analysis demonstrated that comammox Nitrospira were more abundant than AOB based on amoA gene abundance  and had metabolic versatility allowing nichepartitioning among ammonia oxidizers . Consequently, ammonia oxidation by comammox Nitrospira was functionally demonstrated by DNA and RNA stable isotope probing (Gülay et al., 2019). Although these culture-independent approaches revealed the occurrence of Nitrospira in oligotrophic environments, little is known about the physiology and genome of a Nitrospira strain from DWTPs. Recently, Nitrospira inopinata, the only comammox pure culture, was isolated from microbial biofilm on the surface of a pipe and characterized kinetically (Kits et al., 2017). Surprisingly, this strain seemed to be adapted to oligotrophic environments with a high affinity for ammonia but low affinity for nitrite. This finding raised additional questions about niche differentiation among NOB and between comammox Nitrospira and nitrite-oxidizing Nitrospira.
To answer these questions, isolation and characterization of novel Nitrospira species from oligotrophic environments is essential. Here, we report a novel nitrite-oxidizing Nitrospira strain that was isolated from a fixed-bed column present at a DWTP. Continuous feeding of ammonia to imitate the natural environment successfully led to the enrichment of diverse nitrifiers, including Nitrospira species belonging to lineage II and AOB. Subsequently, nitrite-oxidizing Nitrospira belonging to lineage II was purified based on a microcolony sorting technique. Finally, the morphology, physiology, kinetics, and genome of a novel Nitrospira strain were characterized.

Sample Source and Continuous Feeding Bioreactor
Zeolite as a water purification material was set up at the drinking water treatment plant. Zeolite has an ability to adsorb ammonium, creating a suitable habitat for nitrifiers. Zeolite was sampled as seed biomass to cultivate the nitrifiers in May 2010. To enrich nitrifiers, a continuous feeding bioreactor to supply substrate in low concentration was set up (Supplementary Figure S1). The volume of the bioreactor used was 1.1 L. Nonwoven fabric materials were used as a biomass carrier. The inorganic medium was comprised of NH 4 Cl (0.0038-0.304 g L −1 ; 0.071-5.683 mM), NaCl (0.116 g L −1 ; 1.985 mM), MgSO 4 ·7H 2 O (0.040 g L −1 ; 0.332 mM), CaCl 2 ·2H 2 O (0.073 g L −1 ; 0.497 mM), KCl (0.038 g L −1 ; 0.510 mM), KH 2 PO 4 (0.034 g L −1 ; 0.250 mM), and trace elements (1 mL). The composition of trace elements was the same as described in our previous research study (Abe et al., 2017). An aerobic culture condition was maintained at a pH of around 6.9 and incubated in the dark to avoid light inhibition. In the drinking water treatment plant, the influent ammonia concentration was below 1 mg-N L −1 (0.071 mM). Therefore, the influent ammonia concentration was also maintained at 1 mg-N L −1 in the primary culture. Flow rate was 4.4 L day −1 . The hydraulic retention time was adjusted to around 6 h. Loading was also carried out at a rate of 4.4 mg-N L −1 day −1 . During a 0-100 days period from the cultivation start, zeolites and non-woven fabric materials were placed in the continuous feeding bioreactor. On days 100-200, only zeolites were removed. After 200 days, influent ammonia concentration was increased gradually to reach up to 70 mg-N L −1 (5 mM) on day 350. During this period, ammonia concentration in the bioreactor was maintained below 1 mg-N L −1 (0.071 mM).

Chemical Analysis of Nitrogen Compounds During Enrichment Process
During the enrichment process, chemical properties were assessed. Ammonia concentration was measured by performing the colorimetric analysis with indophenol (Kandeler and Gerber, 1988). Additionally, nitrite and nitrate concentrations were measured according to the protocol described in a previous study (Fujitani et al., 2014).

Sorting of Nitrifiers' Microcolonies, Pure Cultivation, and Purity Check
Nitrifiers' microcolonies from the enrichment samples were separated using the cell sorter as described in previous research studies (Ushiki et al., 2013;Fujitani et al., 2014Fujitani et al., , 2015Abe et al., 2017). The sorted microcolonies were incubated for 1-2 months in the 96 well microtiter plates containing inorganic medium as described previously. However, it is important to note that the medium contained both, 5 mg-N L −1 ammonia (NH 4 Cl; 0.36 mM) and 5 mg-N L −1 nitrite (NaNO 2 ; 0.36 mM), and the pH was adjusted to 7.5. Incubation was performed at 23 • C under dark and static conditions. Cell growth and identification were confirmed by fluorescence in situ hybridization (FISH) and 16S rRNA gene sequence analyses. Pure strains identified as Nitrospira were transferred into test tubes and Erlenmeyer flasks, and were supplemented with 10 mg-N L −1 of nitrite (0.71 mM) in the medium every month for further characterization. Likewise, pure strains identified as Nitrosomonas were sub-cultivated with 10 mg-N L −1 of ammonia (0.71 mM) in the medium. Purity checks were carried out by microscopic observations based on FISH analysis and by using the heterotrophic media. The strain KM1 was transferred to 200-fold diluted Luria-Bertani (BD, NJ, United States), Nutrient Broth (BD) and R2A [polypeptone 500 mg L −1 , casamino acid 500 g L −1 , soluble starch 500 mg L −1 , yeast extract 500 mg L −1 , sodium pyruvate 300 mg L −1 , KH 2 PO 4 500 mg L −1 (3.67 mM), MgSO 4 ·7H 2 O 50.0 mg L −1 (0.415 mM)] media in the form of agar plates and liquid cultures.

Fluorescence in situ Hybridization (FISH)
Fluorescence in situ hybridization was performed according to a previously described standard protocol (Amann et al., 1990). Oligonucleotide probes were fluorescently labeled with the hydrophilic sulfoindocyanine dye (Cy3) (Fasmac Co. Ltd., Atsugi, Japan) and are listed (Supplementary Table S1). Observations were carried out according to a previous study (Fujitani et al., 2014).

DNA Extraction, PCR, and Cloning
DNA was extracted from the enrichment samples according to the standard protocol of the ISOIL extraction kit (Nippon Gene, Tokyo, Japan). Fragments of 16S rRNA and amoA genes were amplified using the total DNA with the listed primer sets (Supplementary Table S1). The PCR mixtures and purification of PCR products were conducted based on previously described protocols . Purified PCR products were cloned according to the protocol provided in the Qiagen PCR cloning plus kit (QIAGEN). To identify 16S rRNA gene sequences of pure strains cultivated in 96 well microtiter plates, bacterial DNA was extracted by heating at 95 • C. The 16S rRNA gene was amplified with the 27f/1492r primer set. The sequences of about 750 bases were obtained by using the primer 27f. The cloning of the comammox Nitrospira amoA genes was performed according to the protocol described in a previous study .

Phylogenetic Analysis
The regions of sequences with low quality were trimmed, and the contaminated sequences were removed. The cloned sequences of 16S rRNA genes were grouped into operational taxonomic units (OTU) with a similarity threshold of 98.7%. The OTUs were classified using the SILVA SINA algorithm (Pruesse et al., 2012). In addition, the closest strain to each OTU was searched using the NCBI Web BLASTn program. The 16S rRNA gene sequences of AOB, NOB, and comammox Nitrospira were aligned using the ClustalW tool of the MEGAX software (Kumar et al., 2018). Maximum-likelihood trees were constructed in MEGAX using the Tamura-Nei model. Bootstrap values were calculated by performing the iteration 1,000 times. The phylogenetic analysis of comammox Nitrospira amoA gene was conducted according to a previously described protocol .

Electron Microscopy
Morphology of a Nitrospira pure strain was observed by scanning electron microscopy (SEM) and transmission electron microscopy (TEM), according to a previous study (Fujitani et al., 2014).

Physiological Activity Test to Determine Optimum Temperature and Generation Time
Mean nitrite oxidation rates (µM day −1 ) were calculated during a 3 days incubation period at 4, 13, 22, 26, 30, 34, and 38 • C using the exponentially growing pure strain. We used a fresh medium containing less than 1 mM nitrite in the pre-incubation step. After the cells were incubated with the medium for a few days, we checked the cells were in the exponential growth phase. The oxidation rates were often measured to evaluate the optimum temperature of NOB pure cultures (Nowka et al., 2015b;Kitzinger et al., 2018). The cell number and the amount of oxidized nitrite showed a good positive correlation (Nowka et al., 2015a;Ushiki et al., 2017;Ishii et al., 2020). The culture experiment for determining the generation time (see below) showed that the isolated cells grew exponentially 2-4 days after nitrite addition. To investigate the generation time, the isolate was incubated at 28 • C after replacing the supernatant in order to remove the nitrate and by-products by centrifugation. Nitrite and nitrate concentrations were measured using the ion chromatography with a TSKgel SuperIC-Anion HS (IC-2010; Tosoh, Tokyo, Japan) according to the protocol described in our previous studies Ishii et al., 2020). Cell numbers were estimated by performing qPCR analysis of the nxrB gene as described in a previous study . The nxrB gene was amplified by using a specific primer set (Supplementary Table S1). The purified nxrB products at known concentrations were diluted to 10-10 8 copies µL −1 to generate a standard curve. The PCR efficiency was 103% with an R 2 value of 0.99. The thermal cycler program used was as follows: an initial denaturing step at 98 • C for 2 min, followed by 40 cycles of denaturation at 98 • C for 10 s, annealing at 52 • C for 10 s, and extension at 68 • C for 40 s. The estimated nxrB gene copy number was divided by two, because the Nitrospira strain possesses two copies of the nxrB gene. The average generation time was determined based on the change in the number of cells in the period between the confirmation of cell growth and the depletion of nitrite. The minimum generation time was also calculated using two time points from the log phase.

Kinetic Parameters for Nitrite Oxidation
As described in previous studies (Ishii et al., 2017;Ushiki et al., 2017), the kinetic parameters were calculated from multiple nitrite oxidation rates using exponentially growing cells. The cells were harvested by centrifugation, and resuspended using fresh mineral nitrite medium. The cell suspensions in glass containers were incubated at room temperature with shaking until the available nitrite content was depleted. Every 5-30 min, aliquots of 50 µL of the samples were collected and heated at 95 • C for 5 min to inactivate the Nitrospira cells. Nitrite concentrations were measured by a colorimetric method using the Griess reagent (Hewitt and Nicholas, 1964). The experimental data were fitted to the following Michaelis-Menten kinetic equation: Here, K m is the apparent half-saturation constant for nitrite (µM NO 2 − ), V max is the maximum nitrite oxidation rate (µmol NO 2 − /mg protein/h), V is the nitrite oxidation rate, and [S] is the nitrite concentration (µM NO 2 − ). For protein measurements, the cells were lysed using 0.15 M NaOH solution at 90 • C for 30 min. Protein concentrations were measured by using the BCA protein assay kit (Takara).

Inhibition Effect With Free Ammonia
The inhibitory constant for free ammonia was determined upon amending the medium with 0-3.6 mM NH 4 Cl. The estimated half-saturation constant (K' m ) values in the presence of free ammonia were fitted to the following inhibitory constant equation: K' m = K m (1 + [I]/K i ). Here, K i is the inhibitory constant for free ammonia (µM NO 2 − ), and [I] is the free ammonia concentration (mg-NH 3 L −1 and µM NH 3 ), which was calculated by assessing the NH 4 Cl concentration, temperature, and pH (Anthonisen et al., 1976). The other materials and methods were in accordance with the procedures that are described above.

Genome Reconstruction and Annotation
DNA was extracted from the purified strain KM1 according to the standard protocol of the NucleoSpin Tissue DNA extraction kit (Takara). DNA sequencing, assembly, and manual curation of the genomes of different strains were performed as previously described (Sekiguchi et al., 2015;Ushiki et al., 2018). Coding sequences (CDS) were automatically predicted and annotated by using the DFAST ver. 1.2.3 (Tanizawa et al., 2018) and a published Nitrospira genome sequence (Koch et al., 2015). Average nucleotide identities (ANI) were calculated using fastANI with the default settings (Jain et al., 2018). Orthologous genes were identified using the OrthoVenn 2 web service tool (Xu et al., 2019) with an E-value of 1e-05. Toxin-antitoxin modules were searched using the TAfinder with E-value of 1e-06 for BLAST and E-value of 1e-03 for HMMer (Xie et al., 2018).

Enrichment of Nitrifiers Using a Continuous Feeding Bioreactor
Nitrifiers from a drinking water treatment plant were cultivated using zeolites as primary samples from a continuous feeding bioreactor with non-woven fabric materials to maintain the biomass (Supplementary Figure S1). Influent ammonia, effluent ammonia, nitrite, and nitrate concentrations were monitored during the enrichment process ( Figure 1A). To increase the biomass in the non-woven fabric materials, zeolites and non-woven fabric materials were set up for an initial 200 days incubation. During this period, ammonia and nitrite concentrations in the bioreactor were maintained at nearly 0 mg-N L −1 . After 200 days, the influent ammonia concentration was gradually increased from 1 to 70 mg-N L −1 , in order to enhance the growth activities of the nitrifiers. Even during this period, ammonia and nitrite concentrations were maintained at nearly 0 mg-N L −1 . Although the ammonia concentration in the bioreactor temporarily increased around the 300 days, the nitrate concentration increased with an increase in the influent ammonia concentration. Therefore, complete nitrification reaction occurred in the bioreactor throughout the enrichment process. FISH observations revealed the growth pattern of nitrifiers in the enrichment samples. Although almost no nitrifiers were identified due to low influent ammonia concentration during the first 200 days, the cells of Nitrosomonas and Nitrospira species were observed on day 250. Finally, the ratios of species belonging to the Nitrosomonas and Nitrospira lineage II on day 350 were around 30 and 40%, respectively, based on the cell counts. The species of Nitrosomonas and Nitrospira that were observed in this study formed spherical and dense microcolonies ( Figure 1B and Supplementary Figure S2A). During this enrichment process, AOA and species belonging to Nitrospira lineage I were not detected.
Subsequently, 16S rRNA gene sequences in the enrichment culture were cloned on day 350. In total, 87 clones were obtained, out of which 5 and 36 clones were identified to belong to the genus Nitrosomonas and Nitrospira, respectively (Figure 2, Supplementary Table S2, and Supplementary Figure S2B). The most abundant OTU, OTU1 (35/87 clones) exhibited high sequence similarity with the Nitrospira sp. strain GC86 (99.9%). OTU2 (1/87 clones) exhibited high sequence similarity with Nitrospira japonica (98.0%). Five clones that were affiliated to the genus Nitrosomonas were grouped into three OTUs. Additionally, a total of 48 clones were subjected to sequencing using the primer set for Nitrospira amoA gene that was designed in a previous study  and grouped as one OTU (Supplementary Figure S3 and Supplementary Table S2).

Sorting of Single Microcolonies, Pure Cultivation, and Identification
Enrichment samples were allowed to pass through the cell sorter for sorting single microcolonies according to the method described in the previous studies (Ushiki et al., 2013;Fujitani et al., 2014Fujitani et al., , 2015Abe et al., 2017). By sorting an area with a large forward scatter (FSC) value and a small side scatter (SSC) value, we could successfully sort the nitrifiers' microcolonies (Supplementary Figure S4). Each microcolony was inoculated and cultured in 96 well microtiter plates containing both, ammonia and nitrite inorganic medium. After 1-2 months of incubation period, the cells in each well were subjected to sequencing. In total, 13 isolates that were cultured with the medium containing both ammonia and nitrite belonged to the genus Nitrospira. The most abundant OTU, OTU34 (11/13 isolates) exhibited high sequence similarity with N. japonica (98%) at the 16S rRNA gene level. OTU35 and OTU36 exhibited over 99% sequence identity with N. japonica and Nitrospira sp. strain GC86, respectively (Figure 2 and Supplementary  Table S2). Additionally, 10 isolates that were cultured with the medium containing both ammonia and nitrite belonged to the genus Nitrosomonas (Supplementary Table S2 and Supplementary Figure S2B). Strains that were identified as either Nitrosomonas or Nitrospira were transferred from the 96 well microtiter plates to test tubes and Erlenmeyer flasks. At that time, the Nitrosomonas strains were cultivated using inorganic medium containing only ammonia and the Nitrospira strains were cultivated using inorganic medium containing only nitrite. Although some strains might be comammox Nitrospira, we identified the organisms with 16S rRNA gene only and could not discriminate comammox Nitrospira from nitrite-oxidizing Nitrospira. After that, we could not sub-cultivate all the 10 strains that were identified as Nitrosomonas in the ammonia-containing medium, and only one strain (OTU34) that was identified as Nitrospira grew successfully in the nitrite medium. This successfully cultivated nitrite-oxidizing strain was designated as Nitrospira sp. strain KM1. Most of the clones cultivated from the enrichment culture (OTU1, 35 clones) were closely related to Ca. N. nitrificans and Ca. N. nitrosa, whereas the identified strains (OTU34, 11 clones) were distant from the lineage including Ca. N. nitrificans and Ca. N. nitrosa (Figure 2). After that, we carefully confirmed the purity of the strain KM1 using heterotrophic media, PCR, FISH.

Physiological Characterization of Strain KM1
Using scanning electron microscopy, we confirmed that dozens of cells aggregated densely (Supplementary Figure S5A) and that each cell formed a spiral shape (Supplementary Figure S5B). The strain KM1 produced low amounts of extracellular polymeric substances (EPS) (Supplementary Figure S5C) and formed weak flocs in a manner similar to that observed for Nitrospira lenta, which also belongs to lineage FIGURE 2 | Phylogenetic tree of the genus Nitrospira based on the 16S rRNA gene sequence. The tree was constructed using the maximum likelihood (ML) algorithm. Bootstrap values at the branch nodes were iterated 1,000 times. Known isolates and enriched cultures are indicated in bold. For the Nitrospira sequences obtained in this study, enrichment clones and isolates are presented in red and blue, respectively. The number of OTUs are specified in brackets. Each lineage belonging to the genus Nitrospira is indicated in green. The scale bar corresponds to 2% of the estimated sequence divergence. Accession numbers are shown toward the right side of the organism's names/descriptions. (Nowka et al., 2015b). Transmission electron microscopy revealed that the strain KM1 formed a cell cluster embedded in the EPS (Supplementary Figure S5D).

II of Nitrospira
The strain KM1 showed nitrite oxidation activity over a temperature range from 4 to 38 • C and the optimum temperature was identified as 30 • C ( Figure 3A). This trend was nearly similar those observed for other nitrite-oxidizing Nitrospira species (Spieck and Lipski, 2011;Ushiki et al., 2013;Fujitani et al., 2014;Nowka et al., 2015b). To evaluate the generation time and growth yield, the strain KM1 was incubated at 28 • C in mineral nitrite medium. It oxidized 0.73 mM of nitrite to nitrate in 117 h. The minimum generation time of the strain KM1 was estimated as 34 h, while the average was estimated as 65 h (Figure 3B and Supplementary Figure S6). Growth yield of the strain KM1 was calculated based on the number of cells and nitrite consumption. The growth yield was estimated as 9.38 log cell mmol NO 2 − . Therefore, the growth rate of the strain KM1 was found to be very similar to that of other Nitrospira pure cultures ( Table 1).
Nitrite oxidation by strain KM1 followed Michalis-Menten kinetics ( Figure 4A and Supplementary Table S3), with a mean apparent half-saturation constant K m(app) = 0.81 ± 0.19 µM of nitrite and without ammonium. The mean of maximum oxidation rate of nitrite (V max ) was 2.46 ± 0.33 µM of nitrite (mg of protein h −1 ). The values of K m(app) and V max were compared among different nitrite-oxidizing Nitrospira species (Table 1). Surprisingly, the measured K m(app) (NO 2 − ) and V max values of strain KM1 were orders of magnitude lower than the published values for other known Nitrospira strains. Note that most other K m(app) values for NOB were determined by measuring nitrite-dependent oxygen consumption. Although Nitrospira inopinata is not strictly a nitrite oxidizer, the genus Nitrospira in oligotrophic environments indicated a wide range of affinity for nitrite. Additionally, the values of K m(app) and V max in strain KM1 were compared with other NOB (Supplementary Table S4). Strain KM1 had the lowest value among cultured NOB from non-marine habitats. According to a recent study, in situ kinetic values for nitrite oxidation in the dark ocean were at nanomolar levels and orders of magnitude lower than the K m(app) values of cultured NOB (Zhang et al., 2020). This would suggest uncultured NOB are well-adapted to oligotrophic environments in nature. NOB in marine samples might have a higher affinity for nitrite than NOB in non-marine habitat. As there is still a gap between in situ kinetic values and kinetic values of the cultured representatives, we should continue efforts to characterize uncultured NOB.
Meanwhile, the value of K m(app) was shown to increase in the presence of NH 4 Cl, which means that in this condition, the affinities for nitrite decreased in the strain KM1 ( Figure 4B). Intriguingly, the strain KM1 seemed to lose the ability to oxidize low concentration of nitrite in the condition where it was exposed to higher concentration of ammonia. The inhibitory constant for free ammonia (K i ) for the strain KM1 was calculated as 0.68 ± 0.03 mg-NH 3 L −1 = 40.0 ± 1.89 µM NH 3 , which was lower than that of lineage I Nitrospira sp. strain ND1 (8.5 ± 0.9 mg-NH 3 L −1 = 500 ± 52.9 µM NH 3 ), and lineage II N. japonica strain NJ1 (16 ± 2.3 mg-NH 3 L −1 = 941 ± 135 µM NH 3 ) . Additionally, the thermophilic NOB that is related to Nitrospira calida showed a tolerance for free ammonia with an IC50 value of 5.0 mg NH 3 -N L −1 (Courtens et al., 2016). Therefore, the strain KM1 was more sensitive to ammonia than other strains, such as ND1, NJ1, and thermophilic NOB. Since the Nitrospira NXR is located in the periplasm, nitrite transporters most likely play no role in nitrite oxidation (Lücker et al., 2010). Nitrite oxidation by the periplasmic NXR itself seems to be inhibited by ammonia (Courtens et al., 2016;Ushiki et al., 2017), and the mechanism is in need of further study.
The different sensitivity to free ammonia might distinguish the niches of different species of Nitrospira. Based on previous microscopic and spatial analysis, Nitrospira sp. belonging to lineage I tend to localize in the vicinity of AOB, whereas such close proximity was not observed for the Nitrospira sp. belonging to lineage II in nitrifying biofilms (Maixner et al., 2006). This unique distribution was explained based on the small-scale nitrite gradient that forms around the AOB. Furthermore, similar analysis revealed that the preferred nitrite concentrations were different among the members of Nitrospira  linage I (Gruber-Dorninger et al., 2015). In this context, a gradient of ammonia concentration in the biofilms might be formed depending on the distance from AOB. If ammonia influenced the nitrite affinity of Nitrospira species, ammonia seems to be a key factor that influences and regulates niche differentiation among different Nitrospira species. Apart from nitrite-oxidizing Nitrospira, N. inopinata, the only comammox pure culture, possesses a high affinity for ammonia and a high K m(app) value for nitrite and therefore appears to be well-adapted to oligotrophic environments (Kits et al., 2017). Considering this point, comammox Nitrospira might flourish more than nitriteoxidizing Nitrospira in low ammonia conditions. If AOA and AOB with a high affinity for ammonia coexist with nitriteoxidizing Nitrospira, the competition between comammox Nitrospira and nitrite-oxidizing Nitrospira seems to maintain the balance in low ammonia conditions. Meanwhile, nitriteoxidizing Nitrospira with high affinity for nitrite would be more abundant than comammox Nitrospira in low nitrite and no ammonia conditions. Additionally, continuous feeding with low concentration of ammonia would be effective to enrich comammox Nitrospira as demonstrated in our previous study . However, there is a big hurdle to isolate uncultured comammox Nitrospira. Although pure microcolonies including comammox Nitrospira were probably sorted in this study, we could not confirm cell growth of comammox Nitrospira. Considering comammox Nitrospira could be adapted to lower oxygen concentrations , ammonia and oxygen concentration should be strictly controlled in 96 well microtiter plates after sorting. Otherwise, small-scale devices to continuously feed substrate such as microfluidics might be effective to monitor cell growth in single cell levels during long-term incubation.

Genomic Composition
Genomic properties of the strain KM1 are summarized in Table 2. The sequence coverage was 119.5-fold genome equivalents. The sequence reads obtained from the strain KM1 were assembled into one contig. The reconstructed complete genome sequence of the strain KM1 is 4,509,223 bp in length and with a 56.0% GC content, 4,318 predicted CDS, 46 tRNA, and single copy rRNA genes (5S, 16S, and 23S). The strain KM1 shares high similarity to N. japonica strain NJ1 with respect to the 16S rRNA (98.9%), and nxrB (93.2%) gene sequences. The genome of strain KM1 contains no amoA gene. The ANI between the genome sequences of Nitrospira strains ranges from 76.7% to 78.4%. The predicted metabolic features encoded by the genome of strain KM1 are schematically represented in Supplementary Figure S7. We rechecked the purity of strain KM1 from whole genome sequencing data and found no contamination.
Like other Nitrospira sp., gene sets belonging to the nitrite oxidation pathway and the tricarboxylic acid cycle were identified in the genome sequence of strain KM1. NXR is a key enzyme that catalyzes nitrite oxidation, and is encoded by nxrA, nxrB, and nxrC genes coding for the alpha, beta, and gamma subunits, respectively (Daims et al., 2015). Three copies of potential nxrA and nxrC genes and two copies of nxrB gene were also found in the genome sequence. One of the three nxrA genes The experiments were performed using biological triplicates. The values outside the mean ± standard deviation (68.3% confidence interval) of three data in each condition were removed, which are shown as gray circles. The shown mean and standard deviation were recalculated using the remaining plots. All the kinetic data is shown in Supplementary Table S3.  has a length of 4515 bp, whereas the others are 3437 and 3446 bp in length. A gene likely encoding a sigma-54 dependent transcriptional regulator is directly located in the upstream region of the two nxrA genes. Considering the fact that the NXR operons in most Nitrospira strains are preceded by sigma-54 (Lücker et al., 2010;Koch et al., 2015), the expression of multicopy nxrA and nxrB genes might be controlled by the regulator. However, recent transcriptional analysis revealed that the sigma-54 dependent transcriptional regulator exhibited low expression levels during nitrite oxidation in Nitrospira moscoviensis (Mundinger et al., 2019). Nitrite-oxidizing Nitrospira utilize nitrite as a major source of nitrogen as well as energy. For reduction of nitrite to ammonia, a gene likely encoding octaheme cytochrome c nitrite reductase (ONR) was found in strain KM1, similar to Nitrospira moscoviensis (Koch et al., 2015). NirA-dependent ferredoxin relevant to nitrite reduction was identified in N. defluvii, Nitrospira sp. ND1 in linage I, and Nitrospira lenta in lineage II (Lücker et al., 2010;Koch et al., 2015;Ushiki et al., 2018). Additionally, Nitrospira inopinata as a comammox species has a potential nrfAH gene, encoding cytochrome c nitrite reductase (Daims et al., 2015). These findings suggest that members of the genus Nitrospira possess versatile nitrogen assimilation ability. Genes likely relevant to nitrogen assimilation include glnA, glutamine synthetase and glnK, and the P-II protein controlling nitrogen assimilation were located close to ONR in the genome of the strain KM1.
In the urease and urea transporter gene clusters, the strain KM1 lacks ureE and ureH genes, which encode for urease accessory proteins. UreDEFG is an accessory protein operon that supports the maturation of urease and controls the activity of urease (Lee et al., 1992). Urease contains Ni in its active center. UreE and UreH function as a Ni donor and Ni transporter, respectively (Mobley et al., 1995). Although N. moscoviensis lacks ureE gene, the hydrogenase maturation factors HypA and HypB are expected to exhibit the same function as UreE (Koch et al., 2014(Koch et al., , 2015. Nitrospira sp. ND1 was shown to perform urease degradation despite the lack of a ureE gene, which indicates that urease might be activated with an increase in Ni concentration in the cells because the strain ND1 possesses a gene likely encoding nickel/cobalt transporter with high affinity (Ushiki et al., 2018). However, no chaperon protein and nickel/cobalt transporter with high affinity was identified in the genome sequence of the strain KM1 and may result in low urease activity.
Cyanate hydratase and three copies of an Amt-type ammonium transporter, which may help utilize ammonium as nitrogen source, were also identified in the KM1 genome. Intriguingly, nitrite-oxidizing Nitrospira encode three homologs of Amt-type transporters clustering separately in phylogenetic analysis . If Amt-type transporters among Nitrospira are functionally different, Nitrospira might possess the potential to survive in environments with fluctuating ammonia concentration. N. japonica reduced transcripts of the gene likely encoding octaheme cytochrome c nitrite reductase (NSJP_2412) in medium amended by nitrite and ammonium compared to medium with nitrite as a sole nitrogen source (Ushiki et al., 2018). Ammonium addition can also promote growth and allow cultivation of NOB from other genera (Sorokin et al., 2012;Sayavedra-Soto et al., 2015;Ishii et al., 2017;Wegen et al., 2019;Spieck et al., 2020). Therefore, some types of NOB might utilize ammonium from their surroundings instead of assimilatory nitrite reduction. Although it is unknown whether the strain KM1 utilizes ammonium as a nitrogen source, ammonium availability in the environment for NOB would be an important factor. Upon comparing the genomes of strain KM1, N. moscoviensis, N. japonica, and N. lenta within Nitrospira lineage II, 2035 out of 3917 clusters were shared among all the four strains as orthologous clusters (Figure 5). Carbonic anhydrase (CA) that catalyzes reversible the reaction between carbon dioxide and carbonic acid was identified in strain KM1 and N. lenta genome sequences (Sakoula et al., 2018). Additionally, two isolated Nitrospira strains in lineage I have CA (Lücker et al., 2010;Ushiki et al., 2018). According to previous studies, the growth of Haemophilus influenza lacking CA was inhibited by atmospheric concentrations (∼0.04% CO 2 ), which indicates that CA plays an important role in CO 2 poor conditions (Langereis et al., 2013). Considering that CA is essential for survival in many microorganisms (Kusian et al., 2002;Aguilera et al., 2005;Burghout et al., 2011), CA might be one of the factors that could facilitate ecological niche differentiation among different Nitrospira species. Additionally, 18 sets of genes likely encoding for a toxin-antitoxin (TA) system in a stress response module (Williams and Hergenrother, 2012) were identified in the genome of strain KM1. In Nitrosomonas europaea, a toxin endoribonuclease designated as MazF NE1181 may improve a resistance to mercury in heavy metal stress environments (Miyamoto et al., 2016). This suggests that strain KM1 might respond to different environmental stresses using its TA system.

CONCLUSION
This study reported the isolation, morphology, physiology, kinetics, and genome of a novel nitrite-oxidizing Nitrospira strain from a fixed-bed column present at a drinking water treatment plant. Although morphological and genomic features were typical of Nitrospira, the values of K m(app) and V max for nitrite of the strain KM1were orders of magnitude lower than the cultured NOB in non-marine habitat. Therefore, the strain KM1 is well-adapted to oligotrophic environments with a high affinity for nitrite. To bridge the gap between in situ kinetic values and kinetic values of the cultured representatives, further studies including the isolation and characterization of novel NOB strains would be required. Additionally, strain KM1 had high sensitivity to free ammonia. Considering ammonia influenced K m(app) for nitrite, ammonia as well as nitrite concentration could be the main determinants for niche differentiation among NOB. These findings broaden known physiological features of nitriteoxidizing Nitrospira, an ecologically important group in the biogeochemical nitrogen cycle.

AUTHOR CONTRIBUTIONS
HF and ST contributed to the conception and design of the study. HF and KM sampled out the zeolite from the drinking water treatment plant and enriched the nitrifiers. KM and NU performed the isolation of nitrifiers and cloning based on 16S rRNA gene. MN and KI carried out the cloning based on amoA gene. KI conducted the physiological experiments. HF and KI analyzed the physiological data. YS performed the genome sequencing and assembly. HF, KI, and SK carried out the genome analysis. HF, KI, YS, and ST wrote the manuscript. All authors contributed to the article and approved the submitted version.