Collaborative metabolisms of urea and cyanate degradation in marine anammox bacterial culture

Abstract Anammox process greatly contributes to nitrogen loss occurring in oceanic oxygen minimum zones (OMZs), where the availability of NH4+ is scarce as compared with NO2−. Remineralization of organic nitrogen compounds including urea and cyanate (OCN−) into NH4+ has been believed as an NH4+ source of the anammox process in oxygen minimum zones. However, urea- or OCN−- dependent anammox has not been well examined due to the lack of marine anammox bacterial culture. In the present study, urea and OCN− degradation in a marine anammox bacterial consortium were investigated based on 15N-tracer experiments and metagenomic analysis. Although a marine anammox bacterium, Candidatus Scalindua sp., itself was incapable of urea and OCN− degradation, urea was anoxically decomposed to NH4+ by the coexisting ureolytic bacteria (Rhizobiaceae, Nitrosomonadaceae, and/or Thalassopiraceae bacteria), whereas OCN− was abiotically degraded to NH4+. The produced NH4+ was subsequently utilized in the anammox process. The activity of the urea degradation increased under microaerobic condition (ca. 32–42 μM dissolved O2, DO), and the contribution of the anammox process to the total nitrogen loss also increased up to 33.3% at 32 μM DO. Urea-dependent anammox activities were further examined in a fluid thioglycolate media with a vertical gradient of O2 concentration, and the active collaborative metabolism of the urea degradation and anammox was detected at the lower oxycline (21 μM DO).


Introduction
The global nitrogen cycle is a pivotal biogeochemical cycle on Earth, and massive nitrogen loss (N-loss) is occurring in the ocean, i.e. >240 Tg of nitrogen has been released in the form of N 2 gas from the ocean into the atmosphere [1].Oxygen minimum zones (OMZs) have been recognized as hotspots for oceanic Nloss, as 25%-35% of the N-loss occurs in the OMZs, even though the OMZs account for <1% of the ocean volume (when defined by dissolved O 2 , DO, ≤20 μM) [1][2][3].The anaerobic ammonium oxidation (anammox) process, where NH 4 + is oxidized to N 2 gas using NO 2 − as an electron acceptor [4], significantly contributes to oceanic N-loss occurring in the OMZs.For example, a previous survey of the Benguela and Eastern Tropical North Pacific OMZs showed that 30%-50% of N 2 gas released from the OMZs has been produced through the anammox process [5][6][7][8].Anammox bacteria are affiliated into a monophyletic group in the bacterial order Brocadiales of the phylum Planctomycetota [9][10][11], and the members affiliated into the genus Candidatus Scalindua are often regarded as marine anammox bacteria because most anammox bacterial 16S rRNA gene sequences retrieved from marine environments were affiliated into the Scalindua group [12,13].NH 4 + availability becomes a rate-limiting factor of the anammox process occurring in oceanic OMZs [14][15][16][17] because NH 4 + is often scarce in OMZ core water as compared with NO 2 − .For example, it is in the range of <1 μM in the Arabian Sea and a few μM in the Eastern Tropical South Pacific OMZ cores, respectively [18][19][20][21].Remineralization of simple organic nitrogen compounds including coupled with microaerobic microbial respiration, denitrification, dissimilatory nitrite reduction to ammonium, and sulfate reduction can supply NH 4 + for the anammox process [8,15,17].Simple organic nitrogen compounds such as urea and cyanate (designated here as OCN − ) are believed to be a source of NH 4 + for the anammox process [8,16,[21][22][23][24].Indeed, the activities of anammox bacterial N 2 gas production in oceanic OMZs were correlated with the availability of organic nitrogen compounds [16,24,25].Urea is the primal product of the dissolved organic matter degradation and can be decomposed to CO 2 and NH 4 + enzymatically by urease (Ure), a Ni 2+ -containing metalloenzyme.Urea is excreted from zooplanktons and animals [23,26], and available at a few 100 μM to submicromolar range in oxycline and OMZ core, respectively [21].Other than urea, OCN − is also available in oceanic OMZs (e.g.oxycline and OMZ cores of the Eastern Tropical South Pacific and Eastern Tropical North Pacific OMZs) [21,23], whereas the concentration ranges of OCN − were lower than those of urea (up to a only few 10 nM).OCN − is produced from algal decomposition and from abiotic urea degradation [21,27], and degraded to CO 2 and NH 4 + by cyanate hydratase (cyanase, Cyn) [28] and also by abiotic hydrolysis [29].The occurrence of urea and/or OCN − -dependent anammox process in oceanic OMZs has been proposed in the early studies.The previous 15 Nurea or -OCN − incubation study using the marine water samples collected from the Eastern Tropical South Pacific OMZ revealed the production of the 14-15 N 2 gas derived from the anammox process [18].However, the urea and ONC − degradation by marine anammox bacteria have not been demonstrated so far, and it is yet to be examined whether Scalindua bacteria degraded urea and/or OCN − or coexisting microorganisms were involved in the degradation.Furthermore, the gene sets involved in the urea or OCN − degradation were not commonly found in the known Scalindua genomes, suggesting that the ability to degrade urea or OCN − is not a common physiological trait of the Scalindua group.Therefore, cultivation-based analysis is definitely required to provide a direct evidence of urea and OCN − degradation by Scalindua bacteria.Therefore, the present study aimed to examine the degradation pathway of urea and OCN − in a marine anammox bacterial consortium.Intriguingly, the marine annamox bacterium enriched from coastal sediment in Hiroshima Bay, Candidatus Scalindua sp.husus a7 (designated as Scalindua sp.), did not have both gene sets involved in the urea or OCN − degradation, although their anammox activities were found when the Scalindua enrichment culture (>98% of total cells) was incubated with the addition of urea or OCN -and NO 2 − .It is hypothesized that urea and OCN -were degraded into NH 4 + by coexisting bacteria and/or abiotically, and the produced NH 4 + were sequentially utilized by the Scalindua sp.To test this hypothesis, the enrichment culture was anoxically incubated with the addition of urea or OCN − , and the activities of urea degradation, OCN − degradation, and anammox process were carefully examined by 15

Scalindua biomass
Planktonic cells of Scalindua sp.obtained from a coastal sediment of the Hiroshima bay were enriched using a membrane bioreactor equipped with a hollow fiber membrane module (pore size 0.1 μm, polyethylene) as previously described [30].The culture media fed into the membrane bioreactor (MBR) contained KH 2 PO 4 (24.4 mg l −1 ), MgSO 4 7H 2 O (60 mg l −1 ), CaCl 2 (51 mg l −1 ), yeast extract (Becton, Dickinson and Company, NJ) (1.0 mg l −1 ), an artificial sea salt SEALIFE (28 g l −1 ) (Marine Tech, Tokyo, Japan) [31], and 0.5 ml of trace element Solution I and II [32].Equimolar amounts of NH 4 (SO 4 ) 2 and NaNO 2 were supplemented into the media at 10 mM, and nitrogen loading rate of the MBR was set at 0.45 kg-N m −3 day −1 .The MBR was operated at 25 • C in dark without pH control, but the pH was in the range of pH 7.6-8.0.Scalindua sp. cells accounted for more than 90% of the total biomass in the MBR, which was further enriched by the buoyant density separation using Percoll solution (Cytiva) as previously described [33].The Percoll-separated Scalindua biomass was washed with the culture media without NH 4 + and NO 2

−
two times, and subjected to the following experiments.The abundance of Scalindua sp. cells in the Percoll-separated biomass increased to >98% of the total cells in the MBR, as determined by the f luorescence in situ hybridization analysis using the Scalindua-16S rRNA gene-specific Sca1129b oligonucleotide probe [31] (Supplementary Fig. S1).

Activity tests
Standard anaerobic techniques were employed in an anaerobic chamber (Coy laboratories Products, Grass Lake, MI) where oxygen concentration was maintained at lower than 1 ppm.The culture media and stock solutions were prepared by purging N 2 gas for >30 min, repeatedly vacuuming and purging He gas, and leaving in the anaerobic chamber for >1 week to remove trace amounts of oxygen dissolved in the media.Scalindua biomass was resuspended in the above culture media without NH 4 + and NO 2 − at concentrations of 0.15-0.2mg-protein ml −1 , and 10-30 ml of the cell suspension was dispensed into 20-100 ml serum glass vials (Nichiden-Rika glass, Tokyo, Japan).
The headspace was replaced with He gas (>99.99995%)after sealing with butyl rubber stoppers and aluminium caps.When the Scalindua biomass was incubated under microaerobic conditions, 0.5-20 ml of ambient air was injected into the headspace of the closed vials using a gas-tight syringe (GL Science) or disposable plastic syringe (Terumo).No air was injected for the anoxic incubations.To convert the volume of the injected air to the initial DO concentrations in the liquid media, the standard curve of the volume of the injected air (ml vial −1 ) versus the measured DO concentrations in the liquid media (μM) was prepared as previously describe by the authors [34].Brief ly, the vials without Scalindua biomass were incubated at 25 • C for 12 h after the air injection, and the DO concentrations in the liquid media were determined using an O 2 microsensor (Unisense oxygen needle sensor OX-N 13621) (Unisense, Denmark).The initial DO concentrations were determined using the regression line of the prepared standard curve (Supplementary Fig. S2).The DO concentrations were not controlled during the microaerobic incubation, and more than half of the injected O 2 remained in the head space at the end of the incubation.The vials were incubated without the addition of substrate in the dark at 25 • C for 12 h, then the anoxic stock solution of the following substrate(s) was added at final concentration of 1-3 mM: Na 14 NO 2 − (Fujifilm Wako), Na 15 NO 2 − ( 15 N content: >98 atom%) (Cambridge Isotope Laboratories, Andover, MA), 14 N-urea (Fujifilm Wako), 15 N-urea (>98 atom%) (Shoko Science, Kanagawa, Japan), Na 14 NCO (Fujifilm Wako).Penicillin G (final concentration; 500 mg/l) (Sigma Aldrich), or ATU (50 mg/l) (Fujifilm Wako) was supplemented to inhibit the activities of coexisting bacteria or aerobic ammonia oxidizers, respectively.The vials were incubated in dark at 25 • C, and liquid samples were collected using a 1-ml plastic disposable syringe, immediately filtered using 0.2-μm cellulose acetate filter, and subjected to determination of urea, OCN − , NH 4 + , NO 2 − , and NO 3 − concentrations.Gas samples were collected using a gas-tight glass syringe and immediately injected to a gas chromatograph.
N-loss (μmol-N vial −1 ) from the liquid phase of the vials during the incubation was calculated using Equation (1) where N: initial amounts of each nitrogen compounds fed into vials (μmol-N vial -1 ), N * : the amounts of each nitrogen compounds at the end of the incubation (μmol-N vial −1 ).The contribution of the anammox process to the N-loss was calculated by using Equation (2), where N 14-15N2 * (μmol vial −1 ) indicated the amounts of 14-15 N 2 gas produced by the end of the incubation.

Chemical analysis
Urea concentration was determined as NH OCN − concentration was determined colorimetrically as previously described by Guilloton et al. [35].Brief ly, 0.5 ml of liquid sample was mixed with 0.5 ml of 10 mM anthranilonitrile solution, and incubated at 40 • C for 20 min.One milliliter of 12N HCl was added into the reaction mixture, and the mixture was further incubated at 99 • C for 5 min.After cooling down to room temperature, the absorbance was measured at 310 nm.Since NO 2 − in the liquid sample increased the absorbance at 310 nm, NO 2 − concentration was determined prior to OCN − determination, and the absorbance derived from the coexisting NO 2 − was then subtracted to calculate the OCN − concentration specifically.NH 4 + concentration was determined colorimetrically by the indophenol method [36] and by f luorometrically using the ophthalaldehyde (OPA) method [37].Colorimetric determination was performed as described elsewhere.Brief ly, liquid samples were mixed with phenol and hypochlorous acid solutions, and absorbance was measured at 635 nm.As for f luorometric determination, liquid samples were mixed with 3.8 mM OPA, and f luorescence intensity was determined at 355 nm of excitation and 460 nm of emission.NO 2 − concentration was determined colorimetrically using the naphthylethylenediamine method [36].Liquid samples were mixed with a naphthylethylenediamine-sulfanilamide solution, and absorbance was measured at 540 nm.NO 3 − concentration was determined using an ion chromatograph IC-2010 equipped with the TSKgel SuperIC-Anion HS column (Tosoh, Tokyo, Japan). 15N atom percent of NO 2 − was determined by matrix-assisted laser desorption ionization-time-of-f light mass spectrometry as previously described by the authors [38].Brief ly, liquid samples were mixed with naphthylethylenediamine reagent to develop NO 2 − -complex azo dyes, loaded onto a MALDI sample plate (MTP 384 target plate, ground steel BC, Bruker Japan, Yokohama, Japan) with α-cyano-4-hydroxy-cinnamic acid matrix.The peak-area ratios of m/z of 371 and 372 were determined using a Bruker Ultraf lex III (Bruker Japan), and the 15 N atom% was calculated using a calibration curve prepared using standard 15 N-labeled compounds.
Protein concentration was determined by the Lowry method using a DC protein assay kit (Bio-Rad, Hercules, CA).Bovine serum albumin (Fujifilm Wako) was used as a protein standard.
Concentrations of 14-15 N 2 and 15-15 N 2 were measured by gas chromatography mass spectrometry [39].About 50 μl of the headspace gas were collected using a 100-μl gas-tight glass syringe, and immediately injected into a gas chromatograph GCMS-QP 2010 SE (Shimadzu, Kyoto, Japan) equipped with a fused silica capillary column (Agilent Technologies, Santa Clara, CA).Peaks at m/z = 29 and 30 corresponding to 14-15 N 2 and 15-15 N 2 were monitored, and the concentrations of the N 2 gas were calculated using a standard curve prepared using the 15-15 N 2 gas (Cambridge Isotope Laboratories).

Enrichment of aerobic ureolytic bacteria
Thirty milliliter of the Percoll-separated Scalindua biomass was dispensed into a 100-ml glass vials with a silicon sponge plug (Silicosen, Shin-Etsu Polymer).After the addition of 1 mM 14 N-urea, the vials were incubated at 25 • C without shaking (i.e.standing culture).Urea degradation in the culture was routinely monitored by determining urea, NH 4 + , NO 2 − , and NO 3 − concentrations.After 32 days of the aerobic incubation, stock solution of 14 N-urea was supplemented as a spike.After the depletion of the added 14 Nurea, 10% of the culture was transferred into the fresh media containing 1 mM 14 N-urea, and the incubation was repeated.This subculturing was repeated eight times, and the enrichment culture after 104 days of incubation was subjected to activity test, amplicon sequencing, and metagenomic analyses.

Incubation of the Scalindua biomass in a fluid media
A f luid media was prepared as the conventional f luid thioglycolate media.The 15 N-urea, 14 NO 2 (each 3 mM), 0.5 g/l Na 2 S, 1 mg/l resazurin, and 0.15% (w/v) gellun gum were added into the inorganic media (pH 7.5), and autoclaved after 30 min of N 2 purge.About 50 ml of the autoclaved media was dispensed into a 69-ml glass vials in a clean bench (i.e.aerobic condition), and capped with butyl rubber cap and aluminum caps.The vials were incubated for 12 h without shaking at 25 • C to make a vertical O 2 gradient inside of the media.The Percoll-separated Scalindua biomass (0.5 ml of the culture corresponding to 4 mgprotein) was inoculated onto the surface of the media using a gas-tight syringe and incubated at 25 • C for 28 days.After the incubation, DO and NH 4 + concentrations in a f luid media were determined using the needle O 2 microsensor and a liquid ion exchange membrane (LIX) microsensor, respectively [40].The vial was uncapped under ambient air, and lifted up using a lab jack to reach the tip of the microsensor fixed on a clamp arm (Supplementary Fig. S3).The O 2 microsensor was calibrated as following the instruction manual provided by the manufacture, and the sensor signals were recorded using a UniAmp Multi Channel.Ammonium-selective LIX microsensor was fabricated and calibrated as previously described by de Beer et al. [41].The steady-state concentration profiles of DO and NH 4 + in a f luid media were measured as previously described [40].
After the microsensor measurements, core samples were collected by inserting a sterile 10 ml plastic disposable pipette ( 7.5 mm) (Supplementary Fig. S4) vertically into the f luid media and dispensed into a sterile 1.5 ml plastic centrifuge tube, which was subjected to DNA extraction using Fast DNA SPIN kit for Soil (MP Biomedicals).Quantitative PCR assay for determining Scalindua 16S rRNA gene copy number was carried out by TaqMan-based real-time PCR assay as previously described [42].Premix Ex Taq (Takara Bio) and ABI prism 7500 sequence detection systems were used for the assay, and the standard curve for quantification was prepared by the 10-fold dilutions of plasmid DNA containing the Scalindua 16S rRNA gene molecule ranging from 10 7 to 10 3 copies μl −1 .
Sequence reads with the used oligonucleotide primer sequences were extracted using the fastx_barcode_splitter tool of the FASTX-Toolkit (ver.0.0.14)(http://hannonlab.cshl.edu/fastx_toolkit/)and imported into the Qiime 2 (ver.2021.4).In the Qiime 2 program [45], the removal of low-quality and/or chimeric sequence reads, as well as the removal of regions corresponding to the oligonucleotide primer and 50 bp of 3 end, and the clustering of sequence reads into OTUs were carried out using the DADA2 plugin [46].Phylogeny of the 16S rRNA gene OTUs was examined by the blastn (ver.2.9.0) program using the Greengene (ver.13_8) database, whereas the fungene (accessed on February 2023) and nr database (February 2021) were used for ureC OTUs.

Accession numbers
Metagenomic sequence data and the sequence reads of 16S rRNA gene and ureC amplicons are available in the DDBJ nucleotide sequence database under the accession number DRA016463, DRA016464, and DRA016465, respectively.

Urea and organic nitrogen compounds such as urea and cyanate degradation under anoxic condition
Highly enriched Scalindua biomass (>98% in total cells, Supplementary Fig. S1) was anoxically (<1 μM DO) incubated with the addition of (a) 14 N-urea and 15  (each 3 mM).The consumption of 15 NO 2 − occurred concurrently with the production of 14-15 N 2 (Fig. 1A), which indicated that 14 Nurea was degraded into 14 NH 4 + , and the produced 14 NH 4 + and 15 NO 2 − were subsequently utilized for the anammox process.
Similar behavior was found when OC 14 N − was fed instead of 14 Nurea (Fig. 1D).The rates of the 14-15 N 2 production in Fig. 1A and D were in the range of 0.1-0.2μmol-N mg-protein −1 h −1 , which was an order of magnitude lower than the maximum specific anammox activity for the Scalindua sp.[34].Other than the 14-15 N 2 gas, 15-15 N 2 gas production also occurred during the incubation, which was likely produced via the denitrification of 15 NO 2 − (see below).
Occurrence of abiotic 14 N-urea and OC 14 N − degradations was examined by repeating the above anoxic incubation without the Scalindua biomass.The 14 N-urea and 15 NO 2 − were not consumed, and no 14-15 N 2 and 15-15 N 2 gas was produced (Fig. 1B); therefore, the 14 N-urea degradation was a biological process.On the other hand, the OC 14 N − degradation occurred with NH 4 + accumulation during the abiotic incubation with the rate of 3.41 μmol-N vial −1 day −1 (Fig. 1E).This abiotic OC 14 N − degradation rate was comparable with the biotic OC 14 N − degradation rate of the Scalindua biomass, i.e. 2.65 μmol-N vial −1 day −1 estimated from the 14-15 N 2 gas production rate in Fig. 1D.Therefore, abiotic degradation was responsible for the OCN − degradation during the anoxic incubation.
The biological urea degradation by the Scalindua biomass was further investigated by adding the penicillin G which is an inhibitor of coexisting bacteria without inhibitory effects against anammox bacteria [53].The addition of 500 μg/ml penicillin G resulted in no consumption of 14 N-urea and 15 NO 2 − and no production of N 2 gas (Fig. 1C), indicating that urea degradation was carried out by coexisting bacteria and not by Scalindua sp.When penicillin G was added with OC 14 N − and 15 NO 2 − , 14-15 N 2 gas but no 15-15 N 2 gas was produced (Fig. 1F).This result further supports that OC 14 N − was degraded abiotically, and the produced 14 NH 4 + and 15 NO 2 − were converted into 14-15 N 2 gas by Scalindua sp., and the 15-15 N 2 gas (Fig. 1A and D) was produced via the vials without Scalindua biomass were also prepared and incubated in parallel as abiotic controls (B and E); penicillin G was supplemented at a final concentration of 500 mg/l to inhibit the activities of coexisting bacteria (C and F); the incubations were performed in duplicate, and the plot symbols represented the mean values; the outcomes of the duplication incubations are shown in Supplementary Figs S5 and S6).
denitrification by coexisting bacteria.The absence of 15-15 N 2 gas production also indicates that Scalindua sp. did not dissimilatory reduce 15 NO 2 − to 15 NH 4 + [54, 55], and did not perform the anammox process using the formed 15 NH 4 + .

Urea degradation under microaerobic condition
The Scalindua biomass was incubated under anoxic (<1 μM DO) and microaerobic conditions (ca.5-42 μM DO) with the addition of 15 N-urea and 14 NO 2 − to examine the inf luence of DO concentration on urea degradation.At <1, 5, and 11 μM DO, 14-15 N 2 gas was produced within 7 days at the rates of 0.57 ± 0.17, 0.86 ± 0.11, and 1.0 ± 0.23 μmol-N 14-15 N 2 vial −1 , respectively (Fig 2A -C).The accumulation of NH 4 + occurred after 6 days of the incubation without the production of 14-15 N 2 gas, which likely resulted from the depletion of NO 2 − required for the anammox reaction.Both NO 2 − and NO 3 − (derived from the ground water used for the preparation of culture media) were consumed within 7 days of the incubation, whereas the 14-15 N 2 and 15-15 N 2 gas production were minor as compared with the NO 2 − and NO 3 − consumption.This difference indicated that the denitrification process producing 14-14 N 2 gas from NO 2 − and/or NO 3 − (NO x − ) was responsible for the observed N-loss.At 32 and 42 μM DO, urea degradation and 14-15 N 2 gas production greatly occurred after 6 days of incubation, with rates of 6.56 ± 0.26 and 8.11 ± 1.71 μmol-N 14-15 N 2 vial −1 , respectively (calculated between 6 and 9 days and 7 and 9 days of the incubation) (Fig. 2D and E, respectively).These 14-15 N 2 gas productions occurred after the NH 4 + or NO 2 − accumulation period, indicating that NO 2 − or NH 4 + was the rate-limiting substrate at 32 and 42 μM DO, respectively.In addition to the 14-15 N 2 gas production, the production of 15-15 N 2 gas also occurred through both the anammox ( 15 NH 4 + and 15 NO 2 − ) and denitrification ( 15 NO x − ) process.The production of 15 NO 2 − from 15 N-urea was examined by determining the 15 N/ 14 N atom ratio of the accumulated NO 2 − at 42 μM DO.The 15 N/ 14 N atom ratio was >40 atom%, which was significantly higher than that of the NO 2 − fed at the batch incubations (i.e.<5 atom%), indicating the production of 15 NO 2 − .
The N-loss and the contribution of the anammox process increased when the initial O 2 concentration was greater than 32 μM (Table 1).The highest N-loss and the greatest contribution of the anammox process were found at 32 μM O 2 , where 57.5 μmol-N/vial of N-loss and 33.3% of N-loss occurred via the anammox process.It is notable that the contribution was calculated from the amounts of 14-15 N 2 gas without 15-15 N 2 gas because the 15-15 N 2 gas could be produced through both the anammox and denitrification process as mentioned above.Thus, our calculation likely underestimated the contribution of the anammox process to the N-loss.

Scalindua biomass
Ureolytic bacteria were enriched by subculturing the Scalindua biomass with urea under aerobic condition.After 104 days of aerobic incubation with eight times subculturing, an enrichment culture showing high urea degradation activity was successfully obtained (Supplementary Fig. S7).Batch incubation of the enrichment culture was performed with the addition of urea under aerobic or anoxic conditions.The enrichment culture aerobically degraded the urea into NO 2 − via NH 4 + , but the urea degradation did not occur under anoxic condition (Fig. 3A and B, respectively).The addition of 500 μg/ml penicillin G inhibited the ureolytic activities of the enrichment culture (Fig. 3C) as previously found in the Scalindua biomass (Fig. 1C).The accumulation of NO 2 − during the above aerobic incubation (Fig. 3A) suggested that About 30 ml of the Scalindua biomass was incubated in 100-ml glass vials under different initial DO concentrations with the addition of 15 N-urea and 14 NO 2 − (each 1 mM).Behaviors of nitrogen compounds in liquid and gas phase are shown in Fig. 2. The values are the mean value and the range of standard deviation derived from three biological replicates.a Anoxic incubation without air injection.
aerobic ammonia oxidizers capable of NH 3 oxidation to NO 2 − were present in the enrichment culture.Therefore, the above batch incubation was repeated by adding allylthiourea (ATU), an inhibitor of autotrophic aerobic ammonia oxidizers [56,57].ATU strongly inhibited the aerobic urea degradation, indicating aerobic ammonia oxidizers were responsible for the aerobic urea degradation (Fig. 3D).

Urea degradation in oxycline and anoxic zone mimicking oxygen minimum zones
Urea degradation of the Scalindua biomass was further examined using a f luid thioglycolate media with a vertical O 2 gradient.The Scalindua biomass was incubated in a f luid media fed with 15 N-urea and 14  appeared at a depth of 7 mm.Anammox activities were examined by determining the production of the 14-15 N 2 gas in the headspace of the vials.The involvement of nitrification-denitrification process in the 14-15 N 2 gas production must be minor because the concentration range of the determined NH 4 + was two orders of magnitude lower than the initial NO 2 − concentration (i.e.<11.5 μM and 3 mM, respectively).A peak of the Scalindua 16S rRNA gene copy numbers was detected at 3 mm from the surface where the first NH 4 + peak was detected, which indicated the occurrence of 15 N-urea dependent anammox activities (Fig. 4B).
When the above incubation was repeated without the Scalindua biomass (i.e.abiotic incubation), no increase in NH 4 + concentration was found in the f luid media (Supplementary Fig. S8A), indicating that urea degradation did not occur.Furthermore, there was no increase in the Scalindua 16S rRNA gene copy number, and no production of 14-15 N 2 and 15-15 N 2 gas was found when incubated with the addition of 15 N-urea but not 14 NO 2 − (Supplementary Fig. S8B and C).

Ureolytic bacteria in the Scalindua biomass
Amplicon sequencing analysis of prokaryotic 16S rRNA gene and ureC and also metagenomic analysis was carried out for the Scalindua biomass and the aerobic enrichment culture.In the Scalindua biomass, Planctomycetota (i.e.Scalindua) 16S rRNA gene was the most abundant (>97% of total reads), and the Pseudomonadota, Cyanobacteriota, Chlorobiota, Bacteroidota, and Acidobacteriota 16S rRNA genes were also detected (Supplementary Fig. S9).Four highquality bacterial bins, Scalinduaceae bin1, Rhodobiaceae bin1, Rhizobiaceae bin1, and Thalassobaculales bin1 were obtained from the Scalindua biomass (Supplementary Table S1), and their metabolic potentials are summarized in Supplementary Fig. S10.As for the Scalinduaceae bin1, the gene sets involved in the anammox process and CO 2 fixation via the Wood-Ljungdahl pathway were located in the bin (Supplementary Fig. S10).However, the gene sets required for urea degradation (i.e.ureCBADEFG, urea ABC transporter, urea amidolyase, and urea carboxylase) and also OCN − degradation (cynS encoding Cyn) were absent in the Scalinduaceae bin1 (Fig. 5).This finding was consistent with the above outcomes of the 15 N-tracer incubation (Fig. 1), i.e. coexisting bacteria other than Scalindua sp. and abiotic reactions were responsible for urea and OCN − degradation, respectively.The gene sets involved in ureolysis were located in the Rhizobiaceae bin1 (ureABCDEFG, and urea ABC transporter), affiliated with the bacterial family Rhizobiaceae, but were not found in the Rhodobiaceae bin1 and Thalassobaculales bin1 (Fig. 5).The ureolytic bacteria in the Scalindua biomass were further examined by the ureC-amplicon sequencing analysis.The ureC reads affiliated into the Rhizobiaceae, Nitrosomonadaceae, or Thalassospiraceae were found in the Scalindua biomass (Supplementary Fig. S11A).The Rhizobiaceae ureC reads were identical to the ureC sequence located in the above Rhizobiaceae bin1.The detection of the Nitrosomonadaceae and Thalassospiraceae ureC reads suggested that those bacteria other than Rhizobiaceae bacteria participated in the urea degradation of the Scalindua biomass.It should be noted, however, that the forward oligonucleotide primer used for the PCR amplification of ureC had some mismatched bases against the ureC sequence found on the obtained bins (Supplementary Fig. S11B); therefore, the abundance of the ureC reads shown in Supplementary Fig. S11A may not ref lect the in situ ureC community structure due to PCR amplification bias.

Discussion
The present study is the first to describe the urea-and OCN − -dependent anammox activities in detail using a highly enriched marine anammox bacterial culture.Our 15 N-tracer incubations have revealed that the collaborative metabolism of the urea degradation by coexisting bacteria and anammox bacteria occurred in the Scalindua biomass (Fig. 1A-C).The Rhizobiaceae, Nitrosomonadaceae, and Thalassospiraceae bacteria were the candidates of the ureolytic bacteria as examined by the ureC amplicon sequencing analysis (Supplementary Fig. S11A).Among those, Nitrosomonadaceae and Thalassospiraceae bacteria were successfully enriched in the aerobic enrichment culture, and the urea degradation of the enrichment culture only occurred under aerobic condition (Fig. 3).Therefore, the rest of the above Figure 3. Urea degradation by aerobic enrichment culture; the Percoll-separated Scalindua biomass was subcultured eight times under aerobic condition with the addition of 1 mM urea as shown in Supplementary Fig. S5; 30 ml of the enrichment culture was dispensed into 100-ml glass vials with the addition of 1 mM 14 N-urea and incubated under the following conditions: aerobic condition (A), anoxic condition (B), aerobic condition with penicillin G (C), and aerobic condition with ATU (D); the symbols and error bars represent the mean value and the range of standard deviation derived from three biological replicates, respectively.candidates, Rhizobiaceae bacterium, was likely involved in the urea degradation of the Scalindua biomass under anoxic condition.
Ureolysis by the Rhizobiaceae bacteria has been demonstrated using some pure cultures (e.g.Rhizobium galegae and Agrobacterium strains) previously [58], while the Rhizobiaceae bin1 was affiliated into a phylogenetic clade without representative isolate.The closest relative of the Rhizobiaceae bin1 was the RICO01_sp004000235 bin, formerly obtained as dinof lagellates-associated bacterium, while the value of average nucleotide identity between the Rhizobiaceae bin1 was only 78.58, indicating that Rhizobiaceae bin1 represents a novel bacterial species; therefore, the metabolic capability of urea degradation by the Rhizobiaceae bin1 bacterium needs to be further investigated.As for the aerobic ureolytic bacteria, aerobic ureolysis was completely inhibited by adding ATU (Fig. 3).ATU is an inhibitor of copper-containing monooxygenases including ammonia monooxygenase and inhibits the growth of aerobic ammonia oxidizers including Nitrosomonadaceae [56].Nitrosomonadaceae bacteria can degrade urea using Ure and oxidize the formed NH 4 + to NO 2 − sequentially [59,60]; therefore, the aerobic ureolysis with NO 2 − accumulation by the enrichment culture (Fig. 3) indicated that Nitrosomonadaceae but not Thalassospiraceae bacteria was responsible for the aerobic ureolysis.Notably, inhibition of urea degradation by aerobic ammonia oxidizers by ATU (Fig. 3D) has not been well characterized, and the questions have been raised as to whether ATU directly inhibited enzymatic activities of Ure, a Ni 2+ -containing metalloenzyme [61], or not.
In contrast to the biotic degradation of urea, OCN -degradation occurred abiotically in the Scalindua biomass (Fig. 1).OCN -can be degraded abiotically into NH 4 + and CO 2 via carbamic acid [29], and the abiotic degradation has been found in a sterile seawater [22].Although the rate of abiotic OCN − degradation (344 μmol-N l −1 day −1 ) (Fig. 1F) was 10 3 order of magnitude higher than that previously determined using the seawater (0.463 μmol-N l −1 day −1 at 25 • C), it was reasonable by considering the initial OCN − concentrations at the incubations (i.e. 3 mM and 10 μM, respectively) because the abiotic degradation rate was logarithmically dependent on the OCN − concentration [29].It should be noted that the OCN − concentrations were found at the nanomolar range in the ocean [21], where the involvement of abiotic OCN − Scalindua 16S rRNA gene copy numbers, the measurements were performed using one vial, and the error bar was not available; (c) production of 14-15 N 2 gas in the head space of the vials; the symbols and error bars of the DO concentrations represent the mean value and the range of standard deviation derived from three biological replicates, respectively.
degradation would be minor.Phylogenetically diverse bacteria including marine Cyanobacteria [22,62] were capable of OCN − degradation using Cyn, and they can contribute to OCN − degradation and supply NH 4 + for anammox process [18,21].
Although Scalindua sp.itself was incapable of urea and OCN − degradation, the genes encoding Ure (ureC) and/or Cyn (cynS) were located in some Scalindua bacterial genomes [23,[63][64][65] and those transcripts were found in the Eastern Tropical North Pacific and Figure 5. Presence/absence of the genes involved in biological ureolysis, urea transport, and OCN -degradation on the metagenomic assembled bins obtained from the Percoll-separated Scalindua biomass and the aerobic enrichment culture; general features and phylogenies of the bins are available in Supplementary Table S1.
Eastern Tropical South Pacific OMZs [19], suggesting that some Scalindua bacteria are capable of direct degradation of urea and/or OCN − .Those metabolic capabilities would expand a habitat range of the Scalindua bacteria because the availability of NH 4 + is typically scarce in OMZ core as compared with NO 2 − , and the remineralization of urea and OCN − into NH 4 + would be advantageous and increase their competitiveness.The Scalindua sp.examined in the present study has been enriched from a sediment of Hiroshima bay, where in situ concentrations of NH 4 + , NO 2 − , and NO 3 − were determined to be 17.9, 2.9, and 6.4 μM, respectively [31,66].The NH 4 + was not a limiting substrate of the anammox process in the sediment, and the metabolic capabilities of urea and OCN − degradation were not beneficial as much as the Scalindua bacteria found in OMZ core; therefore, the Scalindua sp.genome did not have the gene sets involved in urea and OCN − degradation.In the sense, the metabolic capability of urea and OCN − degradation can define a niche of specific anammox bacteria.So far, nine anammox bacterial genera have been described in the order Candidatus Brocadiales (NCBI taxonomy database accessed on 13 November 2023), and different anammox species rarely coexisted in the culture [67,68].Several environmental factors including substrate concentrations and salinity (see the reviews, and the references in the review) [69,70] have been identified as a driving force shaping the specific niche, whereas urea and OCN − have not been recognized so far.It will be interesting to enrich anammox bacteria from environmental samples using urea or OCN − instead of NH 4 + as a substrate to examine whether urea or OCN − -degrading anammox bacteria can be enriched or not.Deeper understanding of urea and OCN − metabolisms by anammox bacteria advances our knowledge of microbial ecology of anammox bacteria and + production through the urea degradation and the growth of Scalindua sp.occurred concurrently in the lower oxycline (∼21 μM DO) (Fig. 4), indicating a vertical partitioning of microbial activities with depth.Although anammox bacteria have been recognized as an obligatory anaerobic bacteria, the recent culture-based analyses have revised this understanding [71].For instance, anammox bacteria and activities have been found from even microaerobic environments (<25 μM DO) [18,[72][73][74].As for the Scalindua sp., this marine anammox bacterium has a genetic potential required for detoxification and tolerance for oxygen, and anammox activity has been found even under microaerobic condition [34].The 50% inhibitory concentration and upper DO limits (IC 50 and DO max , respectively) for the anammox activity of the Scalindua sp. have been determined to be 18.0 and 51.6 μM, respectively, and thus this anammox bacterium has been recognized as a microaerotolerant bacterium.In the lower oxycline, the Scalindua sp.depended on the Rhizobiaceae, Nitrosomonadaceae, and Thalassospiraceae bacteria for feeding NH 4 + from urea degradation, and the possible collaborative metabolism occurred in the f luid media is shown in Fig. 6.In the oxycline, a variety of oxidation and reduction metabolisms can contribute to nitrogen conversion [15,71].When the Scalindua biomass was incubated at 5 and 11 μM was the rate-limiting step of the anammox activities at 32 or 42 μM DO, respectively (Fig. 2D and E).Therefore, the f luctuation of DO concentration even in the narrow range significantly inf luenced on the metabolisms of nitrogen compounds.In summary, a collaborative metabolism of the urea degradation by the Rhizobiaceae, Nitrosomonadaceae, and/or Thalassospiraceae bacteria and the anammox by the Scalindua sp. has been demonstrated under both the anoxic (<1 μM DO) and microaerobic (at 5-42 μM DO) conditions.In the case of OCN − , the abiotic degradation occurred under anoxic condition when OCN − was fed at 3 mM.The presence of DO (32-42 μM) accelerated the occurrence of the urea-dependent anammox, and the oxic/anoxic interface became a hotspot of N-loss in the f luid media.Apart from the above collaborative metabolism, direct utilization of urea and OCN − by some Scalindua bacteria is still required to be explored in other physiological studies using the Scalindua bacteria with ureC and/or cynS.

Figure 2 .
Figure 2. Urea degradation by the Scalindua biomass under anoxic and microaerobic conditions; 30 ml of the Percoll-separated Scalindua biomass was incubated in 100-ml glass vials at 25 • C in dark with the addition of 1 mM 15 N-urea and 1 mM 14 NO 2 − ; (A) anoxic incubation without air injection;(B-E) ambient air was injected into the headspace of the vials to set the initial DO concentrations to be 5, 11, 32, and 42 μM; the symbols and error bars represent the mean value and the range of standard deviation derived from three biological replicates, respectively.

NO 2 − 4 +
(each 3 mM), and vertical profiles of DO and NH 4 + concentrations were determined using the O 2 and LIXtype NH 4 + microsensors after 28 days of incubation, respectively.Oxycline was present to a depth of 5 mm, and the NH 4 + concentration increased in the lower oxycline (Fig. 4A).The highest NH concentration (11.5 μM) was found at a depth of 3 mm where DO concentration was 21 μM.The NH 4 + concentration decreased in the bottom oxycline, and then the second NH 4 + peak (10.4 μM)

Figure 6 .
Figure 6.Substrate cross feeding from ureolytic bacteria to Scalindua sp. in f luid thioglycolate media; in the f luid thioglycolate media, a vertical gradient of DO concentration is available due to diffusion of O 2 from atmosphere in the headspace of the vial; ureolytic bacteria degrade urea into NH 4 + , which is subsequently utilized by Scalindua sp. for the anammox process.
DO, the reduction reactions (i.e.denitrification) were prominent, and the urea degradation into NH 4 + was the rate-limiting step of the anammox activities (Fig 2B and C).In contrast, the oxidation reaction (NO 2 − oxidation to NO 3 − ) and urea degradation actively occurred under 32 and 42 μM DO, and the supply of NO 2 − or NH 4 +

Table 1 .
N-loss and the contribution of anammox process under different initial DO concentrations.